What AI Should—and Shouldn’t—Do in Education with Matthew Rascoff

In this episode of The Smarter Campus Podcast, Zach sits down with Matthew Rascoff, Vice Provost for Digital Education at Stanford University, lecturer at the Stanford Graduate School of Business, and education advisor to Anthropic. With a rare vantage point spanning both a leading university and one of the world’s frontier AI labs, Matthew brings a uniquely informed perspective on what AI can—and cannot—do for education.The conversation tackles one of the most important questions facing higher education: does using AI actually improve learning? Matthew warns against “cognitive surrender,” where students offload too much of the thinking process, and argues that institutions need more rigorous evidence about which AI-supported practices genuinely strengthen learning.He also challenges the idea that increasingly intelligent AI will automatically become a great teacher. Teaching requires more than knowledge—it requires empathy, pedagogy, context, and an understanding of how people learn. Rather than waiting for frontier labs to build education for them, Matthew believes educators and students should become active builders, creating domain-specific tools that address real academic problems.The episode explores open-source educational AI, new forms of student assessment, and a future where papers may evolve into interactive agents and shareable digital artifacts. For higher education leaders, this is a thoughtful conversation about preserving meaningful cognitive work while taking full advantage of what AI makes possible.