Samotsvety Forecasting is a distributed collective of elite forecasters focused on producing accurate, calibrated predictions on high-impact questions. Founded by Misha Yagudin, Nuno Sempere, and Eli Lifland, the group won the CSET-Foretell (now INFER) forecasting competition in both 2020 and 2021, outperforming the next-best team by roughly twice the relative Brier score. The team publishes influential forecasting reports on AI risk, nuclear risk, and geopolitical developments, and offers consulting services to organizations seeking expert probabilistic analysis. Arb Research is described as their sister organization.
Funding Details
- Annual Budget
- -
- Monthly Burn Rate
- -
- Current Runway
- -
- Funding Goal
- -
- Funding Raised to Date
- -
- Fiscal Sponsor
- -
Theory of Change
Samotsvety's theory of change is that producing high-quality, well-calibrated forecasts on existential and catastrophic risk questions directly improves the information environment available to funders, researchers, and decision-makers. By aggregating the views of elite forecasters using rigorous methods (such as geometric mean of odds) and publishing results openly, they aim to reduce overconfidence and anchoring in the EA and AI safety communities, surface cruxes in risk debates, and enable better-prioritized resource allocation. Accurate public forecasts on questions like AI timelines and nuclear escalation risk serve as reference points for the broader community and can influence funding, policy, and research direction.
Grants Received
from Long-Term Future Fund
Projects
No linked projects.
People
No linked people.
Discussion
Sign in to join the discussion.
No comments yet. Be the first to share your thoughts.
Details
- Last Updated
- Apr 2, 2026, 10:08 PM UTC
- Created
- Mar 19, 2026, 10:42 PM UTC