hybridforecasting.com

What It Is

The core project introduced on the Cultivate Forecasts page is the Hybrid Forecasting Competition (HFC), a multi-year research competition sponsored by IARPA and funded by the U.S. Intelligence Community. It is not a typical commercial AI SaaS product, but rather a human-machine hybrid research platform for forecasting geopolitical and geoeconomic events. Its goal is to test whether combining human analyst judgment with machine algorithms and statistical models can improve forecasting accuracy and timeliness.

Core Capabilities and Workflow

HFC’s key idea is “Human + Machine”: participants register online, accept the participation terms, complete background/onboarding questionnaires, and are then assigned to a research team’s forecasting system to submit forecasts for real-world questions published on an ongoing basis. The page emphasizes that hybrid methods can combine the respective strengths of human and machine forecasting systems while mitigating the weaknesses of relying on either approach alone. However, the main text does not disclose specific AI models, algorithmic architectures, data sources, automation levels, or output interfaces, making it difficult to assess its technical maturity.

Pricing, Chinese Support, and Integrations

The page does not mention subscriptions, fees, or enterprise purchasing information; the context is closer to a free-to-join research project. There is no disclosed information about Chinese-language support, APIs, system integrations, data export, or enterprise access management. On data privacy, it only states that participants must accept the participation terms and complete a background survey, without explaining how personal data is stored, used, shared, or anonymized.

Pros and Cons

Its strengths are a clear positioning, a focus on high-value geopolitical/geoeconomic forecasting, and the backing of IARPA, making it suitable for studying human-machine collaborative forecasting mechanisms. Its weaknesses are the lack of productization details, the absence of actual accuracy metrics, historical performance, and current availability information. The page also mentions that the competition launched in 2018, so its timeliness needs further verification.

Who It’s For and Access from China

It is better suited to researchers or forecasting enthusiasts in intelligence, policy, international relations, and risk analysis than to enterprise users looking for a ready-made AI decision-making system. There is no information about access from mainland China or supported payment methods, so availability can only be considered unknown. If you need alternatives, consider prediction market or crowdsourced forecasting platforms such as Metaculus, Good Judgment Open, Manifold Markets, Kalshi, or Polymarket.