AI doomerism: will artificial intelligence spiral out of control?

Fears that AI could spiral out of human control have broken into the mainstream. The worry has been building for some time but last week an AI researcher’s public resignation over safety concerns pushed it into the wider public conversation.

Jacob Coxon, who worked at OpenAI and Anthropic, quit, writing on X that the two companies are “racing straight to self-improving superintelligence and gambling with our lives.” He added that the people building the AI “earnestly believe that it could kill us all by the end of the decade.”  Tens of millions have viewed his posts. Then a current Anthropic scientist Evan Hubinger agreed, posting, “Jacob is correct here – we really do earnestly believe AI could kill all humans! I personally think it’s greater than 10% within the next decade.”

These comments broke through and now the media, tech and political leaders are all weighing in. Anthropic CEO Dario Ambio has called for a slow down in development and Sam Altman and Elon Musk have agreed, and support establishing an international oversight organization.

But the political response has been split. President Donald Trump has called the fears a “hoax” and Speaker Mike Johnson has said he wants to have talks with AI executives but warned against Congress jumping in with “red tape and hyper-regulation.” Democrats appear to take it more seriously  and Barack Obama has urged the party to make it a central issue.

So this episode, does artificial intelligence pose an existential threat to humanity? Are the fears overblown? Can we build guardrails that work? And if so, what about China, and the risk of unbounded AI development there?  

Guests: