Well, because that's in near mode.
This is more or less the distinction I was going for.
If I do something for other people, there is always an obvious result: I feel better about myself.
Why isn't this equally true for doing things for oneself?
Because other people reward you socially for doing things for other people. If you do something good for person A, it makes sense for a person A to reward you -- they want to reinforce the behavior they benefit from. But it also makes sense for an unrelated person B to reward you, despite not benefiting from this specific action -- they want to reinforce the general algorithm that makes you help other people, because who knows, tomorrow they may benefit from the same algorithm.
The experimental prediction of this hypothesis is that the person B will be more...
If it's worth saying, but not worth its own post (even in Discussion), then it goes here.