.png)
February 16, 2025
The rapid development of Artificial Intelligence (AI) is increasingly changing the academic landscape. AI-powered chatbots are being used more and more at universities and colleges, acting as digital learning assistants for students. These virtual tutors offer around-the-clock support, answer questions, provide feedback, and enable personalized learning. But how do these systems work in practice, and what opportunities and challenges do they bring?
The International University of Applied Sciences (IU) has been using the AI learning assistant "Syntea" since 2023. This chatbot is based on "Natural Language Processing" models and has been trained with IU course materials and learning content. This allows "Syntea" to answer student questions, create quizzes, and conduct exam simulations. According to IU, over 64 percent of students already use the chatbot regularly, with around two million chat requests per month. Quintus Stierstorfer, responsible for "Synthetic Teaching" at IU, sees AI learning assistants as an opportunity to personalize studies and make them more flexible. "Syntea" is intended to give students a feeling of real communication and function as a mix of tutor and mentor, adapting to individual needs.
The Technical University of Munich (TUM) is also experimenting with AI chatbots in teaching. "Iris," TUM's chatbot, supports students with programming tasks, for example, by providing thought-provoking impulses instead of directly presenting the solution. Students appreciate the chatbot as a helpful supplement to human tutors. Alexander Braun, Vice President for Digitalization and IT Systems at TUM, emphasizes the importance of critical evaluation of AI models by users. The university sees its task as training students in the use of AI so that they can effectively use the technology as a tool.
The use of AI chatbots in university teaching offers a variety of possibilities. Virtual availability around the clock enables flexible learning, regardless of location and time. Personalized feedback and adapted learning paths can improve learning success. The automation of tasks such as answering standard questions relieves teachers and creates space for more intensive support.
Nevertheless, there are also challenges to overcome. The quality of the answers depends on the training data. "Hallucinations," i.e., the generation of false or misleading information by the AI model, represent a known problem. The development of strategies to avoid hallucinations and to ensure quality is therefore essential. Data protection aspects must also be considered. The integration of AI chatbots into existing teaching structures also requires careful planning and didactic concepts.
The development of AI chatbots in university teaching is still in its early stages. In the future, AI agents could support students even more proactively in their learning, for example, through personalized learning plans and reminders. The integration of gamification elements could increase motivation. The use of AI-based assessment systems is also being discussed. The ongoing development of AI promises further innovations in the field of education that could make learning more individual, flexible, and effective.
Bibliography: - https://www.zdf.de/nachrichten/wissen/kuenstliche-intelligenz-agent-hochschule-universitaet-100.html - https://x.com/ZDFheute/status/1890822140447842443 - https://www.zdf.de/nachrichten/thema/kuenstliche-intelligenz-ki-100.html - https://www.zeit.de/2024/21/kuenstliche-intelligenz-studium-syntea-lehre-hochschule - https://ki-echo.de/europa-ringt-um-die-zukunft-der-ki-regulierung-oder-innovationsbremse/ - https://www.vdi-nachrichten.com/karriere/studium/professor-chatbot/ - https://www.sueddeutsche.de/bayern/kuenstliche-intelligenz-hochschulen-bayern-chatgpt-herausforderungen-betrug-1.6288316 - https://news.vm.uni-freiburg.de/de/newsarchiv/wie-intelligent-ist-ki - https://www.uzh.ch/de/explore/basics/ai/recommendations.html