Can AI think? A conversation with Peter Gärdenfors
With the rise of large language models like ChatGPT in recent years, almost every area of society stands to be impacted by these new innovations collectively termed 'AI', or artificial intelligence. But how intelligent actually are these technologies, and what might the future bring? Researchers at Lund University have already been grappling with these questions for some time, and in this episode, we showcase the research of Peter Gärdenfors, professor of cognitive science at Lund University, who this year published a new book, "Kan AI tänka?" ("Can AI think?"). Together with host Timothy Parker (LU International Office), Peter discusses his book and his comparisons between human and animal cognition, and his exploration of the differences between human cognition and AI systems. Also discussed are the difficulties in defining consciousness, the benefits and potential dangers of AI systems, the biases that can occur as a result of datasets used in training AI, and Peter's view on how AI might ultimately affect our working lives.