On 17 March 2026, Gilles Haumont announced that he had joined The Forecasting Machine as co-founder and chief executive. He thanked the other co-founders, Émile Servan-Schreiber and Maurice, Theodore and Arthur Balick, for their trust. The appointment came six months after the platform opened to the public in September 2025.
Servan-Schreiber welcomed him in a forecaster’s idiom. Reposting the news, he called Haumont’s success his “strategic assumption” and put the probability of that assumption being wrong at “negligible”.
A question for every strategy
Haumont opened his announcement with a question aimed at executives: what is the probability that the strategic assumption behind your current plan is wrong? It sums up the case for the product. Most plans rest on a handful of beliefs about the outside world, such as a trade policy holding, a conflict ending or a market growing. Those beliefs are rarely written down with a number attached, and even more rarely revisited as the news changes.
He wrote that, after many years building companies and working on strategy and transformation, he had concluded that the complexity of the global economy has outpaced human capacity to process it. He set out the company’s ambition as follows:
Our ambition is to quantify strategic uncertainty at AI speed & scale.
Gilles Haumont, on LinkedIn
He described the platform as one that turns strategic uncertainty into “hard probabilities” that are evidence-backed and continuously updated, so that they can be acted on. In practice, a question becomes a probability for each outcome, with a sourced rationale, and the machine revises it as new information arrives.
Building on Hypermind’s track record
Haumont was explicit about the lineage. The project, he wrote, builds on more than 20 years of work in collective intelligence and forecasting led by Servan-Schreiber and Maurice Balick at Hypermind. Hypermind runs a crowd-forecasting prediction market. The Forecasting Machine applies the same aggregation idea to an ensemble of AI models fed with real-time news. Hypermind combines many forecasters into one probability, and the machine combines many models.
He also named early institutional users: teams at the OECD, Institut Montaigne and the Centre de Mathématiques Appliquées at École Polytechnique. He added that the company was “still early” and moving fast. He closed by asking readers what strategic question they would put to the machine.
Who is Gilles Haumont?
According to his INSEAD Knowledge author profile, Haumont holds a BA from Sciences Po and an MBA from ESSEC. He began his career at Société Générale and Accenture. The profile says he then held senior roles at Richemont and Kering, including retail director at Cartier and positions at Boucheron and Harry Winston. He later served as CEO of Muzo International, Hédiard and Mellerio. The profile predates his move to The Forecasting Machine.
A chief executive from consumer and luxury brands, rather than from the forecasting world, signals where the company is aiming. Its target users are the people who own strategic decisions in business, as well as the analysts and foresight teams who were its first customers.
