Also, the point of EA is not just numbers. Numbers are a tool to help make you the most ethical, rational judgement, to help the most. If you could provide a co

Also, the point of EA is not just numbers. Numbers are a tool to help make you the most ethical, rational judgement, to help the most. If you could provide a convincing rational, ethically coherent argument that (if we have a binary choice) we should choose 1 guide dog over curing 2000 people of blindness, we should accept it. When speaking of large evils, and large impacts, I don't see how you could build an argument without numbers to support any decision, that wouldn't be obtuse, incoherent, unfounded. The only problem with numbers is oversimplification: when you use some metric (it could even be lives saved) without taking into account more complex factors like quality of life. Of course, when you can cheaply cure a person of blindness, save a man's life or address extreme poverty (hunger, lack of education, sanitary and material conditions), that seems to overwhelm fine considerations or quality of life... addressing extreme poverty is the first step in providing a rich and good life (and we should be always mindful of such goals).