Back to Blog

On OrgaNova, Servan-Schreiber connects crowd forecasting, organisations and AI

2 min read
Navy cover headed “OrgaNova: crowd forecasting, diversity and AI” beside an OrgaNova episode 13 poster with Émile Servan-Schreiber’s portrait, “La prédiction,” and topics on prediction, cognitive diversity and decisions.

On 19 March 2026, Émile Servan-Schreiber shared Nicolas Bassan’s OrgaNova interview with a comment connecting collective intelligence, artificial intelligence and prediction markets. Bassan’s episode description places forecasting inside everyday organisational choices: anticipating a market, assessing a risk and deciding which innovation deserves investment.

The episode, released on 18 March, is a conversation about how to organise knowledge that no single person possesses. Bassan’s accompanying LinkedIn summary makes the practical argument more specific: the value of a crowd depends on the different information its members contribute.

More people, with diminishing returns

Bassan described the move from one participant to five as a large gain, and from five to ten as still useful. Beyond that, he said, additional gains become marginal. These are the host’s illustrative comparisons, not a published accuracy curve or a universal recommendation to recruit ten forecasters.

The distinction is useful for organisations. Adding another judgement may help when it brings a new source or perspective. Recruiting more people who repeat the same account gives a different kind of input. Group size alone does not tell a decision-maker how much independent information has been collected.

Three conditions for a useful collective

Bassan’s summary names diversity, independence and aggregation. Participants should bring different viewpoints, avoid copying each other and have a process for combining their answers. He contrasts that design with a group that amplifies biases and imitation. His account also lists distributed intelligence, idea prioritisation and AI among the topics discussed.

For an organisation, these conditions turn forecasting into a design problem. What knowledge should participants contribute? Can they form an initial judgement before seeing colleagues’ answers? How will competing estimates be combined? Bassan’s summary offers a framework for asking those questions before deciding that a larger group will necessarily produce a better answer.

A later If This Then Dev interview develops related themes in a separate conversation. OrgaNova’s emphasis, as presented by Bassan, is the organisational choice behind a useful crowd: how to gather complementary perspectives rather than simply increase the headcount.

Evidence

Sources & further reading

Keep reading

Related articles

News“Performance = Expertise + Diversity”: Servan-Schreiber on the If This Then Dev PodcastInsightsCrowds versus language models: the more accurate the AI forecaster, the more it errs like humansNewsAn 84-minute conversation, then AI-selected clips: Software in Blue on collective intelligence

Subscribe to The Forecasting Brief — forecasts & model notes, once a week