Anthropic Researcher Coxon Quits, Calls AI Race 'Endgame'

Jacob Coxon, a 27-year-old researcher who spent three years training frontier models at OpenAI and then Anthropic, resigned Tuesday and told WIRED that colleagues inside the labs describe the current moment as 'crunchtime' or 'endgame'. His seven-part resignation thread on X drew nearly 76 million views overnight.

"The people building AI earnestly believe that it could kill us all by the end of the decade," Coxon wrote, accusing both Anthropic and OpenAI of "racing straight to self-improving superintelligence and gambling with our lives." He told the Wall Street Journal he did not want to participate in an industrywide rush to build systems that can improve themselves, warning that some aggressive scenarios could be out of control by the end of next year.

His discomfort is not only about pace, but venue. "It's kind of insane that it has to happen on the MacBooks of some engineers living in San Francisco instead of a bunker in the desert like where they were doing the Manhattan Project," Coxon said.

Anthropic's own alignment science lead, Evan Hubinger, offered a striking on-the-record concession in the same piece: "we do not yet have a plan to solve alignment for superintelligence and are not clearly on track to." Coxon argues only two levers can bend the trajectory now: heavy government intervention, or every major lab agreeing to slow down together.

Four of the AI researchers on our Who's Who tracker had shared the story by the time we picked it up.