This feels quite important, especially in the context of “dying Less Wrong”:
Bottom line: I know of only one minimum viable architecture for turning individuals into a superorganism — for making sure it’s in everyone’s self-interest to work together. I call it the Prestige Economy, and it runs on a deceptively simple rule:
Individuals should grant social status to others for advancing the superorganism’s goals.
Sure, but assuming you have people willing to contribute to “many things”, it can help to say explicitly what behavior is helpful and what is harmful. Especially among nerds.
Karma system itself doesn’t specify what you get most karma for. It could be “for advancing the superorganism’s goals”, but it also could be for “making snarky comments about those who try to advance the superorganism’s goals”, or simply for “posting a lot of trivial stuff”.
Also, social status is perceived not only by how much karma one has, but also how they are treated (replied to, talked about) by other people in comments.
Sometimes what would be easier to do in real life is more difficult to do online, because the functionality is different or missing. For example, the real-life equivalent of “karma”—displays of admiration or friendship—is public.
Assuming voluntary association, one way to “enforce a given behavior” is to only associate with people who agree to behave like that. So if you want to have a group that does X, you associate with people who give social status to people working on X (as opposed to e.g. people who talk a lot about X, but give low status to people actually working on X).
Online… I guess banning people for not expressing proper attitudes would be perceived by many as very problematic (even if that’s what most of us more or less do in real life), so the realistic solution seems to be an invite-only club. Or a two-tiered system, where the outer forum is open to everyone, and only people with the desired behavior get an invitation to the inner forum.
So if you want to have a group that does X, you associate with people who give social status to people working on X
You are glossing over the practice of giving social status. In real life, as you said, it is basically public displays of admiration (friendship is a bit different). So you are going to build a community with the inclusion criterion of being willing to publicly admire a particular set of people. Presumably if you stop admiring, you are no longer welcome in the community. That doesn’t strike me as a way to build a healthy community—there are obvious failure modes looming.
with the inclusion criterion of being willing to publicly admire a particular set of people
That isn’t quite what Viliam is proposing. He says (emphasis mine):
you associate with people who give social status to people working on X
so what membership in this community commits you to is not admiring specific people but admiring people who do specific things, whoever those people are.
This still seems kinda dangerous, but I don’t think it has the same failure modes.
commits you to is not admiring specific people but admiring people who do specific things, whoever those people are.
I suspect that distinction is not going to be as clear-cut when you are dealing with a bunch of actual humans.
“Bob is an asshole but he did the specific thing X so I’m supposed to publicly praise him? I’ll pass.”
“Alice is such a great person, so what that she skipped the specific thing X this time, she’s the best and I’m going to sing hosannas to her really loudly”.
No, my question was more abstract, in the sense of “can you ever design a system that grants a prestige economy that doesn’t de-evolve into a karma system”?
Karma is not granting social status, it is community moderation of discussion. Imagine if the aggregate karma numbers were removed from display (arguably this is maybe something that should be done).
What do you think is the best example of this kind of principle at work? My first guess would be academia, but I can also think of a dozen reasons why the prestige system in academia is flawed.
My intuition is that there is really no shining example of the prestige economy in the real world—but whether this has to do with the difficulty of implementation, or a flaw with the idea itself, I’m not sure.
There can be many flaws in implementation, and we probably won’t find a perfect one, but academia seems like a decent example. It has a goal (research and education), and it assigns status (academic functions) to people who contribute to that goal, and the status comes with certain benefits (salary) and powers (over students) which means other people will recognize it as a status.
If you spend a lot of time hanging out with professors and learn all their buzzwords, but you will do no research nor teaching, you are not going to become a professor, i.e. you are not able to out-status them within the academia.
Another example would be meritocratic open-source projects, where people are respected according to their contributions to the project.
Or perhaps a sales department, where people are rewarded depending on their sales.
It has a goal (research and education), and it assigns status (academic functions) to people who contribute to that goal, and the status comes with certain benefits (salary) and powers (over students) which means other people will recognize it as a status.
How is that different from pretty much any job? Let’s take ditch-digging.
It has a goal (digging ditches), and it assigns status (becoming a foreman, then a manager) to people who contribute to that goal, and the status comes with certain benefits (salary) and powers (over workers) which means other people will recognize it as a status.
If you spend a lot of time hanging out with ditch-diggers and learn all their buzzwords, but you will do no digging of ditches, you are not going to become a ditch-digger, i.e. you are not able to out-status them within the ditch-digging world.
I don’t think anyone was claiming it doesn’t apply to pretty much any job. (In the original context, the point was that it does apply to pretty much any job, and to a host of other things besides.)
It will apply better to some jobs than others. It needs
the people doing that job to form a community in which social status is actually meaningful
mere membership of the community to be seen by its members as conferring status
doing the job effectively to lead to promotion to a position in which one can do it better
(preferably) status within that community to have some currency elsewhere.
Those are all true for academia. There is definitely such a thing as the academic community, its members relate to one another socially as well as professionally, academics tend to think highly of themselves as a group, promotion means better opportunities for furthering academic research (oneself or by organizing subordinates and taking some credit for their work), and—at least in the nice middle-class circles in which academics tend to move—there’s some broader cachet to being, say, a professor.
I think they’re less true for ditch-digging. So far as I know, there isn’t the same sort of widely-spread confraternity of ditch-diggers that there is of academics. I’ve not heard that ditch-diggers see themselves as having higher status than non-ditch-diggers. I think a lot of ditch-diggers are casual labourers with no real prospects of promotion, though I confess I don’t exactly have my finger on the pulse of ditch-digging career progression. And there are few social circles in which introducing yourself as a ditch-digger will make people look up to you.
So yeah, this structure applies to things other than academia, but it does seem like it applies better to academia than to the other example you offered.
We could swing in the other direction and consider hedge fund managers instead of ditch-diggers if you worry that ditch-diggers are too low status :-)
However I think the issue is a bit different. The original question was how to build a community successfully driven by status. Once we switch to jobs we are talking about money and power—pure status becomes secondary.
Besides, academia seems to me to be a poor example. Its parts where advancement doesn’t give you much in the way of money and power—that is, social sciences—became quite dysfunctional and the chasing of status leads to bad things like an ever-growing pile of shit research being published as “science”.
Once we switch to jobs we are talking about money and power—pure status becomes secondary.
Possibly, though I suspect status is more important relative to money in motivating employees than is commonly thought.
Its parts where advancement doesn’t give you much in the way of money and power—that is, social sciences [...]
I’m not sure you get a lot more money and power from advancement in other areas of academia. (Unless you count coaching the football team, in US universities. Plenty of money there.) It seems to me that there are better explanations if the hard sciences are less dysfunctional than the soft.
Whether things like
an ever-growing pile of shit research being published as “science”
are evidence that the academic community isn’t driven successfully (whether by status or by something else) depends on what you take to be the actual goals of academia as an institution. I agree that if we take those goals to be research that actually uncovers truths, and education that actually improves the minds and the lives of the educated, it’s debatable how well academia does.
I have no experience with professional ditch-digging, so I can’t comment on that. My job experience is mostly software development, and in many companies the people who actually create the software are near the bottom of the status ladder. (Which I guess is okay, because the actual goal is making money; creating quality software is just an instrumental goal which can be sacrificied for the higher goals at any moment.)
Awesome link, and a fantastic way of thinking about how human institutions/movements/subcultures work in the abstract.
I’m not sure the quote conveys the full force of the argument out of that context though, so I recommend reading the full thing if the quote doesn’t ring true with you (or even if it does).
True. Maybe we could still make celebrate our minor celebrities more, along with just individual good work, to avoid orbiting too much around any one person. I don’t know what the optimum incentive gradient is between small steps and huge accomplishments. However, I suspect that on the margin more positive reinforcement is better along the entire length, at least for getting more content.
(There are also benefits to adversarial review and what not, but I think we’re already plenty good at nitpicking, so positive reinforcement is what needs the most attention. It could even help generate more long thoughtful counterarguments, and so help with the better adversarial review, improving the dialectic.)
This. Different people need different advice. People prone to worship and groupthink need to be told about the dangers of following the herd. People prone to nitpicking and contrarianism need to be told about how much power they lose by being unable to cooperate.
Unfortunately, in real life most people will choose exactly the opposite message—the groupthinkers will remind themselves of the dangers of disagreement, and the nitpickers will remind themselves of the dangers of agreement.
This feels quite important, especially in the context of “dying Less Wrong”:
-- source
“Individuals should” is not an architecture.
A great many things become possible conditional on “everyone should so as I say”.
Sure, but assuming you have people willing to contribute to “many things”, it can help to say explicitly what behavior is helpful and what is harmful. Especially among nerds.
How is this different from the karma system?
Karma system itself doesn’t specify what you get most karma for. It could be “for advancing the superorganism’s goals”, but it also could be for “making snarky comments about those who try to advance the superorganism’s goals”, or simply for “posting a lot of trivial stuff”.
Also, social status is perceived not only by how much karma one has, but also how they are treated (replied to, talked about) by other people in comments.
Indeed, but how would you enforce a system of points granted if and only if someone advances the goal of this site?
This I think is unavoidable in a human society.
Sometimes what would be easier to do in real life is more difficult to do online, because the functionality is different or missing. For example, the real-life equivalent of “karma”—displays of admiration or friendship—is public.
Assuming voluntary association, one way to “enforce a given behavior” is to only associate with people who agree to behave like that. So if you want to have a group that does X, you associate with people who give social status to people working on X (as opposed to e.g. people who talk a lot about X, but give low status to people actually working on X).
Online… I guess banning people for not expressing proper attitudes would be perceived by many as very problematic (even if that’s what most of us more or less do in real life), so the realistic solution seems to be an invite-only club. Or a two-tiered system, where the outer forum is open to everyone, and only people with the desired behavior get an invitation to the inner forum.
You are glossing over the practice of giving social status. In real life, as you said, it is basically public displays of admiration (friendship is a bit different). So you are going to build a community with the inclusion criterion of being willing to publicly admire a particular set of people. Presumably if you stop admiring, you are no longer welcome in the community. That doesn’t strike me as a way to build a healthy community—there are obvious failure modes looming.
That isn’t quite what Viliam is proposing. He says (emphasis mine):
so what membership in this community commits you to is not admiring specific people but admiring people who do specific things, whoever those people are.
This still seems kinda dangerous, but I don’t think it has the same failure modes.
I suspect that distinction is not going to be as clear-cut when you are dealing with a bunch of actual humans.
“Bob is an asshole but he did the specific thing X so I’m supposed to publicly praise him? I’ll pass.”
“Alice is such a great person, so what that she skipped the specific thing X this time, she’s the best and I’m going to sing hosannas to her really loudly”.
You seem to have hit on (one reason) why the prestige economy is different from the karma system.
No, my question was more abstract, in the sense of “can you ever design a system that grants a prestige economy that doesn’t de-evolve into a karma system”?
Karma is not granting social status, it is community moderation of discussion. Imagine if the aggregate karma numbers were removed from display (arguably this is maybe something that should be done).
Karma was not designed to grant social status, but people look at it as if it does.
What do you think is the best example of this kind of principle at work? My first guess would be academia, but I can also think of a dozen reasons why the prestige system in academia is flawed.
My intuition is that there is really no shining example of the prestige economy in the real world—but whether this has to do with the difficulty of implementation, or a flaw with the idea itself, I’m not sure.
There can be many flaws in implementation, and we probably won’t find a perfect one, but academia seems like a decent example. It has a goal (research and education), and it assigns status (academic functions) to people who contribute to that goal, and the status comes with certain benefits (salary) and powers (over students) which means other people will recognize it as a status.
If you spend a lot of time hanging out with professors and learn all their buzzwords, but you will do no research nor teaching, you are not going to become a professor, i.e. you are not able to out-status them within the academia.
Another example would be meritocratic open-source projects, where people are respected according to their contributions to the project.
Or perhaps a sales department, where people are rewarded depending on their sales.
How is that different from pretty much any job? Let’s take ditch-digging.
It has a goal (digging ditches), and it assigns status (becoming a foreman, then a manager) to people who contribute to that goal, and the status comes with certain benefits (salary) and powers (over workers) which means other people will recognize it as a status.
If you spend a lot of time hanging out with ditch-diggers and learn all their buzzwords, but you will do no digging of ditches, you are not going to become a ditch-digger, i.e. you are not able to out-status them within the ditch-digging world.
I don’t think anyone was claiming it doesn’t apply to pretty much any job. (In the original context, the point was that it does apply to pretty much any job, and to a host of other things besides.)
It will apply better to some jobs than others. It needs
the people doing that job to form a community in which social status is actually meaningful
mere membership of the community to be seen by its members as conferring status
doing the job effectively to lead to promotion to a position in which one can do it better
(preferably) status within that community to have some currency elsewhere.
Those are all true for academia. There is definitely such a thing as the academic community, its members relate to one another socially as well as professionally, academics tend to think highly of themselves as a group, promotion means better opportunities for furthering academic research (oneself or by organizing subordinates and taking some credit for their work), and—at least in the nice middle-class circles in which academics tend to move—there’s some broader cachet to being, say, a professor.
I think they’re less true for ditch-digging. So far as I know, there isn’t the same sort of widely-spread confraternity of ditch-diggers that there is of academics. I’ve not heard that ditch-diggers see themselves as having higher status than non-ditch-diggers. I think a lot of ditch-diggers are casual labourers with no real prospects of promotion, though I confess I don’t exactly have my finger on the pulse of ditch-digging career progression. And there are few social circles in which introducing yourself as a ditch-digger will make people look up to you.
So yeah, this structure applies to things other than academia, but it does seem like it applies better to academia than to the other example you offered.
We could swing in the other direction and consider hedge fund managers instead of ditch-diggers if you worry that ditch-diggers are too low status :-)
However I think the issue is a bit different. The original question was how to build a community successfully driven by status. Once we switch to jobs we are talking about money and power—pure status becomes secondary.
Besides, academia seems to me to be a poor example. Its parts where advancement doesn’t give you much in the way of money and power—that is, social sciences—became quite dysfunctional and the chasing of status leads to bad things like an ever-growing pile of shit research being published as “science”.
Possibly, though I suspect status is more important relative to money in motivating employees than is commonly thought.
I’m not sure you get a lot more money and power from advancement in other areas of academia. (Unless you count coaching the football team, in US universities. Plenty of money there.) It seems to me that there are better explanations if the hard sciences are less dysfunctional than the soft.
Whether things like
are evidence that the academic community isn’t driven successfully (whether by status or by something else) depends on what you take to be the actual goals of academia as an institution. I agree that if we take those goals to be research that actually uncovers truths, and education that actually improves the minds and the lives of the educated, it’s debatable how well academia does.
I think power is more important than commonly thought, but I accept that it’s not easy to disentangle it from status.
You do in some, notably law and business. And yes, I’m not saying that’s the main explanation why soft sciences do so much worse than hard ones.
I have no experience with professional ditch-digging, so I can’t comment on that. My job experience is mostly software development, and in many companies the people who actually create the software are near the bottom of the status ladder. (Which I guess is okay, because the actual goal is making money; creating quality software is just an instrumental goal which can be sacrificied for the higher goals at any moment.)
I’m not sure that “social status” on an online forum like this is an important currency.
Awesome link, and a fantastic way of thinking about how human institutions/movements/subcultures work in the abstract.
I’m not sure the quote conveys the full force of the argument out of that context though, so I recommend reading the full thing if the quote doesn’t ring true with you (or even if it does).
Lesswrong doesn’t celebrate heroes much. I think that’s on purpose though…
Yes, but...
I feel like LW.com has the problem but our local LW Berlin community doesn’t.
True. Maybe we could still make celebrate our minor celebrities more, along with just individual good work, to avoid orbiting too much around any one person. I don’t know what the optimum incentive gradient is between small steps and huge accomplishments. However, I suspect that on the margin more positive reinforcement is better along the entire length, at least for getting more content.
(There are also benefits to adversarial review and what not, but I think we’re already plenty good at nitpicking, so positive reinforcement is what needs the most attention. It could even help generate more long thoughtful counterarguments, and so help with the better adversarial review, improving the dialectic.)
This. Different people need different advice. People prone to worship and groupthink need to be told about the dangers of following the herd. People prone to nitpicking and contrarianism need to be told about how much power they lose by being unable to cooperate.
Unfortunately, in real life most people will choose exactly the opposite message—the groupthinkers will remind themselves of the dangers of disagreement, and the nitpickers will remind themselves of the dangers of agreement.