INFER (formerly Cultivate) vs Good Judgment Open
| Attribute | INFER (formerly Cultivate) | Good Judgment Open |
|---|---|---|
| Founded | 2019 | 2015 |
| HQ | Washington, D.C., USA | New York, USA |
| Type | U.S. Government Early Warning & Policy Forecasting Platform | Crowdsourced Geopolitical Forecasting Platform |
| Regulated | ❌ No | ❌ No |
| Valuation | — | — |
| Status | Active | Active |
| Prestige | 80% | 88% |
INFER (INtegrated Forecasting and Estimates of Risk)—originally launched as Foretell by Georgetown University's Center for Security and Emerging Technology (CSET) in 2020 with seed funding from Open Philanthropy—was created to provide quantitative early-warning signals for U.S. policy and technology decision-makers. The program transitioned to the University of Maryland's Applied Research Laboratory for Intelligence and Security (ARLIS) under the INFER banner before evolving into the RAND Forecasting Initiative (RFI), operated by the RAND Corporation in partnership with software developer Cultivate Labs.
Operating as a non-monetary, reputation-scored crowdsourced forecasting ecosystem, INFER utilizes Cultivate Labs' enterprise software engine to capture probability distributions on critical policy questions. The platform decomposes complex geopolitical, artificial intelligence, biosecurity, and microelectronics challenges into discrete, measurable indicators. Forecasts are gathered from a hybrid crowd composed of vetted public forecasters, university researchers, and U.S. government analysts, with individual contributions evaluated using strict Brier score calibration.
Full profile: INFER (formerly Cultivate) →Good Judgment Open (GJ Open) was launched in 2015 as the public-facing forecasting portal of Good Judgment Inc, a commercial spin-off originating from the landmark Good Judgment Project (GJP). Co-founded by Wharton professors Philip Tetlock and Barbara Mellers alongside UC Berkeley professor Don Moore, the platform stems from research funded by the US Intelligence Advanced Research Projects Activity (IARPA) Aggregative Contingent Estimation (ACE) tournament, which demonstrated that well-calibrated crowd ensembles systematically outperformed professional intelligence analysts.
Unlike financial orderbook prediction exchanges that rely on real-money or cryptocurrency pricing, GJ Open operates as a non-monetary, probability-based forecasting engine. Participants submit percentage probabilities on complex geopolitical, macroeconomic, public health, and environmental questions. Individual predictions are tracked over time and evaluated using Brier accuracy scores—a mathematical metric measuring probabilistic calibration and resolution accuracy.
Full profile: Good Judgment Open →