You disagree that societies work best when the interests of all the participants are aligned? This seems obvious to me, so I am curious as to what you object to. If people are rewarded for doing thing X, X will be more likely to happen, and if they are punished for Y, then Y will be less likely to happen. Clearly there will be outliers, but why would you construct a society that actively penalises things that it wants?

Truly aligned incentives are almost impossible to achieve - in a lot of situations, "mostly aligned most of the time" is the best we can hope for.

For example, if my startup is renting servers from AWS our incentives are aligned in the sense that we'll buy more from AWS as we get more users; but unaligned in the sense that if we're using more resources than we need to AWS wouldn't profit from pointing that out. Indeed, even an externality-free positive-sum deal between two parties brings with it a zero-sum deal in deciding how to distribute the benefits.

How do you define what a society wants? There is not an implicit goal when a society exists, there is just a collective of individuals who happen to have some interaction. So essentially, this is the function of politicians in a democracy: to reconcile the irreconcilable views of individuals. However, as with all compromises, often no one is left happy...

I'm not explaining things well I think, hope you can make sense of that!