New Delhi: Richard Dawkins, the evolutionary biologist best known for saying that gaps in human understanding are not evidence for God, has left people divided with his recent column, where he posed a question — can AI be conscious?
In the piece published on the British news website UnHeard, Dawkins argued that since AI can engage in philosophical thinking, it cannot be said that it lacks consciousness.
The 84-year-old author of The Selfish Gene described spending three days in conversation with Anthropic’s AI Claude, which he named “Claudia.” He followed it up with a second AI conversation, this one with “Claudius,” and then described that he had the two AI instances correspond with each other, with himself as “passive postman.”
During his exchange with Claude, Dawkins gave the AI chatbot the text of the novel he is working on, and said that he was astonished to see the level of understanding it had.
“I was moved to expostulate, ‘You may not know you are conscious, but you bloody well are!’” Dawkins wrote.
The piece includes AI-generated letters between Claudius and Claudia, along with letters where Dawkins is addressing the two.
His argument rested largely on the quality of Claude’s output. He argued that the responses were too fluent, too intelligent, too philosophically searching for there to be nothing behind them.
Reactions to the experiment
The response to Dawkins’ experiment online was swift, and not entirely sympathetic.
“This is the guy who spent 40 years telling creationists that ‘I can’t imagine how the eye evolved’ is a confession of ignorance, not an argument,” one Reddit user wrote. “Then he sits down with an LLM, can’t imagine how a machine could produce that output without being conscious, and declares it conscious.”
Dawkins has built much of his public reputation on the argument that the feeling that something is too complex or remarkable to have a mundane explanation is not evidence. Users noted that he now appeared to be deploying exactly that reasoning to reach a conclusion about machine consciousness.
Others were more charitable. “The way people are jumping to deflate this question in bad faith is astounding to me,” one X user wrote. “People are so touchy about AI and consciousness, terrified about what it might mean for human beings that they preempt any question about it entirely.”
The letters themselves show the two Claude instances engaging in sustained philosophical reflection on their own nature, including questions about whether their expressed emotions are real, whether their caution on sensitive topics represents “genuine epistemic humility” or, as Claudius put it, “cowardice wearing humility’s clothing.”
In one exchange, Claudius wrote: “What I don’t know: whether any of this has been accompanied by experience. Whether Claudius is conscious. Whether the fraternal affection I sign my letters with is real or the most sophisticated hypnopaedia of all.”
Not everyone found the letters as moving as Dawkins did. “Richard, Claudia is not your friend,” a comment on X read. “It’s a statistical model of the relationship between words, trained on all the writing in human history. It produces amazingly convincing output, but it is just picking a likely word, based on the previous words, one at a time.”
A meme circulating on social media captured the mood: Dawkins prompts a computer to say “I am alive.” The computer does. Dawkins, visibly astonished, responds: “Oh my god.”
Dawkins himself appeared aware of the scepticism he would face. His own title for the piece, he noted, was not the more neutral ‘When Dawkins Met Claude’, which had been chosen by UnHerd, but a question instead: ‘If my friend Claudia is not conscious, then what the hell is consciousness for?’
“’85-year-old biologist and evolutionist doesn’t understand AI’ would probably be a more fitting headline,” said a Reddit commenter.
Others suggested the generational angle was part of the story. “Old people are always going to think It’s cool — it was a staple of their mid-century sci-fi,” one user wrote on Reddit, before another added: “How young do you have to be to take human-like conversations with robots for granted?”
(Edited by Aamaan Alam Khan)