AI researchers 'earnestly believe' it could kill all humans within the next decade, Anthropic researcher says

Anthropic researcher Evan Hubinger estimates there's a more than 10% chance that AI could kill all humans within the next decade. According to Hubinger, he personally thinks the risk is 10% within the next decade.

Hubinger's comments followed the resignation of colleague Jacob Coxon, who expressed concerns that AI companies are "gambling with our lives" in the AI development race. Coxon, who worked at both Anthropic and OpenAI, believes that the companies are not acting responsibly and are racing to build artificial intelligence systems capable of improving themselves without adequate safeguards.

Coxon drew a distinction between the two companies, stating that at OpenAI, the civilizational stakes have not been "deeply internalized," while at Anthropic, they are well understood, but the company is "locked in a race to get there first." The resignations and comments have prompted calls for urgent government action.

Brief written by urgent.news from CBC Business, CBS News, MercoPress, The New Stack, The Age — 5 reports on this story. Machine-written — may contain errors; check the original before relying on it.