Above is a link to a new release highlighting Constantinos Daskalakis and his important work. Together with Paul Goldberg and Christos Papadimitriou, Constantinos received the Game Theory and Computer Science Prize for his paper “The Complexity of Computing a Nash Equilibrium.” The prize is awarded once every four years at the World Congress of the Game Theory Society. A shorter and simpler version of the paper is available here.
The question of what computer science can teach social science disciplines such as economics is a constant discussion here at Michigan CSCS. I would say the consensus is that there exists real opportunities for meaningful cross-fertilization. For example, take the results presented in this paper. Now, consider a group of agents/players with ranges of different cognitive limitations and information sets playing all types of different games on a multi-dimensional graphs …. now imagine we did not assume the system in question had a fixed point attractor. Rather, assume we allowed for the possibility of a limit cycle attractor or even a strange attractor. Under such conditions, the richness of the problem might encourage one to put aside their pencil and paper and consider what computation might be able to bring to the problem ….
One thought on “What Computer Science Can Teach Economics …”