These were the first types of experiments done in algorithmically determining the value of what we refer to as the concept of reciprocal altruism, which is a concept in human sociology often used to describe the basis of most of our ethical and moral behaviors. It is also a concept often devalued for its simplicity. It just seems intuitively to be too simple a ways of describing how our behaviors work and play out in the real world. But our intuitions can fail us, as we all know too well.
The prisoners dillemma is fascinating in it's simplicity, and also in it's ability to confound what one would assume is a winning strategy in a game.
In the prisoners dilemma you are to imagine a scenario where you are a prisoner with a cellmate, and you are left with the option in a given turn of the game to either give evidence against your fellow inmate or not.
The scoring is really simple. If you give evidence and your cellmate does not, you get 10 points, and your cellmate gets 1 point. If you give evidence and your cellmate does as well, you each get 5 points for being mutually not nice. If you decide not to give evidence and so does your cellmate, you each get only 3 points for mutually deciding to be nice.
On the surface of it this would seem to be easy to figure out. Intuitively most people can see the point system and realize that to win is to not be nice. And indeed in multiple trials with one individual this is the winning strategy.
But our lives never involve merely interacting with one other person. Our lives are a construct of continual interactions with astonishing numbers of people throughout our lives.
And ironically in this prisoners dillemna when you take the iterations of it and expand it out into more and more interactions with more and more prisoners, the odds flip...and rather drastically. So drastically that over those many iterations the strategy that is a winning one on a purely one to one basis becomes the one guaranteed to lose you points over the course of many games.
The strategy that wins? And more importantly always wins. And even more importantly always wins in every iteration of game theory that has been imagined in computational mechanics over the last 40+ years? Tough but fair. In other words, simply deciding to be nice from the get go, until you encounter those not nice back to you, and being not nice back until THEIR BEHAVIOR CHANGES so that you can go back to being nice again.
In other words a preference towards being genuinely good and giving towards others, whilst being mindful that this tendency can be taken advantage of (and the strength of self to react accordingly), always wins.
ALWAYS
It's a lesson we all need to internalize and cherish for the truth that it is.
The strategy that wins? And more importantly always wins. And even more importantly always wins in every iteration of game theory that has been imagined in computational mechanics over the last 40+ years? Tough but fair. In other words, simply deciding to be nice from the get go, until you encounter those not nice back to you, and being not nice back until THEIR BEHAVIOR CHANGES so that you can go back to being nice again.
In other words a preference towards being genuinely good and giving towards others, whilst being mindful that this tendency can be taken advantage of (and the strength of self to react accordingly), always wins.
ALWAYS
It's a lesson we all need to internalize and cherish for the truth that it is.
No comments:
Post a Comment