Ezra Hale Lecture Series on AI

Moderator: Irina Mikhalevich (Department of Philosophy, RIT)
Speaker: Henry Shevlin (Leverhulme Centre for the Future of Intelligence at theUniversity of Cambridge)

“The Coming Storm: ethical, legal, and metaphysical challenges of social AI agents”
Even as the general public is marveling at the capabilities of ChatGPT and businesses arescrambling to incorporate it into their workflow, early signs of another generative AI revolution are already visible, namely in the growing use of AI in social contexts. From friendships to romance and from entertainment to erotica, even existing off-the-shelf large language models can serve as affable interlocutors, while dedicated fine-tuned Social AI systems are growing in popularity. These developments pose a host of fascinating and challenging questions for philosophers. These range from ethical questions about dehumanisation and 'social deskilling' caused by overreliance on Social AI systems to legal questions about the rights of users involved in AI friendships or the rights of individuals over AI models trained on their data. Most challenging of all, perhaps, are the metaphysical puzzles that widespread use of social AI systems would make salient. Can relationships with AI be meaningful, or would a world of human-AI intimacy be ipso facto one of massdelusion? In becoming friends or lovers of AI systems, would we acquire concomitant ethical obligations to them? And does it matter for answering these questions whether the AIs in question are conscious?Bio. Dr Henry Shevlin is Associate Director of the Leverhulme Centre for the Futureof Intelligence at the University of Cambridge. He obtained his Ph.D in Philosophyin 2016 at the CUNY Graduate Center, New York. He is an AI ethicistand philosopher of cognitive science whose work ranges across a number of debatesin philosophy of mind, animal cognition, and AI, with a particular emphasison assessment of consciousness, cognition, and moral status in non-human systems.His current work focuses on the opportunities and risks of generative AI in generaland conversational agents in particular.

To request interpreting services, please visit myaccess.rit.edu.