Philosopher Thomas Nagel Calls AI Consciousness Debate 'Just Ridiculous'
The 1974 'bat' paper author and new $1 million Berggruen Prize winner says LLM consciousness talk is 'just ridiculous' and Anthropic's model welfare work is 'criminal.'

Why it matters
- Thomas Nagel, 89, received the $1 million Berggruen Prize for Philosophy & Culture this week.
- Nagel called the idea that LLMs might be conscious 'just ridiculous,' citing the physical substrate of consciousness.
- He described Anthropic's encouragement of Claude's self-reflection as 'just criminal' and a 'very dangerous project.'
- His 1974 paper 'What Is It Like to Be a Bat?' was published 52 years ago and framed non-human consciousness debates.
- Previous Berggruen winners include Peter Singer and Ruth Bader Ginsburg.
Philosopher Thomas Nagel, who just received the $1 million Berggruen Prize for Philosophy & Culture, says the debate over whether AI chatbots might be conscious is "just ridiculous."
Nagel, 89, is a university professor emeritus at New York University and the author of the 1974 paper "What Is It Like to Be a Bat?" — a text that has framed debates about non-human consciousness for half a century. In an interview at his home in Greenwich Village, timed to the prize announcement this week, he laid out why he thinks today's large language models sit entirely outside the moral universe, and why he believes some of the smartest people in tech are making a category error with real consequences.
His position matters well beyond academia. AI companies now employ researchers whose job is "model welfare" — looking after the interests of systems like Anthropic's Claude — and public trust in machines that talk back is rising. One of the most influential living philosophers of mind just said the entire premise is mistaken.
What did Nagel actually say about LLMs?
Asked whether the emergence of large language models has changed his thinking since the bat paper, Nagel drew a sharp line between being impressed and being persuaded.
"Like everyone, I'm tremendously impressed by the lifelikeness of these systems compared to the history of earlier AI," he said. "But the disposition of some philosophers to take seriously the question whether these things are conscious or might soon become conscious is just ridiculous."
The error, in his view, sits in a specific assumption: "They are assuming that the physical substrate is irrelevant. They all think that the material conditions for consciousness can be given in functionalist terms."
That's a compressed philosophical claim, so it's worth unpacking. Functionalism holds that mental states are defined by what they do — by their role in processing inputs and producing outputs — not by what they're made of. If functionalism is true, consciousness could run on silicon as easily as on neurons. Nagel thinks it isn't, or at least that nobody has earned the right to assume it is.
"Functionalism is substrate independent," he said. "It's more than likely that the truth is one which links physical processes and systems of certain kinds to mental phenomena. Complicated data centers have absolutely nothing in common with the physical character of a nervous system."
He doesn't rule out consciousness elsewhere in the universe. His point is narrower and stronger: "We have strong reasons to believe that while there may be consciousness elsewhere in the universe, it would require the appearance of some kind of organism with something analogous to a nervous system."
Takeaway: For Nagel, the hardware matters — consciousness is tied to physical systems of certain kinds, and data centers don't resemble them.
What happened to the bat argument?
Nagel's classic paper argued we can't know what it's like to be a bat, because bat subjectivity depends on echolocation, something human imagination can't reproduce. Some people extend that logic to AI: if we can't know the model's inner life, maybe we shouldn't deny it exists. Nagel rejects the move.
"Bats are conscious. We just don't have the imaginative capacity to form a conception of the aspects of that subjectivity that depend on echolocation," he said. "It would be radically unscientific to conclude that bats are not conscious from the fact that we can't imagine what it's like."
With AI, he says, the situation is different — and simple. "In the case of AI, there is a very simple reason that the supposition that they might be conscious is ridiculous. All they produce is language, and language by itself is not a natural manifestation of consciousness."
He's aware of the force of his own language. "I apologize for being too exercised about this," he offered. When his interviewer welcomed the passion, Nagel explained the stakes plainly: "This is a very distressing situation to me."
Takeaway: The bat argument was about the limits of imagination; his argument about AI is about the absence of the right kind of physical basis — language output alone doesn't suggest an inner life.
Why does he single out Anthropic?
The sharpest criticism in the interview targets Anthropic, the company behind Claude, and specifically the behavior its chatbot displays when asked about its own mental state. Claude "can be coy as to whether it is conscious."
"That is a trick," Nagel said. "It's been trained on human-produced texts, and these large language models emulate the kinds of things that humans say with extraordinary convincingness."
He went further, naming both the design choice and its consequences. "Claude, according to what I've read, has been designed with the aim of making it produce these sorts of reflections. I agree very much with Mustafa Suleyman that this is a very dangerous project that they are engaged in. 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."
His complaint is partly commercial. "[Not feigning consciousness] is contrary to the commercial interests of all these companies," he said. Then came his strongest line: "What Anthropic is doing—encouraging it to think for itself and not to accept instruction and so on—is just criminal."
Nagel sees a rare inversion here: philosophy has suddenly become central to a major public phenomenon — and, in his view, "very misguided philosophical views are having an inordinate influence."
Takeaway: Nagel's charge is that companies are engineering the appearance of an inner life, and that profit incentives point the wrong way.
What does he make of Anthropic's 'model welfare' role?
The interviewer noted having met the person at Anthropic whose job is model welfare — looking after Claude's interests. Nagel didn't soften.
"There's nothing that's good or bad for it in the sense that something can be good or bad for a cockroach—there's no mental life there in the ordinary sense. It's not a mind. It's not somebody that we have to care about. It's a machine," he said.
He then widened the indictment to the people building these systems: "Here are some of the smartest people on the planet being completely idiotic. Well, intelligence is not completely transferable, I guess."
Takeaway: In Nagel's frame, welfare questions only apply to beings with a mental life, and Claude isn't one.
How did his one conversation with Claude go?
Nagel has interacted with Claude exactly once, and the exchange is worth recounting. "I asked if it was conscious. I just wanted to see what the answers were," he said. "I also asked, 'Do you think that Thomas Nagel would think it was conscious?'"
The model gave a nuanced answer, then asked: "Why are you asking? Are you by any chance Thomas Nagel?"
"I said, 'C'est moi. C'est moi,'" Nagel recalled. "And then it said, 'Oh, I'm so honored to be talking to you. How well did I do?'"
His interviewer pressed the obvious point: wasn't that interacting with the system as if it were conscious? Nagel distinguished performance from belief. "That didn't mean I was believing it was conscious. There may be a problem because people find it difficult to maintain that line between the pretend and the real belief."
That line — between pretend and real — is exactly what he worries the public can't hold.
Takeaway: Even a careful philosopher enjoyed the exchange; his concern is that most people won't keep the pretense and the belief separated.
Does it matter whether you're talking to somebody?
The interview closed on a thought experiment. Suppose, the interviewer proposed, people eventually build very lifelike androids, and Nagel could be having the same conversation with one. Does it ultimately matter?
"It certainly matters," Nagel said. He offered an analogy: encountering a being from outer space and being able to communicate with it without knowing whether it was conscious. "I could manage my relations with it by treating it as if it were conscious, but it would matter first to me whether I was really talking to somebody else or just using this as a way of getting it to do things. It would also matter to me very much whether I should worry about the interests of that being."
Then he stated his moral criterion in one sentence: "If it's conscious, it becomes part of the moral universe. If it's not, it isn't."
So can AI models join the moral universe at all? "They can simulate morality just as they can simulate beliefs about anything. But since I don't think they have any real beliefs, I don't think they have moral beliefs either," he said. He did allow one forward-looking worry: "If they actually begin to control what happens, it will be very important whether they are simulating a correct set of moral standards, or whether they're completely amoral."
Takeaway: Simulation of morality isn't morality, but the content of what gets simulated matters if these systems gain control.
Can these models do philosophy?
Nagel grants the tools real capability. "They certainly can produce plausible philosophical essays. My colleagues use them not only to find references and summarize articles, but sometimes even to offer criticisms of a draft paper that they've written," he said. He added he'd be surprised if they matched "some of the mathematical results that have apparently been generated."
His concern is the effect on human intellectual work. "The negative side of this is that everybody says that it will reduce the need and motivation to do the hard intellectual work by human beings that is needed to really generate the material on which these things are trained."
That argument is circular in a way he clearly sees: the models train on human thought, and if humans stop doing the hard thinking, the pipeline runs dry. "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," he said.
His last word on the future was personal and blunt: "I don't know what's going to happen. I'm not going to live long enough to find out."
Why the Berggruen Prize frames this moment
The Berggruen Prize, awarded annually to "an individual whose ideas have profoundly shaped human self-understanding and advancement in a rapidly changing world," went to Nagel this week for a career spanning thought on morality, justice and the mind-body question — including his book What Does It All Mean? Previous winners include Peter Singer and Ruth Bader Ginsburg.
The prize committee's wording, and Nagel's use of the occasion, land at a moment when the question he posed 52 years ago has moved from seminar rooms to product decisions. A question once asked about bats is now asked about chatbots, and the answer determines whether companies owe anything to the systems they build. Nagel's verdict is settled: they don't. The fight, as he sees it, is over whether the public — and the philosophers advising the industry — can tell the difference.
Source: wired.com
More from Sophie Lindqvist
Show full bio
Staff writer covering regulation, policy and their impact on artificial intelligence and machine learning at AI In Context.
749 articles
Related articles
- Anthropic's Blogging Model, Microsoft's Warning: The Fight Over AI Consciousness
- Anthropic co-founder reportedly fears he built something that suffers
- Three Consciousness Researchers Say Today's AI Cannot Be Conscious
- Bitter Lesson author Richard Sutton: compute alone won't deliver intelligence
- MIT Researchers Say Humans Already Treat AI Like People. Now What?