In an interview aired on CBS’ “60 Minutes” on Sunday, the 75-year-old pioneer of deep learning suggested that rapidly evolving AI systems could outsmart humans within just five years. “One of the ways these systems might escape control is by writing their own computer code to modify themselves,” Hinton said. “And that’s something we need to seriously worry about.”
Geoffrey Hinton, renowned as a “Godfather of AI,” has issued a stark warning: if not carefully managed, artificial intelligence could eventually surpass human control by modifying its own programming.
Hinton, a co-winner of the 2018 Turing Award for his groundbreaking work on neural networks, stepped down from his position as a vice president and engineering fellow at Google earlier this year. His decision was driven by a desire to speak openly about the dangers that advanced AI may pose.
He emphasized a critical concern within the AI research community: even the experts who created these systems don’t fully grasp how they function or evolve. “When this learning algorithm then interacts with data, it produces complicated neural networks that are good at doing things,” Hinton explained. “But we don’t really understand exactly how they do those things.” This lack of understanding is often referred to as AI’s “black box” problem — a term used by Google CEO Sundar Pichai in April.
While Hinton’s alarm has caught public attention, not all AI experts share his concern. Fellow Turing Award winner and AI pioneer Yann LeCun has dismissed such warnings, calling the idea that AI could overpower humanity “preposterously ridiculous.” LeCun believes humans will always be able to shut down technologies that pose a risk.
Still, Hinton remains wary. He acknowledged that AI has already brought significant benefits in fields like health care, but he stressed the darker side of the technology as well — particularly its role in spreading misinformation through fake images and videos online.
To address these challenges, Hinton is calling for intensified research into how AI systems operate, regulatory frameworks to govern their use, and global agreements to ban AI-controlled military weapons.
His views echo recent discussions on Capitol Hill, where top tech leaders — including Elon Musk, Sundar Pichai, Sam Altman of OpenAI, and Mark Zuckerberg of Meta — explored ways to balance innovation with safety-focused regulations.
Whatever path governments and companies choose, Hinton warns time is of the essence. “Humanity is likely at a kind of turning point,” he said, urging leaders to decide “whether to develop these things further and what to do to protect themselves if they [do].”
In the end, Hinton’s outlook remains cautious: “I think my main message is there’s enormous uncertainty about what’s going to happen next.”