Researcher leaves Google DeepMind for METR, puts AI takeover odds above 20%
sjgadler · x · 2026-09-10
Former Google DeepMind researcher Josh Engels explains why he joined safety org METR over a frontier lab: he's increasingly convinced "we don't have things under control" and we should slow down, putting AI takeover probability above 20%.
He quotes Evan Hubinger saying Anthropic researchers genuinely believe AI could kill all humans — Hubinger estimates >10% within the next decade — and that despite Anthropic trying its best, there is not yet a plan to solve alignment for superintelligence.
More from AGI Musings
- AI safety discourse is replaying 2020's pandemic expert free-for-all — SanhEstPasMoi · 2026-09-10
- Train on Frontier Papers or Build RL Envs? An Insider Debate on Math Model Training — ctjlewis · 2026-09-10
- AI extinction drama: researchers believe in the risk yet keep racing, says viral thread — AIandDesign · 2026-09-10
- François Fleuret asks: what intellectual endeavor can humanity still claim from AI in two years? — francoisfleuret · 2026-09-10
- That Proof Was Downstream of Mathematicians' Chats — but Is 'Direct Training' Even Real? — shaurizard · 2026-09-10
- Debate sparks: Banks' Culture series already wrote the definitive AGI utopia — Promptmethus · 2026-09-10