Hinton, Bengio and 20+ experts urge governments to track how much AI research is done by AI
rohanpaul_ai · x · 2026-10-07
A policy paper by 20+ experts including Geoffrey Hinton and Yoshua Bengio asks governments to start tracking how much AI research is already done by AI. Key claims: AI task horizons jumped from seconds-long tasks in 2023 to expert-days now; OpenAI alone has enough inference compute to generate 10^13 tokens/day; at expert level, one frontier developer's compute could run an AI workforce equal to millions of top researchers, versus a few thousand humans frontier labs employ.
Related event: Hinton, Bengio and 22 Experts Warn on AI Automating AI Research(3 posts)→
More from AGI Musings
- OpenAI reasoning models went from basic arithmetic to decades-old math breakthroughs in two years — daniel_mac8 · 2026-10-07
- The New Skill in the Math-Proof Era: Spending Verification Effort Where Errors Cost — dbreunig · 2026-10-07
- Richard Ngo's hard sci-fi Tinker sketches how AI designs chips for its successors — teortaxesTex · 2026-10-07
- tszzl: Don't be an uncomprehending audience clapping at the AI magic show — Afinetheorem · 2026-10-07
- VraserX: Personal AI agents like Grok Bot, Muse and Dot will soon be everywhere — VraserX · 2026-10-07
- Predicting a 'third thing' RSI: capability leap in ~9 months via automated R&D, short of full FOOM — teortaxesTex · 2026-10-07