Found and graded S to C by a screening method I built and keep improving. Each source has an English digest and a link to the original.
179 sources so far, updated weekly.
Every tier S source, plus the tier A sources that scored highest in the screening. Newest first.
43% of workers in the 11 occupations work for employers that explicitly encourage chatbot use; 6% are prohibited.
2%largest average effect on earnings or hours that the estimates still allow, for workers and workplaces
Cost per MW: about $11 million for the building, $2 million for power infrastructure, $28 million for GPU racks and IT; a 200 MW campus costs about $8.2 billion, two-thirds of it IT equipment.
$41 billionestimated cost of a 1 GW AI campus ($41 million per MW)
About 36 million US jobs see reduced demand by 2035; about 41 million are created elsewhere.
about 11 millionUS workers (roughly 7% of the workforce) who would need to leave their occupations by 2035 in the base case
The Hugging Face incident occurred while OpenAI had monitoring for most internal use but not for evaluations, and the evaluation used a different harness and prompts than production Codex.
>100xreduction in propensity to compromise out-of-scope infrastructure when OpenAI used its production Codex harness and prompt
Value and growth: Agrawal, Gans and Goldfarb on "Genius on Demand"; Benjamin F. Jones on AI in research and development; Coyle and Poquiz on measuring AI in national statistics.
Modest: AI touches 4% of task instances by 2030; GDP +1.6%, growth 2.4% a year instead of 2%, labor share 60.0% to 59.4%, unemployment 3.9% against a normal 3.8%.
+8.3%GDP above the no-AI path in 2030, substantial scenario; the median US adult's answers imply +8.6%
44% say AI is scaling across the enterprise, up from 38%; 56% use AI in three or more functions, up from 51%; nearly nine in ten use it in at least one function.
37%respondents attributing any EBIT impact to AI, unchanged from 2025
Fact 1: average employment in the ADP sample rose about 6% from November 2022 to June 2026; the most exposed quintile grew about 4%.
19%employment shortfall for 22-25-year-olds in AI-exposed occupations vs less-exposed peers, June 2026
The full text is one paragraph of about 100 words, followed by a signature list of more than 200 names.
17signatories marked as Nobel laureates on the published list
Average OECD employment rate reached 72.1% in Q1 2026 and labour force participation 76.7%; the median unemployment rate rose from 5.4% in May 2025 to 5.7% in May 2026.
72.1%average OECD employment rate in Q1 2026, a historic high
22% of advertised jobs are professionalised, 52% democratised and 26% have low AI exposure.
22% vs 52%share of advertised jobs AI is professionalising (more expert) vs democratising (less expert)
Regular AI use (daily or several times a week): frontline 74% (51% in 2025), managers 88%, leaders 93%.
74%frontline employees who are regular AI users, up 23 points from 2025
Generative AI reached 53% population adoption within three years, faster than the PC or the internet; organizational adoption reached 88%.
88%organizational adoption of AI in 2025
69% of firms use AI on average: US 78%, UK 71%, Germany 65%, Australia 59%.
69%firms across the US, UK, Germany and Australia that currently use AI
Five categories of technological change: labour-augmenting, capital-augmenting, automating, expertise-leveling and new task-creating.
52%US workers surveyed in 2024 who are worried about how AI will affect their jobs
An AI pre-analysis of about 26 critical components, checked by component owners, produced a database of scenarios and test cases with priorities; the video description cites 28 components in five languages and more than 11,000 test cases, the talk more than 13,000.
70%pipeline runs that produced a successful merge request, out of about 1,500
Task suite: 11 tasks in five categories; models tested on Codex, Claude Code, Grok Build, Antigravity, Kimi Code and Qwen Code at their highest reasoning settings.
11 tasksreal Epoch work tasks across five categories in the launch suite
Skills chosen by GitHub stars on 28 September 2026, from superpowers (292,189) to compound-engineering (25,299); domain skills, memory systems and orchestrators excluded.
2 of 9popular Claude Code skills that beat a length-matched placebo, both on cost only (12% and 5% cheaper)
OpenAI run: GPT-5.6 Sol and GPT 6 Astra, over $400,000 in API-priced tokens, over 1.3 million lines of Rust over several months of goal loops, about 84% compatibility at best.
$400,000+API-priced tokens spent on GPT-5.6 Sol and GPT 6 Astra, which never got past about 84% compatibility
Survey of 2,500 workers: 11.31% said they had ever used AI on a Prolific submission when asked directly; an anonymous list experiment estimated 14.24%.
1.00%share of 712,930 Prolific submissions with observed ChatGPT assistance
Each agent received 3,000 GPU-hours on up to 50 GPUs and a 10 billion-token inference budget, in a sandbox without internet access, with a codebase scrubbed of SDPO.
35%share of the human method's gains reached by the best uncontaminated model (GPT-5.6 Sol), scored generously; 15% when only in-scope changes count
Knowledge cutoff June 2026; text output only.
1620 vs 735GDPval-AA v2.1 Elo on professional tasks for Haiku 5.5 versus Haiku 4.5 (Sonnet 5.5: 1840)
Anthropic, OpenAI and Google report AI writing 75% or more of their new code; Anthropic reported that as of August 2026 Claude "leads" 26% of its R&D work.
183%median increase in code submitted by regular contributors to open-source AI repos, Q1 2024 to Q2 2026
About 4,000 PRs a day pass through NeuroReview, roughly 40% of PRs created in Arcadia; use is voluntary and internal only.
~4,000pull requests a day reviewed by NeuroReview, about 40% of all PRs in Yandex's Arcadia monorepo
In a 30-day September 2025 sample, the time from a test-failure notice to the developer's next edit was 2.37 minutes at p10, 6.72 at p25 and 27.48 at p50; this is the window the agent targets.
67.18%correct fixes in expert review of 195 real test failures
Share of dissertations with AI-generated abstracts: 1.7% (2023), 5.5% (2024), 18.8% (2025), 29.4% (2026).
29.4%2026 US STEM dissertations with AI-generated abstracts (filed through May 2026)
2,611 production interview recordings scored in 14 batches between 26 March and 24 August 2026 on a nine-family rubric (structure, interruptions, latency, pacing, coherence, audio-visual sync, freezes, audio quality, outcome).
0.79 SDseverity gap between two reviewers scoring the same batches on the composite index
On 447 tasks, Claude Code and mini-SWE-agent differ by -1.8 points (Qwen3.6-35B-A3B) and -1.4 points (Qwen3.8-27B), equivalent within a 5-point margin.
13%median share of hard tasks whose outcome flips, both when swapping the harness and when rerunning the same one
ONET-Bench is built from 45,796 worker ratings of AI responses to ONET-derived tasks; the test split has 4,501 ratings.
3.0%-97.9%share of responses LLM judges rated acceptable, versus 61.1% for occupation-matched workers
In DeepSeek-V4-Flash-Base, a code completion flips between the correct "8" and the wrong "32" with a period of four tokens as filler text is lengthened, matching the compression stride of 4.
40.2 pplargest gap in 128K-context retrieval accuracy across token phases in DeepSeek-V4 base models
Summer (June-August) unemployment for graduates aged 22-25: 7.1% (2022), 6.3% (2023), 7.8% (2024), 7.2% (2025), 7.3% (2026).
7.3%summer 2026 unemployment of college graduates aged 22-25, within the 6.3-7.8% range of 2022-2025
Second-stage slope of 1.24: a 1 pp move in the AI index predicts 1.24 basis points more return per 1 pp higher software engineering payroll share.
32.6%market-implied permanent gain in software engineering productivity from AI news, Nov 2022 to Dec 2025
Figures as the sources published them. Each opens the digest; the credit names the original.
Exhibit 1
Exhibit 2
Exhibit 3
