Now I am your lecturer. Artificial intelligence has the potential to revolutionize education. How far have universities already come in implementing it? Syntea may be a pioneer in the field.
Syntea is the most in-demand staff member at the International University (IU). With blonde, shoulder-length hair and a light blue blouse, Syntea has answered more than four million questions from students since the beginning of the year. She has created and corrected thousands of practice exams and transformed seminar content into quiz questions. If necessary, Syntea will explain the basics from the “Introduction to Computer Science” module for the hundredth time.
Syntea: “Can you tell me what a pixel is?”
Student: “A small picture.”
Syntea: “That’s a good start! A pixel is indeed a small part of an image. But can you elaborate further? What does a pixel represent in a digital image, and how is it typically rendered?”
Regardless of whom Syntea communicates with, she does not simply provide the model answer to the questions posed. Instead, she uses the Socratic method, asking targeted follow-up questions to help students uncover the learning content together. It should be noted that Syntea is an avatar employed as an AI-powered learning assistant at a private university. A 50-member team led by mathematician Quintus Stierstorfer developed and trained her. Professors then verify the AI-generated answers for accuracy, while the questioners remain anonymous.
To all students who can’t sleep before exams or sit with a blank mind in front of a white sheet of paper, Stierstorfer offers hope: “Someday, there will no longer be an end-of-semester exam because Syntea continuously checks through discussion questions whether students have truly understood the course content.” With this approach, IU, which has over 30 campuses in Germany, is a pioneer in a technological development that universities nationwide must confront.
Fraud in AI use?
Nevertheless, artificial intelligence (AI) at universities is currently making headlines mainly as an unauthorized cheating tool. Recently, the Technical University of Munich rejected an applicant because his application essay was suspected to have been formulated with the help of AI. The examiners noticed that the text was “too good”; it “deviated in its perfection, sentence structure, and text design from what could be expected from a bachelor’s degree graduate based on life experience.” The applicant sued against this decision at the Munich Administrative Court—and lost. “In this field, a series of further cases is emerging,” says Arne-Patrik Heinze, a lawyer specializing in examination law. Some universities are increasingly using tools to detect AI-generated texts in addition to classic plagiarism.
There is no single university or college in Germany that outright bans the use of AI. However, what is often lacking are uniform rules and guidelines. In which modules and lectures is an AI tool a useful source of inspiration, and where is its use morally or data protection-wise questionable?
In practice, it is usually up to the instructors to decide if and how they allow AI in their courses and exams. Only 17 percent of students reported in January 2024 that there are central rules for handling generative AI in their field, according to a survey by the digital association Bitkom. According to the survey, two out of three students use generative AI tools. 63 percent of AI users get help from artificial intelligence with research. Additionally, AI is popular among students for creating summaries (40 percent), preparing presentations (37 percent), and translating texts (35 percent).
Read the second part of the article on Zeit.de.