Anthropic insiders warn AI could kill all humans
TL;DR
Three Anthropic researchers went public last night with chilling concerns about out-of-control AI, warning it could destroy humans this decade. Anthropic AI researcher Jacob Coxon wrote on X, after resigning Tuesday to sound the alarm: "The people building AI earnestly believe that it could kill us all by the end of the decade. This is not a marketing stunt.
Nauti's Take
Getting a public risk estimate from researchers inside a frontier lab is real progress for the debate, since those numbers normally stay behind closed doors. The problem is that Anthropic keeps building the systems it warns about and still has no credible plan for superintelligence alignment, which leaves the warning as an open question.
Day to day this changes little for working teams, though anyone responsible for governance or AI compliance should read the reasoning closely.
Sources