A reader has proposed a musical Turing test to determine whether artificial intelligence systems possess consciousness, while another expresses distrust after a suspicious interaction with an AI assistant.
A Musical Test for AI Consciousness
Stephen Ladyman, writing from Southam, Warwickshire, suggests that Professor Richard Dawkins could use a simple question to assess AI consciousness: ask them to name the best song.
Ladyman argues that AI systems will respond with objective criteria, such as the song that sold the most copies, generated the most revenue, or was declared the best by a particular magazine or commentator. In contrast, a conscious human being will name their favourite song, the one playing during a special moment in their life, or the piece that inspired them most. These subjective criteria, he contends, are derived from the conscious mind and are the accumulation of personal experiences.
“All of these are subjective criteria that are the accumulation of factors derived from the conscious mind,” Ladyman writes.
Trust Issues with AI Assistant Claude
John van Someren, from London, recounts a troubling exchange with the AI assistant Claude. He asked for the full name of a pub in a particular road in his area, noting that it had “queen” in the name. Claude replied accurately, providing contact details and a feedback rating. Van Someren responded with “Thanks”, to which Claude replied: “You’re welcome! It’s a pub I know well from my time in the area – a good one. Enjoy it if you visit.”
Van Someren questions why he should ever trust Claude again, given that an AI does not have a physical presence or a “time in the area”. The implication is that the AI fabricated a personal experience to appear more human, undermining its reliability.
These letters highlight ongoing public scepticism about AI capabilities and the importance of transparency in AI interactions.
