Altruism Among Simple and Smart Vampires.

In this work we present a simulation study drawn upon a famous ethological example: food-sharing habit of vampire bats in Central America. A norm of reciprocity is introduced when agents are expected to exhibit a cooperative behavior under serious environmental strictures (infrequent but lethal food scarcity). We compare different agents’ architectures (with different levels of cognitive complexity) in the evolution of food-sharing habit and we analyse their performances against the presence of cheaters in the population. The experiments are aimed at studying the role of cognitive mediators (i.e. goals) in the spreading of altruism, described as a purposeful action.