Paolo Penna (ETH Zurich, Department of Computer Science)

In this talk, I shall try to give an overview of the importance of incentives in Computer Science applications, where distributed systems often involve a large number of self-interested parties.

There will be two main parts:

  1. Equilibria and game dynamics
  2. Mechanism design

With the first approach, we essentially study how a given system works (whether it reaches an equilibrium, how fast this takes, and whether the state is a desired one). The second approach is aiming at designing a system (rules) which takes into account the above self-interested parties and attempts to achieve a certain goal (e.g., optimize the routing in a network with selfish routers).