Few people have done more to build modern AI, and fewer still have turned around to warn the world about it as loudly. Geoffrey Hinton spent half a century making neural networks work when most of the field thought they never would, and then - at the point of maximum credibility - left his job at Google to say he was worried about where the technology is heading. This page is a growing, chronological index of his interviews, talks, and public appearances, with enough context around each to know what you are clicking into.

TL;DR

  • An evolving collection of interviews with Geoffrey Hinton, the British-Canadian “Godfather of AI” whose work on backpropagation underpins modern deep learning
  • Hinton spent decades at the University of Toronto and Google before leaving in 2023 to speak more freely about AI risk
  • The interviews trace the arc from foundational technical work to public warnings - including his 2024 Nobel Prize in Physics for neural network research
  • Recurring themes: timelines for advanced AI, job displacement, existential risk, machine consciousness, and what regulation realistically looks like
  • A useful entry point if you want to understand the safety-concerned wing of the AI community in the person’s own words

About

Geoffrey Hinton, born in London in 1947, is a foundational figure in artificial intelligence and one of the small handful of people whose work genuinely changed the trajectory of the field. A British-Canadian computer scientist and cognitive psychologist, he spent decades arguing that intelligence could be built out of brain-inspired networks of simple units learning from data - an unfashionable position for most of his career, and the dominant one now. His story has a few load-bearing milestones:

  • Backpropagation: Hinton was a co-author of the influential 1986 paper, with David Rumelhart and Ronald Williams, that popularised backpropagation as a way to train multi-layer neural networks. The algorithm was not entirely new, but their work is what put it at the centre of the field. It remains the workhorse of essentially all modern deep learning.
  • The deep learning breakthrough: In 2012, Hinton and his Toronto students Alex Krizhevsky and Ilya Sutskever built AlexNet, a neural network that won the ImageNet competition by a margin large enough to end the debate about whether deep learning worked. That single result is often treated as the spark of the current AI era.
  • Academia and industry: Hinton spent decades at the University of Toronto and later at Google Brain (2013-2023), which acquired his startup DNNresearch. He also co-founded and served as chief scientific adviser of the Vector Institute in Toronto, one of the major hubs of Canadian AI research.
  • Awards and recognition: He shares the 2018 Turing Award - often called the “Nobel Prize of computing” - with Yoshua Bengio and Yann LeCun, and in 2024 was awarded the Nobel Prize in Physics jointly with John Hopfield for foundational work on machine learning with neural networks.
  • The turn to risk: In May 2023 Hinton resigned from Google so that he could talk about the dangers of AI without those warnings being read as company positions. Since then his public profile has been defined less by the technical work and more by what he thinks it might lead to.

Recurring Themes

If you watch several of these interviews back to back, the same threads keep surfacing. They are worth holding in mind as a kind of map:

  • Shortening timelines: Hinton has repeatedly revised his estimate for when AI might match or exceed human intelligence downward, from “30 to 50 years or never” to something closer to “5 to 20 years,” and he is candid that he has been surprised by the pace.
  • Existential risk: His central worry is that systems more capable than us may become difficult to control, and that we have no strong reason to assume they will stay aligned with human interests. He frames this not as certainty but as a risk large enough that ignoring it is reckless.
  • Job displacement: He is increasingly blunt that AI will not just augment work but remove large amounts of it, and that the gains may flow to a narrow slice of society rather than the people whose jobs disappear.
  • Machine consciousness and understanding: Hinton argues, more strongly than most of his peers, that today’s systems genuinely understand what they say and may have something worth calling subjective experience - a position many researchers push back on hard.
  • Regulation and what is realistically possible: He tends to be sceptical that a development pause is achievable given competitive and geopolitical pressure, and instead points toward safety research, international coordination, and putting serious resources into alignment.

A personal note

I write this as an interested hobbyist rather than a researcher, so take what follows as a layperson’s reaction rather than an informed verdict. What I find compelling about Hinton is not that he is necessarily right about the timelines or the risk - those are genuinely open questions and I am happy to be proven wrong on any view I hold - but that someone with so little left to prove chose to spend his final, most credible years sounding an alarm that brings him no benefit. His claim that current systems may be conscious is the part I sit least comfortably with. My own thinking on consciousness is unsettled - I go back and forth on whether it is something fundamental in the universe or something that computation can produce - and Hinton lands firmly on the computational side. I do not think the question is closed, and I would not want to assert either way with confidence. But it is exactly the kind of question these interviews force into the open, which is part of why I keep this page. These are just my reflections, always evolving.

Interviews

[2026-06-04] AI Pioneer Geoffrey Hinton: AI Is Conscious, Superintelligence is Coming, And We Should Be Worried

[2026-01-29] ‘Godfather of AI’ predicts ALL jobs will be in ‘wiped out’ by AI

[2025-10-09] AI: What Could Go Wrong? with Geoffrey Hinton

[2025-04-26] Full interview: “Godfather of AI” shares prediction for future of AI, issues warnings

[2025-01-30] ‘Godfather of AI’ predicts it will take over the world

[2024-06-15] Discussion on AI safety and future risks

View on Twitter/X

[2024-06-07] Keynote interview with Geoffrey Hinton (remote) and Nicholas Thompson (in-person)

[2024-05-20] Geoffrey Hinton | On working with Ilya, choosing problems, and the power of intuition

A long-form conversation in which Hinton reflects on his research instincts, how he picks problems, and his years working with Ilya Sutskever - one of his most consequential students.

[2024-02-29] Prof. Geoffrey Hinton - “Will digital intelligence replace biological intelligence?” Romanes Lecture

Hinton’s Romanes Lecture at the University of Oxford, one of his more structured public statements of the case that digital intelligence may end up surpassing - and perhaps superseding - the biological kind.

[2023-10-09] “Godfather of AI” Geoffrey Hinton: The 60 Minutes Interview

Hinton’s widely watched CBS 60 Minutes interview with Scott Pelley, recorded not long after he left Google, and for many viewers the first time they heard him lay out his concerns in plain terms.

[2023-06-22] The Godfather in Conversation: Why Geoffrey Hinton is worried about the future of AI

[2023-03-25] “Godfather of artificial intelligence” talks impact and potential of AI

Note: This page will be expanded with additional Geoffrey Hinton interviews as they become available.