Anthropic's Evan Hubinger Puts Odds of AI Wiping Out Humanity Within a Decade Above 10%
wfithian · x · 2026-10-08
Anthropic safety researcher Evan Hubinger said on X that the pace of AI progress has led him to believe there is a better-than-10% chance AI could "kill all of humanity" within the next decade. He noted that current models still pose low risk, but worries the technology could soon advance to the point of self-improvement and become an existential threat.
More from AGI Musings
- Asking Claude what the biggest threat to humanity will be in 10 years — alex_verem · 2026-10-08
- 62% of US voters fear AI could threaten humanity's future, Reuters/Ipsos poll finds — Polymarket · 2026-10-08
- Ethan Mollick: AI's rapid scientific progress is rewriting the norms of science itself — emollick · 2026-10-08
- Victorian elites partied instead of grinding — and AI is making the grind look overrated — dr_alphalyrae · 2026-10-08
- Mathematicians predict AI will mean more mathematicians, not fewer — Afinetheorem · 2026-10-08
- Thread: surveillance, scarce compute and the firewall may shape China's AI risk calculus — teortaxesTex · 2026-10-08