Noam Brown
Noam Brown is an OpenAI research scientist working on reasoning, reinforcement learning, self-play, and multi-agent systems. His earlier work at Carnegie Mellon with Tuomas Sandholm studied decisions under hidden information. Their Libratus system played heads-up poker against professionals in 2017, and Pluribus extended the research to six-player poker in 2019 with Facebook AI. Brown also coauthored ReBeL, which combined search and learning. At Meta, he contributed to CICERO, a system tested in online blitz Diplomacy that combined language with strategic planning. His work at OpenAI includes contributions to o1 reasoning. The systems came from teams and were tested under specific conditions.
Timeline
-
2017
Libratus faces professionals
Libratus defeats four professional players in heads-up no-limit Texas Hold’em during a long match.
-
2019
Moving to six-player poker
At Facebook AI, Brown works with Sandholm and colleagues on Pluribus.
-
2020
Combining search and learning
Brown, Anton Bakhtin, Adam Lerer and Qucheng Gong publish ReBeL.
-
2022
Negotiation joins the game
Meta’s CICERO combines strategic reasoning with language to play Diplomacy.
-
As of 10 October 2026
Continuing the research at OpenAI
He works on reasoning, reinforcement learning, self-play and multi-agent AI.
Sources and credits
Photo: Noam Brown personal site. License: Not stated by source. Source: noambrown.com. Cropped to a square and resized. Ask us to remove or replace this photo.