In this episode of the fAQ Podcast, host Tai-Danae Bradley sits down with Daniel Litt, an assistant professor of mathematics at the University of Toronto, to discuss the rapidly evolving intersection of AI and mathematics. Bypassing the public speculation and social media hype, this conversation provides a practical look at some ways large language models are actively being used in high-level mathematical research. Daniel shares his daily workflow, detailing his transition from testing early AI models in a Dungeons and Dragons game to currently utilizing advanced tools for literature search, "vibe coding," and finding counterexamples. The discussion also covers the new challenge of "proof abundance," practical tips for prompting AI effectively, and how these powerful tools might force a necessary restructuring of mathematical education and the ac ...