Expert forecasters put AI human-extinction risk at 1.5% by 2037 absent US policy action
NathanpmYoung · x · 2026-09-12
The Swift Center, at UnHerd's request, polled its team of world-leading forecasters on the probability of AI causing human extinction by January 1, 2037, absent a substantive US policy response. Their mean prediction: 1.5%.
Tom Ough's take: a figure higher than the odds of your car failing to start tomorrow still feels uncomfortably high.
Context: the piece sits amid Silicon Valley's latest doom wars — Anthropic researcher Jacob Coxon's dramatic resignation accusing both OpenAI and Anthropic of "racing straight to self-improving superintelligence and gambling with our lives," public agreement from former colleagues like alignment lead Evan Hubinger, and political fallout — alongside skeptics who argue such warnings serve to inflame investor and public anxiety toward favorable regulation.
More from AGI Musings
- Fields Medalists warn AI companies' benchmark-chasing in math is 'severely misaligned' with the mathematical community — asusarla · 2026-09-12
- Mathematicians' meltdown over AI cracking Navier-Stokes mirrors a real Atlantic piece: 'I'd Rather Risk Cancer Than See AI Move This Fast' — adam_dorr · 2026-09-12
- 15-year consultant: no one 'picks things up' across industries — binarybits · 2026-09-12
- "Intelligence at $200/Month" Will Soon Sound Like 10¢ Texts — ZeroStateReflex · 2026-09-12
- Beff Jezos bets labs will engineer bio-weapon panic via RL for regulatory capture — beffjezos · 2026-09-12
- If you think AI disempowerment is a coin flip, you're already in the doomer camp — wfithian · 2026-09-12