On April 23, 2026, we were once again part of the Girls’ Day.

Girls’ Day is a nationwide event aimed at presenting girls with various study and career opportunities. Universities, companies, and businesses across Germany take part to introduce students to science, technology, and many other fields in which the proportion of women is less than 40 percent. This format also exists for boys as Boys’ Day, which focuses on career paths in which the proportion of men is lower.

FU Berlin participated with a range of workshops designed to give participants insight into different fields of study and professions. As the Human-Centered Computing Research Group, we welcomed students to our workshop, “Hey ChatGPT, how do I cook my favorite recipe? – A look behind the scenes of virtual assistants,” which builds on formats developed in previous years.

The goal of the workshop was to introduce participants to the fundamentals of chatbots and virtual assistants through a mix of discussion, hands-on activities, and reflection.

We began with a short brainstorming session, collecting associations with the terms “chatbot” and “virtual assistant.” Participants shared their thoughts and prior experiences. They were already familiar with systems such as ChatGPT, Siri, and Alexa, which they had primarily used to support everyday tasks, such as understanding school assignments or adjusting device settings.

Before diving into how chatbots actually work, we provided a brief introduction to virtual assistants and their development over time, highlighting key examples such as Radio Rex, ELIZA, Siri, Alexa, and ChatGPT.

The central part of our workshop consisted of an interactive group activity in which students simulated a recipe chatbot. Participants were divided into two groups: “users” and “chatbots.” The user group formulated potential questions one might ask a recipe chatbot, while the chatbot group developed possible answers. The questions and answers, written on cards, were then combined to simulate interactions between users and a chatbot. During this exercise, participants explored how their answers could be used in response to a variety of questions and discovered that some of them worked across multiple situations. They also noticed that simple replies, such as “yes” or “no,” were sometimes missing but would have been helpful. This highlighted the importance of formulating answers in a clear and flexible way.

After a short break, participants engaged in an interactive Kahoot activity in which they completed sentence prompts on their mobile devices. The aggregated responses were visualized in real time, with frequently occurring answers displayed more prominently. This activity offered a hands-on way to explore how chatbots learn from large amounts of data to predict likely responses.

In a final discussion, we reflected on potential applications of chatbots as well as associated challenges. In particular, we addressed the importance of protecting personal data and critically evaluating generated information, given that chatbots may produce incorrect or misleading outputs.

Overall, the workshop provided an engaging introduction to the principles behind virtual assistants and encouraged participants to critically reflect on their use in everyday contexts. We greatly enjoyed the exchange with the students and hope that the workshop sparked further interest in computer science and human-centered technologies.

Workshop by Malte Heiser with Laura Jürgensmeier and Yasemin Mutlugil