Le 30 octobre 2026, à la Société de Lecture, j’ai prononcé l’allocution d’ouverture de la soirée « From Lab to Market » du Geneva Center for Neutrality, consacrée à une question : comment la Suisse peut-elle transformer sa recherche de pointe en intelligence artificielle en entreprises de premier plan ? Voici le texte de mon discours, prononcé en anglais.
Excellencies, distinguished guests, ladies and gentlemen,
Good evening and welcome.
The Geneva Center for Neutrality is an independent, non-partisan think tank. We believe that in an open, democratic society like Switzerland, it matters to bring people together and exchange ideas. We study every facet of neutrality, in particular digital neutrality and our sovereignty. These questions will shape Switzerland’s future, and technology now sits at the heart of them. Thank you to the Société de Lecture for having us.
This Société was founded in 1818 because Geneva’s public library was short of scientific publications. Two centuries later, Switzerland’s shortage is not research; we produce some of the best in the world. Schumpeter distinguished invention, which belongs to knowledge, from innovation, which needs a company to carry an idea to market. Our shortage lies in between.
Artificial intelligence sharpens the question. One new Swiss deep-tech company in four now works on it. Last year, EPFL and ETH Zurich counted a record 85 new start-ups and spin-offs between them. With the Swiss National Supercomputing Centre, they released Apertus, a language model trained on the Alps supercomputer and open down to its training data.
We also know how to keep companies alive: more than nine ETH spin-offs in ten are still active five years on. That is reassuring, and it is also a question. In venture capital, most young firms fail; that is the price of the few that become very large. Many of our spin-offs become solid, profitable small firms. Few set out to lead the world.
Now follow the money. When Swiss deep-tech companies raise between 15 and 100 million dollars, the rounds that turn a start-up into a real company, 82% of the capital comes from abroad. The European average is 60%. In 2024, 30 of the 41 Swiss start-ups that were sold went to foreign buyers. Foreign capital is welcome. The problem begins when value created with public research money grows elsewhere.
So tonight’s question is not whether Switzerland can do research. It is whether we can turn it into companies that grow here.
A paper is not a product. A product is not a paying customer. A customer is not a company, and one company is not an economy. Each step needs entrepreneurs ready to stake years on an idea that may fail, commercial skill that few doctorates teach, a first customer willing to bet on a newcomer, and years of execution. None of this diminishes fundamental research; much of what matters most was never meant to be sold, and not every professor should found a company. But when a result is ready to leave the laboratory, the people, the money and the customers should be there to meet it.
Geneva has done this before. In 1835, a young Pole, Antoine Norbert de Patek, arrived here as a refugee. He was a watch merchant. In Paris in 1844, he met a French watchmaker, Jean Adrien Philippe, who had found a way to wind a watch by its crown instead of a key. Patek brought him to Geneva: Philippe supplied the invention, Patek the commercial drive, Geneva its workshops. In 1854, Patek sailed to New York to find customers himself.
Our banks tell a slower story. Henri Hentsch, born in Geneva to a pastor from Lower Lusatia, founded in 1796 the bank to which Lombard Odier traces its origins. In 1805, two descendants of Protestant refugees from France, Jacob-Michel-François de Candolle and Jacques-Henry Mallet, founded the bank that became Pictet.
In Zug, the newcomers were American. In 1866, the brothers Charles and George Page set up the Anglo-Swiss Condensed Milk Company in Cham. They brought a process learned in America; Cham supplied water, a railway, workers and milk from some 1,300 farms. At its peak, the factory employed about one inhabitant of Cham in seven. In 1905, it merged with the firm of another immigrant, the German-born Henri Nestlé.
Different centuries, one lesson. An invention creates value when talent, entrepreneurship, capital and institutions meet in one place, and when that place lets them put down roots. Today, that meeting depends on three resources, the themes of tonight’s panel: capital, computing power and trust.
Capital first. Our pension funds manage more than 1,200 billion francs. One per cent, invested gradually in diversified portfolios of young Swiss companies, would be more than four times all the venture capital invested in Switzerland last year. No one should force the funds’ hand; their fiduciary duty comes first. But it is our responsibility to ask what we want to do with our savings, which could share in the growth our research makes possible.
Computing power next. A single campus in Texas is designed for 1.2 gigawatts; all of Switzerland’s commercial data centres together offer about 340 megawatts. We will not win the race on volume. In January 2025, Washington came close to capping our access to advanced chips. Sovereignty here is not self-sufficiency; it is choosing our dependencies rather than suffering them. Computing has become infrastructure, like energy or rail, and young companies need access to it.
And trust. Hospitals, companies and administrations rightly hesitate to hand their data to opaque systems under someone else’s jurisdiction. Switzerland can offer open, verifiable models, stable law, and data protection recognised by the European Union. But rules must be clear and proportionate; vague or shifting rules hit the smallest players hardest. Proton, founded in 2014 by scientists who met at CERN and still headquartered in Geneva, sells privacy, so the law it works under is part of its product. Last year, citing legal uncertainty over proposed surveillance rules, it chose to host its new AI assistant in Germany. Companies read our institutions as carefully as their contracts.
Geneva can bring these three together: Swiss institutional stability, the trust that comes with neutrality, and an international community that brings the world to our door. In September, Freedom24, the European platform of the Nasdaq-listed Freedom Holding, said it is exploring a presence here. We welcome it. Next June, Geneva hosts the international AI summit, under the theme « Bridging Innovation and Trust ». But neutrality is not a business model. Trust does not write code or sign purchase orders; it must be maintained, not simply inherited.
Three questions, then, for our panel. What is missing first: capital, computing power, customers, or people who have built a company before? How can a Swiss company raise its next 100 million at least partly at home? And what can Geneva do, concretely, before the world arrives next June?
In 1845, a refugee and an inventor began building a company in this city that still bears their names. Our ambition is just as simple: that the companies born in our laboratories today are still growing here, under their own names, a century from now.
We have the privilege of welcoming tonight’s panel:
– Professor Anna Fontcuberta i Morral, President of EPFL;
– Professor Marilyne Andersen, Director General of GESDA;
– Professor Srdjan Čapkun, Associate Vice President for Digital Transformation and Governance at ETH Zurich;
– Jay Woods, Chief Global Strategist at Freedom Capital Markets;
– and Nicholas Niggli, Executive Chairman of Oⁿ Venture Studio.
The debate is moderated by Raimondo Pictet, Head of Digital Initiative at the Geneva Center for Neutrality. The floor is yours. Thank you.