Dynamic Contracting with AI

This paper investigates a dynamic agency problem that includes the AI-style agent and principal. We develop a quantitative bionic approach to dynamic contracting based on calibrating the incentive properties of a workhorse contracts model in which self-awareness and memory are endowed. We further solve the bionic model quantitatively via specific algorithms powered by Artificial Intelligence. In different scenarios, the bionic framework can endogenize various stylized preferences inspired by psychology studies and neuroscience. Note that the model theoretically rationalizes the neuro-foundations of a strand of conventional contract models in microeconomics, shedding new light on widely-debated issues surrounding the dynamic contracting problems