It's a good question. I think the core of it is to reframe how we think about risk, the KPIs that we're using to measure these things.
Currently, most of the programs that are interested in supporting commercialization, both at the institutional and the public sector level, try to pick winners. They try to evaluate the success or failure on an individual project basis. I think to be successful in commercializing research—which, as you mentioned, has this very high failure rate—we need to think about aggregate impact. We need to think about risk at the level of an entire portfolio of projects, as opposed to individual projects.
Les Deeptech in France is a good example. It's a relatively new initiative that's deploying three billion euros over 10 years, aiming to create 500 new deep-tech companies every year over that time. They're targeting a 2% success rate, accepting that 98% will fail and that the 2% that succeed will more than pay for all of those failures.
That's the kind of portfolio-level thinking that needs to happen in order to be able to address this here.