Kokotajlo: GFF more likely false, and how much METR speedup would signal imminent AI R&D automation
DKokotajlo · x · 2026-11-12
Daniel Kokotajlo confirms Chalmers read him right: the General Factor Hypothesis is serious but likelier false than not, and a real 2x acceleration happening now would be a medium-sized (not conclusive) chunk of evidence. He also asks @bshlgrs, @ajeyacotra and @RyanGreenblatt: what evidence would convince you GFF is false, and at what METR speedup number (3x, 5x, …) would you conclude full AI R&D automation is imminent or already underway?
Related event: Kokotajlo vs Chalmers debate over a missing general factor for ASI(6 posts)→
More from AGI Musings
- Talent vs. luck model reproduced: equal funding is ~9x more efficient than funding winners — ChenhaoTan · 2026-11-12
- Stanford HAI dashboards show employment in AI-exposed jobs down over 4% this year — StanfordHAI · 2026-11-12
- 'AGI is three months away' has been said for two years: Diogo's 10 sharpest quotes on software — TheTuringPost · 2026-11-12
- Why is it harder to verify alignment than capability? A thread on evaluation asymmetry — sandersted · 2026-11-12
- Shun-ichi Amari on how 'do as you like' freedom birthed the math behind AI's 60 years — FrnkNlsn · 2026-11-12
- Pedro Domingos: with equal data and compute, AI is not even close to the human brain — pmddomingos · 2026-11-12