The Geometry of Reasoning and Learning in the Age of Agentic AI with Stefano Soatto

From University of California Video Podcasts (Video) by UCTV

May 6, 2026 · 44 min

About this episode

Stefano Soatto discusses the impact of artificial intelligence on understanding machine learning systems through the lens of information geometry.

Artificial intelligence affects how we understand the behavior of machine learning systems. Stefano Soatto, VP of Applied Science, Amazon Web Services, explains how ideas from information geometry shape emerging theories of how these artifacts work. Soatto examines the natural gradient, the connections between geometry and concepts such as probability distributions, entropy, mutual information, and KL divergence, and the challenge of defining information in trained models, helping clarify how reasoning and learning can be understood in the era of AI. Series: "Kyoto Prize Symposium" [Science] [Show ID: 41494]

People in this episode

Guest: Stefano Soatto

Topics covered

artificial intelligence
machine learning
information geometry
probability distributions
entropy
mutual information
KL divergence

Keywords

AI
machine learning
information geometry
natural gradient
entropy
mutual information
KL divergence

Mentioned in this episode

Organizations: Amazon Web Services, UCTV