Back
Forecasting Research Institute
webforecastingresearch.org·forecastingresearch.org/
FRI is a key institutional player in forecasting science; their methods inform how AI safety researchers and policymakers estimate probabilities of AI-related risks and make decisions under deep uncertainty.
Metadata
Importance: 55/100homepage
Summary
The Forecasting Research Institute (FRI) is a nonprofit research organization dedicated to improving the science and practice of forecasting, with applications to high-stakes policy and risk domains. They develop methodologies, run experiments, and collaborate with governments and nonprofits to make predictions more accurate and actionable. Their work has relevance to AI risk assessment and decision-making under uncertainty.
Key Points
- •Conducts empirical research on forecasting methods, including superforecasting, prediction markets, and aggregation techniques.
- •Partners with policymakers and nonprofits to translate forecasting science into practical decision-support tools.
- •Relevant to AI safety through improving risk estimation, scenario planning, and collective intelligence for existential threats.
- •Bridges academic forecasting research and real-world application in governance and high-stakes domains.
- •Contributes to epistemic infrastructure needed for rational responses to emerging technological risks.
Review
The Forecasting Research Institute (FRI) represents an important evolution in predictive methodology, building on the foundational work of Philip Tetlock in establishing rigorous prediction standards. Their approach moves beyond traditional forecasting by emphasizing practical applications and developing novel techniques for addressing complex, long-term challenges. FRI's research strategy concentrates on four key areas: generating high-quality forecasting questions about complex topics, creating methods for resolving seemingly unresolvable questions, testing forecasting techniques across different contexts, and developing tools to support organizational decision-making. This comprehensive approach demonstrates a sophisticated understanding of predictive science's potential to impact critical global issues, with particular relevance to domains like existential risk, biosecurity, and emerging technologies.
Cited by 3 pages
| Page | Type | Quality |
|---|---|---|
| Forecasting Research Institute (FRI) | Organization | 55.0 |
| AI-Augmented Forecasting | Approach | 54.0 |
| ForecastBench | Project | 53.0 |
Resource ID:
46c32aeaf3c3caac | Stable ID: M2ViMDBmND