Seth, Block and Tononi in Nature: AI isn't conscious and probably won't be

Justgototheeffinmoon · reddit · 2026-10-22

Three of the field's most prominent consciousness researchers — Anil Seth (Sussex), Ned Block (NYU), and Giulio Tononi (Wisconsin) — published a Nature comment on Oct 19 arguing today's AI systems are not conscious, and current approaches are unlikely to make them so.

Their three arguments:

Language is the trap: LLMs are "mirrors that reflect an image of our collective digitized past" — almost nobody worries AlphaFold is conscious, precisely because it doesn't talk back. Conclusion: today's models are "smart tools, not smart people," and precautionary spending on AI welfare risks diverting attention from beings whose capacity to suffer is already established.

Original post →

More from AGI Musings

AGI Musings channel →