Team members: Bahador Bahrami, Jimmy Esmaily, Mustafa Yavuz, Jurgis Karpus, Dardo Ferreiro (associate), Zahra Rezahadeh

Funding from:

Improving collective decisions by eliminating overconfidence: mental, neural and social processes 

Overconfidence is a frequent bias in human decisions: we overestimate how right we are, or how likely we are to succeed. Can overconfidence be reduced in interactive decision-making processes involving both human interactions and interactions with AI systems? This project delves into the cognitive and neurobiological basis of these decisions, with a particular emphasis on understanding metacognition and post-decisional attributions of responsibility or regret. The research integrates various methodologies, including behavioral psychological testing, brain imaging, and psychopharmacology techniques.

The project seeks to identify triggers of overconfidence and explore interventions to modulate metacognitive processes and reduce overconfidence, thereby improving the accuracy of decision-making.

The expected outcomes of this research have broad implications. Insights gained from the study of interactive decision-making may inform various fields, including decision science, where improved understanding can lead to more informed choices in economics, politics, and business. Moreover, the findings may contribute to the design of AI systems that interact with humans by accounting for and mitigating overconfidence in their decision-making abilities. The investigation may also offer new perspectives on the role of metacognition in mental health disorders and education.