As someone knowledgeable in these areas, what is your take on this article?
Oops, I thought I had already posted this early on, but apparently not:
===========================================
Note to libertarians: cooperation, not just self-interest is hardwired.
Note to Nebullis: Cooperation is in our long term self-interest - which is why it's hardwired!
Even when unemotional software agents play the Prisoners' Dilemma (the recurring kind where you'll be encountering the same players in the future), the ones that adopt a variation of tit-for-tat always win out in the end. It doesn't surprise me that there was a selection pressure among humans for empathy & cooperation.
What's unhealthy, IMO, is the kind of altruism that says you have a moral obligation to help people, just because they say so. There's a subtle difference between making a long term investment in the well-being of those you value, and willingly playing host to a parasite.
Long-term cooperation requires a certain level of brainpower, since you have to be able to remember who you've interacted with in the past and who's done whom wrong, & who owes whom a favor. And putting yourself in the other person's shoes (empathy) also requires you to be able to step back & construct a model with yourself & the other person, & relate that model to the real world where you are in it. Which helps explain why only humans use cooperation to such an extreme as we do.
===========================================
Note to Nebullis: Cooperation is in our long term self-interest - which is why it's hardwired!
Cooperation is in the interest of the group and may or may not result in long-term reciprocation for the self. In the Prisoner’s Dilemma, selfish action always results in an immediate and greater return to the individual. Cooperation results in a reduced return which is not immediate or even certain; it depends entirely on the action of others—it’s no wonder that society has built in a moral obligation for this reciprocation.
Even when unemotional software agents play the Prisoners' Dilemma (the recurring kind where you'll be encountering the same players in the future), the ones that adopt a variation of tit-for-tat always win out in the end.
In this study, when humans played against the unemotional software agents, humans chose non-cooperative behavior.
It doesn't surprise me that there was a selection pressure among humans for empathy & cooperation.
I’m sure there is a good reason for empathy and cooperation, but it is not as immediately obvious to me, especially in the context of kin selection and the distructive hostility to non-kin. This seems to me non-adaptive behavior for the species as a whole.
What's unhealthy, IMO, is the kind of altruism that says you have a moral obligation to help people, just because they say so.
It makes perfect sense to me, as the long-term benefit of cooperation is dependent on the reciprocal behavior of others. I’m inclined to think it’s codified natural behavior.
There's a subtle difference between making a long term investment in the well-being of those you value, and willingly playing host to a parasite.
It’s not given to the players in the Prisoner’s Dilemma whether the other player is a parasite or not. That is, cooperation occurs even with the expectation that the other player might be a parasite. ( I’ll submit that playing may trick the normal evolutionary strategy of players who possibly assume that any other player is a member of the in-group, a group to value.)
In real life, many rounds are played, and the eventual pay-off for the individual is in the long-term benefit of group cohesion and tit-for-tat behavior of others. A common example is a group that lunches together and shares the cost of the meal. The person who habitually underpays or overeats is eventually abandoned or exiled from the group (unless he's so much fun he's allowed to freeload!). However, the apparently obvious long-term disincentive for “cheating” the friends is actually dependent on repeat behavior and group size, as well as other factors. In a large group, the cost of one freeloader to the other individuals in the group is minimal and easily ignored. A one-time freeloader can similarly be unaffected by tit-for-tat removal from the group. In fact, in game theory, the winning strategy in this situation is for the freeloader to pay as the others do until the last lunch at the most expensive restaurant in town.
In voluntary non-contractual situations, the freeloader is simply acting in his own self-interest. There is no force or fraud involved. The other members of the group accept the risk of a non-payer in the decision to share. In fact, the expectation of the other individuals in the group is a shared burden of uneven payment or eating on the part of some.
If nobody cooperates, nobody wins. If only one or a few don’t cooperate, they win big. The incentive for the one-time freeloader is enormous. It is a winning strategy for the individual at a cost to the group. If everybody cooperates, it comes at a determined cost to the individual, a benefit to the group, and an indirect, uncertain, long-term gain to the individual. The shared well-being of the group comes at a cost to the individual and is dependent on the behavior of others in the group as well as group size, number of interactions with the group and other parameters.
So, cooperation is not so obviously in the interest of the individual (long-term or short-term) as you make it sound.
Long-term cooperation requires a certain level of brainpower, since you have to be able to remember who you've interacted with in the past and who's done whom wrong, & who owes whom a favor.
Coooperation does require a certain level of brainpower, but it’s not, necessarily, to keep tally in the way you describe. The overall best interest of the group is the reward for cooperation. This may or may not result in a direct reciprocation to the individual. As the behavior of players in the Prisoner’s Dilemma points out, cooperation is altruistic in the sense that the reciprocation is not certain even though the cost is. It is really not a value transaction for the individual.