Good Judgment Open - Forecasting Platform

Summary

Good Judgment Open is a public forecasting platform where participants make probabilistic predictions on geopolitical, economic, and other real-world questions. It applies the superforecasting methodology developed from IARPA's research, aggregating crowd wisdom to produce well-calibrated probability estimates. The platform is relevant to AI safety for its work on forecasting AI-related developments and demonstrating structured uncertainty quantification.

Key Points

• •Public platform for probabilistic forecasting on real-world questions using superforecasting methodology
• •Aggregates predictions from large crowds to produce calibrated probability estimates on uncertain outcomes
• •Relevant to AI safety as a tool for forecasting AI capabilities timelines and policy developments
• •Based on research showing that structured forecasting with feedback significantly improves prediction accuracy
• •Used by researchers and analysts to track and quantify uncertainty around emerging technology risks