Denver-based Science Markets has launched a new platform that allows participants to engage in prediction markets focused on significant scientific questions. Announced on November 17, 2026, the platform enables users to buy and sell YES or NO positions on well-defined, falsifiable hypotheses, potentially aiding in the broader public understanding of science.
The initial markets cover four scientific hypotheses: the composition of dark matter, the nature of neutrinos, the Riemann hypothesis, and the P versus NP problem in computer science. Each market provides a clear statement of the hypothesis and the criteria required for a YES or NO outcome, allowing participants to understand what they are evaluating before trading begins. These hypotheses are specifically designed to be falsifiable and well-defined.
Unlike traditional markets, these prediction markets do not have set end dates, reflecting the uncertain nature of scientific inquiries that may remain unresolved for extended periods. Participants can adapt their trading in response to new findings as research progresses, with transactions occurring on a decentralized blockchain.
The aim of Science Markets is to incentivize individuals to critically examine research and surface overlooked evidence, thus creating a decentralized method for aggregating insights on pressing scientific questions. Weekly assessments conducted by three AI models from OpenAI, Anthropic, and xAI are published on the platform, offering evaluations of the ongoing research and citing sources relevant to each hypothesis. By comparing these assessments with market prices, users can investigate differing opinions within the scientific community.
Each market will settle only when all three AI models provide a unanimous YES or NO based on the predetermined criteria; if agreement is not reached, the market will remain unresolved. The platform aims to foster deeper engagement with scientific research, though it does clarify that agreement among the AI models does not equate to established scientific truth.
Trading is currently not available to participants in the United States and certain other jurisdictions, with the platform advising that participants may potentially lose their entire stake and that markets can remain unresolved indefinitely. Once a market does settle, those holding the winning position will be eligible to claim their payout per the established rules.
Conner Lambden, the founder and a former computational scientist at Brigham and Women’s Hospital, leads Science Markets, which is operated by Conner Productions LLC. He holds degrees in molecular and cell biology and computer science from the University of California, Berkeley.


