Episode Details

What if machine learning could uncover particle tracks physicists have been trained to miss? In this episode of Into The Impossible, host Brian Keating talks with Daniel Whiteson, UC Irvine professor and ATLAS physicist, about how artificial intelligence is changing particle physics. This condensed summary distills the full conversation into a quick listen, saving you time while covering the core ideas: why modern collider data is so hard to interpret, how traditional tracking algorithms rely on hidden assumptions, and how graph neural networks can search for smooth but unexpected trajectories. Whiteson explains how his team tested these ideas on hypothetical “quirks” and found that AI can learn more than just standard helix patterns—it can detect unconventional paths that may point to new discoveries. You’ll also hear how this approach could improve data reconstruction, reduce false positives, and speed up precision measurement. Listen now to get the key ideas in minutes.