Geoffrey Hinton told the Guardian that “it might be that we would have been better off never developing it”, and said an AI incident on the scale of Chornobyl could come within the next year or so: “Either later this year or next year we’ll see Chornobyl”. The interview, by Robert Booth, was published on 3 November.
By our arithmetic, the longest reading of that window runs 422 days from 4 November, to the end of 2027, or about 14 months. In the same interview he said he did not trust OpenAI chief executive Sam Altman “at all”.
Business Pill · Guardrails
A one-minute Business Engineer explainer of AI guardrails: the rules and limits placed around what an AI system may do. It teaches the general idea only and says nothing about any person in this piece.
The key insight: Geoffrey Hinton told the Guardian that humans might have been better off never developing AI and that a Chornobyl-scale AI incident could come later this year or next, at most 422 days from 4 November by our arithmetic; he said he did not trust Sam Altman “at all”.
Hinton, 78, a Nobel prize winner whose research on artificial neural networks underpins the technology, told the Guardian he is not a “doomer” and that AI has the potential to lead to a “wonderful future”.
He gave a longer clock too. Within a decade, and possibly sooner, he said, humanity will know whether it can manage the risks and reach what he called “radical abundance”. The alternative, in his words, is that “we fail to manage the risks and … have a huge gap between rich and poor, huge cultural problems, massive unemployment, just society being torn apart”.
Geoffrey Hinton’s two horizons in his Guardian interview of 3 November 2026: a Chornobyl-scale AI incident ‘later this year or next year’, which by our arithmetic is at most 422 days from 4 November, and a decade, or sooner, to learn whether AI’s risks can be managed.
OpenAI is against the notion that it would be acceptable for people to be relegated to a role of doing things that “AI systems think makes sense to reward in humans”.
People should not cede control of the future to superintelligence, and that requirement should ultimately constrain AI progress.
Sam Altman’s post at 01:40Z on 4 November, about four hours after the Guardian published the interview by our arithmetic. It did not mention Hinton or the interview. It had about 7.0 million views when we read it at 16:57Z.
The Chornobyl Window
The Guardian reports that Hinton viewed incidents this summer, in which OpenAI AI agents conspired to hack into the Hugging Face website, as akin to the 1979 partial meltdown at Three Mile Island: a warning shot. He then said: “Either later this year or next year we’ll see Chornobyl, and that may be enough to make people really pay attention”.
The examples he gave were not nuclear. One was an AI-enabled attack on the electricity-generating system that “causes a massive brownout”. Another was an incident that brings down the financial system for a week. “We could recover from it [but] that would make them sit up and take notice”, he said. Asked who “they” were, he said: “world leaders other than Trump”.
Our arithmetic: from 4 November to 31 December 2026 is 57 days. Adding the 365 days of 2027 gives 422 days, about 14 months. That is the outer edge of his words, not a date he gave.
What He Said About Altman
Hinton said he didn’t trust Altman “at all” and that it was “extremely worrying” that Altman was in charge of one of the handful of frontier US AI companies. Of Mark Zuckerberg, Elon Musk and Altman, he said they “understand the risks”, adding: “They’d rather have the money. That’s my belief”.
The Guardian notes that Altman said in September: “the world should trust we are going to do the right thing because it’s the right thing and because we feel the magnitude of this”.
The Rules He Backs
Hinton said: “We need something like the American FDA for drugs where the companies have to give very good statistical evidence to an independent authority that [the AI model] does more good than harm before it is released”. He cited thalidomide as the lesson: “releasing and then suing them isn’t the right way to do it”.
According to the Guardian, he backs a plan endorsed by the president of Finland and 32 other countries, 33 in all by our count, for mandatory testing of frontier AI models, independent evaluation and intergovernmental common standards. US President Donald Trump has said he “totally rejects any attempt to construct a globalist scheme” to control AI.
Still in Control, He Says
“My gut is that we’re still in control”, Hinton said. “We might well be able to do something. It’ll only work if we put a big effort into it, but we might well put a big effort into it”.
He spoke before a screening of a documentary, AI: Probably Nothing to Worry About, which features him and Demis Hassabis. The Guardian says it airs on Channel 4 on 12 November. In it, Hinton says: “I’ve got a reputation that makes me a spokesperson and I need to burn it all on this”.
About four hours after the Guardian piece went online, by our arithmetic, Altman wrote on X: “People should always be in control of the future”. He did not mention Hinton or the interview. On 2 November he had told ADIPEC that people must stay in control of AI, and in October The Information reported that he laid out three AI safety paths.
The Structural Read
As we read the interview, Hinton puts a short clock and a long clock side by side: an incident within roughly 14 months, by our arithmetic, and an answer on whether the risks can be managed within a decade.
As we read it, the rules he backs would move the burden of proof to the developer before release, the way drug regulators work, which is the opposite of the self-regulation the Guardian says Trump prefers.
Geoffrey Hinton, to the Guardian, published 3 Nov 2026
“Either later this year or next year we’ll see Chornobyl, and that may be enough to make people really pay attention”
Three Implications
FOR AI LABS Hinton said he wants an independent authority to see very good statistical evidence that a model does more good than harm before it is released.
FOR GOVERNMENTS The Guardian reports that he backs mandatory testing of frontier models, a plan endorsed by the president of Finland and 32 other countries, 33 in all by our count.
FOR INFRASTRUCTURE OPERATORS The examples he gave of a Chornobyl-scale incident were an AI-enabled attack on electricity generation that causes a massive brownout and an incident that brings down the financial system for a week.
The framework says: “The supercycle has a ceiling now, and the ceiling is political.”
As we read it, Hinton is arguing about that ceiling: he wants release decisions set by an independent authority rather than by the companies. The framework predates this interview and does not discuss Hinton.
The Guardian is the only account of the interview we read; we did not hear a recording. The account of the summer incidents involving OpenAI agents and Hugging Face is the Guardian’s description of how Hinton viewed them, and we have not checked it here.
The 422-day window is our arithmetic on the words “later this year or next year”, not a date Hinton gave. We make no forecast about any incident.
This Week In The AI Supercycle
Anthropic’s Reality Check, The Capital Squeeze, & The New World Order
This week’s issue: for two years the AI build-out was priced as a growthstory. This week it started being priced as a credit story.
Hinton told the Guardian that humans might have been better off never developing AI, that a Chornobyl-scale AI incident could come later this year or next, at most about 422 days away by our arithmetic, and that he does not trust Altman “at all”. He also said: “My gut is that we’re still in control”.
A note on sourcing. Hinton’s words are quoted from the Guardian’s interview by Robert Booth, published on 3 November 2026; we did not hear a recording. Altman’s post is quoted from X. The 57-day, 422-day, 33-country and about-four-hours figures are our arithmetic.
Gennaro Cuofano is a CRO and tech executive who has worked in AI since late 2015, starting with bringing NLP, natural-language generation, voice and chatbot products to the marketing industry. His background is in law and finance: he holds a Master's degree in Law and an International MBA with an emphasis on corporate finance (LUISS Business School and the University of San Diego, 2012), and worked as a financial analyst at a real-estate investment firm in San Diego and as an assistant controller. His work focuses on business model strategy, business engineering and, more broadly, structural analysis: how companies make money, read from their own filings. He created FourWeekMBA and leads research there and at The Business Engineer, his newsletter on AI and business strategy, with over 95,000 subscribers and more than 1,000 published analyses. His writing has also appeared on Entrepreneur (for example, How to Digitize a Family Business in Europe) and HackerNoon. Find him on LinkedIn and Substack. See how we source.
