Software Developer Graduates 2031

The computer science students starting their five-year degree in the fall of 2026, submit their master thesis in the spring of 2031. At universities around the country, there are enthusiastic and heated debates every week on what we should teach our students, as we’re gazing into the crystal ball. Student applicant numbers are plummeting, more or less LLM-generated solutions are increasingly submitted in coursework projects, and graduates are struggling to find an entry level job. This is the hard reality of computer science education in 2026. So where do we go from here? In academia, the discourse is centered around two core questions. Firstly, what are the skills and competencies we should teach the students that enroll in computer science studies this fall? And secondly, how do we get them there, to that desirable set of skills and competencies? There is a tradeoff between building a solid foundation along the lines of traditions in the field, versus bringing in agentic workflows, or other technological specs that can be found directly in job listings. This tradeoff is not entirely new, but the educational and technological landscape of today permeated by GenAI is forcing the issue. This talk attempts to give a peek into how the computer science education branch of academia is reacting in these times of uncertainty. How can we bring back the euphoria and passion that strikes when you finally figure that thing out after a week of debugging?