21st Ethical Forum of the University Foundation
ChatGPT & Co in higher education: to be cheered or feared ?
Artificial Intelligence takes many forms. One of those of widest relevance for our universities is the passionately discussed chatbot application ChatGPT, launched in November 2022 by OpenAI. In a matter of seconds, it can produce texts that sound like well-informed answers to any possible question, react to users’ objections and summarize long texts. Its potential is evident in many fields, including all aspects of university life.
But it also constitutes, along with related AI applications, a formidable challenge, not least for the way we teach and evaluate our students. The texts ChatGPT generates, whether concise or richly developed, often seem to provide students with all they need to perform many tasks currently assigned to them as homework, tests or exams. Should we regulate, restrict or even forbid their use of ChatGPT?
Should we on the contrary encourage our students to use it and radically redesign how we evaluate them? Should we even rethink in depth what we need to teach them: why bother trying to teach them to write texts in their own and other languages if ChatGPT can do that far better for them? Should we concentrate instead on teaching our students critical thinking skills?
Or should we rather be primarily worried about the unwarranted trust which our students will unavoidably nourish for ChatGPT’s answers, or about the many biases that affect unwittingly the information that is reaching them, or about the huge ecological cost of the worldwide operation of AI applications?
In a nutshell: as teachers and as evaluators, should academics fear or cheer ChatGPT and its cognates?
P. Van Parijs (coordinator), J.M. Chaumont, E. De Keuleneer, M.C. de Marneffe, B. De Munck,
H. Garmyn, P. Goethals, J.P. Lambert, Q. Michel, B. Seghers, D. Willems, J. Willems.