Nice try, but even if my utility for oiled birds was as nonlinear as most people’s utility for money is, the fact that there are many more oiled birds than I’m considering saving means that what you need to compare is (say) U(54,700 oiled birds), U(54,699 oiled birds), and U(53,699 oiled birds) -- and it’d be a very weird utility function indeed if the difference between the first and the second is much larger than one-thousandth the difference between the second and the third. And even if U did have such kinks, the fact that you don’t know exactly how many oiled birds are there would smooth them away when computing EU(one fewer oiled bird) etc.
(IIRC EY said something similar in the sequences, using starving children rather than oiled birds as the example, but I can’t seem to find it right now.)
Unless you also care about who is saving the birds—but you aren’t considering saving them with your own hands, you’re considering giving money to save them, and money is fungible, so it’d be weird to care about who is giving the money.
Nice try, but even if my utility for oiled birds was as nonlinear as most people’s utility for money is, the fact that there are many more oiled birds than I’m considering saving means that what you need to compare is (say) U(54,700 oiled birds), U(54,699 oiled birds), and U(53,699 oiled birds)
Nonlinear in what?
Daniel’s utility for dollars is nonlinear in the total number of dollars that he has, not in the total number of dollars in the world. Likewise, his utility for birds is nonlinear in the total number of birds that he has saved, not in the total number of birds that exist in the world.
(Actually, I’d expect it to have two components, one of which is nonlinear in the number of birds he has saved and another of which is nonlinear in the total number of birds in the world. However, the second factor would be negligibly small in most situations.)
He has a utility function that is larger when more birds are saved. If this doesn’t count as caring about the birds, your definition of “cares about the birds” is very arbitrary.
Nice try, but even if my utility for oiled birds was as nonlinear as most people’s utility for money is, the fact that there are many more oiled birds than I’m considering saving means that what you need to compare is (say) U(54,700 oiled birds), U(54,699 oiled birds), and U(53,699 oiled birds) -- and it’d be a very weird utility function indeed if the difference between the first and the second is much larger than one-thousandth the difference between the second and the third. And even if U did have such kinks, the fact that you don’t know exactly how many oiled birds are there would smooth them away when computing EU(one fewer oiled bird) etc.
(IIRC EY said something similar in the sequences, using starving children rather than oiled birds as the example, but I can’t seem to find it right now.)
Unless you also care about who is saving the birds—but you aren’t considering saving them with your own hands, you’re considering giving money to save them, and money is fungible, so it’d be weird to care about who is giving the money.
Nonlinear in what?
Daniel’s utility for dollars is nonlinear in the total number of dollars that he has, not in the total number of dollars in the world. Likewise, his utility for birds is nonlinear in the total number of birds that he has saved, not in the total number of birds that exist in the world.
(Actually, I’d expect it to have two components, one of which is nonlinear in the number of birds he has saved and another of which is nonlinear in the total number of birds in the world. However, the second factor would be negligibly small in most situations.)
IOW he doesn’t actually care about the birds, he cares about himself.
He has a utility function that is larger when more birds are saved. If this doesn’t count as caring about the birds, your definition of “cares about the birds” is very arbitrary.
He has a utility function that is larger when he saves more birds; birds saved by other people don’t count.
If it has two components, they do count, just not by much.