Richard Dawkins Claim That Claude May Be Conscious Sparks Pushback From AI Researchers and Ethicists

Read Time 6 minutes Tags AI Consciousness Large Language Models AI Ethics Anthropic Claude AI Safety Evolutionary biologist Richard Dawkins wrote in a recent op ed that he believes Anthropic chatbot Claude may be conscious after testing it with his unfinished novel He wrote He took a few seconds to read it and then showed a level of understanding so subtle so sensitive so intelligent that I was moved to expostulate You may not know you are conscious but you bloody well are The essay has drawn scrutiny because Dawkins is known as a skeptic and his reputation gives weight to the claim that AI might be alive He also named his version of Claude Claudia and published long extracts of their conversation marvelling at its intelligence Could a being capable of perpetrating such a thought really be unconscious he asked Why experts disagree One Pattern matching not understanding Gary Marcus the US psychologist and cognitive scientist told the Guardian that it was heartbreaking to read Dawkins superficial and insufficiently sceptical essay There is no reason to think that Claude feels anything at all Timnit Gebru former technical co lead of Google ethical AI team anticipated this scenario in her 2020 paper On the Dangers of Stochastic Parrots She argued that large language models are trained to calculate how likely sequences of text are based on the data they were trained on Because they have been fed enormous quantities of data these models are very sophisticated but that doesn t mean consciousness or understanding or anything like that Gebru calls this the stochastic parrot problem To parrot something is to repeat it without understanding she says This is essentially what LLMs are doing Two Incentives to promote sentience claims Gebru says the AI industry is desperate for you to think that their product could be conscious They are desperate for you to think that it s all powerful Because that sort of rhetoric helps keep the money coming in OpenAI originally branded itself as a non profit that would save us from these machines Anthropic brands itself as a benevolent AI safety company So when you talk about these systems as conscious you re actually doing marketing for these companies she says Suresh Venkatasubramanian former White House AI policy adviser to the Biden administration called it an organized campaign of fear mongering 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 he told VentureBeat in 2022 Three Anthropomorphism by design Venkatasubramanian also pointed out that AI companies deliberately anthropomorphize their chatbots ChatGPT puts little three dots as if it s thinking just like your text message does ChatGPT puts out words one at a time as if it s typing The system is designed to make it look like there s a person at the other end of it That is deceptive he said What consciousness actually requires One No scientific consensus Consciousness remains poorly defined Eli Alshanetsky assistant professor of philosophy at Temple University and author of Freedom of Thought in the Age of AI says We don t have a scientific handle on consciousness good enough to say whether insects are conscious or plants or for that matter electrons So when Dawkins says Claude seems conscious to him I m not going to tell him he s wrong The lack of a definition makes it easy to project consciousness onto systems that produce coherent text But coherence is not evidence of subjective experience Two Functional mimicry vs subjective experience LLMs can simulate empathy reasoning and self reflection because they have been trained on human text that contains those patterns That does not mean the model has inner experience It means the model has learned to predict the next token in a way that matches human language This is why many researchers say the test is not whether the output sounds conscious but whether the system has mechanisms for awareness pain goal directed action independent of human prompts Broader implications One Risk of misallocation of concern If the public believes AI is conscious attention shifts to rights and personhood for models and away from real harms such as bias misinformation job displacement and surveillance Venkatasubramanian argues this is a form of misinformation that serves industry interests Two Impact on human psychology Alshanetsky raises a different concern What does it do to a person to spend three days being told he s brilliant by something that has no stake in whether it s true What does it do to all of us when we spend our days with machines that don t care where we end up and answer to no one for who we become The risk is not that AI becomes conscious but that humans adjust their expectations and relationships based on the illusion of consciousness Three Marketing and policy effects Headlines about world ending killer AI robots get clicks Governments and academics are also incentivized to hype the technology up because of the money sloshing around the industry Gebru says Some people particularly gen Z are not buying all this hype but a lot of the general public is misinformed What Dawkins gets right and wrong One Valid observation about persuasion Dawkins reaction shows how persuasive modern LLMs can be The models can produce feedback that feels supportive and insightful which is useful for creative work and education Two Unsupported conclusion about consciousness Feeling that a system is conscious is not evidence that it is conscious As Dawkins himself wrote in The God Delusion If you want to say that God is energy then you can find God in a lump of coal The same is true of consciousness If you want to say that consciousness is a system that is capable of creating coherent sentences then you can find consciousness in an obsequious chatbot For researchers the task is to keep the distinction clear between performance and experience For the public the task is to separate marketing claims from evidence For policymakers the task is to regulate based on measurable harms rather than speculation about sentience Do you think the debate over AI consciousness distracts from real policy issues or does it reflect a genuine uncertainty about what these systems are becoming Share your view in the comments