In the article “What we should still be learning in 2025” in Forschung & Lehre, Professor Ute Schmid, Chair of Cognitive Systems at Otto Friedrich University Bamberg and Member of the bidt’s Board of Directors, explains how universities can use artificial intelligence in teaching. It is important that generated content is critically reviewed and corrected, which requires specialised skills. However, this requires clarifying which skills must be retained, which are becoming less relevant and which new skills need to be added.
We must ensure that the widespread use of generative AI does not lead to a loss of grounding in reality. The extreme case here would be that exercise sheets solved with generative AI are corrected using generative AI.
Prof. Dr. Ute Schmid To the profile
In order to overcome these challenges, Ute Schmid is conducting research in her project “Human-AI co-creation of programme code with different levels of prior knowledge: Effects on Competence and Confidence” (pAIrProg) at the bidt on this in the context of computer science degree programmes. It is assumed that learning by doing is still the only way to learn basic programming concepts & high-quality programme code. This is why new exercise and examination formats are needed, especially at the beginning of the degree programme, to encourage “do-it-yourself”.
To the article (german only)
research project