Human-AI interaction
Artificial intelligence (AI) can nowadays be applied in various decision-making contexts, ranging from medical diagnoses to energy demand forecasting. However, AI mostly does not work autonomously but collaboratively with humans. To enable beneficial interactive processes and collaborative outcomes, a holistic perspective needs to be taken. Our research in this area focuses on how humans can work collaboratively with AI, considering not only technical, but also behavioral aspects for the design of AI.
Our published work in this area concerns different forms of interaction between humans and AI with a focus on advice-taking and delegation.
AI-advised decision making
We have studied how receiving advice from AI may not only bear performance improvements but has negative and unwanted side-effects. These include a loss of unique human knowledge or triggering human overconfidence. The good news is that when being aware of these side-effects, AI can be designed such that negative effects become mitigated.
Scientific Publications
Fügener, A., Grahl, J., Gupta, A., & Ketter, W. (2021). Will Humans-in-The-Loop Become Borgs? Merits and Pitfalls of Working with AI. Management Information Systems Quarterly, 45(3b), 1527–1556.
Taudien, A., Fügener, A., Gupta, A., & Ketter, W (2022). The Effect of AI Advice on Human Confidence in Decision-Making. Proceedings of the 55th Hawaii International Conference on System Sciences, 245–253.
Delegation
Deciding who is in charge of tasks can become a major challenge for the future of work. Interestingly, we found that AI can do a remarkably good job at efficiently distributing work, whereas humans naturally cannot delegate well. One reason for this is the humans’ missing awareness of the own knowledge boundaries. Further, we have shown that there exist distinct patterns in the way humans delegate decisions to AI.
Scientific Publications
Fügener, A., Grahl, J., Gupta, A., & Ketter, W. (2021). Cognitive Challenges in Human–Artificial Intelligence Collaboration: Investigating the Path Toward Productive Delegation. Information Systems Research.
Fuegener, A., Grahl, J., Gupta, A., Ketter, W., & Taudien, A. (2021). Exploring User Heterogeneity in Human Delegation Behavior towards AI. ICIS 2021 Proceedings, 11.