I think the impartiality really helps. The default thing that happens if we threaten moderation action on any specific individual with a long history on LW is that they feel personally persecuted, complain about it publicly, and try to generally rile up a bunch of social momentum to defend against the prosecution, which then produces a lot of distrust and paranoia and stress for everyone involved.
A nice thing about automatic rate limits is that it’s really transparent we are not doing some kind of strategic purging of dissenters or are trying to find the most politically convenient pretense by which to ban someone, which many people tend to be worried about (I think not without reason given the outside view on the realities of human politics). I think for many people it is much less stressful to interact with a deterministic machine than a human who could potentially be pulling some kind of galaxy brained strategic moves at each step.
they probably shouldn’t be here to begin with.
Many people get triggered for a while. LessWrong commenters change in quality. People get caught in some horrible demon-thread where they feel like they have to keep saying things or lose face. Temporary rate-limits do actually catch many of those cases reasonably well.
The basic problem is that it’s not treating the person as a person, like a human moderator actually talking to them and going “hey, we think you’re not helping here, here’s why … in the future could you …” (and then proceeding to a ban if there’s no improvement) would be.
To be clear, the thing I would do instead of a ban in most cases is an intense rate limit. They just have much better properties in terms of not completely banning certain viewpoints from the site for most cases.
I also think you vastly overestimate our ability to give people constructive feedback. New content review and moderation currently already takes up around one full-time equivalent on average. We don’t have time to do much more of that.
And lastly, I also think you just underestimate the internet’s tendencies to desperately freak out if you ever try to ban anyone. Every time we consider banning any long-term contributor, no matter how obviously harmful they seem for the site, we have dozens of people who otherwise leave good comments come out of the wood work, strong-vote downvote anything even remotely adjacent to the discussion that tries to explain the rationale, complain in like 15 different places, threaten to leave the site, threaten to become enemies of LessWrong forever, and all kinds of things. I think some of that instinct is healthy, but I really think you are vastly vastly underestimating the cost associated with banning a long-time contributor.
I think the impartiality really helps. The default thing that happens if we threaten moderation action on any specific individual with a long history on LW is that they feel personally persecuted, complain about it publicly, and try to generally rile up a bunch of social momentum to defend against the prosecution, which then produces a lot of distrust and paranoia and stress for everyone involved.
A nice thing about automatic rate limits is that it’s really transparent we are not doing some kind of strategic purging of dissenters or are trying to find the most politically convenient pretense by which to ban someone, which many people tend to be worried about (I think not without reason given the outside view on the realities of human politics). I think for many people it is much less stressful to interact with a deterministic machine than a human who could potentially be pulling some kind of galaxy brained strategic moves at each step.
Many people get triggered for a while. LessWrong commenters change in quality. People get caught in some horrible demon-thread where they feel like they have to keep saying things or lose face. Temporary rate-limits do actually catch many of those cases reasonably well.
To be clear, the thing I would do instead of a ban in most cases is an intense rate limit. They just have much better properties in terms of not completely banning certain viewpoints from the site for most cases.
I also think you vastly overestimate our ability to give people constructive feedback. New content review and moderation currently already takes up around one full-time equivalent on average. We don’t have time to do much more of that.
And lastly, I also think you just underestimate the internet’s tendencies to desperately freak out if you ever try to ban anyone. Every time we consider banning any long-term contributor, no matter how obviously harmful they seem for the site, we have dozens of people who otherwise leave good comments come out of the wood work, strong-vote downvote anything even remotely adjacent to the discussion that tries to explain the rationale, complain in like 15 different places, threaten to leave the site, threaten to become enemies of LessWrong forever, and all kinds of things. I think some of that instinct is healthy, but I really think you are vastly vastly underestimating the cost associated with banning a long-time contributor.