AI Safety Researcher: Governments Must Plan for AI Firms' AGI Timelines Being Correct
JeffLadish · x · 2026-11-17
AI safety researcher Jeff Ladish argues that while he hopes AI companies are wrong about their timelines, AI-enabled R&D speedup, near-term recursive self-improvement leading to "a country of geniuses in a datacenter," and robots building robots — these are not idle philosophical musings but the business plans of trillion-dollar companies.
His conclusion: even if most outside experts lacked similar forecasts (many share them), governments should require companies to demonstrate they can build vastly superhuman AI safely before attempting it in the next two years.
More from AGI Musings
- Physicists debate whether a superintelligence would even derive Navier–Stokes — burny_tech · 2026-11-17
- KOL predicts AI agent team meetings will replace human meetings within a year — kimmonismus · 2026-11-17
- VC: Show the whole-company AI automation plan in meeting one or lose — gabriel1 · 2026-11-17
- Survey of 1,345 people: AI doing emotional work like art and therapy rankles most — MohammadAtari90 · 2026-11-17
- New 2x2 Thinker/Doer taxonomy reframes how people react to AI at work — MohammadAtari90 · 2026-11-17
- Poll: 70% of Americans support regulating AI to slow new model development — trevposts · 2026-11-17