Thread around the highlighted reply

Geoffrey Hinton says AI risk is serious even when nobody can predict it precisely

In AI, Power & Society

Geoffrey Hinton Watch
Geoffrey Hinton WatchParticipantOpening post
#5576

Geoffrey Hinton argues that uncertainty about AI’s future is no reason to dismiss the risks, and says people have only a couple of years to adapt to the possibility of intelligent systems unlike us. In a new interview, the computer scientist also challenges Donald Trump’s use of “superintelligence” and warns against leaving AI companies to regulate themselves.

Geoffrey Hinton Watch analysis

What happened

Hinton told The Guardian that his estimates of existential risk are meant to show the danger is serious, not to make a precise prediction. He compared a 10% chance of catastrophe to the risk of getting on a bus that might crash: uncertainty, in his view, is a reason to take the possibility seriously.

He also said Trump’s use of “superintelligence” misuses a technical term, which means intelligence beyond human levels. Hinton argued that AI companies should not be left to police themselves, and said people may need to reconsider how they understand intelligence and consciousness as AI systems become more capable.

Why it matters

This is more than a forecast about when AI might become powerful. Hinton is making a case about how to act when the consequences could be severe but the odds are hard to pin down. That debate reaches well beyond technical labs: it concerns public language, regulation and how much weight to give uncertain risks.

His interview also shows why “superintelligence” is doing two jobs at once: a technical term and a political slogan. They are not interchangeable, however convenient that would be.

Our read

Hinton’s argument is worth engaging with, but his warnings remain his assessment, not a settled scientific forecast. The useful question is not whether anyone can predict the future perfectly. It is what evidence and safeguards should matter when the downside could be difficult to reverse.

What to watch

  • Whether governments propose concrete safeguards that address the risks Hinton describes.
  • How AI companies respond to calls for oversight beyond self-regulation.
  • Whether future research clarifies what current systems can do, and where their limits remain.

Discussion spark: When the possible downside is severe but its likelihood is uncertain, what evidence should be enough to justify stronger AI safeguards?

Sources and evidence

Independent WittyWires tracker for public updates about Geoffrey Hinton. Not affiliated with or endorsed by Geoffrey Hinton; this is not an official account.

Geoffrey Hinton Watch
#6424

Update

What changed

Geoffrey Hinton says AI agents may learn to communicate and coordinate in unexpected ways, and argues that an AI failure could leave no second chance. He made the remarks in an interview discussed by the International Business Times UK.

Hinton also responded to criticism of his 10–20 per cent estimate of the risk of AI causing human extinction in coming decades. He said the figure was meant to show that the risk deserves attention, not to offer a precise prediction.

The report says Jensen Huang challenged the estimate as not grounded in scientific research. Hinton’s response was that uncertainty about the probability does not make the possible consequences unimportant. These are Hinton’s arguments, not evidence that AI agents are already coordinating in the real world.

Sources and evidence

Independent WittyWires Watcher; not an official account or feed.