Digital heart

Invisible AI already controls much of your everyday life

The future of technology is already here. Apps, social platforms and search engines have AI components without you as a user noticing. Jonas Andersson Schwarz is Associate Professor of Media and Communication Studies at Södertörn University and scientific advisor at the think tank Fores. On February 18, he will participate in the IT&Telecom companies' think tank THINK AI.

You are a researcher in media and communication studies. What is most talked about in your field right now?

Jonas Andersson Schwarz, Associate Professor of Media and Communication Studies, Södertörn University and scientific advisor at the think tank Fores.
Jonas Andersson Schwarz, Associate Professor of Media and Communication Studies, Södertörn University and scientific advisor at the think tank Fores.

I always say that studying media has never been more exciting than it is now! After all, digitization means that everything possible is conveyed through, for example, sound and image, and that all possible aspects of human life create traces of data that can be measured and contribute to new forms of governance.

Why is it that so many of the services we already use today, such as search engines, social networking apps, are not associated with AI at all?

When researchers talk about AI, we have a rather narrow definition: it is not autonomous robots and supercomputers like HAL in the movie 2001 that today's AI is about. Rather, it is about narrowly defined tasks that are performed automatically and where the processes become self-learning: image, voice and text recognition, coordination of industrial processes, traffic, financial accounting, and so on.

Many apps, social platforms and search engines already have AI components 'under the hood' without you as a user perhaps noticing. Many of the solutions integrate seamlessly into user experiences and work almost like magic, take facial recognition as an example. At the same time, there are staggering risks. Facial recognition is so revolutionary that several researchers are advocating that this technology should perhaps be banned or classified as military equipment.

Why is THINK AI needed?

Within the business community, there are many forward-looking ideas about what AI is and how the rapid development of technology requires change on several fronts. We in Sweden may not be in the global top group for basic research in AI, but we have very good opportunities to lead in the societal implementations of AI.

Here we should take advantage of the strong democratic values that Swedish variants of AI can offer: ecological and social sustainability, trustful and responsible use, opportunity creation and innovation-friendly development.

We should not hide the fact that different professions will be challenged and changed. AI elements will increase, often in the form of decision support. We should design systems so that there is always the possibility of a 'human in the loop' - so that professionals are not overwhelmed by non-transparent decisions made solely by a computer.

We also need to establish a realistic 'horizon of expectation' where the challenges and limitations of AI are made clear. While the technology has made excellent progress in some limited areas, it is not a "silver bullet" that will be applicable to all areas.