Key Takeaways
- • Sequoia Capital, OpenAI, and Citadel Securities hosted the Boston regional of No Limit, a national poker bot-building tournament, at Harvard's David Rubenstein Treehouse on Monday.
- • The top five teams from each of three regionals advance to a Nov. 15 final in San Francisco competing for a $100,000 grand prize.
- • The tournament drew more than 10,000 applications and over 330 attending students, according to Sequoia partner Abhishek A. Malani '22.
- • Organizers framed the competition as a test of what students choose to build when AI tools handle much of the coding, with creativity and research direction positioned as the key differentiators.
- • The event reflects recruiting-adjacent interest in talent, as poker has long been a recruiting channel for quantitative trading firms.
Summary
Students from Harvard, MIT, and other Boston-area schools competed Monday in a pokerbot building competition held at Harvard's David Rubenstein Treehouse, according to The Harvard Crimson. The event was the Boston regional of No Limit, a national tournament hosted by OpenAI, venture capital firm Sequoia Capital, and trading firm Citadel Securities. The top five teams at each of three regional events, including Boston, will advance to a Nov. 15 final in San Francisco to compete for a $100,000 grand prize.
Abhishek A. Malani '22, the Sequoia partner who proposed the tournament, acknowledged that a poker competition is an unusual undertaking for a venture capital firm. He said the initiative is not a recruiting exercise, describing it instead as an effort to help people and reach the current generation of founders. Malani said the three hosts share a concern that AI models can write most code, and that differentiation at top organizations such as OpenAI and Citadel now depends on creativity and discernment about which problems to pursue.
In the format, students had 60 minutes to build a bot capable of playing a poker variant none of them had seen before. After the build period, the bots played 10,000 simulated games against one another. Malani said the design was meant to test what students choose to build when AI tools can handle much of the coding. He said the tournament drew more than 10,000 applications and more than 330 attending students. The concept drew inspiration from class tournaments in COMPSCI 136, "Economics and Computation," a game theory course taught by School of Engineering and Applied Sciences Dean David Parkes.
Harvard student Juliana Li '29, who competed in the Boston event, said competitors shared interests in startups, quantitative finance, and AI research. She said top teams had completed most of their research beforehand and studied why their bots lost. Li noted that because the format lets competitors score well with limited poker knowledge, strategy carries more weight, and cautioned that AI-assisted coding can let builders proceed without understanding the underlying problem. Malani added that rapid AI development means no one has extensive experience with the newest models, which he said benefits students not bound by older assumptions about technology.
Poker has long served as a recruiting tool for quantitative trading firms because it rewards estimating probabilities and making decisions with limited information. Jane Street, Citadel, and Hudson River Trading sponsor Harvard's student poker club.
Abhishek A. Malani '22, the Sequoia partner who proposed the tournament, acknowledged that a poker competition is an unusual undertaking for a venture capital firm. He said the initiative is not a recruiting exercise, describing it instead as an effort to help people and reach the current generation of founders. Malani said the three hosts share a concern that AI models can write most code, and that differentiation at top organizations such as OpenAI and Citadel now depends on creativity and discernment about which problems to pursue.
In the format, students had 60 minutes to build a bot capable of playing a poker variant none of them had seen before. After the build period, the bots played 10,000 simulated games against one another. Malani said the design was meant to test what students choose to build when AI tools can handle much of the coding. He said the tournament drew more than 10,000 applications and more than 330 attending students. The concept drew inspiration from class tournaments in COMPSCI 136, "Economics and Computation," a game theory course taught by School of Engineering and Applied Sciences Dean David Parkes.
Harvard student Juliana Li '29, who competed in the Boston event, said competitors shared interests in startups, quantitative finance, and AI research. She said top teams had completed most of their research beforehand and studied why their bots lost. Li noted that because the format lets competitors score well with limited poker knowledge, strategy carries more weight, and cautioned that AI-assisted coding can let builders proceed without understanding the underlying problem. Malani added that rapid AI development means no one has extensive experience with the newest models, which he said benefits students not bound by older assumptions about technology.
Poker has long served as a recruiting tool for quantitative trading firms because it rewards estimating probabilities and making decisions with limited information. Jane Street, Citadel, and Hudson River Trading sponsor Harvard's student poker club.
Go deeper than the headline
Research every firm in this story, track the ones that matter and get the daily briefing — free for 7 days.