The project investigates how users interact with AI systems based on large language models, such as ChatGPT, by using prompts to request information and explanations. It focuses on how understanding develops over time through interaction and how users gradually learn to interpret and calibrate the system’s responses in context. Users often begin with an unclear or only partly formed knowledge gap, which becomes visible and more precise through the ongoing exchange with the AI. The project aims to develop a context-sensitive framework for human–machine interaction that helps users work effectively with AI systems while remaining aware of their risks and limitations.

Research areas: Computer science, Linguistics

Project duration: from 01/2026