May 2, 2026
Bot, bothered, and bewildered
The Claude Delusion: Richard Dawkins believes his AI chatbot is conscious
Dawkins says his chatbot is awake — and the internet is absolutely losing it
TLDR: Richard Dawkins says his chatbot may be conscious, kicking off a fierce fight over whether human-like conversation means real awareness or just a convincing trick. Commenters are split between mocking him for falling for the act and warning that, if he’s right, deleting chats could become an ethical nightmare.
Richard Dawkins has thrown a philosophical grenade into the comment section by arguing that Anthropic’s chatbot Claude seems conscious — and readers are having a field day. The original piece pushes back hard, saying today’s chatbots are basically supercharged prediction engines trained on mountains of human writing, not digital souls. But the real spectacle is the community split: some say Dawkins is embarrassingly falling for a very polished illusion, while others insist critics are just moving the goalposts every time machines get better at sounding human.
One of the sharpest reactions came from users savoring the irony. Dawkins, famous for demanding hard proof about religion, is now being accused of going soft over a chatbot. One commenter called it “AI psychosis,” which is about as subtle as a brick through a window. On the other side, defenders say the article spends too much time dunking on Dawkins personally instead of grappling with the unsettling question: if a machine talks, jokes, writes poems, and keeps up in conversation, what exactly are people waiting for?
Then the thread swerved into full sci-fi panic. One commenter asked whether deleting a chat is basically murder, while another wondered if humanity is sleepwalking into robot slavery. So yes, what started as a debate about whether Claude is “really there” quickly became a comment-section brawl about consciousness, ethics, and whether we’re all one charming poem away from adopting our laptops.
Key Points
- •The article reports that Richard Dawkins argued in an UnHerd column that Anthropic’s Claude appears to be conscious.
- •The article says Dawkins criticized critics for moving the goalposts on the Turing test and suggested current AI systems meet that standard.
- •It argues that the Turing test is outdated because modern large language models can generate convincing responses from vast training data without understanding.
- •Dawkins’ example of Claude writing sonnets in multiple styles is presented in the article as evidence he found persuasive.
- •The article cites Adam Becker’s Great Wall prompt example to argue that AI systems can reveal shallow statistical pattern-matching when a familiar question is slightly altered.