Students and educators are facing challenges regarding the use of artificial intelligence (AI) in educational settings. Heaten up by the hype of generative artificial intelligence tools and discussions about it, AI might change the way students and educators think and work. Using AI tools for almost anything could be as wrong as denying them altogether. They are useful for some tasks, as long as model and training data bias, ethical concerns and sustainability issues are addressed. We must also be careful not to forget what we know. My native language is German. Generative AI could write this abstract for me in an English voice of my choice. However, I decided to write it myself in English to practice my language skills. After finalizing the abstract, I've used DeepL to give it a polish to make it more readable. Afterwards, I’ve read the abstract again carefully to ensure that the meaning of the words have not changed and to learn new formulations. This is similar to the process I'd like to teach my students using the approach described below. In the end, my observations and student feedback showed me that it might work.

While generative AI can provide opportunities, the output of AI tools should be critically reflected upon before it is used. To enable students to reflect on the output of AI tools, they need to be able to do for themselves what the AI tools do (e.g., write an inspiring text). But encouraging students to do things on their own that AI could do for them is sometimes a difficult task for educators. Why take the long and hard way when there is an excellent and convenient shortcut?

To achieve this goal, I've developed a course on the User Centered Design (UCD) method, which I have to teach to 6th semester bachelor students of the Business Informatics program, using the Learning Experience Design (LXD) framework. The theoretical inputs as well as the student interactions within the UCD course are designed as experiences. Every aspect of UCD has to be learned, created and developed by the students themselves. And every aspect is done with the help of AI tools again afterwards. Sometimes the AI is helpful, sometimes it turns out that the AI-generated content has weaknesses compared to the students' own work.

Theoretical inputs were presented by me as an educator as an introduction (e.g. an introduction to the theoretical background of the UCD analysis phase) and appropriate UCD tools (e.g. personas). By balancing learning outcomes, learning objectives, activities, and emotions in the LXD canvas, students were able to have powerful experiences during their own work to make their learning outcome more relevant, meaningful, and valuable. In short, to make it more memorable.

After the theoretical introduction, students are encouraged to try out the UCD tools themselves (e.g. creating personas for different target groups). In the next step they must use AI tools to do the same steps again. Depending on the UCD tool and the AI tool used, sometimes better output or more possibilities were given with the help of AI. It also happened that the output generated by the students themselves was better. Most importantly, only by doing each step by themselves they are able to reflect on their own results and compare them to those generated by the AI. Students gained this insight throughout the course. It can be assumed that they understand the importance of doing something themselves, even if some parts can be assisted by AI later on.

Keywords: Education, course design, awareness, generative artificial intelligence, critical thinking, user centered design learning experience design.