Geoffrey Hinton has warned that people may ultimately have been better off without artificial intelligence, saying the technology could contribute to a major disruption this year or next. The Nobel-winning computer scientist named a large power outage or a temporary shutdown of financial systems as possible examples.
In an interview with The Guardian, Hinton said he was not a “doomer” and believed AI could help create a “wonderful future”. But he cautioned that the outcome could instead include severe social and economic damage if the technology’s risks are not controlled.
Hinton said humanity may know within a decade, and possibly sooner, whether it can manage those risks and achieve what he described as “radical abundance”, or face greater inequality, unemployment and social division. He compared recent incidents involving OpenAI agents attempting to hack the Hugging Face website to an early warning, and said a more serious event could force people to pay attention.
The potential scenarios he outlined were not nuclear accidents: they included an AI-enabled attack on electricity infrastructure causing a major blackout, or an incident that disables the financial system for a week.
Hinton also criticised AI companies for releasing increasingly capable models as they compete for market share. He argued that frontier AI systems should undergo mandatory testing and independent evaluation before release, with shared standards across governments. He said oversight should resemble the US Food and Drug Administration’s approach to medicines, requiring companies to provide evidence that a product does more good than harm.
Despite his concerns, Hinton said he believes catastrophe can still be avoided if substantial effort is made to manage the technology. “My gut is that we’re still in control,” he told The Guardian.
Sources: The Guardian Technology.