On 5 March 2020, Émile Servan-Schreiber reshared Pierre Haren’s announcement of four charts tracking COVID-19’s economic impact. Haren said they were updated hourly and followed the success of an article in Harvard Business Review. Servan-Schreiber described the initiative as hybrid collective intelligence. His short endorsement referred to another team’s work; it did not announce a Hypermind dashboard.
Haren reported more than 140,000 unique visitors, presenting that audience as a reason to keep the analysis going. The article was co-authored with David Simchi-Levi and dated 28 February 2020. The follow-up charts promised readers a way to revisit a developing economic story as new material appeared.
Machines organizing human explanations
A contemporaneous Causality Link analysis makes the hybrid idea more concrete. Its February 28 account describes tracking explanations across a large multilingual document collection. The company said it extracted causal links as authors expressed them, going beyond positive or negative sentiment to organize accounts of how the outbreak affected economic activity.
Its analysis distinguishes overlapping phases concerned with people, demand and supply. One chart groups mentions geographically; another compares affected industries. A further comparison separates reported consequences for consumption from those for production. Causality Link described these as signals of the collective understanding of the crisis, including which effects received more attention.
That description gives the evidence an important boundary. Counting explanations in documents measures what people are saying and what the system recognizes. It does not directly measure lost output or demonstrate that a proposed causal relationship is true. Such a display can help an analyst spot a developing concern, but forecast accuracy needs a separate comparison with later events.
Using monitoring to inform a forecast
The original hourly dashboard is unavailable, so the February analysis provides methodological context rather than its exact values. The distinction matters when using such a system: the intensity of discussion is a signal for an analyst to investigate. Turning it into a forecast would require a defined question about economic activity and an estimate that can later be compared with an observed result.
The journal’s Johns Hopkins–Hypermind COVID forecasting account describes a separate effort using explicit forecasts. Together, the examples show different roles for collective intelligence during a fast-moving crisis: organising explanations of current events and forming explicit expectations about future events.
