Virtual teams, in which humans and AI systems work together virtually as human-AI teams, are becoming increasingly important in the modern workplace. AI is no longer just a tool, but takes on tasks in communication, coordination, and cooperation as an active team member, requiring a deep understanding of the interaction between humans and AI to enable trustworthy and effective virtual human-AI collaboration. The design of human-AI collaboration is complex at both the individual and team levels and requires knowledge from disciplines such as business & economics, psychology, computer science, and information systems. It includes a variety of skills ranging software development to quantitative experimental research. This makes a well-thought-out prototypical AI system, research design and empirical evaluation strategy indispensable.

The Research Course builds on the fundamentals of existing lectures on the design of experimental studies and interactive systems with human-AI interaction, offering students the opportunity to put their own ideas into practice. Working in a team of 3-4 students, students have the possibility to experience and actively engage in hands-on research and independently – supervised by a research assistant – design AI agents for virtual team work. They integrate these agents into an experimental environment that provides basic virtual meeting functionalities and experimental control. Students then design their own experiment to investigate the potential and limitations of human-AI collaboration. The entire research cycle is covered, from formulating the research question and hypotheses to design, prototype implementation, and execution, to data analysis and the write-up of a working paper.

This approach not only teaches students how to apply the theoretical principles from existing lectures in a practical manner, but also how to master a rigorous scientific approach. They are encouraged to submit their results to research conferences, thereby strengthening both their scientific competence and their scientific communication skills – skills that are of central importance for both academic careers and professional practice in the design of human-AI collaboration.

Participation in the lecture “Designing Interactive Systems: Human-AI Interaction” or similar courses is recommended but not required. Familiarization with the lecture content is recommended if the course has not been attended.

Learning Objectives:  

• learn to design a AI agent prototype sing state-of-the-art AI technologies and to implement it in practice in existing experimental environments in a human-AI collaboration context
• learn to run a research project independently (incl. experimental design, conducting the experiment, data analysis, communication of  results).
• learn the standards and best practices of scientific work in human subject research and apply them in practice, in particular open science, data protection and ethical requirements.
• acquire in-depth knowledge of the design of experimental studies and learn to apply this in practice. This also includes learning how to work in an experimental laboratory such as the KD2Lab. This also includes administrative (e.g., participant management) and operational (e.g., setting up a server  infrastructure) activities.
• learn to work together as a research team and to distribute tasks within the team in a structured manner.
• engage actively in scientific discourse by discussing their own work in the team as well as the work of others , to prepare scientific findings, and to reflect on scientific  work.

This course is funded by the Federal Ministry of Research, Technology and Space (BMFTR) and the Baden-Württemberg Ministry of Science as part of the Excellence Strategy of the German Federal and State Governments from the ”Research Infrastructures in Research-Oriented Teaching (RIRO)“ initiative at KIT.