Co-operate as default if you don’t have a good read on whether the other person is co-operative is included in my suggestion, is that not enough for it to feel non-aggressive to you? Tit-for-tat, with co-operate first turn. If they come in at you sufficiently rudely without giving a chance for a reasonable request, you can take that as them playing defect preemptively.
And it’s not specifically expected reciprocation I’d look for, but whether the person would be as helpful to a person in general if they were on the other side of the situation.
In general the problem with Tit-for-tat is that it requires repeated interactions with the same person. In the neighbor example, that’s not necessarily a problem.
Of course that raises the question of what previous interactions with the neighbor were like something the OP hasn’t specified, e.g., does he have a history of making unreasonable demands of the other members of the apartment? Has he been otherwise a nice guy?
What if your estimate of other person’s defection is 90%? 95%? 99%? What if he explicitly stated that he won’t cooperate? What if you are Omega, and can make the prediction with 100% accuracy?
“Tit-for-tat is a better strategy than Cooperate-Bot.”
Can you use this premise in an explicit argument that expected reciprocation should be a factor in your decision to be nice toward others. How big a factor, relative to others (e.g. what maximises utility)? If there’s an easy link to such an argument, all the better!
If the problem you are trying to solve is how to motivate morality at the societal level, to a random bunch of people with varying preferences, then expected reciprocation is very important.
Under other assumptions, it isnt: for instance, forms of egoism, where you never risk any possible loss, and forms of altruism where only acts performed without expectation of reciprication are truly good.
Did you edit your original comment? When i first read it, I thought it was saying the opposite of what it now seems to say… I actually agree with it now—should is not universal, it depends on your goals.
P.S That paper you provide actually argues hedonism, not utlilitarianism :).
The simple explanation is that it it’s what should means. Beyond that, it would be helpful to have a more specific question..,whether you are questioning the rationality of morality,.or the morality of altruism ,. or whatever,
Expected reciprocation shouldn’t be a factor in your decision to be nice toward others.
Why not? Tit-for-tat is a better strategy than Cooperate-Bot.
Tit-for-tat makes sense. Preemptive tit-for-tat is unnecessarily aggressive.
Co-operate as default if you don’t have a good read on whether the other person is co-operative is included in my suggestion, is that not enough for it to feel non-aggressive to you? Tit-for-tat, with co-operate first turn. If they come in at you sufficiently rudely without giving a chance for a reasonable request, you can take that as them playing defect preemptively.
And it’s not specifically expected reciprocation I’d look for, but whether the person would be as helpful to a person in general if they were on the other side of the situation.
In general the problem with Tit-for-tat is that it requires repeated interactions with the same person. In the neighbor example, that’s not necessarily a problem.
Of course that raises the question of what previous interactions with the neighbor were like something the OP hasn’t specified, e.g., does he have a history of making unreasonable demands of the other members of the apartment? Has he been otherwise a nice guy?
What if your estimate of other person’s defection is 90%? 95%? 99%? What if he explicitly stated that he won’t cooperate? What if you are Omega, and can make the prediction with 100% accuracy?
“Tit-for-tat is a better strategy than Cooperate-Bot.”
Can you use this premise in an explicit argument that expected reciprocation should be a factor in your decision to be nice toward others. How big a factor, relative to others (e.g. what maximises utility)? If there’s an easy link to such an argument, all the better!
If the problem you are trying to solve is how to motivate morality at the societal level, to a random bunch of people with varying preferences, then expected reciprocation is very important.
Under other assumptions, it isnt: for instance, forms of egoism, where you never risk any possible loss, and forms of altruism where only acts performed without expectation of reciprication are truly good.
Can you give an explicit argument for why you”should” maximize utility for everyone, instead of just for yourself?
Some people offer arguments—eg http://philpapers.org/archive/SINTEA-3.pdf - and for some people it’s a basic belief or value not based on argument.
Did you edit your original comment? When i first read it, I thought it was saying the opposite of what it now seems to say… I actually agree with it now—should is not universal, it depends on your goals.
P.S That paper you provide actually argues hedonism, not utlilitarianism :).
Not that I recall
The simple explanation is that it it’s what should means. Beyond that, it would be helpful to have a more specific question..,whether you are questioning the rationality of morality,.or the morality of altruism ,. or whatever,