Geoffrey Hinton Warns of 10% AI Extinction Risk, Narrowing Timeline to Under a Decade

In a recent BBC Newsnight appearance, Geoffrey Hinton reiterated that a 10% probability of AI-driven human extinction is not far-fetched, noting that the window for such catastrophes has shrunk from decades to perhaps under ten years. Citing threats like automated cyberattacks and synthetic pathogens, Hinton acknowledged the inherent difficulty of quantifying risks from systems smarter than humans. Against the backdrop of escalating model scaling and internal safety dissent at top labs, he emphasized that binding international cooperation remains the only viable hedge against existential fallout.