Students in the course “Coding Evolution of Cooperation” (Philosophy 51 C) have been developing a simulator for Iterated Prisoner’s Dilemma games.
If the rewards for “cheating” or “cooperating” in a Prisoner’s Dilemma situation have a certain structure, then agents have an incentive to cooperate over the long term, even though it is always rational to cheat in a single interaction. This structure of rewards is extremely common in the natural world, from the interaction of groupers and wrasse fish to mutual cooperation in human societies. Showing that mutual cooperation is indeed the best policy requires examining a lot of strategies for dealing with other agents over time, over many generations. We can do this using computer simulations.
We will talk about what we did in this class and what we think we learned. There will be a light lunch.
Where: Heger Hall 101