I actually think that utility monsters (of a limited form) do exist: they’re called children.
People will often go out of their way to satisfy the preferences of children, even at some inconvenience to themselves. Fortunately, childrens’ preferences are often pretty easy to satisfy, but in general it seems that we weight them higher than we do those of most adults, suggesting that they are at least minor utility monsters.
My understanding of ‘Utility Monster’ is someone who gets increasing utility per resource unit when greater resources are spent (greater than linear return). For example it would get utility of 1x when getting one cupcake, and utility of >2x when getting 2 cupcakes.
If such a monster existed, you would increase the average utility by giving ALL of everyone’s resources to it.
A child may get more enjoyment from the same amount of resource than an adult in some situations, but you don’t raise the average utility by giving the entire cupcake to just one of your 2 children.
I thought the point of a utility monster was simply that they were more efficient at turning resources into utility—I suspect both notions serve to make similar points! I agree that children aren’t utility monsters in the sense that you describe, but I think they are at least agents that turn resources into utility more efficiently than adults. I think the claim that such beings could exist is a conclusion that some have claimed is a dodgy consequence to be avoided—which obviously isn’t the case if such things already exist.
I actually think that utility monsters (of a limited form) do exist: they’re called children.
I agree, kind of. Reading this article changed my intuition about “utility monsters” by describing their opposites, the “dysutility monsters”. Then I realized that if we convert money to utilons, the difference is not only in a pleasure derived from the results (which could be maximized by a psychopath), but also in ability to convert the money into meaningful results. Also in real life the “utility monster” does not need the ability to convert money to utilons infinitely, only in the range we consider now.
Therefore, “utility monster” could simply be someone who is able to translate the resources into growth more efficiently than others. Such as a child. The child can efficiently use only a limited amount of resources, and then the “utility monster” effect turns off, but within this range, it makes utilitarian sense to make some sacrifices.
Another example could be a poor family with many children, one of them exceptionally gifted, others average. It would make sense to allocate more resources to the gifted child, especially under assumption that after the child successfully converts resources into status and income, they will support their siblings, so this arrangement is even in the other siblings’ net benefit. Assuming that exceptional people are a net benefit for their society, it could be useful to give more resources to education of gifted people. (Though in real life it would be difficult to check that the resources are really efficiently used for education, and not e.g. burned in a signalling contest.)
But this is all based on the assumption that for a human, the “utility monster” effect is only temporary, and that the boundaries are reasonable. And also it assumes our ability to measure other people’s utility. In real life, people sometimes spend of lot of money for their children, but children use the money for immediate pleasure instead of education—so they may seem like “utility monsters” to their parents, even when they are not.
I actually think that utility monsters (of a limited form) do exist: they’re called children.
People will often go out of their way to satisfy the preferences of children, even at some inconvenience to themselves. Fortunately, childrens’ preferences are often pretty easy to satisfy, but in general it seems that we weight them higher than we do those of most adults, suggesting that they are at least minor utility monsters.
My understanding of ‘Utility Monster’ is someone who gets increasing utility per resource unit when greater resources are spent (greater than linear return). For example it would get utility of 1x when getting one cupcake, and utility of >2x when getting 2 cupcakes.
If such a monster existed, you would increase the average utility by giving ALL of everyone’s resources to it.
A child may get more enjoyment from the same amount of resource than an adult in some situations, but you don’t raise the average utility by giving the entire cupcake to just one of your 2 children.
I thought the point of a utility monster was simply that they were more efficient at turning resources into utility—I suspect both notions serve to make similar points! I agree that children aren’t utility monsters in the sense that you describe, but I think they are at least agents that turn resources into utility more efficiently than adults. I think the claim that such beings could exist is a conclusion that some have claimed is a dodgy consequence to be avoided—which obviously isn’t the case if such things already exist.
I agree, kind of. Reading this article changed my intuition about “utility monsters” by describing their opposites, the “dysutility monsters”. Then I realized that if we convert money to utilons, the difference is not only in a pleasure derived from the results (which could be maximized by a psychopath), but also in ability to convert the money into meaningful results. Also in real life the “utility monster” does not need the ability to convert money to utilons infinitely, only in the range we consider now.
Therefore, “utility monster” could simply be someone who is able to translate the resources into growth more efficiently than others. Such as a child. The child can efficiently use only a limited amount of resources, and then the “utility monster” effect turns off, but within this range, it makes utilitarian sense to make some sacrifices.
Another example could be a poor family with many children, one of them exceptionally gifted, others average. It would make sense to allocate more resources to the gifted child, especially under assumption that after the child successfully converts resources into status and income, they will support their siblings, so this arrangement is even in the other siblings’ net benefit. Assuming that exceptional people are a net benefit for their society, it could be useful to give more resources to education of gifted people. (Though in real life it would be difficult to check that the resources are really efficiently used for education, and not e.g. burned in a signalling contest.)
But this is all based on the assumption that for a human, the “utility monster” effect is only temporary, and that the boundaries are reasonable. And also it assumes our ability to measure other people’s utility. In real life, people sometimes spend of lot of money for their children, but children use the money for immediate pleasure instead of education—so they may seem like “utility monsters” to their parents, even when they are not.