Martin Jakobsson: Discussions about AI from multiple perspectives

There is currently a lot of discussion about AI (artificial intelligence) and machine learning at Stockholm University. For example, it is included in various calls for proposals, such as in one of the Swedish Research Council’s strategic research areas – “Health, life sciences, and artificial intelligence”. It also comes up in discussions about cheating, where the unauthorised use of AI stands out due to a sharp surge in such use. However, the issue is much bigger than this, and will affect us going forward, both positively and negatively. We also discuss AI from different perspectives in management, including whether or not a policy for AI use is needed. AI is a broad field that is developing rapidly. I therefore believe it is more important to emphasise the importance of always being able to stand behind what you produce, regardless of how it was created, than to try to develop a permanent policy.

Machine learning is a subcategory of AI that involves training computers to identify patterns and correlations in existing data, rather than being controlled solely by pre-programmed rules. I myself work with machine learning in my research to recognise objects on the seabed that have been mapped using advanced sonar. However, the discussion about how AI is used mainly concerns what is known as generative AI, a specific category that involves the creation of new content.

AI has enormous potential in many respects and will contribute to many new research breakthroughs. I am absolutely convinced of this. At the same time, we need to reflect on how we use the technology. What will happen if we turn a blind eye and let our students, the researchers, teachers and citizens of the future, rely on generative AI to produce most of what they need for all kinds of assignments? How will we then be able to move forward in terms of development? Are we putting our entire future in the hands of the major corporations? Yet, I am personally convinced that what we could call “domain knowledge” will become more important than ever. We must ensure that our students acquire in-depth knowledge in the subjects they study, so that in the future they will be able to critically evaluate the output of generative AI.

Work is currently underway at Stockholm University to establish an AI centre. It started as a pilot project in the natural sciences, but there are now discussions about making it university-wide – something that has proven to be in high demand. Some research groups have already come a long way in their use and development of AI, while others are in the start-up phase. During a so callled Town Hall meeting this spring, organised by the pilot project leader, a clear need for networking emerged. Researchers want to meet and learn from each other. I look forward to seeing a university-wide centre take shape. It could be very valuable.

This text is written by Martin Jakobsson, Deputy Vice President. It appears in the section ”Words from the University’s senior management team”, where the management take turns to write about topical issues. The section appears in News for staff.