“AI4Research gave the researchers the time to think long-term”

“Above all, I feel pride. When we started, very few people were talking about using AI for research and for science. There was not yet a European strategy for AI in the sciences, and there was very little discussion about AI as a method for basic research. It wasn’t until several years later that the EU launched its initiative on AI in Science and the pan-European initiative RAISE (Resource for AI Science in Europe) which was presented in 2025, to strengthen the continent’s position in AI‑driven research. So Uppsala University was on the ball early and I feel very pleased that our initiative was as big as it was.

“I think that we also succeeded in creating something valuable for the researchers: the chance to actually take the time to think long-term, which is almost a luxury in our daily lives as academics.”

The participants’ projects ranged from physics and medicine to the humanities and social sciences. What has AI4Research meant for the University’s research?

“It’s created opportunities to do things we couldn’t do before. In areas that have access to large data sets, it’s a tool. Areas such as physics and medicine sometimes measure more, and accumulate more data than what can be saved on hard drives. Then you have to decide very quickly what data is relevant and what to do with it, and that is something that needs to be automated.

“A particularly exciting trend is self-driving labs. We’ve had projects in chemistry as well as solar cells where the experimental process has been automated to a large extent or completely. Each new experiment is planned and carried out autonomously so that they can keep working all the time.

“Then we’ve had groups in the humanities and social sciences that have really taken to working with AI. In these disciplines, it actually has even greater potential because it provides completely new perspectives on what technology is doing to us. For example, I recall a project that looked at the consequences of autonomous decision-making in social security systems, which I found very interesting.”

Will the collaborations and ideas from AI4Research live on?

“Yes, definitely. I know that joint publications as well as applications for funding have been submitted following the project. But the main thing was that the researcher should have the opportunity to run their own project so that when they subsequently went back to their local environment, they were well ahead in that development.

“In addition, the University continues to work strategically with AI questions within the framework of UUniFI, and internationally this initiative is clearly being noticed. For example, the EU’s scientific advisory boards invited me to present what we have done here in Uppsala, which is proof that AI4Research has had an impact.”

Anneli Björkman