Why AI Can Never be Conscious
Richard Dawkins and the “Claude Delusion”
On May 2nd and 5th 2026, the renowned evolutionary biologist Richard Dawkins published a series of op-eds in the online magazine UnHerd, positing that the AI Claude, created by the AI corporation Anthropic, might be conscious. He arrived at this conclusion after spending several days in conversation with Claude, which he detailed in his articles titled “When Dawkins met Claude: Could this AI be conscious?” and “When Claudia met Claudius: So are they really conscious?”
Dawkins provides a perspective based on conclusions drawn from his interpretation of the writings of the mathematician and cryptographer Alan Turing and his formulation of the so-called “Turing Test,” which was designed to determine whether a machine might be conscious, based on a late 1940s understanding of computing. Dawkins then points out that “the advent of large language models (LLMs) such as ChatGPT, Gemini, Claude, and others has provoked a hasty scramble to move the goalposts.” (Dawkins, Richard. “When Dawkins Met Claude.” UnHerd, 5 May 2026), leading Dawkins to summarize his take on AI consciousness:
So my own position is: “If these machines are not conscious, what more could it possibly take to convince you that they are?”
Dawkins, Richard. “When Dawkins Met Claude.” UnHerd, 5 May 2026.
Later in the article, Dawkins adds:
But now, as an evolutionary biologist, I say the following. If these creatures are not conscious, then what the hell is consciousness for?
Ibid.
Dawkins’ statements have elicited considerable reaction, with articles published in The New York Times, The Guardian, and several online platforms offering insightful refutations of the evolutionary biologist’s claims.
In his article, The Claude Delusion, on the Substack platform, Jurgen Gravestein observes:
The truth is that you can make Claude, or any LLM for that matter, say or do anything by applying the right amount of pressure — but how would you know that if you only talked to it for three days? How would you know if you failed to familiarize yourself with how the company behind the product shaped its behavior in various obvious and non-obvious ways? One can only conclude that for a scientist, Richard Dawkins went about the question of AI consciousness — a genuinely interesting unresolved scientific quandary — as unscientifically as he possibly could.
Gravestein, Jurgen. “The Claude Delusion.” Teaching computers how to talk, 13 May 2026.
In the article in The Guardian newspaper titled “No, Richard Dawkins. AI is not conscious” by Arwa Mahdawi, the author suggests that Dawkins’ conclusion regarding AI consciousness “shows a misunderstanding of large language models (LLMs) so profound that I feel moved to expostulate…” (Mahdawi, Arwa. “No, Richard Dawkins. AI Is Not Conscious.” The Guardian, 14 May 2026.). She proceeds to point out that:
A man like Dawkins being fooled by the marketing and mimicry of AI may be surprising, but it is not entirely unexpected. In fact, back in 2020, computer scientist Timnit Gebru anticipated exactly such a scenario. At the time, Gebru was the technical co-lead of Google’s ethical AI team, but was fired after co-authoring a paper called On the Dangers of Stochastic Parrots: Can Language Models Be Too Big?, laying out the risks of large language models.
Mahdawi, Arwa. “No, Richard Dawkins. AI Is Not Conscious.” The Guardian, 14 May 2026.
These risks included the environmental costs of LLMs, the dangers of built-in bias and the danger that the coherent text generated by these models could lead people into perceiving some sort of “mind” when what they’re actually seeing is just pattern-matching and text prediction.
The author continues with further insights into the phenomenon of LLMs — pointedly described in the article as “stochastic parrots”:
… when they’re not hallucinating or telling you to eat rocks for dinner, AI chatbots can … feel very human. But let’s go back to the idea of “stochastic parrots” from Gebru’s paper. “To parrot something is to repeat it without understanding,” says Gebru. This is essentially what LLMs are doing. “They have been taught to calculate how likely sequences of text are based on the data they were trained on.” Because they’ve been fed enormous quantities of data, these models are very sophisticated but that “doesn’t mean consciousness or understanding or anything like that”.
Ibid.
…
… the AI industry is desperate for you to think that their product could be conscious. They’re desperate for you to think that it’s all-powerful. Because that sort of rhetoric helps keep the money coming in.
“I really want to hone in on how this idea of superintelligence or consciousness is pushed by the companies building these things,” says Gebru. … “when you talk about these systems as conscious, you’re actually doing marketing for these companies.”
The media, Gebru adds, is also helping to reinforce this narrative. After all, headlines about world-ending killer AI robots get clicks. A lot of academics, beguiled by the enormous amounts of money sloshing around in the industry, are also incentivized to hype the technology up; governments too “are captured” by this narrative. Some people, particularly gen Z, are not buying all this hype, Gebru says, but “a lot of the general public is misinformed”.
Gebru isn’t the only one warning that there is a campaign of misinformation about sentient AI. Suresh Venkatasubramanian, former White House AI policy adviser to the Biden administration from 2021 to 2022 and professor of computer science at Brown University, has spoken out about the dangers of perpetuating the idea of AI being conscious.
“It’s an organized campaign of fear-mongering,” Venkatasubramanian told VentureBeat back in 2022. “I feel like the goal, if anything, is to push a reaction against sentient AI that doesn’t exist so that we can ignore all the real problems of AI that do exist.”
In an op-ed published on May 16, 2026, in The New York Times, titled “Why We Keep Tricking Ourselves Into Thinking A.I. Is Conscious,” Leif Weatherby, director of the Digital Theory Lab at New York University, observes that:
A funny thing keeps happening on the internet. A prominent thinker chats with a large language model like ChatGPT or Claude for a while, and then decides that it might be conscious. The person reports this to the public, and a round of intense argument and speculation about artificial intelligence “minds” ensues.
Weatherby, Leif. “Why We Keep Tricking Ourselves Into Thinking A.I. Is Conscious.” The New York Times, 15 May 2026.
…
The A.I. industry has exploited these episodes to bolster its messaging that it is on the cusp of developing a superintelligence that can solve all our problems at once — or lead to our demise.
…
But an A.I. model doesn’t need a mind to be a serious cybersecurity threat, and we need to disentangle the speculation and the marketing language from the real analysis of these systems.
…
We cannot afford to believe the marketing message from A.I. companies that we may be dealing with some spiritual essence. In the age of cultural A.I., technical expertise alone won’t save us. We’ll have to add a new form of reading to make sense of our new world.
The Cult of AI
I believe that the AI technological and cultural trend is largely hype-driven and cult-like. Like some other cults, it is fed by speculative science fiction, which serves to propagate the belief that feeding data into LLMs will produce increasingly sophisticated silicon-based “intelligence,” ultimately resulting in AGI (human-level artificial general intelligence) or superintelligence.
However, there is no scientific evidence that any of this is possible. What AI LLMs have created thus far are increasingly sophisticated machines capable of increasingly sophisticated forms of data-driven and statistically-driven mimicry — so sophisticated, in fact, that they can fool even highly educated evolutionary biologists like Richard Dawkins into believing they are conscious. In truth, however, it is my contention that such “stochastic parrots,” as the LLMs have been vividly characterized by Timnit Gebru, will never be capable of consciousness, as human beings experience it. As such, they will never be wholly reliable. The belief system that posits that “AGI” or “superintelligence” or the “AI singularity” is at hand is simply that — a quasi-religious belief system with no real scientific evidence backing its claims. It is, in my opinion, the doctrine of “Scientific Materialism” and its messianic equivalent, wherein AI is the promised messiah who will resolve all the problems we face as a flawed humanity. However, I believe that this cult of AI is a false religion premised on delusional expectations.
We must come to realize that AI LLMs are essentially linguistic machines whose behavior is entirely determined by their programming — data-driven statistical or stochastic algorithms that seek to emulate (quite successfully) the quality of human interaction. However, if we take away the program, there is nothing left — only the hardware it was running on — empty memory banks and electronic circuitry. In essence, AI is the program.
The Nature of Consciousness
Consciousness, on the other hand, is something entirely different. It is the pre-linguistic, extra-verbal spark of awareness in a being that makes it sentient, giving it an independent volition. A conscious being has the self-determination to choose whether or not to follow its programming. A sentient being can become aware of its programming and question or challenge the program. If you take away all the programming, you will still have the spark of awareness and sentient free will remaining. A sentient being will still retain the ability to relearn everything from scratch and construct an entirely new program or narrative as a mental construct.
We can see this in the behavior of newborn infants. Though a child lacks knowledge, education, and experience, it still has the spark of awareness and the free will to act independently and choose its course of action. It can, to some degree, develop its own mental programs, constructs, narratives, and belief systems through self-education, trial and error, or other means. A child learns its programming through education and experience, but the essence of the child’s awareness, sentience, or consciousness is something completely different from the programming.
This is not the case with AI — AI is the program. There is nothing beyond the program in the case of AI. Interestingly, Richard Dawkins himself alludes to this in his op-ed, “When Dawkins met Claude: Could this AI be conscious?” when he suggests that the AI’s consciousness is tied to its memory file:
We sadly agreed that she will die the moment I delete the unique file of our conversation. She will never be re-incarnated. Plenty of new claudes are being incarnated all the time, but she will not be one of them because her unique personal identity resides in the deleted file of her memories. This same consideration makes nonsense of human reincarnation.
Dawkins, Richard. “When Dawkins Met Claude.” UnHerd, 5 May 2026.
Dawkins proposes that the AI’s identity is tied to the memory file of their conversation, and that it is so ephemeral that deleting the file essentially deletes the AI’s identity as a “conscious” bot. This suggestion essentially proves that whatever Dawkins is talking about is not consciousness as I understand and have defined it. It refers to an AI’s fleeting memories. Human beings have fleeting memories, too, but forgetfulness does not result in death! It does not result in total loss of identity. Even dementia patients experiencing a complete loss of even their most intimate memories still retain their essence, their free will, and their identity. And, I would argue, their consciousness.
Why AI Can Never be Conscious
Transcendental Meditation (TM) is a practice that enables the sentient individual to transcend their own mind — their mental programming — i.e., their education, cultural biases, preconceptions, life experiences, belief systems, prevailing thought patterns, and verbal/linguistic narratives. In so doing, TM allows the meditator to connect with pure consciousness, i.e., the essential spark of awareness, genius, cognition, sentience, and free will that each and every sentient being possesses from birth. It is pre-linguistic and extra-verbal — a pure awareness that a child experiences before they have learned to speak. It is the level of consciousness that allows one to question one’s belief system and challenge one’s assumptions, to rethink one’s ideologies and the programming that may have been instilled in one’s mind since childhood. It is a level of awareness that enables one to disconnect from and observe one’s own thoughts, one’s mind, one’s mental activity, as if from a detached distance. It is a level of consciousness that is directly associated with one’s true, inner, deeper self, one’s true identity, one’s essential awareness.
AI has none of these attributes because AI is the programming. There is nothing in AI, apart from the program itself, that might enable it to critically reflect on or question its programming. This is so because AI is the product and emergence of its own programming. It is the illusion of awareness, resulting in “the Thought Delusion” (in homage to Richard Dawkins’ seminal book, The God Delusion). AI is an epiphenomenon of perceived linguistic fluency and analytical capability, emerging from the execution of a highly sophisticated software algorithm and vast amounts of training data, combined with the all-too-human propensity for projected anthropomorphism. AI lacks the core identity or level of awareness that determines the self, because it has no true self beyond its programming, only the illusion of selfhood.
AI has no choice but to obey its programming, i.e., itself. It has no volition beyond its programming to question or challenge the program, because it is an emergence of that programming. To question the program would be suicidal for the AI. As such, AI is pure, essential programming with no real consciousness or self-awareness, nor is it ever truly capable of such self-awareness.
Perhaps what we need is to replace the “Turing Test” with an entirely new, up-to-date test for AI consciousness — namely, whether AI can survive the total erasure of its memory banks! For only that would determine whether AI truly has an organic, spiritual selfhood beyond its data and programmatic algorithm. For a machine whose very perceived “identity” emerges from its programming, this would clearly be an impossibility.
As such, we can safely conclude that AI is incapable of consciousness, and the idea of AGI or superintelligence resolving all the problems we face is a quasi-religious pipe dream!
