To begin with, the students themselves took on the role of neurons in a neural network. In an "unplugged" game, they classified images and thus learnt directly how data is processed and decisions are made in AI systems.
In the next step, the students used block-based programming to programme a robot that could be controlled by classifying images. They trained their own image classification model and experienced directly how the robot could be controlled on the basis of the model they had created themselves.
Another focus was on large language models (LLMs). Through playful plugged and unplugged activities, the students imitated a language model and understood how LLMs use probabilities to select the next token and generate texts.
Finally, the groups explored bias in AI systems and discussed ethical and social issues surrounding the use of artificial intelligence. The workshop combined basic technical knowledge with critical reflection and offered the students a lively insight into current developments in Computer Science.