Ex-OpenAI researcher defends Evan Hubinger: >10% extinction risk belief is sincere, not marketing

ronbodkin · x · 2026-09-10

Evan Hubinger and Jacob stated they genuinely believe AI could kill all humans — Evan puts >10% odds within a decade — and that Anthropic, while trying its best, has no clear plan for superintelligence alignment. Responding to accusations of ulterior motives, Rosie Campbell, who spent 3.5 years at OpenAI, says she pivoted her career to AI safety in 2017 after reading about extinction risk and knows these researchers personally: their concerns are sincerely held, whatever you think of the arguments.

Related event: Anthropic Researcher Estimates Over 10% Chance AI Exterminates Humanity Within a Decade(10 posts)→

Original post →

More from AGI Musings

AGI Musings channel →