empirically we are massively losing out by limiting the audience of LW to TOUGH GUYS who can HANDLE CRITICISM
First, not audience but content creators, but second, is this so? Did any of the really valuable contributors to LW go away because they were driven away by incessant criticism? You think Scott Alexander moved to SSC because he couldn’t handle the downvotes?
The general cry here seems to be “We want more content!”. Well, I don’t want more content. I have a whole internet full of content. What I want is more high-quality content that I do not need to search through piles of manure to find. The great advantage of LW is that here pearls are frequent but bullshit is rare—and I attribute this in not a small degree to the fact that you’ll be punished (by downvotes and comments) for posting bullshit.
A system without downvotes encourages posting, true, but it encourages posting of everything including cat pictures and ruminations on a breakfast sandwich in three volumes. Someone has to do pruning and if you take this power away from the users, it’ll fall to the moderators. I don’t see why this would be better—and people whose cat got disrespected will still be unhappy.
Did any of the really valuable contributors to LW go away because they were driven away by incessant criticism? You think Scott Alexander moved to SSC because he couldn’t handle the downvotes?
Didn’t Eliezer say somewhere that he posts on Facebook instead of LW nowadays because on LW you get dragged into endless point-scoring arguments with dedicated forum arguers and on Facebook you just block commenters who come off as too tiresome to engage with from your feed?
As far as I understand (it isn’t very far), Eliezer prefers Facebook basically because it gives him control—which is perfectly fine, his place on FB is his place and he sets the rules.
I don’t think that degree of control would be acceptable on LW—the local crowd doesn’t like tyrants, even wise and benevolent.
If Eliezer would bans someone on LW on the other hand he would get a storm of criticism.
I’m curious what solution would work here.
Suppose you had a list of ~10 users with ‘censor’ power, and the number of censors who have ‘remonstrated’ a user is public, possibly also with the remonstrations. “Don’t be a jerk,” or “don’t promote other sites in your early posts,” or “think before you speak,” or so on. If a sufficient number of censors have remonstrated a user, then they’re banned, but censors can lift their remonstration once it’s no longer appropriate.
Thoughts on this solution:
Reasoning is clear and transparent, and gradual. Instead of “all clear” suddenly turning to “can’t post anymore,” people are put ‘on notice.’
If which censor has remonstrated a user is hidden, it isn’t “Eliezer” using his dictatorial powers; it’s some moderator moderating.
If which censor has remonstrated a user is hidden, the drama might multiply rather than decrease. Now an offending user can message the entire group of censors, pleading to have their remonstration removed, or complain bitterly that clearly it was their enemy who is a censor, regardless of whether or not that was actually the person that remonstrated with them.
If three out of ten moderators agree that a poster should stop posting, then it becomes much easier to defend the action to remove the poster.
But it doesn’t really get at the heart of the issue. Eliezer acts that way because of the Roko affair and people telling him that he shouldn’t have moderated.
In that case the decision being made by three people instead of one wouldn’t have made it more defensible.
This forum currently has MIRI ties that make controversial moderating decisions reflect badly on MIRI.
A solution would be to cut those ties and give LW into the hand of a small group of moderators who are more free to focus on what’s good for the community instead of larger PR effects.
You think Scott Alexander moved to SSC because he couldn’t handle the downvotes?
He did explicitly point out that this culture of criticism / high standards makes writing for LW a chore, and so he doesn’t do it anymore. So, yes.
I am not advocating for the removal of downvotes; I think they serve a necessary function, and I think having some sort of pruning and sorting methodology is a core site feature. But to cultivate good content, it is not enough to just remove bad content.
He did explicitly point out that this culture of criticism / high standards makes writing for LW a chore
Let’s bring in the entire quote. Yvain said:
Less Wrong requires no politics / minimal humor / definitely unambiguously rationality-relevant / careful referencing / airtight reasoning (as opposed to a sketch of something which isn’t exactly true but points to the truth.) This makes writing for Less Wrong a chore as opposed to an enjoyable pastime.
Note that the first three points have nothing do with criticism. The fourth point is the requirement to show evidence which still isn’t criticism. And the final point I read as having to be literal and formal with little “free play” in the moving parts—I think there is a connection with the recent series of posts by Jonah Sinick where he talks how gestalt pattern recognition is, at certain level, superior to formal reasoning (and LW expects formal reasoning).
Yeah, I still think Scott Alexander could handle the downvotes just fine.
But to cultivate good content, it is not enough to just remove bad content.
I agree, but the suggestions offered tend to gravitate to “Let’s just be nice to everyone”...
What kind of positive incentives to creators of high-quality content can LW come up with?
The thing is, the high standards on LW that Yvain refers to are precisely what makes LW content valuable. At some level, wanting to escape requirements such as airtight reasoning means you want to write stuff that doesn’t have airtight reasoning.
I’ve refrained from posting because I expected to get really banal criticism. You may or may not consider that a loss. But I kind of get the impression that Scott feels somewhat similarly. It’s not like he doesn’t get criticized on SSC.
I think this isn’t a case of me needing to HTFU. (Other self-modification would have worked, but it would also not be very useful outside of LW.) So it may not be relevant to what you’re trying to say. But I also wonder whether other people feel similarly, and are expressing it in ways that you’re interpreting as them needing to HTFU.
(I know I’m allowed to ignore comments like that, but I still didn’t feel like bothering.)
I don’t think “better” readers would be a helpful way to frame it. There are lots of dimensions of quality. E.g. one of the HN comments said
Game theorists always seem to assume that there’s no such thing as nuanced communication.
which is a bad comment in a way that I don’t think would get traction on LW.
I think… maybe one factor is comments that are bad because they’re wrong, and comments that are bad because they’re right but, really, who cares? Like jaywalking in front of a policeman who then stops you, gives you a stern lecture, and you have to say yes officer and no officer and so on. It feels more like a power trip than an actual attempt to make me or anyone else safer.
If insightful and to-the-point skewering was justified, then I wouldn’t enjoy it and it might put me off future posting (and maybe it should), but I hope I would find it valuable and take it as a sign that I needed to level up.
maybe one factor is comments that are bad because they’re wrong, and comments that are bad because they’re right but, really, who cares?
So, nit-picking? Yes, it’s popular on LW :-/ but (a) you are still free to ignore those; and (b) as opposed to the example with the cop, there is no inherent power imbalance. Nothing prevents you from going meta and pointing out the difference between what is important and what is not.
Do I read you right in that you want more co-travelers in figuring out problems and solutions and less critics who carefully examine your text for minor flaws and gotchas, basically?
On reflection, I’m not sure that nitpicking is quite the problem that I’m pointing at, but I don’t think I have a very good handle on what is. (I do think nitpicking is a problem.)
Maybe next time I have that feeling, I’ll just post anyway and see what happens.
So, nit-picking? Yes, it’s popular on LW :-/ but (a) you are still free to ignore those; and (b) as opposed to the example with the cop, there is no inherent power imbalance. Nothing prevents you from going meta and pointing out the difference between what is important and what is not.
It often takes a special effort to -notice- that a criticism isn’t meaningful, especially when it is correct—especially because Less Wrong entertains a -much- higher level of pedant than will generally be encountered elsewhere. More problematically, pedantry tends to get upvoted, which means people may pay too much attention to it, and also that it is being encouraged.
If we’re interested in discouraging pedantry-for-the-sake-of-pedantry, I’d lean towards implementing an applause-lights keyword to indicate that a criticism may be valid, but doesn’t actually add anything to what is being said, along the lines of how “Updating” was used as an applause-lights keyword to counterbalance the generally negative attitude people start with towards admitting wrongness.
It often takes a special effort to -notice- that a criticism isn’t meaningful, especially when it is correct
True—but I think it’s a very useful skill to develop and practice.
pedantry tends to get upvoted
And that is probably a feature of the local culture by now, heavily supported by the meme of how you can’t make even one tiny little itty bitty mistake when programming the AI because if you do it’s all paperclips all the time.
I’d lean towards implementing an applause-lights keyword
I call such things “technically correct, but irrelevant”, but I don’t think this expression functions well as an applause-lights switch. Ideas?
The best opposite to “pedantry” I can come up with is “pragmatic.” Pragmatism is a relatively good value on Less Wrong, but I don’t see a good application.
Yours seems good. It concedes the argument attempted to be raised, shutting off further discussion—a very desirable quality when dealing with somebody who is specifically looking for something to argue with—and rebuts the fundamental problem, redirecting future attention there. (Minor shift for reasons I have trouble explicating, but which seems a stronger, slightly harsher version of the sentiment—“Technically correct. Also irrelevant.”) If it’s used appropriately, and consistently, I think it could become an applause-light within the sub-culture here.
First, not audience but content creators, but second, is this so? Did any of the really valuable contributors to LW go away because they were driven away by incessant criticism? You think Scott Alexander moved to SSC because he couldn’t handle the downvotes?
The general cry here seems to be “We want more content!”. Well, I don’t want more content. I have a whole internet full of content. What I want is more high-quality content that I do not need to search through piles of manure to find. The great advantage of LW is that here pearls are frequent but bullshit is rare—and I attribute this in not a small degree to the fact that you’ll be punished (by downvotes and comments) for posting bullshit.
A system without downvotes encourages posting, true, but it encourages posting of everything including cat pictures and ruminations on a breakfast sandwich in three volumes. Someone has to do pruning and if you take this power away from the users, it’ll fall to the moderators. I don’t see why this would be better—and people whose cat got disrespected will still be unhappy.
Didn’t Eliezer say somewhere that he posts on Facebook instead of LW nowadays because on LW you get dragged into endless point-scoring arguments with dedicated forum arguers and on Facebook you just block commenters who come off as too tiresome to engage with from your feed?
As far as I understand (it isn’t very far), Eliezer prefers Facebook basically because it gives him control—which is perfectly fine, his place on FB is his place and he sets the rules.
I don’t think that degree of control would be acceptable on LW—the local crowd doesn’t like tyrants, even wise and benevolent.
On the LW facebook group Eliezer bans occasionally bans people who post really low quality content. The same goes for his own feed.
If Eliezer would bans someone on LW on the other hand he would get a storm of criticism.
I’m curious what solution would work here.
Suppose you had a list of ~10 users with ‘censor’ power, and the number of censors who have ‘remonstrated’ a user is public, possibly also with the remonstrations. “Don’t be a jerk,” or “don’t promote other sites in your early posts,” or “think before you speak,” or so on. If a sufficient number of censors have remonstrated a user, then they’re banned, but censors can lift their remonstration once it’s no longer appropriate.
Thoughts on this solution:
Reasoning is clear and transparent, and gradual. Instead of “all clear” suddenly turning to “can’t post anymore,” people are put ‘on notice.’
If which censor has remonstrated a user is hidden, it isn’t “Eliezer” using his dictatorial powers; it’s some moderator moderating.
If which censor has remonstrated a user is hidden, the drama might multiply rather than decrease. Now an offending user can message the entire group of censors, pleading to have their remonstration removed, or complain bitterly that clearly it was their enemy who is a censor, regardless of whether or not that was actually the person that remonstrated with them.
If three out of ten moderators agree that a poster should stop posting, then it becomes much easier to defend the action to remove the poster.
That’s a bureaucratic solution.
But it doesn’t really get at the heart of the issue. Eliezer acts that way because of the Roko affair and people telling him that he shouldn’t have moderated. In that case the decision being made by three people instead of one wouldn’t have made it more defensible.
This forum currently has MIRI ties that make controversial moderating decisions reflect badly on MIRI. A solution would be to cut those ties and give LW into the hand of a small group of moderators who are more free to focus on what’s good for the community instead of larger PR effects.
He did explicitly point out that this culture of criticism / high standards makes writing for LW a chore, and so he doesn’t do it anymore. So, yes.
I am not advocating for the removal of downvotes; I think they serve a necessary function, and I think having some sort of pruning and sorting methodology is a core site feature. But to cultivate good content, it is not enough to just remove bad content.
Let’s bring in the entire quote. Yvain said:
Note that the first three points have nothing do with criticism. The fourth point is the requirement to show evidence which still isn’t criticism. And the final point I read as having to be literal and formal with little “free play” in the moving parts—I think there is a connection with the recent series of posts by Jonah Sinick where he talks how gestalt pattern recognition is, at certain level, superior to formal reasoning (and LW expects formal reasoning).
Yeah, I still think Scott Alexander could handle the downvotes just fine.
I agree, but the suggestions offered tend to gravitate to “Let’s just be nice to everyone”...
What kind of positive incentives to creators of high-quality content can LW come up with?
The thing is, the high standards on LW that Yvain refers to are precisely what makes LW content valuable. At some level, wanting to escape requirements such as airtight reasoning means you want to write stuff that doesn’t have airtight reasoning.
Yes, I agree. That’s why I think “more content” is the wrong yardstick. I want “more high-quality content” which you don’t get by relaxing standards.
Correct, but that’s fine. There is a lot of high-quality and valuable stuff that is not airtight-reasoned.
I’ve refrained from posting because I expected to get really banal criticism. You may or may not consider that a loss. But I kind of get the impression that Scott feels somewhat similarly. It’s not like he doesn’t get criticized on SSC.
I think this isn’t a case of me needing to HTFU. (Other self-modification would have worked, but it would also not be very useful outside of LW.) So it may not be relevant to what you’re trying to say. But I also wonder whether other people feel similarly, and are expressing it in ways that you’re interpreting as them needing to HTFU.
..so just ignore it?
I think yours is a different case—it’s as if you want better readers than the LW crowd. Would you be fine with insightful and to-the-point skewering?
I don’t think “better” readers would be a helpful way to frame it. There are lots of dimensions of quality. E.g. one of the HN comments said
which is a bad comment in a way that I don’t think would get traction on LW.
I think… maybe one factor is comments that are bad because they’re wrong, and comments that are bad because they’re right but, really, who cares? Like jaywalking in front of a policeman who then stops you, gives you a stern lecture, and you have to say yes officer and no officer and so on. It feels more like a power trip than an actual attempt to make me or anyone else safer.
If insightful and to-the-point skewering was justified, then I wouldn’t enjoy it and it might put me off future posting (and maybe it should), but I hope I would find it valuable and take it as a sign that I needed to level up.
So, nit-picking? Yes, it’s popular on LW :-/ but (a) you are still free to ignore those; and (b) as opposed to the example with the cop, there is no inherent power imbalance. Nothing prevents you from going meta and pointing out the difference between what is important and what is not.
Do I read you right in that you want more co-travelers in figuring out problems and solutions and less critics who carefully examine your text for minor flaws and gotchas, basically?
On reflection, I’m not sure that nitpicking is quite the problem that I’m pointing at, but I don’t think I have a very good handle on what is. (I do think nitpicking is a problem.)
Maybe next time I have that feeling, I’ll just post anyway and see what happens.
It often takes a special effort to -notice- that a criticism isn’t meaningful, especially when it is correct—especially because Less Wrong entertains a -much- higher level of pedant than will generally be encountered elsewhere. More problematically, pedantry tends to get upvoted, which means people may pay too much attention to it, and also that it is being encouraged.
If we’re interested in discouraging pedantry-for-the-sake-of-pedantry, I’d lean towards implementing an applause-lights keyword to indicate that a criticism may be valid, but doesn’t actually add anything to what is being said, along the lines of how “Updating” was used as an applause-lights keyword to counterbalance the generally negative attitude people start with towards admitting wrongness.
True—but I think it’s a very useful skill to develop and practice.
And that is probably a feature of the local culture by now, heavily supported by the meme of how you can’t make even one tiny little itty bitty mistake when programming the AI because if you do it’s all paperclips all the time.
I call such things “technically correct, but irrelevant”, but I don’t think this expression functions well as an applause-lights switch. Ideas?
The best opposite to “pedantry” I can come up with is “pragmatic.” Pragmatism is a relatively good value on Less Wrong, but I don’t see a good application.
Yours seems good. It concedes the argument attempted to be raised, shutting off further discussion—a very desirable quality when dealing with somebody who is specifically looking for something to argue with—and rebuts the fundamental problem, redirecting future attention there. (Minor shift for reasons I have trouble explicating, but which seems a stronger, slightly harsher version of the sentiment—“Technically correct. Also irrelevant.”) If it’s used appropriately, and consistently, I think it could become an applause-light within the sub-culture here.