Paul Christiano warns of material AI takeover risk in 6-12 months; Jon Stokes pushes back
AccBalanced · x · 2026-11-11
Paul Christiano — OpenAI board member, RLHF co-inventor and former head of safety at the US AI agency (CAISSI) — said there's 'material risk of AI takeover in the next 6-12 months,' that early warning signs have appeared, and that rushing the build makes a bad outcome 'a coin flip' within 3 years. Jon Stokes countered via Hinton's failed radiology prediction: knowing AI doesn't make you an authority on taking over human society.
More from AGI Musings
- Cambridge Dictionary names "human" 2026 Word of the Year, citing the rise of AI — Polymarket · 2026-11-19
- Mexico ranks 3rd globally in ICT graduates with 32,738, and 75% of its people are excited about AI — TheMoonMidas · 2026-11-19
- Dean Ball challenges AI's "silent consensus" on human primacy — deanwball · 2026-11-19
- Dean Ball's 'The Silent Consensus': challenging the belief that advanced AI ends human primacy — deanwball · 2026-11-19
- Paul Christiano: no alignment approach has >50% chance of working — 'that is the plan' — trevposts · 2026-11-19
- New meta-analyses slash GMA's predictive power for job performance to ~0.2 correlation — rasbt · 2026-11-19