AI researchers discuss why large language models excel at connecting ideas but struggle with business strategy and ecosystem design. They're testing real-world applications like detecting sepsis in newborns at Wageningen and Utrecht Medical Universities, where regulatory constraints matter more than raw intelligence.
Where one of the crazy things about a large language model is it naturally sees everything as interconnected.
Not only what's desirable, what's feasible, what's viable, figuring out the right business model, figuring out the ecosystem strategy around how to launch something effectively, actually building and testing solutions in real world contexts. That's not really so easily disrupted by AI. And that's actually where a lot of our focus is shifting as well as we, as we look at that, then.
We got like empowered by AI assistants, even AI powered. And I feel we're somewhere there right now. Furthermore, we've worked together with Wageningen University and Utrecht Medical University to figure out how we can detect sepsis in newborns infants. And we don't want to be overly broad in the regulation that we're crafting, but we also don't want to be too specific in where it's going to be just outdated and not even applicable. Why is not widely adopted across a lot of companies, a lot of institutions, a lot of countries and. Well, there's a lot of reasons why. But I'm going to tackle two of the most key challenges that we have noticed by working with a lot of institutions.