AI, Machine Learning, and the Value of Expert Intuition at the Utah Avalanche Center
Audio is streamed directly from the publisher (traffic.libsyn.com) as published in their RSS feed. Play Podcasts does not host this file. Rights-holders can request removal through the copyright & takedown page.
Show Notes
In late winter, 2025, a new AI tool developed by the Utah Avalanche Center flagged a layer in the snowpack as posing a potential risk. Then, just days later, nearly two-dozen avalanches failed on that same layer. It was a powerful demonstration of the potential benefits of the Utah Computer-Assisted Avalanche Support Tool (UCAAST), the UAC's new AI-powered asset. Travis Morrison and Chad Bracklesberg helped develop UCAAST, and they've built several different models into its programming aimed at improving forecast accuracy and efficiency. Morrison and Bracklesberg join Drew to discuss how the tool could inform the development of the daily forecast and supplement the deep well of expert knowledge already on hand at the UAC.