On 14 September 2026, Émile Servan-Schreiber started a new run of his IE University seminar, “How to predict the future with collective and artificial intelligence”. He wrote that students had rated the spring edition 4.95 out of 5, which left “still room for improvement”. He added that he hoped the upgraded Hypermind prediction market and the latest version of The Forecasting Machine would help him get there.
What the seminar covers
The course description starts from a simple premise: smart decisions require accurate predictions. It notes that cognitive scientists and AI engineers see prediction as the essence of intelligence, so improving forecasting skills should make people smarter, individually and collectively. The course sets out to teach, through theory and practice, how to use collective and artificial intelligence to forecast better, both personally and inside the businesses or public bodies students go on to join.
Practice comes from forecasting itself. Students take part in online prediction tournaments on economics and geopolitics, so they make probability forecasts on real questions and are scored when those questions resolve. The IE+ page lists the skills the course aims to build:
- telling collective intelligence apart from groupthink;
- understanding when and why crowds forecast accurately;
- the individual habits associated with superforecasters, the small share of people whose forecasts prove consistently accurate;
- applying these methods to real-world decisions;
- the comparative strengths of collective intelligence and AI, and how to combine them.
Format
The seminar runs over six sessions between 14 September and 19 October 2026, mostly on Mondays from 12:30 to 14:00, with one Wednesday-morning session on 14 October. According to the IE+ FAQ, advanced seminars are optional short courses of six sessions worth one ECTS credit, offered at no extra cost. Students must attend at least 80% of sessions, the first one is mandatory, and the course appears on the transcript without counting towards the grade point average.
The IE+ page introduces Servan-Schreiber as a French-American cognitive scientist who co-founded Hypermind in 2000 and is a founding member of the School of Collective Intelligence at Mohammed VI Polytechnic University (UM6P). He holds a B.S. in applied mathematics and a Ph.D. in cognitive psychology from Carnegie Mellon University.
An executive version in Morocco
Servan-Schreiber teaches the same subject to executives. In April 2025 he gave a seven-hour class on crowd forecasting in the “Collective Intelligence and Decision Making” module of the Executive MBA at UM6P’s Africa Business School, a 24-month programme run with Columbia Business School. He described a mix of theory, hands-on practice and case studies drawn from Hypermind’s experience, covering:
- prediction markets, where participants trade on outcomes and prices act as probabilities;
- prediction polls, where forecasters give probabilities directly and are scored on accuracy;
- the Bayesian truth serum, a scoring method that rewards honest answers even when the truth cannot be checked;
- superforecasting and AI forecasting;
- crowdstorming, a way of pooling ideas from many people.
He closed with a line for his students:
Prediction without decision is wastefulness, and decision without prediction is foolishness!
Émile Servan-Schreiber, LinkedIn, 27 April 2025
A longer trail of classrooms and public lectures
The IE and Executive MBA courses continue a teaching record visible across Servan-Schreiber's earlier posts. In November 2019, he taught an in-depth prediction-markets class at the 1337 coding school in Morocco with forecaster and researcher Pavel Atanasov as a guest. A 2021 UM6P profile described his graduate teaching as a combination of prediction markets, the psychology of forecasting, collective intelligence and computer science.
The same material also appeared in public and professional settings. At Oxford Saïd's Risk Management Symposium in November 2022, his talk connected crowd forecasting with Francis Galton's early account of the wisdom of crowds. In June 2023, he made a 15-minute case for the “crowd-forecasting superpower” at the State of Foresight Studies in the World symposium in Rabat; the organizer's full day-two recording preserves the session.
In April 2024, Sorbonne University Abu Dhabi hosted an interactive conference on group intelligence, diversity, open-mindedness and prediction markets. Its official recap says that live digital polls were used alongside examples to show how diversity can improve group outcomes. These appearances are not separate research results, but they show a consistent curriculum: understand why groups fail, structure independent judgements, express uncertainty numerically, and learn from resolved forecasts.
