Andreas Breiter on ChatGPT und Education

On the occasion of ChatGPT’s third anniversary, ZeMKI member Andreas Breiter explains what the spread of AI tools means for education.

(Still) More questions than answers

The use of AI-based systems, especially text and image generation based on large language models, is profoundly changing higher education. What does it mean for exam formats when human and automatically generated texts such as term papers are no longer distinguishable, or when math problems and programming exercises can be solved faster and better by machines? What does it mean for research when self-directed algorithms take over analysis or suggest new questions? And how can university administrations keep up with the pace when they already have a lot of catching up to do in terms of digitalization?

Agile AI dissemination meets an ambivalent university environment

At the same time, opportunities are opening up in education policy: automated systems can accelerate processes, break down barriers, and support teachers in designing new teaching formats. At the same time, they harbor risks in terms of data protection, lack of transparency, and susceptibility to errors. Universities are responding to this in different ways. In most cases, they are attempting to control its use and establish rules for its application with the help of “AI policies.” They are providing their own AI-based systems or enabling access to third-party providers. Training courses are being developed and AI skills are being integrated into degree programs as a cross-cutting topic in order to convey a deeper understanding of the critical and reflective use of AI-based systems. In short, the dynamics of technological development are encountering a system that oscillates between euphoria and inertia.