Geoffrey Hinton, often called the "Godfather of AI, " was awakened last year in the middle of the night to learn he had won the Nobel Prize in physics. He said he never anticipated receiving such an honor. “I dreamt of winning one for figuring out how the brain works. But I didn't figure out how the brain works, yet I won one anyway, ” Hinton remarked. The 77-year-old scientist received the award for his groundbreaking contributions to neural networks—introducing in 1986 a method for predicting the next word in a sequence—which has become the fundamental principle behind today’s large language models. While Hinton believes artificial intelligence will revolutionize fields like education and medicine and possibly help address climate change, he is growing increasingly uneasy about the rapid pace of its development. “The best way to grasp it emotionally is that we are like someone who has a really cute tiger cub, ” Hinton explained. “Unless you can be absolutely sure it won’t want to kill you once it’s grown, you should be worried. ” “People haven’t got it yet” The AI pioneer estimates there is a 10% to 20% chance that artificial intelligence will eventually take control away from humans. “People haven’t got it yet, people haven’t understood what’s coming, ” he cautioned. His worries align with those of industry leaders such as Google CEO Sundar Pichai, X-AI’s Elon Musk, and OpenAI CEO Sam Altman, who have similarly voiced concerns.

Yet Hinton also criticizes these companies for prioritizing profit over safety. “If you look at what the big companies are doing right now, they’re lobbying to reduce AI regulation. There’s barely any regulation to begin with, but they want even less, ” Hinton stated. He seems especially disheartened with Google, where he previously worked, for reversing its policy regarding military uses of AI. Hinton argues that AI companies should allocate much more of their resources to safety research—“like a third” of their computing power—compared to the small portion they devote now. CBS News reached out to all the mentioned AI labs to ask what percentage of their computing resources go to safety research. None provided specific figures. All emphasized the importance of safety and support for regulation in principle but have largely opposed the regulatory measures that lawmakers have proposed so far.