Tiffany Matej Hrkalovic
PhD Student at Vrije University Amsterdam & Delft University of Technology
The enigma of human willingness and ability to cooperate has been a topic of interest for millennia. However, due to the recent technological developments in designing intelligent systems and their potential usage in cooperative settings with humans, newer research is steered towards understanding the underlying factors of human-AI (HAI) interactions. Thus, addressing the question of why (or why not) do people cooperate with intelligent systems and which aspects of the setting impact their cooperative decision is timely and important. Here, we orient on investigating the role of incentives on cooperative behavior. Addressing the problem of adjusting incentive adjustment between HHI and HAI is important for a couple of reasons. Firstly, recent research has shown that human players do not think that systems need money, which has been used in certain studies. Similarly, humans have a bigger financial risk (i.e. skin in the game) in sharing money compared to an artificial agent. For instance, when we ask a participant how much they would give to the other person/agent, we cannot ignore the different risks they are taking when making the cooperative decision. In contrast, intelligent systems will never have this financial risk. This asymmetry in personal investment between intelligent systems and humans can also result in decreased human cooperativeness towards intelligent systems. Thus, this project will focus on addressing the question of the role of the type of incentives (e.g., money) in cooperative tasks that have been used so far in cooperative HAI and address some implications (theoretical and experimental) for its improvements.
Keywords: Human-ai interaction; incentives; collaboration
Scientific area: Human-AI Interaction
Bio: I am a Postdoctoral research at Jheronimus Academy of Data Science. Currently, I am finishing my PhD at the Vrije University of Amsterdam and Delft University of Technology which was focused on understanding how people select partners for cooperative and collaborative task with the future goal of investigating if we can create reflective hybrid intelligence systems that help people make better partner choices, but also to reflect on their behavior in social interactions. My research interests include social signal processing, social decision making, person perception, social information processing and human-AI collaboration. I obtained my Master’s Degree of Arts in Psychology, Master’s Degree of Science in Cognitive Science, and Bachelor’s Degree in Psychology.
Visiting period: July-August 2024 at Research Center Trustworthy Data Science and Security, Lab: Human Understanding of Algorithms and Machines