Robots and artificial intelligence (AI) are increasingly serving as collaborative partners for humans. Text-based chatbots, which enable humans to communicate with a technical system through natural language, have gained popularity in a variety of application areas. So far, little is known about the processes involved in teams consisting of humans and AI. Therefore, the focus of this dissertation project is to explore relationships between team composition and team performance and dynamics.

Advances in machine learning have brought about a rapid proliferation of computer-based dialog systems. Previous research has shown that people interact differently with automated agents than with their fellow humans. Moreover, the willingness to interact with automated agents is influenced by the nature of the task, such that people in analytically perceived contexts show a higher willingness to collaborate with automated agents than in moral contexts.

Human-AI-teaming is a relatively new area of research in the field of human-AI collaboration. Little is known about team processes in human-AI teams, especially when collaborating on different tasks, and research on characteristics of human collaboration with chatbots is still scarce. This dissertation project therefore investigates potential relationships between team composition, task type, various user characteristics, user-related variables (e.g., trust in the AI-system, satisfaction with the collaboration) and group-level variables such as team performance and product quality. The focus of the planned studies will be on human-AI dyads and teams of humans and AI.