AI Ethics: Cooperating with Stochastic Parrots?

Abstract:

Large Language Models are a significant new technology both technologically and socially, but what they are, exactly, and how we use them in society is still unclear. After reviewing what LLMs are, and why they are considered "Stochastic Parrots," this talk will present a definition of cooperation, then make a case that cooperation is a relevant concept for interacting with these models, even if they are not moral actors. It will conclude with notes on the emerging capabilities and the near future, and some thoughts on what would be needed if and when they are considered moral actors socially, ethically, and societally.

Bio:

David Manheim is a policy researcher and visiting lecturer at the Technion's department of humanities and the arts, as well as founder and director of research and policy for ALTER Israel, which does policy advocacy and research focusing on long term resilience and emerging technologies. His current project is a book, tentatively entitled Cooperation and the Future: Artificial Intelligence, Biology, Companies, Disinformation, and Everywhere Else.