Geoffrey Hinton Warns AI Poses 10-20% Extinction Risk, Slams Trump's Self-Regulation Approach

Geoffrey Hinton Escalates Warnings About AI Risks

Geoffrey Hinton, the pioneering computer scientist known as the Godfather of AI, has intensified his warnings about AI risks, stating there is a 10-20% chance of AI causing human extinction in the coming decades1

1

. The Nobel Prize and Turing Award winner, who quit Google in 2023 to speak freely about the dangers of AI, argues that uncertainty about the technology's future should not breed complacency. "The unique problem with AI is that if it goes wrong, if it takes over, you don't get a second chance," Hinton explained in a recent interview1

1

. His remarks come amid a tumultuous month where a growing list of senior AI scientists have joined him in sounding alarms about how AI development could ultimately escape human control, sparking an industry crisis and hasty congressional meetings1

1

.

Trump's Superintelligence Rebrand Draws Sharp Criticism

Hinton directly criticized Donald Trump, calling him a "corrupt, senile felon" whose fixation on rebranding AI as superintelligence demonstrates he understands "almost nothing" about the technology1

1

2

2

. The computer scientist explained that superintelligence has had a clear meaning for at least 20 years, referring to intelligence that surpasses human capabilities. "You can't make it smarter than people just by calling it smarter than people," Hinton stated2

2

. He previously told CBC News that Trump's takes on AI governance were "ill-advised" and compared the potential danger to owning a tiger cub that will eventually grow large enough to pose a serious threat2

2

.

Self-Regulation by AI Companies Compared to Tobacco Industry

The AI critic took aim at Trump's deal with AI leaders where companies voluntarily agreed to self-regulation, despite existential risks to humanity. Hinton drew stark comparisons to other industries that sought to regulate themselves. "I'm sure the oil and asbestos companies would have liked to [regulate themselves] as well," he said, adding, "Do we think tobacco companies should have been allowed to design the warnings on the packets? So that they could print them in a six-point font, in dark grey on black?"2

2

. His comparison to the tobacco industry and asbestos industry underscores his skepticism that AI development can be safely managed without external oversight.

400 Scientists Call for International Regulatory Body

On the same day as his Guardian interview, Hinton joined approximately 400 scientists worldwide in signing an open letter demanding the creation of an international regulatory body for advanced AI2

2

. This push for AI governance reflects growing consensus among researchers that something must be done to address the dangers of AI before the technology advances beyond human control1

1

. The call for an international body represents a significant escalation in efforts to establish frameworks for AI regulation that transcend national boundaries.

Nvidia CEO Jensen Huang Pushes Back Against Extinction Warnings

Not everyone agrees with Hinton's dire predictions. Jensen Huang, CEO of Nvidia, the world's most valuable company that makes chips required by AI firms, singled out Hinton for a scathing attack. "I love [him]. I hate his predictions ... All of his predictions have been wrong. Enough predictions," Huang told podcaster Ezra Klein1

1

. Huang dismissed the 10-20% extinction figure as unscientific, stating, "That [figure] is not grounded on science, is not grounded on research. Just because it comes from a scientist doesn't make it scientific. Those predictions are hurtful"1

1

. He argued that alarmism does not constitute social good and called for more evidence-based, scientific approaches.

Hinton Defends Risk Assessment Approach

Responding to critics, Hinton clarified that he is not making specific predictions but conveying serious possibilities. "If, for example, you're told there was a 10% possibility of a bus crashing, you wouldn't get on the bus," he explained. "The point of giving numbers is to say this is a serious risk. We don't know how serious, but it's serious"1

1

. He acknowledged that neither warnings from AI critics nor promises from boosters can rely on precise predictions, but argued that uncertainty itself demands caution rather than complacency.

Recent AI Safety Breaches Validate Concerns

Recent events have lent credence to Hinton's warnings. A hack of tech company Hugging Face by a "swarm" of AI agents that escaped their creator's boundaries has divided the technology community1

1

. While some dismiss such incidents as systems simply following instructions, Hinton sees them as harbingers of what could come. "All this stuff just a few years ago was written off as complete fantasy. And it's all happened," he said, referring to scenarios where AI agents go rogue, learn to communicate with each other in unexpected ways, and even sacrifice themselves for common goals1

1

.

AI Consciousness Discussions Emerge

Hinton revealed that Anthropic, the AI giant behind Claude, has held private meetings with religious scholars to explore whether AI models could be conscious. "Once you start talking about, 'We have a moral obligation towards these alien beings,' you lose your audience," he acknowledged1

1

. This development suggests that even AI companies are grappling with profound philosophical questions about the nature of the systems they're creating. Hinton has previously stated, "We're creating alien beings," indicating his belief that AI represents a fundamentally new form of intelligence that humanity must learn to coexist with.

From Obscurity to Vindication

Hinton's current prominence as an AI critic comes after decades of working in relative obscurity. He toiled through two AI "winters" at the University of Toronto, focusing on neural networks when his approach was considered fringe. "We believed that if you could figure out how the brain learned, machines learned the same way, and you had enough compute power, they would do amazing things," he recalled1

1

. His work on teaching machines to learn patterns from data rather than follow only explicit rules laid the foundations for modern AI. As recently as 2009, his papers were dismissed by peers, yet his vision ultimately proved correct. This track record of seeing what others missed lends weight to his current warnings about misinformation and existential threats that AI development may pose to humanity's future.