So there's this argument that utilitarianism doesn't work because (per work by behavioral economists) preferences are non-transitive. That is, you can't make a top to bottom list of preferences for each person where the higher position is always a higher preference.
But, like, surely preferences are *pretty* transitive? You can construct weird non-transitive cases. But for most purchase you could construct a list.
@ZachWeinersmith
There's a money-pump argument that goes like this: if you're happy with your intransitive preferences then I will upsell you.
You should pay me €1 to exchange your X for my Y, my Z for your Y, and my X for your Z.
@ZachWeinersmith
The argument tries to show that preferences can be logical or illogical, so some sets of preferences are simply 'wrong'.
My intuition (as the Philosophers like to say) is that anyone who disagrees has spent too long in academia.
@ZachWeinersmith I suspect you'd get a pretty good value system merely by adhering to: do the Good thing when it's clear most people would agree about what the Good thing is, and flip a coin otherwise.
Not very satisfying for the folks tied to the trolley's tracks, of course.
You'd be right, *if* in the real world you didn't find both passionately for better material circumstances independent of social other and passionately for a higher social position independent of material circumstances.
These are completely incompatible.
I suspect there are many others.
@dancer_storm
I'm not the right professor to ask, but there are good works on preferences favouring "collective harmony" over the individual benefit.
However, your specific question is not well posed, because by definition, utility is the final value, and may well include "making other people happy".
@ZachWeinersmith
@iinavpov @ZachWeinersmith what do you mean by well posed?
I'm thinking of something like \[u_a = U_a([C_{ax}, C_{ay}, C_{az}, ...], [u_b, u_c, u_d, ...])\] and \[u_b = U_b([C_{bx}, C_{by}, C_{bz}, ...], [u_a, u_c, u_d, ...])\]
Where \(a\), \(b\), \(c\) are individuals and \(C\) is consumption
Since \(u_a\) depends on \(u_b\) and vice versa, aggregation becomes more challenging I imagine. Do you have something specific that I can read?
@ZachWeinersmith isn't that more of a math issue? If you can't have a utility function (because it requires transitivity), what do you sum?
But we estimate utility functions all the time and they're pretty useful. Do we have any specific examples of a violation that would really mess up an estimate and subsequently really mess up an aggregation?
@dancer_storm @ZachWeinersmith the 2024 American presidential election?
It's amazing in retrospect how much and how effectively the Republican party was able to elevate secondary concerns to primary just for a moment, resulting in a government that no one really likes.
@buckfiftyseven @ZachWeinersmith that's not an issue with transitivity, it's more about persistence ("I like broccoli now") and mismatches between stated/revealed preferences and experienced utility ("I bought this treadmill for a fair price but I hate running" or "I bought this thing from Temu for what I thought was a fair price but the thing that arrived was not what I thought it was and I can't return it")
Both are important issues too, but they don't mean there is no utility function, just that the system failed to reveal it.
@dancer_storm @ZachWeinersmith I'm not sure it's that different, if people thought they were voting their preferences.
*Our* perception is that they were deceived, but certainly the moment theirs was not. For many still.
@buckfiftyseven @dancer_storm @ZachWeinersmith
I don't understand. If you sort a list by preference, it won't be sorted by preference? 🤔
@Phosphenes @dancer_storm @ZachWeinersmith someone linked Kenneth Arrow’s “impossibility” theorem.
I think that applies, and "it's even worse" for reasons described above
@ZachWeinersmith
Well, one major issue here is that we manage multiple preferences at the same time, and these change their position in the list in response to stuff.
So like, say cheeseburgers are near the top and have a value of +27, but pickle chips have a value of -28. Most of the time, as much as you like the burger, you won't accept it if it has pickles, or you'll remove them before eating it.
Now let's say you're really hungry. Haven't eaten in days. Now that pickle number might go to -14. Still don't like pickles, but you're starving and won't readily reject the burger on that basis alone. Not unless maybe mustard is at -14 or more, tipping the scales again.
I'm not convinced. There are people (NOT TO GET POLITICAL ABOUT IT) for whom conformity and uniformity are strong societal preferences, to the extent that they will vote against their own economic interests for people who promise to enforce them, and others (yeah, hi) for whom they are problems we as a civilization are attempting to solve.
Zach Weinersmith
Rackhir
Nick Judd, Chicago dad
Nick
Malin
MiSt
Mans R
economística
Vincent Tunru
Nemo
John
Phosphenes
Willow Brook 💜
Ronan
Tobias Klausmann
John Kaniarz
Steve Gisselbrecht
outofthenorm
standev