Jiro comments on Humans are utility monsters - Less Wrong

67 Post author: PhilGoetz 16 August 2013 09:05PM

You are viewing a comment permalink. View the original post to see all comments and the full post content.

Comments (213)

You are viewing a single comment's thread. Show more comments above.

Comment author: PhilGoetz 16 August 2013 10:04:57PM *  6 points [-]

The qualification that a utility monster is not susceptible to diminishing marginal returns is made only because they're still assuming utility is measured in something like dollars, which has diminishing marginal returns, rather than units of utility, which do not. Removing that qualification doesn't banish the utility monster. The important point is that the utility monster's utility is much larger than anybody else's.

Comment author: Jiro 17 August 2013 01:23:48AM 3 points [-]

Removing that qualification does banish the utility monster. If the utility monster gets greater utility from dollars than someone else (let's say nematodes), but is still subject to diminishing marginal returns (at a slower rate than nematodes), then the utilitarian result is to start giving dollars to the utility monster until its utility-per-dollar has diminished enough to match the starting utility-per-dollar of the nematodes, and then to give to both the utility monster and the nematodes in a proportion which keeps them at the same rate. The "utility monster" has ceased to be a utility monster because it no longer gets everything. It still gets more, of course, but that's the equivalent of deciding that the starving person gets the food before the full person.

Comment author: satt 17 August 2013 01:39:44AM *  6 points [-]

The "utility monster" has ceased to be a utility monster because it no longer gets everything. It still gets more, of course, but that's the equivalent of deciding that the starving person gets the food before the full person.

This sounds like it could be almost as repugnant as a utility monster that gets literally everything, depending on precisely how much "more" we're talking about.

Edit: if I were the kind of person who found utility monsters repugnant, that is. I'd already dissolved the "OMG what if utility monsters??" problem in my own mind by reasoning that the repugnant feeling comes from representing utility monsters as black boxes, stripping away all of the features of theirs that make it intuitively obvious why they generate more utility from the same inputs. Put another way, the things that make real-life utility monsters "utility monsters" are exactly the things that make us fail to recognize them as utility monsters. When a parent values their child's continued existence far more than their own, we don't call the child a "utility monster" if the parent sacrifices themselves to save their child, even though that's exactly the child's role in that situation.

Comment author: PhilGoetz 17 August 2013 03:34:11PM 1 point [-]

Re. "black box", nice way of putting it. This post just gives an example where we can look inside the black box.

Comment author: TGM 18 August 2013 10:29:35PM 1 point [-]

I want to criticise either the idea that diminishing returns is important, or, at least, that dollar values make sense for talking about them.

Suppose we have a monster who likes to eat. Each serving of food is just as tasty as the previous, but he still gets diminishing returns on the dollar, because the marginal cost of the servings goes up.

We also have nematodes, who like to eat, but not as much. They never get a look in, because as the monster eats, they also suffer diminished utilons per dollar.

So the monster is serving the 'purpose' of the utility monster, but still has diminishing returns on the dollar. If we redefine diminishing returns to be on something else, I'm not sure it could be well justified or immune to this issue.

And, although humans are not an example of this sort of monster, the human race certainly is.