Project Grant 2505006

This Project Grant award from the National Science Foundation (CFDA 47.070 - Computer and Information Science and Engineering) provides $900,000 to the University of California, Irvine to develop new methods for measuring and optimizing human-AI decision-making partnerships. The project focuses on three core research activities: (1) creating Bayesian inference frameworks to dynamically evaluate human and AI agent abilities; (2) developing adaptive optimization algorithms to manage decision policies under uncertainty for real-world human-AI collaboration; and (3) integrating subjective metrics like fairness and agency into optimization frameworks. The research aims to ensure AI systems respect and reinforce human values while effectively collaborating with people across tasks like image classification, language processing, and navigation. Behavioral studies will validate the theoretical models, and the project will create openly accessible datasets to support further research in this area. This award reflects the National Science Foundation's efforts to advance scientific knowledge and technological innovation in computing and information science.