Researcher(s)
- Francesca Tero, Neuroscience, University of Delaware
Faculty Mentor(s)
- Jennifer Kubota, Psychological and Brain Sciences, University of Delaware
- Jasmin Cloutier, Psychological and Brain Sciences, University of Delaware
Abstract
Abstract:
Human decision-making is not purely rational, but can be influenced by extraneous factors, including social factors such as partner competency status and humanness. Previous research has investigated how financial status impacts impression formation, trust, and cooperation between humans, finding that humans positively evaluate, trust, and cooperate more with high-status humans. But it is unclear how competency status (either being competent at a task or not) influences impression formation, trust, and cooperation. Additionally, less is known about impression formation, trust, and cooperation when humans interact with artificial intelligence (A.I.) agents. The present study examines how competency (high or low) and agency (human or A.I.) influence spontaneous cooperation. Participants in this study complete an impression formation task and economic games (Trust Game, to measure trust, and an Escape Room, to measure cooperation) with partners varying in competency (high and low) and humanness (human or A.I.). In the escape room, participants complete tasks ostensibly with a team of human and A.I. partners with varying competency levels, whom they had to solicit clues from to succeed. They complete team cognitive assessments (i.e., Reading the Mind in the Eyes (RME) Test, Posner Attention Network Test, and Stroop Word and Color Test) and team puzzle tasks (i.e., word searches, crosswords, and spot-the-difference puzzles). We also collected electroencephalography (EEG), physiological data (e.g., heart rate), audio, and video during these interactions. Data collection is still in progress. By understanding factors that influence human A.I. cooperation, we can better understand ways in which cooperation between humans and A.I. could be better fostered.