Innovation Policy Colloquium: Feng Fu

The Innovation Policy Colloquium focuses each year on different aspects of the law’s role in promoting creativity, invention, and new technology. This year, we will discuss the the implications of complexity for law and policy related to innovation, privacy and AI. Complexity science is a cutting-edge multi-disciplinary field that studies a wide variety of systems comprised of numerous interacting components. The human social network, the internet, social media applications, cities, biological systems and financial networks are all examples of complex systems. Complexity can lead to non-linear and surprising responses to policy initiatives, such as tipping points and feedback effects. Policymaking that is insensitive to these possibilities can go drastically awry.

Feng Fu, Associate Professor, Dartmouth College
Resilience of Cooperation in Perturbed Social Networks: Hybrid Human–AI Systems and the Optimal Steering of Collective Behavior (paper available upon request if you plan to attend the talk)

Abstract: Hybrid social and socio-technical networks increasingly face exogenous disruptions, including user dropouts, account deactivations, and interventions by automated agents. Yet how such heterogeneity shapes the resilience of cooperation remains poorly understood. Using evolutionary graph theory, we model perturbations as random site dilution and derive closed-form conditions (under weak selection) for the evolutionary stability and long-run resilience of cooperation. We extend this framework to include empty sites with strategy-dependent effects, capturing the role of LLM-based bots in human–AI systems. Our analysis reveals a Goldilocks effect in the prevalence of embedded intelligent agents: too few bots fail to shield cooperators from defectors, while too many disrupt cooperative clustering and block pathways for behavioral conversion. An intermediate level of AI presence maximizes cooperation. Our work offers insights into the optimal alignment of human populations for the greater social good using intelligent agents.

Email Nicole Arzt if you would like to attend the colloquium. If you are outside of NYU and do not have an NYU Id then I will need to add your JRNY for building access.