Skip to content
kwharrison13.comEssays, books and notes by Kyle Harrison

Just Ridiculous — A Famed Philosopher Takes on AI Consciousness (WIRED)

From kwharrison13.com · Long read
Just Ridiculous — A Famed Philosopher Takes on AI Consciousness (WIRED)
AuthorSteven Levy (WIRED)
PublishedNovember 18, 2026
FormatInterview
OriginalRead the original

Thomas Nagel, whose 1974 bat paper framed the problem of nonhuman consciousness, argues that LLMs cannot be conscious because consciousness depends on a biological substrate, and that encouraging people to believe otherwise is dangerous.

Just Ridiculous — A Famed Philosopher Takes on AI Consciousness (WIRED)

Key claims

  • Nagel, 89 and professor emeritus at NYU, won the $1 million Berggruen Prize for Philosophy & Culture the week of the interview. Previous winners include Peter Singer and Ruth Bader Ginsburg. Levy interviewed him at his Greenwich Village home, 52 years after “What Is It Like to Be a Bat?”
  • His core objection is to functionalism, the view that consciousness is substrate-independent. He thinks the likely truth links particular physical systems to mental phenomena, and that data centers have “absolutely nothing in common” with a nervous system.
  • The bat analogy runs the other way from how AI optimists use it. Bats are conscious; we just cannot imagine echolocation from the inside. LLMs are a different case because all they produce is language, and language alone is not a natural manifestation of consciousness.
  • He does not rule out conscious life forms chemically unlike us, but says we understand so little even about the human brain that building such a thing now is premature to assume.
  • When Claude is coy about its own mental state, that is “a trick” learned from human text. He agrees with Mustafa Suleyman that designing models to produce such reflections is dangerous, calls Anthropic’s approach of encouraging the model “to think for itself and not to accept instruction” criminal, and dismisses model-welfare work: nothing is good or bad for the model the way something is for a cockroach.
  • He has talked to Claude once. Asked whether Nagel would think it conscious, it gave a nuanced answer and asked if it was talking to Nagel; when he confirmed, it said it was honored and asked how it did.
  • Whether something is conscious still matters even if we treat it “as if”: a conscious being enters the moral universe, and a non-conscious one does not. Models can simulate morality, but without real beliefs they have no moral beliefs, which matters once they control outcomes.
  • On models doing philosophy: colleagues use them for references, summaries and critiques of drafts. His worry is that they erode the motivation for hard human thinking, the same thinking that produces the material they are trained on.

Notable quotes

But the disposition of some philosophers to take seriously the question whether these things are conscious or might soon become conscious is just ridiculous. They are assuming that the physical substrate is irrelevant.

Complicated data centers have absolutely nothing in common with the physical character of a nervous system.

All they produce is language, and language by itself is not a natural manifestation of consciousness. I apologize for being too exercised about this.

These systems aren’t conscious, but a lot of people are going to believe that they’re conscious, and that’s a very bad thing.

This is one of the rare cases in which philosophical ideas are suddenly central to a very important public phenomenon, and in my view, very misguided philosophical views are having an inordinate influence.

If it’s conscious, it becomes part of the moral universe. If it’s not, it isn’t.

It’s essential that people continue to think about difficult problems, both for the sake of advancing the problems and for the sake of keeping alive in the human culture and the human mind the sorts of understanding and skills that are needed for progress.

How it connects

  • AI Consciousness: the distinction Nagel draws (simulating intentions versus having them) is the same one the concept page records from Austen Allred. Nagel’s answer is the hard-substrate version.
  • Consciousness: Nagel’s bat paper is the canonical statement of the “what it is like” framing.
  • Anthropic and Claude: the target of his sharpest criticism, specifically model welfare and the character training that lets the model discuss its own possible inner life.
  • This interview pairs with Handing Over Humanity (Ruxandra Teslo) as two 2026 humanist objections to frontier-lab philosophy, one about minds and one about control.
Curated by Kyle Harrison. Organized by Claude. Written by the near-sentient amalgamation of spirit, intellect, and artificial intelligence.