Will Brown
Will Brown, who publishes as William Brown, studied algorithmic game theory at Columbia UniversityColumbia UniversityColumbia University, formally Columbia University in the City of New York, is a private research university in New York City and a member of the Ivy League. Founded in 1754 by royal charter from King George II as King's College, it is the oldest institution of higher education in New York State and the fifth oldest in the United States. It was renamed Columbia College in 1784 and adopted the university name and moved to Manhattan's Morningside Heights in 1896. It has 17 schools and four affiliated colleges, with 35,978 students enrolled in fall 2025. Legal scholar Jennifer L. Mnookin became president on July 1, 2026.Open the full entry and worked on language models at Morgan StanleyMorgan StanleyMorgan Stanley is an American global financial services firm that grew from an investment banking partnership into a publicly traded financial group spanning investment banking, securities trading, wealth management, and investment management.Open the full entry. As head of applied research at Prime Intellect, he created Verifiers, an open-source library for sharing environments used to train and evaluate models.
Contents6 sections
Key facts
Undergraduate years: Discovering algorithms at Penn
Brown studied computer science, data science and philosophy at the University of PennsylvaniaUniversity of PennsylvaniaThe University of Pennsylvania, commonly called Penn, is a private research university and Ivy League member in Philadelphia. It traces its origins to a 1740 charity-school project; Benjamin Franklin proposed his academy in 1749, classes began in 1751, and a 1791 merger established the institution under its present name. Penn has 12 schools, with undergraduate education concentrated in arts and sciences, engineering, nursing and Wharton. Enrollment totaled 29,384 in fall 2025, and the fiscal 2025 endowment was $24.81 billion. As of October 2026, J. Larry Jameson was president, with his term scheduled to end on June 30, 2027.Open the full entry. He credited Sanjeev Khanna and Rajiv Gandhi with sparking his interest in algorithms and theory. Michael KearnsMichael KearnsMichael Kearns is an American computer scientist and University of Pennsylvania professor who works on computational learning theory, algorithmic game theory, and algorithmic fairness and privacy; he formerly headed the AI department at AT&T Bell Labs, co-wrote The Ethical Algorithm with Aaron Roth, and was elected to the National Academy of Sciences in 2021.Open the full entry and Aaron Roth guided his early research into fairness, privacy and game theory. His summer internship supervisors included Larry Rudolph and David Daly. He held research or engineering roles at AWSAmazon Web ServicesAmazon Web Services (AWS) is the cloud computing business of Amazon. Launched in July 2002, it made its public debut with the Amazon S3 storage service on March 14, 2006, and has since added hundreds of services, including Amazon EC2; its segment net sales were about $46.0 billion in the third quarter of 2026.Open the full entry, Two Sigma, MongoDBMongoDBMongoDB, Inc. is a New York-based database software company that develops the document database of the same name and sells it as the cloud-hosted service MongoDB Atlas, the self-managed commercial edition MongoDB Enterprise Advanced and the free Community Server. Formerly called 10gen, the company listed on Nasdaq in 2017 under the symbol MDB.Open the full entry and AmFam.[1][2][3]
2019–2024Doctoral research at Columbia
After a summer experimenting with multi-agent reinforcement learning in simulated trading markets, Brown entered graduate school in 2019 under Christos Papadimitriou and Tim Roughgarden. One question that occupied him was what a recommender might do over time if users’ preferences changed in response to its recommendations. He published a solo paper on multiplayer stochastic games in 2021 and coauthored papers with Arpit Agarwal on adaptive preferences in 2022 and 2024.[3][2][4]
In 2023 he published work with Jon Schneider and Kiran Vodrahalli asking whether the people who use learning algorithms in games benefit from doing so. In 2024 he and his advisers presented a paper on online optimization at COLT. He completed his PhD in computer science that year with the dissertation Learning in the Presence of Adaptive Behavior. He dedicated it to his parents Megan and David, his brother Ian and his cat Bingo, who kept him company as he wrote late at night.[3][4][5]
By 2025: From financial research to open-source reinforcement learning
At Morgan Stanley’s Machine Learning Research group, Brown worked on large language models. His publication list includes a 2024 report on clustering and entity matching and a 2025 collaboration with Brendan R. Hogan and others on tuning models for the Q programming language. By the time Delta interviewed him in August 2025, he was leading research at Prime Intellect and had created Verifiers. Developers could define tasks and feedback in the library and publish their environments in the Environments Hub for others to use.[4][5][1]
February 2026Discussing shared environments with Johannes Hagemann
In their February 10 interview with Sequoia CapitalSequoia CapitalSequoia Capital is a venture capital firm founded by Don Valentine in 1972, investing from the seed stage through the growth stage. Its historical investments include Apple, Cisco, Google, YouTube, Airbnb, and Stripe; in 2021 it introduced the Sequoia Capital Fund structure, which can continue to hold public securities, and in 2023 its regional businesses became independent brands and partnerships.Open the full entry, Brown explained how community members could find existing environments and make their own work reusable. He also drew a distinction between measuring and training. Repeatedly using a test task to train a model can teach it the particulars of that test, making its score a less informative guide to performance on new tasks.[6][7]
Related Organizations, Websites and Public Accounts
Morgan Stanley: Former member of its Machine Learning Research group. Official website: https://www.morganstanley.com/ (opens in a new window).[1]
Prime Intellect: Head of applied research. Official website: https://www.primeintellect.ai/ (opens in a new window).[1]
Personal website: https://willcb.com/ (opens in a new window). He uses the handle willccbb on GitHub and X.[1][8]
Sources8 sources
- About — Will Brown (opens in a new window)willcb.com↑
- Five years of feedback loops (opens in a new window)willcb.com↑
- Learning in the Presence of Adaptive Behavior (opens in a new window)academiccommons.columbia.edu↑
- Research — Will Brown (opens in a new window)willcb.com↑
- Will Brown — Prime Intellect Research Lead (opens in a new window)joindelta.org↑
- Building the GitHub for RL Environments: Prime Intellect’s Will Brown & Johannes Hagemann (opens in a new window)sequoiacap.com↑
- Building the GitHub for RL Environments: Prime Intellect's Will Brown & Johannes Hagemann (opens in a new window)youtube.com↑
- willccbb (will brown) (opens in a new window)github.com↑