COMPAS Colloquium: Generative Artificial Intelligence and the Alignment Problem

Artificial intelligence (AI) has taken over many parts of our lives. With the release of generative AI systems like ChatGPT, it will soon take over even more of our lives. But is that a good thing? The history of the development and implementation of AI systems has shown that even when computer scientists create AI systems with the best of intentions, those systems fail to behave in ways that align with what we value most. How can we bring the behavior of generative AI systems into line with what we value?

This COMPAS Colloquium featured Kevin Scharp (Illinois Urbana-Champaign, Philosophy), Micha Elsner (Ohio State, Linguistics), and moderator Bryan Weaver (Ohio State, Computer Science and Engineering; CEHV Steering Committee).