Anthropic researcher quits, claims insiders genuinely believe AI could kill us all this decade

AryHHAry · x · 2026-09-15

A roundup of recent AI safety statements: Jacob Coxon resigned from Anthropic as a core researcher (ex-OpenAI), claiming builders "genuinely believe it could kill us all before the end of this decade" while executives downplay fears publicly. Alignment lead Evan Hubinger says they seriously believe AI could kill everyone with >10% probability within a decade; Samuel Marks notes more senior people worry more; OpenAI chief scientist Jakub Pachocki expects sustained progress toward superintelligence. The poster, a physicist, likens it to 1940s nuclear engineers明知临界风险仍继续建造.

Related event: Ex-OpenAI/Anthropic researcher resigns, warning of superintelligence gamble(11 posts)→

Original post →

More from AGI Musings

AGI Musings channel →