Better Societies Podcast

Masterclass conversations with the world's leading AI safety experts.

Each episode is a deep-dive masterclass with a researcher, ethicist, or practitioner working on the hardest problems in AI safety. Free. For researchers, founders, policymakers, and anyone serious about AI safety.

Listen on SpotifyListen on Apple PodcastsWatch on YouTubeSubscribe via RSS

Listen now

Episodes

001

Risto Miikkulainen — UT Austin + Cognizant

Neuroevolution vs. Gradient Descent: What 30 Years of Research Tells Us About Safe AI

Risto shares why the field is rediscovering evolutionary approaches to neural network training, and what this means for AI alignment and safety properties.

52 min

More episodes releasing soon. Subscribe on Spotify to be notified.

Help accelerate AI safety

Tell us who you are and what you're working on. We'll route you to the right program.

Tell us about youJoin the free community