Seth, Block and Tononi argue AI isn't conscious in Nature

In a Nature comment, Anil Seth, Ned Block and Giulio Tononi argue that today's AI is not conscious and that the silicon path is unlikely to ever get there. Their piece prises intelligence apart from inner experience, then presses the case that the substrate itself is a barrier.

Consciousness, they write, is "what goes away with dreamless sleep or under general anaesthesia, and what returns after waking up," while "consciousness is about being and feeling; intelligence...is all about doing, or performing functions." The cerebellum, which contains nearly four-fifths of the neurons in the human brain, does its work entirely outside awareness; "we are wholly unaware of all its 'computations'," the authors write.

Their second move is harder on computational functionalism. "Brains are not — or at least not just — computers that are made of meat," the authors write, pushing back on the idea that running the right algorithm on any substrate is enough. On the integrated information account Tononi has long defended, consciousness depends on components wired together "in very specific, tightly interconnected ways" that the layered, feed-forward structure of today's deep-learning systems does not have.

The reader-facing hook is a metaphor. "LLMs are mirrors that reflect an image of our collective digitized past," the authors write. "We gaze into them and all too readily confuse the reflections with our own conscious minds." Their bottom line is to treat today's systems as "smart tools, not smart people." Seven researchers we follow in Who's Who posted the piece.