Day of Teaching 2025: “AI is a Desire Path”
2025-01-29 How will the widespread availability of generative artificial intelligence (AI) change exams and learning assessments at universities? The keynote by Prof. Dr. Marlit Annalena Lindner (EUF - Europa-Universität Flensburg | IPN Kiel - Leibniz Institute for Science and Mathematics Education) addressed the far-reaching implications and potentials.
“Chat GTP is now available for everyone to try,” tweeted founder Sam Altman on November 30, 2022. Within days, more than 100 million people worldwide were using the program. Students were among them. Some of them have already predicted the end of written term papers. Despite the possibility of cheating with artificial intelligence in exams, psychologist Marlit Annalena Lindner rejects a strict ban: “Prohibiting generative AI at universities is like trying to catch the wind with nets,” argues the researcher. For exam formats such as term papers, she therefore advises a signed declaration of originality. “This assurance can provide a basis for action in clear cases of suspicion,“ she advises.
Overall, Marlit Annalena Lindner sees more opportunities than risks in AI: ‘AI is a ’desire path',” she explains. Like a trail that arises as a shortcut between paths, AI can accelerate the (cognitive) path to a goal, even if it is sometimes less perfect than one's own intelligence, but significantly faster.
AI can be used not only to create exam questions and sample answers more efficiently. “Students can also use this method to prepare for their own exams,” says Marlit Lindner.
Examinations would have to be adapted to the new reality. The Flensburg researcher advises integrating AI into teaching and examinations: “Students should acquire certain skills, and AI is one of them from now on. It will play an important role in their future professional lives. Students must not be left behind. Training AI literacy in a subject-specific way is therefore a good idea.” These skills also include a reflective approach to AI: ‘Why and how did I use AI? What would have been a better method? Where does data protection come in? We need skills-based assessments,’ Marlit Lindner summarizes.
To ensure the autonomy of the services, creative examination formats are essential, such as podcasts, learning diaries or practical questions that are answered with the help of theoretical knowledge: “In a computer science seminar, for example, students had to build a given model from a children's construction kit, applying basic knowledge of computer science,” reports Marlit Lindner.
One thing is certain: the progressive integration of AI technologies into everyday university life makes a reorientation of the examination culture inevitable in order to meet the challenges of the digital transformation and to establish contemporary forms of examination, says the psychologist: “We will come up against formal limits with new creative ideas, but examination regulations are not set in stone.”