Hinton calls 10% chance of AI killing all humans ‘not unreasonable’

Nobel Prize-winning computer scientist Geoffrey Hinton told BBC Newsnight on Tuesday that a 10% probability of artificial intelligence killing all humans within the next decade "seems not an unreasonable estimate," a remark that left presenter Victoria Derbyshire visibly stunned on air.ndtvprofit+1

"We've never created beings that may soon be smarter than us. We don't know what's going to happen," Hinton said when asked by Derbyshire whether the figure was credible. "A 10% chance seems not an unreasonable estimate to me. But nobody really knows how to give a sensible estimate." When Derbyshire repeated his words back to him, Hinton replied with a single word: "Yes." Her response — "Wow… oh my God" — quickly circulated online.ndtv+1

How AI Could Cause Devastation

Hinton, who won the 2024 Nobel Prize in Physics for his work on machine learning, said advanced AI would not need physical control of the world to pose an existential threat. "It would be able to cause complete chaos just by talking to people," he said. "But it would also be able to design very nasty viruses — biological viruses — as well as computer viruses. It could do devastating cyber attacks." He added that it would be "very foolish" to assume the risk was as low as 1%.ndtvprofit+1

Stuart Russell, a UC Berkeley professor, echoed those concerns, telling Business Insider that AI or AI-enabled actors could attack critical infrastructure including electricity, water, transport, and financial systems.businessinsider

The Anthropic Resignations That Sparked the Debate

Hinton's interview followed a chain of events that began on September 8 when Jacob Coxon, a former pretraining researcher at both OpenAI and Anthropic, resigned publicly. "Neither company is acting responsibly," Coxon wrote on X. "They are racing straight to self-improving superintelligence and gambling with our lives."ibtimes+1

Anthropic's alignment science lead, Evan Hubinger, then responded publicly that Coxon was "correct" and that he personally believes there is a greater than 10% chance AI could kill all humans within the decade. "We do not yet have a plan to solve alignment for superintelligence and are not clearly on track to," Hubinger wrote.businessinsider+1

A Sharp Divide Among Experts

Not all researchers share the alarm. Gary Marcus, an AI researcher, told Business Insider he sees cyberattacks, biological weapons, and disinformation as serious risks but does not consider extinction a realistic near-term outcome. "There are some very serious harms to worry about, but no realistic scenario that I am aware of for actually 'killing all humans,'" he said. Meta's ↘ 2.43% Yann LeCun has previously dismissed such probability estimates as gut feelings — a characterization Hinton himself has partly accepted, telling CNN that "anybody who estimates probabilities like that is really just making a wild guess".businessinsider+1

ITV presenter Robert Peston added his voice to the debate, writing on X: "I am slightly amazed it hasn't attracted more attention, including from inside the government."ladbible