Skip to main content

Using Crowdsourced Forecasting to Augment Collective Intelligence


Collective intelligence is the “shared or group intelligence that emerges from the collaboration, collective efforts, and competition of many agents and individuals” [1] and which enables an organization or group to leverage human diversity, strengths, and capabilities to make better decisions than could be made by any single individual or agent.  While cooperation is commonplace among humans, collective intelligence can also be deliberately fostered through technology that structures more effective, larger-scale collaboration and aggregates the results to create high-quality outputs.  Collective intelligence tools are of value in making decisions, in particular, where:

  • There is a lack of clean, historical data to build comprehensive models;
  • The future may have little to no resemblance to the past;
  • Numerous variables and outcomes must be taken into account along with complex, hard-to-measure causal factors;
  • There are regular streams of new information that must be weighed and considered; or
  • There is a possibility of surprise events that likely wouldn’t be found in existing data patterns.

Making predictions about the future is one area where enhanced collective intelligence could dramatically improve the state of the art. Accurate forecasts could provide significant decision advantage, but — in spite of significant technological advances for predicting certain types of outcomes — they still depend heavily on human judgment.  While most decision-makers are under constant pressure to make critical decisions, the evidence suggests that most people are poor at making individually accurate forecasts.  However, collective intelligence — in the form of aggregating many individual forecasts using networks and platforms — has demonstrated success in enhancing the accuracy of forecasting.  Accordingly, ARLIS is leveraging technology and a decade of lessons learned from the Intelligence Community Prediction Market to launch the Crowd Security and Intelligence Forecasting Tool (CSIFT) to “augment” collective intelligence by eliciting and collecting probabilistic forecasts from cleared experts across agencies; aggregating forecasts to show changing consensus over years, months, or days in advance; and providing analyses and results for integration into reports and products for decision advantage.

Applied AI for Performance, Performance Augmentation, Cognitive Security, Collective Intelligence, Mitigating Insider Risk Research Team

Collective Intelligence Research Team