My doctoral research addresses two fundamental obstacles to beneficial outcomes from strategic interactions between multiple parties: strategic incentives against cooperation (as in the Prisoner's Dilemma) and the multiplicity of strategic solutions (sometimes called the equilibrium selection problem). As AI systems are increasingly involved in consequential decision making processes on behalf of human principals, understanding how to achieve desirable outcomes in multi-agent AI settings becomes critical. My research leverages unique features of AI systems – including their transparency, reproducibility, and malleability – to develop novel game-theoretic approaches that enable better, more cooperative outcomes.

Three primary research directions form the core of this dissertation. First, the concept of "safe (Pareto) improvements" provides a rigorous framework for improving outcomes without resolving equilibrium selection problems. Unlike traditional solution concepts, safe Pareto improvements make qualitative assumptions about pairs of games rather than individual games. This sometimes allows us to prefer playing one game over another, without any judgment about how each of the individual games is played. Second, my research on so-called Newcomb-like decision problems takes inspiration from philosophical branches of decision theory concerning, for example, how one should reason when interacting with a copy of oneself. I investigate how cooperation can be achieved when different parties deploy similar AI systems. Third, the concept of program equilibrium explores how the use of mutually transparent decision-making algorithms can allow for cooperation.

501 pages

Thesis Committee:
Vincent Conitzer (Chair)
Fei Fang
Tuomas Sandholm
Ben Levinstein (University of Illinois / Anthropic)
Stuart Russell (University of California, Berkeley)

Jignesh Patel, Interim Head, Computer Science Department
Martial Hebert, Dean, School of Computer Science

Creative Commons: CC-BY (Attribution)

Return to: SCS Technical Report Collection
School of Computer Science

This page maintained by reports@cs.cmu.edu