nature.com web signal

Seth, Block and Tononi: AI likely isn't conscious, maybe never

TL;DR

  • Three leading consciousness researchers argue in Nature that current AI systems are not conscious and are unlikely to become so along current trajectories.
  • They separate intelligence from consciousness: the cerebellum holds roughly four-fifths of the brain's neurons yet its removal does not affect subjective experience.
  • On Tononi's integrated information theory, the layered, feed-forward architecture of today's deep-learning systems lacks the tight interconnection consciousness is said to require.

Three of the heaviest names in consciousness research have gone on the record saying today's AI isn't conscious, and probably won't be. In a comment published in Nature on 19 October 2026, neuroscientists Anil Seth and Giulio Tononi, with philosopher Ned Block, set out three reasons they think the AI-consciousness talk is overstated.

They start by prising intelligence apart from consciousness. 'Consciousness is about being and feeling; intelligence — very broadly — is all about doing, or performing functions,' they write. The cerebellum, which contains roughly four-fifths of the neurons in the human brain, runs sophisticated computation without being part of experience at all. Smart behaviour and inner experience come apart in the biology itself.

The second reason is a swipe at 'computational functionalism', the assumption that any substrate running the right algorithm would wake up. 'Brains are not — or at least not just — computers that are made of meat,' the authors write. Running the program is not the same as being the thing.

The third reason leans on Tononi's integrated information theory. Consciousness, on that account, depends on a system's components being wired together 'in very specific, tightly interconnected ways' that the layered, feed-forward architecture of today's deep-learning systems does not have.

The authors' bottom line is blunt: today's AI are 'smart tools, not smart people.' Their worry is that false positives on machine consciousness could pull welfare attention away from beings whose capacity to suffer is already certain. Richard Dawkins and Geoffrey Hinton get a mention as counter-voices who think AI might be conscious; the authors dispute them without engaging at length.

'LLMs are mirrors that reflect an image of our collective digitized past,' Seth, Block and Tononi write. 'We gaze into them and all too readily confuse the reflections with our own conscious minds.' Seven researchers in our Who's Who tracker shared the piece in the days after it ran.

Shared on Bluesky by 7 AI experts (top 5 by trust)