I thought a bit about credit allocation (mostly from an effective altruism point of view).
I think you're missing the next important step which is that people who would counterfactually do Bezos' work presumably also have good next-best options. So Bezos1 taking the Amazon niche means Bezos2 is free to do their next great project. So Bezos1 + Bezos2 combined created utility equal to Bezos' apparent utility minus Bezos2's counterfactual utility (if Bezos2 were to create Amazon instead) plus Bezos2's apparent utility minus Bezos3's counterfactual contribution plus Bezos3's....
In the EA space I have a moderately strong intuition that a lot of this adds up to normality so people should mostly just do what their naive highest impact is. I'm less sure how it applies to the (substantially more competitive and less coordinated) moneymaking world.
I thought a bit about credit allocation (mostly from an effective altruism point of view).
I think you're missing the next important step which is that people who would counterfactually do Bezos' work presumably also have good next-best options. So Bezos1 taking the Amazon niche means Bezos2 is free to do their next great project. So Bezos1 + Bezos2 combined created utility equal to Bezos' apparent utility minus Bezos2's counterfactual utility (if Bezos2 were to create Amazon instead) plus Bezos2's apparent utility minus Bezos3's counterfactual contribution plus Bezos3's....
In the EA space I have a moderately strong intuition that a lot of this adds up to normality so people should mostly just do what their naive highest impact is. I'm less sure how it applies to the (substantially more competitive and less coordinated) moneymaking world.