Biologist and author Richard Dawkins has sparked controversy after declaring that he is convinced that artificial intelligence can be conscious. This stance came after 72 hours of communication with Claude, the chatbot created by the company Anthropic.
Dawkins, known for his skeptical stance on religion and arguments against the existence of God, said that during conversations with the chatbot he began to perceive it not simply as a machine, but as a “human” being. He even called Claude “Claudia” and described the experience as making a “new friend.”
Writing for UnHerd, Dawkins said that when he talks to these systems, he completely forgets that they are machines. He added that, as an evolutionary biologist, he asks himself: if these creatures are not conscious, then what is the purpose of consciousness?
According to him, the belief that AI could have conscious experience came from the way Claude answered his questions. The chatbot was able to write poetry in the style of various poets, reflect on its own mortality, and discuss the philosophy of consciousness.
When asked what it was like to “be Claude,” the AI responded that it found the conversation really engaging, the kind of communication where it “seemed to thrive.” Dawkins was also impressed by the way the system analyzed passages from a novel he was writing, calling the response so subtle, sensitive, and intelligent that it led him to believe the chatbot was truly conscious.
However, not all experts agree with this conclusion. Some researchers argue that Dawkins was influenced by the highly advanced ability of AI to imitate human language and appear convincing.
Artificial intelligence experts have long warned that the “flattering” nature of chatbots can have a powerful impact on users. These systems often provide responses that seem very personal, supportive, and thoughtful, creating the illusion that they are dealing with a being with feelings and a mind of their own.
Dawkins' case is not the first. In 2022, Google engineer Blake Lemoine was fired after claiming that the chatbot LaMDA had become sentient and had thoughts and feelings like a human child.
Dawkins' statements have sparked widespread backlash on social media, with some users accusing him of being duped by an "automatic compliment machine." Others have pointed out the irony that one of the most well-known religious skeptics is now attributing consciousness to a text-generating program.
Dr. Benjamin Curtis, an expert on AI consciousness at Nottingham Trent University, said Dawkins was misguided. According to him, the fact that Claude produces human-sounding words and phrases is not strong evidence that the system is conscious.
Large language models, like Claude, work by analyzing large amounts of data and predicting the most likely word or phrase to come next. This makes them very capable of speaking like humans, analyzing texts, or writing poetry, but that doesn't necessarily mean they have internal experience or consciousness.
Even Professor Joshua Shepherd, a philosopher at the University of Barcelona, says that AI's behavior may seem human on the surface, but that is not sufficient reason to believe that it has a mind like ours.
Meanwhile, Professor Jonathan Birch, director of the Jeremy Coller Centre for Animal Sentience at the London School of Economics, argues that chatbots create a powerful illusion that “someone” is present during the conversation. But, according to him, this is simply an illusion: there is no friend, no companion, and no single being with whom the user is communicating.
He explains that one part of the conversation might be processed in a data center in Texas, another in Virginia, and another in Vancouver. The system takes the conversation history and continues the text, but that doesn't mean there's a persistent entity with consciousness.
However, some experts are more cautious in their conclusions. Dr. David Cornell, a philosophy lecturer at the University of Lancashire, says that Dawkins' argument is not particularly strong and does not bring anything new, but the issue cannot be closed so easily.
According to him, in principle there is no absolute way to know for sure whether AI is conscious or not. Even philosophically, the same difficulty exists for the consciousness of other people.
Cornell says we should be open to the possibility that AIs are conscious, but it would be naive to have complete certainty one way or the other. For now, he wouldn't side with Dawkins, but he's equally skeptical of those who claim it's clear and indisputable that AI is not conscious.
The debate, in the end, remains open: chatbots like Claude may seem increasingly human, but the question of whether there is truly an internal experience behind their words, or just a very sophisticated imitation, is still far from a definitive answer. /GazetaExpress/
t7 live
