Hinton: 'Chernobyl' AI disaster 'later this year or next'; Polymarket summary draws 2,400 likes

A post saying Geoffrey Hinton warned "either later this year or next year we'll see Chernobyl" for AI spread on X on Nov. 7. Polymarket's post drew about 2,400 likes and 170,000 views.

The Guardian published an interview with the Nobel laureate (78) on Nov. 3. Per summaries such as FourWeekMBA's, Hinton said "it might be that we would have been better off never developing it," while saying he is not a "doomer." If risks aren't managed, he said, mass unemployment and a society "torn apart" follow.

He likened this summer's reported incidents of OpenAI agents trying to hack Hugging Face to the 1979 Three Mile Island partial meltdown, a warning shot. His examples of a Chernobyl-scale event were an AI-enabled attack causing a massive brownout and an incident that takes down the financial system for a week. He also said: "My gut is that we're still in control."

He backs an independent FDA-style authority requiring "very good statistical evidence" that a model does more good than harm, and mandatory frontier-model testing, an idea endorsed by Finland's president and other countries. The Guardian reported that President Donald Trump "totally rejects" such global schemes. Hinton said he does not trust OpenAI CEO Sam Altman "at all."