People tend to act in their own perceived individual self-interest.

I consider that there are two types of self-interest we can talk about. Interest of the self (IOS), and interest for the self (IFS). The first one is axiomatically defined as anything a person does. An act I do, I do because the result of it will be of psychological benefit to me, in some way.

The later type is defined as anything a person does so that he gets a tangible benefit for himself. Wether that is more money, power etc.

So while all IFS persons act for their IOS, All IOS is not IFS (ie a doctor working for the poor).

Greed is an interest for the self which is unequenchable. It cannot be sustained in Communism because accumulation of power is impossible and accumulation of wealth will be problematic to say the least.

You are not talking about greed in your later arguments but about IOS. But other than that, I agree.

I'll reply more later

I'm not sure there's a distinction between "tangible" and "intangible" benefits except to in that the fulfillment of a desire might have more or less connection to the physical world: some desires can be fulfilled just by thinking, others just by speaking and listening, others, such as the desire to quickly move between Oakland and San Francisco, might require massive effort by thousands of people to fulfill.

I think a better way to distinguish desires is by their social implication: some desires are mutually fulfillable: satisfying my desire also satisfies another's desire; other desires are exploitative: satisfying my desire entails another person's desires will go unsatisfied.

It's possible that given enough time, people will simply evolve not to have exploitative desires. But… for this to happen, there will have to be real social selection pressures.

The problem is not that people are "greedy" in the sense of acting in their own self-interest; the problem is that capitalism institutionalizes the value of acting in one's own immediate, material self-interest, and institutionalizes the negative value of acting with regard to mutual self-interest.

My argument is not that federated decision-making cannot determine their mutual self-interest. My argument is that in Prisoner's Dilemma situations, two or more parties must coercively bind themselves to acting in their mutual self-interest to escape the "rational" Nash Equilibrium of mutual defection. The problem is not primarily about information, but about motivation.

The only penalty that federated agents can impose on each other is their own defection, the "Tit-for-Tat" solution to the iterated Prisoner's Dilemma. But when the lag between one player's defection and the retaliation exceeds the agents' lifetime, then it's difficult to see how this solution could be operative.