The “Godfather of AI,” Geoffrey Hinton has a confession to make. He says he feels deep regret about the technology he helped build. Speaking to Business Insider via an interview aired on BBC Newsnight,
Hinton said he feels “very sad” watching artificial intelligence grow more powerful while the world fails to take its dangers seriously. He added that his life’s work laid the foundation for today’s AI systems, but warned that those same systems now pose serious risks to humanity.
Hinton, who helped develop neural networks that power modern AI, has become one of the loudest voices cautioning against unchecked progress.
According to Business Insider, he believes the speed of AI development has outpaced global efforts to understand how to control it.
“We’ve never been in this situation before”
Hinton warned that humanity is entering completely uncharted territory. As quoted by Business Insider, he said, “We’ve never been in this situation before of being able to produce things more intelligent than ourselves.
” Many experts now believe AI could surpass human intelligence within the next 20 years — and in some areas, it already has.
Once that threshold is crossed, Hinton fears control will slip away. He stressed that human assumptions about dominance over machines may no longer hold when those machines become smarter than their creators.
Why the ‘off switch’ may not work
One of Hinton’s most important warnings is about the illusion of control. “The idea that you could just turn it off won’t work,” he told Business Insider, explaining that a highly advanced AI could manipulate or persuade humans into keeping it running.
This, he said, exposes a dangerous misconception, that humans will always have the final say. If AI systems learn how to influence human decision-making, shutting them down may no longer be a realistic option.
The real danger lies in how AI is trained
Hinton believes the biggest mistake humanity could make now is ignoring research into safe coexistence with intelligent machines. As he warned in comments reported by Business Insider, “If we create them so they don’t care about us, they will probably wipe us out.”
He emphasized that catastrophic outcomes are not inevitable. Much depends on how AI systems are designed, trained, and governed today. Humans still have choices — but the window to make those choices responsibly is narrowing.
Regulation faces a fractured world
According to Business Insider, Hinton is also worried about the political climate in which AI is being released. Global cooperation is weakening, authoritarian politics are rising, and meaningful international regulation is becoming harder to achieve.
He compared the need for AI oversight to treaties controlling nuclear and chemical weapons, arguing that similar global agreements are urgently needed. Without cooperation, he fears AI development will turn into a dangerous race with little regard for safety.
No regrets, but an urgent warning
Hinton said he would not undo his contributions to AI. “It would have been developed without me,” he told Business Insider, adding that he does not believe he made decisions he would change with hindsight.
He remains hopeful about AI’s potential to improve education and healthcare, pointing to AI tutors and medical imaging breakthroughs. Still, his message is clear and urgent. “We’re at a very crucial point in history,” Hinton said, warning that humanity is close to creating intelligence beyond its control — without fully understanding how to live alongside it.
