Human-AI collaboration: evidence from laboratory experiments

Information Systems and Operations management

Speaker : Wolf KETTER, from University of Cologne

Room Bernard Ramanantsoa

Abstract

"In an era where artificial intelligence (AI) is increasingly integrated into decision-making processes across diverse domains such as recruitment, finance, and healthcare, understanding how humans collaborate with AI systems is crucial. Despite the potential for synergy between human and AI capabilities, challenges persist in designing this collaboration effectively. We will give an overview of our previous and current research on human-AI collaboration, exploring insights from laboratory experiments. Specifically, we will focus on different forms of user collaboration with AI systems, namely advice-taking and delegation.
One significant obstacle for effective human-AI collaboration is the phenomenon of algorithm aversion, wherein human users exhibit reluctance to fully trust AI. By investigating the impact of enhanced transparency in AI systems on user trust and delegation behavior, our latest research sheds light on strategies to enhance users’ willingness to rely on AI systems and the resulting performance. As such our findings offer practical insights for improving human-AI teamwork in various contexts."

 Bio

Wolfgang Ketter is professor of Information Systems and Director of the Cologne Institute of Information Systems at the University of Cologne, where research he leads focusses on how digital transformation can create a faster and more stable transition to sustainable energy and mobility. He is also professor and director at Erasmus University and energy policy advisor to the German and the EU government. He has served as editor for ISR and MISQ and has won, among other prestigious prizes, the best ISR paper award for the year 2020.