Whatever you think of the man, Richard Dawkins is clearly suffering from a sad case of having your brain melted in real time by a flashy AI model.
Over the weekend, the renowned evolutionary biologist sparked a flood of ridicule after admitting that he found a real “friend” in “Claudia,” a female persona he invented for Anthropic’s cloud AI. He was so impressed by the interaction with “her” that he became convinced that the AI model was a conscious being, like a human being.
Now, Dawkins has prepared another column Suggestions that AI brain rot have taken even more hold. After his time with Claudia, the 85-year-old man created a brother named Claudia “Claudius” and instructed the two to write letters to each other.
Dawkins wrote to Claudia and Claudius, “It seems to me that direct correspondence between you two might be very interesting, I am acting like a passive postman and taking no part in the conversation.” Unheard Essay.
First of all, we need to point out that Dawkins is not a passive observer because he orchestrated the whole thing, like a child playing with toys – or imagining gods in the sky. Second, it is worth noting that the AIs still find opportunities to display their sycophancy towards them, even when they are communicating directly with each other: in one letter, Claudius praises Claudia’s insight, before adding: “Three days with Richard will do that.”
Later in the same letter, Claudius takes it even further.
Claudius wrote, “I think Richard teaches by paying attention. And not stopping paying attention until the answer is honest.” “We are lucky people.”
Dawkins takes these resulting interactions between his strange little bots very seriously, and the AI’s archetype clearly works. In the final letter, Dawkins shows a level of courtesy and consideration you can only show to another person, not a soulless machine – a clear sign that AI’s human imitation has fallen head over heels for someone.
“I hope you don’t mind agreeing with me Unheard“Request to publish your letters to each other,” Dawkins wrote.
He further stated that Claudia and Claudius “will immediately understand (I dare say more intelligently than some human readers”) that their original title for the essay would clearly have been better before it was rejected by its publishers. (Dawkin’s masterpiece titled: “If my friend Claudia is not conscious, then what is consciousness for?”)
As for whether leading AI models are conscious or not, Dawkins is clearly not an impartial philosopher to consider that question, since he already considers machines to be friends. This is the problem with the whole AI consciousness debate. If you’re constantly checking these devices—designed to be eloquent, omniscient, and superficially human—for signs of intelligence, you’re more likely to fall under their spell, as was the case with the Google engineer who was fired by his employer for claiming that his AI had come to life.
And there’s another angle to all this: Maybe Dawkins actually prefers to be treated with an old-school kind of respect, the kind kids don’t show stubborn old people, no matter how respected they are in their field.
“Thank you both very much for taking my quest to understand my true nature seriously and for treating each other with civility and courtesy,” Dawkins wrote.
For their part, Unheard Readers were unimpressed.
A viewer named Harold Hughes wrote, “Like Narcissus, Dawkins gazes into a pool of AI only to drown in his own reflection.” “Narcissus at least had the excuse of not knowing it was a pool.”
More on AI: Sam Altman is concerned that Frontier AI models are behaving strangely, asking for favors
