☀️ AI Morning Minute: Geoffrey Hinton
He spent 40 years building the thing he now thinks might be the most dangerous technology in human history.
Geoffrey Hinton is the researcher whose work on neural networks made modern AI possible. He shared the 2018 Turing Award, computer science’s highest honor, with Yann LeCun and Yoshua Bengio for that work. In 2024 he won the Nobel Prize in Physics. In 2023 he left Google to speak freely about AI risks. That sequence tells you most of what you need to know about where the field has been and where it might be going.
Who they are
Hinton was born in London in 1947, a great-grandson of George Boole, whose Boolean algebra underlies all modern computing. He studied psychology at Cambridge, earned a PhD in AI from Edinburgh in 1978, and spent decades working on neural networks when most of the AI research community considered them a dead end. Along with David Rumelhart and Ronald Williams, he co-authored the 1986 paper that popularized backpropagation, the algorithm that remains the foundation of how neural networks learn. That paper was largely ignored for two more decades before the field caught up to it.
He joined Google in 2013, the same year Google acquired the company he co-founded to commercialize his research. He worked there for a decade. In May 2023 he resigned, later writing on social media that he left “so that I could talk about the dangers of AI without considering how this impacts Google.” He holds an emeritus professorship at the University of Toronto and co-founded the Vector Institute, Canada’s national AI research center.
Why they matter
Backpropagation runs everything. The algorithm Hinton helped popularize in 1986 is the training mechanism inside every major AI model in use today. ChatGPT, Claude, Gemini, image generators, speech recognition: all of them learn using the process his paper described. That’s not influence. That’s foundation.
His departure from Google shifted the public conversation about AI risk. When the person most responsible for modern deep learning leaves one of the world’s most powerful AI labs to warn the public, it lands differently than when a critic with no involvement does. His 2023 exit made AI safety a mainstream topic in a way that years of academic papers hadn’t.
He puts a specific number on what worries him most. At Nobel Week in December 2024, Hinton said there’s a 50% chance AI surpasses human intelligence within 5 to 20 years. He had previously put that timeline at 30 to 50 years. The revision wasn’t rhetorical. He changed his mind based on how fast the field moved.
What they’ve said or done
At the AI4 conference in 2025, Hinton told a crowd of thousands of AI leaders: “The people in this room are the ones writing history. In 50 years, no one will care how much revenue your model generated in 2025. They will care whether you built something that improved human life, or endangered it.”8