One of those repeated adversaries is one of the creators and maintainers of GreaterWrong, which I believe many people have found valuable as a different UI on top of the LW backend; that seems pretty constructive to me
I think if LW’s stance is “we’re going to let this person drive away high-quality writers on the regular because we like the fact that they made GreaterWrong,” that’s worth being explicit about.
(That person is more responsible than any other single individual for Eliezer not being around much these days.)
To be clear, I wasn’t at all saying that. I was saying “this person is clearly not interested only in destructive nitpicking; here’s a substantial high-effort constructive thing they have done”.
Whether their presence on LW is net positive or negative is a completely separate question. (My current opinion is “clear net positive” but of course I could be wrong, and if e.g. it were the case that we could have Said or Eliezer but not both then that would be an argument on the other side—though I really don’t like the idea of banning user A because user B dislikes having them around even if user B is someone we would very much like to be here.)
I agree it’s a high-effort, constructive contribution.
I don’t know how to contrast it with making the business of actually writing LW content far worse than it otherwise would be.
I agree it’s a bad idea to ban user A because user B dislikes having them around. I think it’s a good idea to ban user A because users B, C, D, G, L, P, Q, R, and W, all of whom were valued contributors, all cited user A on their way out the door.
Have those people in fact said “I am leaving LW because I dislike interacting with Said”? Or is it something more like “I am leaving LW because I dislike the over-critical culture, of which Said is the clearest example”? Because if it’s the latter then it could be simultaneously true that (1) lots of people abandoning LW gave Said’s interaction style as a reason for leaving and that (2) banning Said wouldn’t actually do much to help.
(I remark that I have no evidence other than your say-so that lots of people have cited Said in explaining why they left LW, and that people’s explanations of such things are not always an accurate reflection of the actual causes. I am not, for the avoidance of doubt, claiming or even conjecturing that either you or they are/were lying.)
Since we’re talking about it, I have also told the mods that Said is one of three people who are readily top of mind at having a net negative impact on my LW experience. I’d just sort of been dealing with it, and personally I do not favor banning people sitewide for making me feel uncomfortable. But one of the mods reached out to me and specifically asked about factors influencing my experience of LessWrong, and the vibe and some of the specific commenters Duncan’s pointing out here are what I had reported was the main negative factor for me.
One of the facets of your comment here is that it puts the emphasis in a frustrating place: instead of saying “ah, n of 1, it’s good to know Said bugs you, I wonder if anybody else feels the same way,” it says “you’re the only person I’ve ever heard of complaining about Said, although I’m not saying you’re lying,” which is like a super negative way of being “neutral” about what Duncan said.
Like, imagine you said that to your significant other if they told you that one of your mutual friends, Bob, bothered your SO a lot during conversations, and that your SO had heard from other friends that Bob was very bothersome as well and made them not want to come and hang out. Would you say something like “I remark that I have no evidence other than your say-so that lots of our mutual friends have cited Bob in explaining why they don’t come to our board game night...?”
That sort of approach is a great way to alienate people and kill relationships. Why not just come right out and say “I don’t trust you enough to update ~at all on your reports about how other people feel about Bob?” That would at least be honest!
I did not intend to say or imply “you’re the only person I’ve ever heard of”, nor to be super-negative, and I don’t think I trust Duncan particularly less (or more) than a typical other prominent LW participant. (But I might trust him less specifically when reporting negative things about Said than I would trust him on other matters, and I would likewise trust Said less specifically when reporting negative things about Duncan.)
If anything I said caused anyone else to trust Duncan less, or caused Duncan any distress, then that was very much not an intended effect and I apologize.
I do, rereading what I wrote, see how it could be interpreted as having a much stronger negative subtext than I intended. What I intended was along the following lines: What Duncan is saying in this thread about “more than a dozen” high-quality ex-LWers citing Said as a major reason for their departure is, unlike the already well-known fact that Duncan himself finds Said intensely unpleasant to interact with, new information for me; but I am aware that when people find one another intensely unpleasant to interact with the things they say about one another are not always perfectly accurate, even when everyone involved is aiming to be truthful and fair, and I want to be cautious in how much I adjust my estimate of Said’s net impact on LW in response to something that comes from a single source I can’t check.
(But, in response to your outright accusation of dishonesty at the end: if I said the thing you would prefer me to say, it would be an untruth; my opinion of Duncan is not what you imply it is. My opinion of my wife, even less so, but I assume the point here is the application to the present case.)
I have heard both, the latter maybe 2x as often as the former?
I think that what’s happening right now, and has been happening over the years, is that the message “this is the kind of engagement we do, here” has been sent over and over again, and the mass of users has responded accordingly. I think if that message stopped being sent, or if a strong countersignal was sent, the mass of users would again respond accordingly.
That person is more responsible than any other single individual for Eliezer not being around much these days.
I can’t model Eliezer very well, but I’d argue that Eliezer is responsible for that decision. Whether it’s because he gets more reach elsewhere, less annoying diagreement/pointing out flaws, or what, it’s on him, not on any individual poster. I give him a ton of credit (and Robin Hanson) for starting LessWrong, but I reject any implication that we should limit or prune LessWrong to make it more appealing to Eliezer. Multiplied by a lot if Eliezer isn’t directly asking for some change, and we’re just cargo-culting what it looked like a decade ago.
I chose Eliezer as one example; it would be a mistake to think that my overall point was about Eliezer.
This single individual (who was recently making a big deal about being non-banned/in good standing) is more responsible than any other single individual for driving away more than a dozen high-quality authors; is the most-frequently cited reason, as far as I can tell, when high-quality people are asked to be specific about why they’re not on LessWrong/what makes LessWrong a miserable place to try to share their ideas and co-think.
Hearing that they’re tolerated because they maintain GreaterWrong makes my CFAR instincts tingle and makes me want to teach people about goal factoring.
Fair enough. I bounce off a lot of posts and comments, so I probably do some amount of victim-blaming if someone over-weights criticism, especially if it’s annoying and wrong. I don’t see how one or a few posters can move the site from well-worth posting to negative value, though it certainly could change the sign for a very marginal decision.
Even so, unless the problematic posts and comments are getting downvoted, but somehow still in-your-face enough to drive someone away, it’s hard to see that banning or otherwise authoritatively controlling the poster is justified.
Er, the (somewhat fuzzy) thesis of the post is that there’s a thing where, like, one or two prominent Socrati set the vibe, and then a bunch of other people follow suit; I find that criticism on LW feels more like [that guy] + [100 emulators of his style of engagement] than it used to.
Ultimately, it isn’t about just that one person’s engagement (although as I note above, they seem to have an outsized direct impact) so much as it is about a … prion disease?
As someone who’s posted about 60 or 70 essays on LW over the past eight years, doing so in 2023 involves a lot of bracing, as if I were hyping myself up to grab an electric fence. I straightforwardly expect the experience of posting to involve a net-negative subsequent four days. This was not the case in e.g. the Conor Moreton days, even though one or two of the Conor Moreton essays ended up in net negative territory.
If it’s that rough for me, when I’m a good writer who’s confident in his insights and has a lot of people who are interested-by-default in his thoughts, I imagine it can get a lot harder for the twenty-year-old not-yet-known version of me.
My recollection is that “Conor Moreton” at least once wrote something along the lines of “I find writing for LW stressful because I get a lot of criticism”. Maybe I’m misremembering, and for sure the obvious guess would be that you remember more clearly than I do, but my recollection now of my impression then is that it was nearer than you’re suggesting to how you describe your present experience of posting on LW.
Oh, I’ve definitely found it somewhat stressful all along; I think I’m in the most sensitive quintile if not decile.
Sorry, didn’t mean to imply that the shift was from [zero] to [large number]. More wanted to gesture at a shift from [moderate number] to [large number].
Like, there’s the difference between viscerally expecting one in four essays to result in a negative experience of magnitude 10 lasting for a day or two, and viscerally expecting two in three essays to result in a negative experience of magnitude 30 lasting for four days.
If the “ban commenter” function had not been implemented, I wouldn’t have posted any of my last five or six essays, and would be already gone.
Ah, I see. I don’t really disagree, but I also don’t think LW is unique in this, nor that there is (or can be) a long-lived growing-popularity group that maintains the feel of the early days. This seems like an evolution that’s plagued old-timers of a medium for all time, from SF fandom to pre-internet BBSs and Usenet, to early-internet special-purpose forums, to LW and rationalist-adjecent fora.
I don’t think Duncan’s gesturing at the “eternal September” problem—I think he’s talking about the “toxic low-grade criticism” problem, which is a related but separate issue. A persistent culture of toxic low-grade criticism exacerbates the eternal September problem by allowing impressionable newcomers to become acculturated to the pre-existing toxic dynamic, or to self-select for compatibility with it, making it that much harder to deal with. But you can work on improving the culture while also admitting that the constant influx of newcomers may make it difficult to go as far as you’d like to in terms of creating a specific and consistent set of norms.
I think there are some differences between this and other instances of degradation of quality by growth and entry of less-hardcore newcomers, and a resulting shift in norms that are generally negative in terms of quality. But I think there are a lot more similarities than differences.
I affirm that you did not say it, and believe that you did not mean it.
I would bet $100 to someone’s $1 that, if we could check the other timeline where that person behaved exactly the same way in comments, but had not made GreaterWrong, they would’ve been banned years ago.
(I believe this in part because I’ve heard multiple mods across multiple years talk about wanting to ban that person specifically, and I’ve heard GreaterWrong raised explicitly as cause for hesitation in something like four out of seven such discussions.)
I can’t think of any way to operationalize that bet—other than maybe noting that you have some of the same objections to Zack as to Said, and Zack also has not been banned; but to my mind their styles of interaction on LW are quite different and I can easily imagine someone finding one of them clearly net positive and the other clearly net negative. I guess we could ask the moderators, but they might very reasonably not want to answer that question, and you might fairly reasonably not trust whatever answer they gave.
If we could operationalize it, though, I think I would be quite happy taking the other side of it. 100:1 feels way overconfident to me.
Oh, to be clear: I don’t think Zack should be banned from LW. I’d prefer to never interact with him at all, but as noted elsewhere, I’m perfectly happy to stay in my corner and leave him alone over in his.
(I also can’t think of any way to operationalize the bet.)
(That person is more responsible than any other single individual for Eliezer not being around much these days.)
Wait, the only thing I remember Said and Eliezer arguing about was Eliezer’s glowfic. Eliezer dropped out of LW over an argument about how he was writing about tabletop RPG rules in his fanfiction?
Eliezer wasn’t posting much on LW before then, unless I’m misremembering badly, so if (1) you’re right about that being the only substantial argument between Eliezer and Said and (2) Duncan’s claim is specifically that Eliezer avoids LW because of unpleasant interactions he’s had with Said (as opposed to e.g. because he’s observed how Said interacts with others and doesn’t want to risk joining their number) then something doesn’t add up.
fyi I don’t have a belief that Eliezer avoids the site because of Said-in-particular (although I think he does generally avoid the site because it’s unhedonic). I think a couple things maybe got conflated here.
I think if LW’s stance is “we’re going to let this person drive away high-quality writers on the regular because we like the fact that they made GreaterWrong,” that’s worth being explicit about.
(That person is more responsible than any other single individual for Eliezer not being around much these days.)
To be clear, I wasn’t at all saying that. I was saying “this person is clearly not interested only in destructive nitpicking; here’s a substantial high-effort constructive thing they have done”.
Whether their presence on LW is net positive or negative is a completely separate question. (My current opinion is “clear net positive” but of course I could be wrong, and if e.g. it were the case that we could have Said or Eliezer but not both then that would be an argument on the other side—though I really don’t like the idea of banning user A because user B dislikes having them around even if user B is someone we would very much like to be here.)
I agree it’s a high-effort, constructive contribution.
I don’t know how to contrast it with making the business of actually writing LW content far worse than it otherwise would be.
I agree it’s a bad idea to ban user A because user B dislikes having them around. I think it’s a good idea to ban user A because users B, C, D, G, L, P, Q, R, and W, all of whom were valued contributors, all cited user A on their way out the door.
Have those people in fact said “I am leaving LW because I dislike interacting with Said”? Or is it something more like “I am leaving LW because I dislike the over-critical culture, of which Said is the clearest example”? Because if it’s the latter then it could be simultaneously true that (1) lots of people abandoning LW gave Said’s interaction style as a reason for leaving and that (2) banning Said wouldn’t actually do much to help.
(I remark that I have no evidence other than your say-so that lots of people have cited Said in explaining why they left LW, and that people’s explanations of such things are not always an accurate reflection of the actual causes. I am not, for the avoidance of doubt, claiming or even conjecturing that either you or they are/were lying.)
Since we’re talking about it, I have also told the mods that Said is one of three people who are readily top of mind at having a net negative impact on my LW experience. I’d just sort of been dealing with it, and personally I do not favor banning people sitewide for making me feel uncomfortable. But one of the mods reached out to me and specifically asked about factors influencing my experience of LessWrong, and the vibe and some of the specific commenters Duncan’s pointing out here are what I had reported was the main negative factor for me.
One of the facets of your comment here is that it puts the emphasis in a frustrating place: instead of saying “ah, n of 1, it’s good to know Said bugs you, I wonder if anybody else feels the same way,” it says “you’re the only person I’ve ever heard of complaining about Said, although I’m not saying you’re lying,” which is like a super negative way of being “neutral” about what Duncan said.
Like, imagine you said that to your significant other if they told you that one of your mutual friends, Bob, bothered your SO a lot during conversations, and that your SO had heard from other friends that Bob was very bothersome as well and made them not want to come and hang out. Would you say something like “I remark that I have no evidence other than your say-so that lots of our mutual friends have cited Bob in explaining why they don’t come to our board game night...?”
That sort of approach is a great way to alienate people and kill relationships. Why not just come right out and say “I don’t trust you enough to update ~at all on your reports about how other people feel about Bob?” That would at least be honest!
Did you know about the option to ban people from your posts? If yes, may I ask why you didn’t ban Said?
I didn’t know about that until I saw somebody mention it in another comment today. I will probably start doing this.
I did not intend to say or imply “you’re the only person I’ve ever heard of”, nor to be super-negative, and I don’t think I trust Duncan particularly less (or more) than a typical other prominent LW participant. (But I might trust him less specifically when reporting negative things about Said than I would trust him on other matters, and I would likewise trust Said less specifically when reporting negative things about Duncan.)
If anything I said caused anyone else to trust Duncan less, or caused Duncan any distress, then that was very much not an intended effect and I apologize.
I do, rereading what I wrote, see how it could be interpreted as having a much stronger negative subtext than I intended. What I intended was along the following lines: What Duncan is saying in this thread about “more than a dozen” high-quality ex-LWers citing Said as a major reason for their departure is, unlike the already well-known fact that Duncan himself finds Said intensely unpleasant to interact with, new information for me; but I am aware that when people find one another intensely unpleasant to interact with the things they say about one another are not always perfectly accurate, even when everyone involved is aiming to be truthful and fair, and I want to be cautious in how much I adjust my estimate of Said’s net impact on LW in response to something that comes from a single source I can’t check.
(But, in response to your outright accusation of dishonesty at the end: if I said the thing you would prefer me to say, it would be an untruth; my opinion of Duncan is not what you imply it is. My opinion of my wife, even less so, but I assume the point here is the application to the present case.)
That’s fair—I should not have accused you of dishonesty, and I apologize.
I have heard both, the latter maybe 2x as often as the former?
I think that what’s happening right now, and has been happening over the years, is that the message “this is the kind of engagement we do, here” has been sent over and over again, and the mass of users has responded accordingly. I think if that message stopped being sent, or if a strong countersignal was sent, the mass of users would again respond accordingly.
I can’t model Eliezer very well, but I’d argue that Eliezer is responsible for that decision. Whether it’s because he gets more reach elsewhere, less annoying diagreement/pointing out flaws, or what, it’s on him, not on any individual poster. I give him a ton of credit (and Robin Hanson) for starting LessWrong, but I reject any implication that we should limit or prune LessWrong to make it more appealing to Eliezer. Multiplied by a lot if Eliezer isn’t directly asking for some change, and we’re just cargo-culting what it looked like a decade ago.
I chose Eliezer as one example; it would be a mistake to think that my overall point was about Eliezer.
This single individual (who was recently making a big deal about being non-banned/in good standing) is more responsible than any other single individual for driving away more than a dozen high-quality authors; is the most-frequently cited reason, as far as I can tell, when high-quality people are asked to be specific about why they’re not on LessWrong/what makes LessWrong a miserable place to try to share their ideas and co-think.
Hearing that they’re tolerated because they maintain GreaterWrong makes my CFAR instincts tingle and makes me want to teach people about goal factoring.
Fair enough. I bounce off a lot of posts and comments, so I probably do some amount of victim-blaming if someone over-weights criticism, especially if it’s annoying and wrong. I don’t see how one or a few posters can move the site from well-worth posting to negative value, though it certainly could change the sign for a very marginal decision.
Even so, unless the problematic posts and comments are getting downvoted, but somehow still in-your-face enough to drive someone away, it’s hard to see that banning or otherwise authoritatively controlling the poster is justified.
Er, the (somewhat fuzzy) thesis of the post is that there’s a thing where, like, one or two prominent Socrati set the vibe, and then a bunch of other people follow suit; I find that criticism on LW feels more like [that guy] + [100 emulators of his style of engagement] than it used to.
Ultimately, it isn’t about just that one person’s engagement (although as I note above, they seem to have an outsized direct impact) so much as it is about a … prion disease?
As someone who’s posted about 60 or 70 essays on LW over the past eight years, doing so in 2023 involves a lot of bracing, as if I were hyping myself up to grab an electric fence. I straightforwardly expect the experience of posting to involve a net-negative subsequent four days. This was not the case in e.g. the Conor Moreton days, even though one or two of the Conor Moreton essays ended up in net negative territory.
If it’s that rough for me, when I’m a good writer who’s confident in his insights and has a lot of people who are interested-by-default in his thoughts, I imagine it can get a lot harder for the twenty-year-old not-yet-known version of me.
My recollection is that “Conor Moreton” at least once wrote something along the lines of “I find writing for LW stressful because I get a lot of criticism”. Maybe I’m misremembering, and for sure the obvious guess would be that you remember more clearly than I do, but my recollection now of my impression then is that it was nearer than you’re suggesting to how you describe your present experience of posting on LW.
Oh, I’ve definitely found it somewhat stressful all along; I think I’m in the most sensitive quintile if not decile.
Sorry, didn’t mean to imply that the shift was from [zero] to [large number]. More wanted to gesture at a shift from [moderate number] to [large number].
Like, there’s the difference between viscerally expecting one in four essays to result in a negative experience of magnitude 10 lasting for a day or two, and viscerally expecting two in three essays to result in a negative experience of magnitude 30 lasting for four days.
If the “ban commenter” function had not been implemented, I wouldn’t have posted any of my last five or six essays, and would be already gone.
Ah, I see. I don’t really disagree, but I also don’t think LW is unique in this, nor that there is (or can be) a long-lived growing-popularity group that maintains the feel of the early days. This seems like an evolution that’s plagued old-timers of a medium for all time, from SF fandom to pre-internet BBSs and Usenet, to early-internet special-purpose forums, to LW and rationalist-adjecent fora.
I don’t think Duncan’s gesturing at the “eternal September” problem—I think he’s talking about the “toxic low-grade criticism” problem, which is a related but separate issue. A persistent culture of toxic low-grade criticism exacerbates the eternal September problem by allowing impressionable newcomers to become acculturated to the pre-existing toxic dynamic, or to self-select for compatibility with it, making it that much harder to deal with. But you can work on improving the culture while also admitting that the constant influx of newcomers may make it difficult to go as far as you’d like to in terms of creating a specific and consistent set of norms.
I think there are some differences between this and other instances of degradation of quality by growth and entry of less-hardcore newcomers, and a resulting shift in norms that are generally negative in terms of quality. But I think there are a lot more similarities than differences.
I repeat that I did not say and did not mean that anyone is “tolerated because they maintain GreaterWrong”.
I affirm that you did not say it, and believe that you did not mean it.
I would bet $100 to someone’s $1 that, if we could check the other timeline where that person behaved exactly the same way in comments, but had not made GreaterWrong, they would’ve been banned years ago.
(I believe this in part because I’ve heard multiple mods across multiple years talk about wanting to ban that person specifically, and I’ve heard GreaterWrong raised explicitly as cause for hesitation in something like four out of seven such discussions.)
I can’t think of any way to operationalize that bet—other than maybe noting that you have some of the same objections to Zack as to Said, and Zack also has not been banned; but to my mind their styles of interaction on LW are quite different and I can easily imagine someone finding one of them clearly net positive and the other clearly net negative. I guess we could ask the moderators, but they might very reasonably not want to answer that question, and you might fairly reasonably not trust whatever answer they gave.
If we could operationalize it, though, I think I would be quite happy taking the other side of it. 100:1 feels way overconfident to me.
Oh, to be clear: I don’t think Zack should be banned from LW. I’d prefer to never interact with him at all, but as noted elsewhere, I’m perfectly happy to stay in my corner and leave him alone over in his.
(I also can’t think of any way to operationalize the bet.)
Lucky, I have a comment saved to that exact effect, as I have been workshopping a community dynamics post for a long time about a related dynamic:
> I don’t intend to ban you any time soon, because I really value your place in this community—you’re one of the few people to build useful community infrastructure like ReadTheSeqeunces.com and the UI of GreaterWrong.com, and that’s been one of the most salient facts to me throughout all of my thinking on this matter. (Ben Pace, 5 years ago now according to the timestamp)
Wait, the only thing I remember Said and Eliezer arguing about was Eliezer’s glowfic. Eliezer dropped out of LW over an argument about how he was writing about tabletop RPG rules in his fanfiction?
Eliezer wasn’t posting much on LW before then, unless I’m misremembering badly, so if (1) you’re right about that being the only substantial argument between Eliezer and Said and (2) Duncan’s claim is specifically that Eliezer avoids LW because of unpleasant interactions he’s had with Said (as opposed to e.g. because he’s observed how Said interacts with others and doesn’t want to risk joining their number) then something doesn’t add up.
fyi I don’t have a belief that Eliezer avoids the site because of Said-in-particular (although I think he does generally avoid the site because it’s unhedonic). I think a couple things maybe got conflated here.
(For anyone else curious, that thread is here.)