Hire a team of well-paid moderators for a three-month high-effort experiment of responding to every bad comment with a fixed version of what a good comment making the same point would have looked like. Flood the site with training data.
Maybe we can start with a smaller experiment, like a group of (paid or volunteer) moderators do this for just one post? I sometimes wish that someone would point out all the flaws in my comments so I can tell what I can improve on, but I’m not sure if that won’t be so unpleasant that I’d stop wanting to participate (or there would be some other negative consequence). Doing a small experiment seems like a good first step to finding out.
Assuming such experiments go well, however, I’m still worried about possible longer term unintended consequences to having a “high standards” culture. One that I think is fairly likely is that the standards will be selectively/unevenly enforced, against comments/posts that the “standards enforcers” disagree with, making it even more costly to make posts/comments that go against the consensus beliefs around here than it already is. I frequently see such selective enforcement/moderation in other “high standards” spaces, and am worried about the same thing happening here.
I posit that “selective enforcement” is a way that unfairness expresses itself with high standards, but that the overall level of unfairness is approximately constant, i.e. raising standards is just good. Reduces some unfairnesses, increases others, but meanwhile you actually have clear communication and good discourse.
I can’t think of an operationalized experiment yet, but if someone comes up with one, I expect I would bet in the ballpark of 5:1 odds (my $5 to their $1) that an actual increase in standards does not cause an increase in unfairness. I’d bet at 2:1 odds that it results in a detectable decrease of it.
Maybe we can start with a smaller experiment, like a group of (paid or volunteer) moderators do this for just one post? I sometimes wish that someone would point out all the flaws in my comments so I can tell what I can improve on, but I’m not sure if that won’t be so unpleasant that I’d stop wanting to participate (or there would be some other negative consequence). Doing a small experiment seems like a good first step to finding out.
Assuming such experiments go well, however, I’m still worried about possible longer term unintended consequences to having a “high standards” culture. One that I think is fairly likely is that the standards will be selectively/unevenly enforced, against comments/posts that the “standards enforcers” disagree with, making it even more costly to make posts/comments that go against the consensus beliefs around here than it already is. I frequently see such selective enforcement/moderation in other “high standards” spaces, and am worried about the same thing happening here.
I posit that “selective enforcement” is a way that unfairness expresses itself with high standards, but that the overall level of unfairness is approximately constant, i.e. raising standards is just good. Reduces some unfairnesses, increases others, but meanwhile you actually have clear communication and good discourse.
I can’t think of an operationalized experiment yet, but if someone comes up with one, I expect I would bet in the ballpark of 5:1 odds (my $5 to their $1) that an actual increase in standards does not cause an increase in unfairness. I’d bet at 2:1 odds that it results in a detectable decrease of it.