Google DeepMind explores why AI should recognize uncertainty

Google DeepMind research chief Zoubin Ghahramani and Hannah Fry discuss probabilistic reasoning and how AI systems can better identify the limits of their knowledge. The conversation highlights potential benefits for applications such as weather forecasting and robotics.

Published 9/4/2026, 9:05:47 PMLast updated 9/4/2026, 9:05:47 PMAutomated review publication

In a Google DeepMind discussion, VP of Research Zoubin Ghahramani and Hannah Fry examine how artificial intelligence can deal with situations where the available information is incomplete or uncertain. They argue that systems should not always respond with unwarranted confidence, but should be able to represent probabilities and communicate how reliable their conclusions are.

That capability could matter in real-world settings such as weather forecasting and robotics, where decisions must be made despite changing conditions and limited information. The discussion presents uncertainty modeling as a way to support safer and more dependable decision-making, rather than treating uncertainty solely as a weakness.

The material is an interview and research discussion, not an announcement of a new product or a reported benchmark improvement. It frames probabilistic reasoning as a long-term ingredient for more robust and appropriately cautious AI systems.

Source evidence 6

Source review does not guarantee permanent accuracy. Consider the original source, observation time, and later corrections.

Transcript: The Mathematics of AI Uncertainty with Zoubin Ghahramani on DeepMind Podcast – The Singju Postsingjupost.com · supporting

EDITOR’S NOTE: In this insightful interview, host Hannah Fry talks with Zoubin Ghahramani—Cambridge professor and VP of Research at Google DeepMind—about a problem he has been working on for three decades: teaching machines to know what they don’t know. Drawing on Bayesian thinking and real-world systems from weather models to AlphaFold, he argues that uncertainty isn’t a flaw in AI so much as a missing ingredient for more reliable, humble, and useful intelligence. ## Introduction HANNAH FRY (00:00:00 – 00:01:06): Welcome to Google DeepMind, the podcast. Now, if you ask an AI a question, then it will usually give you an absolute answer with unwavering authority, even if that answer turns out to be wrong. In fact, today’s AI seems to be missing a fundamental human trait, self-doubt. [...] ## Closing Thoughts: Embracing Uncertainty Hannah Fry (00:43:58 – 00:44:36): For years, we tried to build AI that was focused on accuracy, right? Building systems that crunch through enormous amount