Brandon Anderson

Brandon Anderson builds AI systems for RNA drug discovery at Atomic AI, where he is listed as Staff Scientist, Machine Learning. He is also a cohost of Latent Space’s AI for Science podcast alongside RJ Honicky.[3][1][2]

Anderson and Honicky became hosts of the new podcast at its January 28, 2026 launch. Its opening episode featured Andrew White discussing scientific agents and automated discovery. The episode introduction distinguishes Anderson’s commentary from his employer’s positions: “Anything said on this podcast is his personal take — not Atomic’s.”[2][1]

Before the podcast launch, Anderson coauthored the ATOM-1 preprint, posted December 14, 2023, under the name Brandon M. Anderson and an Atomic AI affiliation. His collaborators included Nicholas Boyd, Brent Townshend, Ryan Chow, Stephan Eismann and Raphael J. L. Townshend. The work addressed a challenge in RNA therapeutic design: rational design requires understanding RNA structure and function, while experimental screening can be costly and time-consuming.[6]

ATOM-1 learns RNA representations from chemical mapping data, using data-collection strategies developed for machine-learning training. The coauthors evaluated those representations by training small probe neural networks on the model’s embeddings with limited additional data. They reported state-of-the-art accuracy on key RNA prediction tasks and described ATOM-1 as the first RNA foundation model trained on chemical mapping data. These are joint research results; the preprint does not specify Anderson’s individual responsibilities.[6]