The other major problems with trying to calculate your way to a neat and orderly table of “the best” interventions:
There is no single absolute objective perspective on value weights. Just listen to conversations at EAG or in the forum. There will never be full consensus on the value of a quadrillion crustaceans vs. billions of mammals, or on the value of future sentient beings vs. current ones.
Preferences matter. And in fact, a diversity of preferences is essential. If we only used the final ranking, in the extreme case we devote 100% of our resources to Program A. But not everyone is passionate about A, and the world probably can not handle the capacity for everyone to productively contribute to A. So we need people to have a preference for all projects, even and especially those further down the list so we have reasonable coverage of interventions across the entire problem space.
The other major problems with trying to calculate your way to a neat and orderly table of “the best” interventions:
There is no single absolute objective perspective on value weights. Just listen to conversations at EAG or in the forum. There will never be full consensus on the value of a quadrillion crustaceans vs. billions of mammals, or on the value of future sentient beings vs. current ones.
Preferences matter. And in fact, a diversity of preferences is essential. If we only used the final ranking, in the extreme case we devote 100% of our resources to Program A. But not everyone is passionate about A, and the world probably can not handle the capacity for everyone to productively contribute to A. So we need people to have a preference for all projects, even and especially those further down the list so we have reasonable coverage of interventions across the entire problem space.