Vladimir Trajkovikj on AI, Education and the Western Balkans

By Aleksandra Todorovikj · Sep 4, 2026

Generative AI is changing more than the tools students use. It is also forcing universities to reconsider what students need to learn, how knowledge should be assessed, and what skills will matter as increasingly capable systems can generate code, text and analysis within seconds. Prof.Dr. Vladimir Trajkovikj For North Macedonia and the wider Western Balkans, the discussion also reaches beyond the classroom. It touches research capacity, digital infrastructure, cooperation between universities and industry, and whether smaller countries can build their own AI expertise and applied solutions rather than remain primarily users of technologies developed elsewhere. AI Balkans spoke with Prof. Dr. Vladimir Trajkovikj about how generative AI is changing computer science education, the position of North Macedonia in the European research landscape, the potential for applied AI, and what the country needs to strengthen its technological and research capacity. AI Balkans: How is generative artificial intelligence changing the way students learn programming and computer science? Vladimir Trajkovikj: Significantly. Generative artificial intelligence (GenAI), in principle, changes the way students (and anyone else) can learn (and not only programming and computer science). The main change is that things which previously (at least according to Bloom’s taxonomy) were the most difficult, creating something new (a program) based on what you have understood and learned, now seem quite simple and easy, provided that you know how to explain it well (that is, provided that you understand the problem). This brings us back to the importance of the “big picture”. Learning programming and computer science in general is no longer aimed at enabling students to program functions or create interfaces in the first year after their graduation. Now they will need to know how to create architectures and more complex solutions. AI Balkans: Should universities change the way they assess students now that students have access to tools that can generate code, text and analyses? Vladimir Trajkovikj: In principle, yes. Assignments can now be more complex and more connected to system solutions (the big picture). In my opinion, students should learn and use the GenAI tools that are available to them, but that process should be transparent, and students should know how to explain how and why those tools were used. Such knowledge can be confirmed through self-reflection or through a practical verbal explanation of the system that was built. Additionally, because of the need to achieve detailed knowledge throughout the educational process (a better understanding of the problem being studied), students can also be assessed in a non-digital way (on paper or verbally, depending on the conditions and the topic being studied). AI Balkans: Are our universities managing to keep pace with the speed at which AI is developing, or is technology moving faster than the education system? Vladimir Trajkovikj: I cannot answer for all universities, or even for all faculties within my university. I believe this is being discussed, I believe that the teaching staff are making an effort, but I do not see that there is a systemic solution (even at the global level), so at the moment it depends on the commitment of professors and possibly on the dean’s offices of the faculties. AI Balkans: From your long-standing experience with e-learning and digital systems, is AI the next big change in education? Vladimir Trajkovikj: Formal education is difficult to change. Although there have been several industrial revolutions, it still (to a very large extent) seems to be stuck in the first one (students sit at desks and try to repeat what they have learned, while assessment comes down to detecting and punishing mistakes, exactly what was needed for the production lines that were created with the first industrial revolution). E-learning and numerous digital platforms provided a greater share of informal education (primarily because of the availability of educational resources and content) and learning that is not necessarily connected to a place (a classroom) or time (a class schedule), but they did not make an additional fundamental change. Artificial intelligence can adapt the material, summarise it, systematise it according to needs, and gradually introduce it (personalisation), as well as deliver it when needed (learning on time), and therefore formally provides an opportunity to change the way learning can take place. AI Balkans: In which areas of AI do you currently see the greatest research potential for smaller countries such as North Macedonia? Vladimir Trajkovikj: In my opinion, that would be the application of AI to increase the efficiency of various research processes, applied AI. But for this, valid data and digitised processes are necessary. The opportunity is there, but in order to use it, we must first make an effort to complete the prerequisites for it. If we immediately start applying AI to invalid data and non-digitised processes, things may be disappointing. If we start applying AI without having critical infrastructure and security mechanisms in place, the results could also be frightening (not only disappointing). AI Balkans: How realistically involved is North Macedonia in the European Research Area when it comes to AI and advanced digital technologies? Vladimir Trajkovikj: The answer to this question depends on several factors and, of course, on the perspective of the person answering it. I think the answer is yes, but unfortunately not equally. Namely, there are research projects in Macedonia, conferences where research related to AI and/or the application of AI is published, but a relatively small percentage of the entire research community is involved in this. From the perspective of that small part of the research community, the answer is YES. From the perspective of the larger part of the research community, the answer would probably…