Gradient Dissent is an interview podcast from Weights & Biases, hosted by its founder Lukas Biewald. It launched in March 2020, comes out roughly every two to four weeks, and episodes typically run 50-68 minutes.
Gradient Dissent is rated 4.8 out of 5 from 70 ratings on Apple Podcasts, as of October 2026.
Lukas interviews AI leaders and practitioners about frontier AI work and bringing models into production. The production angle sets it apart from research-only shows.
Guests range from AI coding startups to chip makers to open-source framework creators.
Our take: it is for ML engineers and AI leaders interested in production ML and AI infrastructure. For more shows like it, see our best machine learning podcasts list.
New episodes come every few weeks, so the back catalog is where most of the value is. The Cursor, Jensen Huang and PyTorch episodes below span AI coding, hardware and open-source tooling.
We have tested Cursor ourselves; see Cursor vs Google Antigravity.
The latest episode in the public feed when we checked was dated August 18, 2026.
If you like Gradient Dissent, these shows are worth a listen:
Show details on this page come from the public listings and feeds below, checked on October 4, 2026.
Lukas Biewald, founder of Weights & Biases, which produces the show.
Roughly every two to four weeks, with episodes of about 50 to 68 minutes.
Interviews with AI leaders and practitioners about frontier AI and bringing models into production.
On Apple Podcasts, Spotify and the Weights & Biases website.
