Nature essay by Seth, Block, Tononi rejects AI consciousness

Three senior consciousness researchers have used a Nature comment to argue that today's AI systems are not conscious, and that current trajectories are unlikely to make them so. Anil Seth of the University of Sussex, Ned Block of NYU philosophy and Giulio Tononi of the University of Wisconsin–Madison published their case in Nature on 19 October, structured around three reasons.

The first is that intelligence and consciousness are different things. "Consciousness is about being and feeling; intelligence is about doing, or performing functions," the authors write. They lean on the cerebellum: it holds nearly four-fifths of the human brain's neurons and performs sophisticated computation, yet its removal "does not affect consciousness directly." Large language models, in their framing, are "mirrors that reflect an image of our collective digitized past" — impressive at functions, silent on experience.

The second reason targets computational functionalism, the assumption that running the right algorithm on any substrate suffices for experience. Brains, they argue, are not merely computers made of meat, and silicon gives no reason to expect subjective states to emerge from the same algorithmic recipe.

The third draws on integrated information theory, which Tononi has long championed. On that view, consciousness requires components "being wired together in very specific, tightly interconnected ways," and the "layered, feed-forward" architecture of today's deep-learning systems does not meet that bar.

The piece names Anthropic's recent work claiming to identify computational signatures of consciousness in its Claude model, alongside public comments from Richard Dawkins and Geoffrey Hinton suggesting today's AI might already feel something. The authors' bottom line: treat current systems as "smart tools, not smart people." Seven of the researchers we follow added the link to their feeds after it ran.