"machiavellian" intelligence as the basis for evolution of cooperative dispositions?
some "fun" friday night reading while waiting for stephen to arrive... actually, a prof suggested it as helpful reading for my research, but it's actually quite interesting if you take the time and energy to read it.
anyway, some crazy smart dudes set up what i think is an extremely clever simulation to model this and found:
In an initial cooperation-unfriendly world, individuals can occasionally be led to enter cooperative games that are played to costly mutual defect outcomes, thus selecting for high mindreading capacities. Once in place, those capacities permit upward drift on cooperative dispositions. Individuals with such dispositions are the frequent targets of exploitative attempts by those who are more defection-inclined, but their mindreading allows them to survive (via alternative choices) no differently from others in the population. At some point, however, two such individuals with higher probabilities of cooperating are likely to play a cooperative game with each other and to do so cooperatively— and, given that the alternative payoff is less than the mutually cooperative payoff, they will prosper by comparison with the rest of the population. Quite rapidly, the descendants of these individuals populate the entire ecology. The equilibrium level of cooperative dispositions after such a transition is high enough to attract skilled mind-readers into cooperative games but low enough to extract some gains from exploiting them once they have entered. Transitions are sustained by selection in favor of individuals who “discover” an optimal mixture of mindreading and mistrust—sufficient to ensure that they accept the true “I will always cooperate” messages characteristic of this environment but low enough to ensure that they reject the false messages that are, nevertheless, still being sent.
heh, have i officially crossed the line into academic nerdom?
