False!—I would miss him. I agree that comments like the grandparent are not great, but Ilya is a bona fidesubject matter expert (his Ph.D. advisor was Judea Pearl), so when he contributes references or explanations, that’s really valuable. Why escalate to banning a user when individual bad comments can be safely downvoted to invisibility?
I’m aware of Ilya’s subject matter expertise (as well as his connection to Pearl), yes. My decision to avoid mentioning said expertise was motivated in part by precisely a curiosity as to whether it would be brought up as a relevant factor in replies (for the record: I predicted that it would), and indeed it seems my prediction was borne out.
Now, recognizing that you (Zack) obviously don’t speak for the moderation team, I’d nonetheless like to ask you (and any other bystanders or—indeed—moderators who might happen to be reading this): what role do you think things like subject matter expertise ought to play in deciding whether to evict a user from an online forum?
Note 1: Despite the somewhat snide-sounding tone of the above, I do intend my question as a genuine, non-rhetorical question; I am open to the answer being something other than “none whatsoever”. I do think, however, that whatever the correct norm is here, it would benefit from being made common knowledge, even if that involves making somewhat ugly-sounding statements like “The LW moderation team will treat you differently if you are a subject matter expert compared to if you are not.”
Note 2: I also don’t mean to imply that, if Ilya were not a subject matter expert who occasionally contributes comments of real value, the comments he made here would in and of themselves be ban-worthy. This is the other part of the reason why I avoided talking about Ilya’s credentials until it was brought up by someone else: I’m entirely open to the answer to my original question (“what does it take to get a LW mod to ban somebody”) being something like, “We have a bunch of red lines, which have nothing to do with credentials, and also which Ilya’s comments entirely fail to cross, such that our decision not to ban (or take any other administrative action against) him would hold even if Ilya was not a subject matter expert.”
Note 3: Having said that, suppose it is the case that whether a commenter is ban-worthy is dependent, not purely on whether they cross some set of red lines, but on some kind of cost-benefit calculation. Then to what degree do a commenter’s non-constructive (or outright destructive) comments have to outnumber their productive contributions before the scales are considered to have tipped? Looking at Ilya’s recent comment history, the ratio of “useless” comments to “useful” comments seems quite heavily skewed in favor of “useless”, and that’s without counting the slew of comments he’s left on this post. Is the argument here that “useless” comments are in some sense “okay”, because they can all be “downvoted into invisibility”, such that the implied ratio is actually infinite, i.e. someone can make as many terrible comments as they want, as long as they’ve made at least one positive contribution in the past? Or is it merely that the ratio is some really large number? Or something else entirely?
Note 4: Perhaps the ratio isn’t the right way to think about it at all. Perhaps the idea is simply that banning a user from LW is a really serious thing to do (which somewhat lines up with Zack calling it an “escalation”), and each instance of a ban requires an in-depth discussion (cf. the decision to ban Brent’s account), such that the effort involved isn’t worth it unless the harms are really huge and obviously visible?
It’s not clear to me what the right way is to think about this. What I do know is that the impulse which triggered my initial comment was a thought along the lines of “If this was my personal blog or Facebook wall, I would consider multiple comments as bad as Ilya’s to be a ban-worthy offense.” To the extent that LW moderation norms differ from those of a personal blog or Facebook wall (and again, I am entirely open to the idea that they do differ, for sensible, important reasons!), I think it’s useful to have an open, transparent discussion of how, where, and why.
If someone’s producing lots of great ideas and posts on the site, but they’re sometimes aggressive or rude or spiky, then I will put in more effort to give them feedback and give them a lot more rope than if (on the other end of the spectrum) they’re a first time poster. If Ilya’s comment was an account’s first comment, I’d ban the account and delete the comment. That sort of new user growth is bad for the site.
Responding to this situation in particular: I had the perception that Ilya had in the past contributed substantially to the site (in large part on the topic of causal modeling), and have (in my head) been giving him leeway for that. Also I met him once at LessWrong thing in Cambridge UK when I was 16 and he was friendly, and that gave me a sense he would be open to conversation and feedback if it came. That said, looking over his past comments was much more heat to light than I expected (lots more random unpleasant and rude comments and way fewer substantive contributions), so I am a bit surprised.
Personally, I lean laissez faire on moderation: I consider banning a non-spam user from the whole website to be quite serious, and that the karma system makes a decently large (but definitely not infinite!) ratio of useless-to-useful comments acceptable. Separately from that, I admit that applying different rules to celebrities would be pretty unprincipled, but I fear that my gut feeling actually is leaning that way.
False!—I would miss him. I agree that comments like the grandparent are not great, but Ilya is a bona fide subject matter expert (his Ph.D. advisor was Judea Pearl), so when he contributes references or explanations, that’s really valuable. Why escalate to banning a user when individual bad comments can be safely downvoted to invisibility?
I’m aware of Ilya’s subject matter expertise (as well as his connection to Pearl), yes. My decision to avoid mentioning said expertise was motivated in part by precisely a curiosity as to whether it would be brought up as a relevant factor in replies (for the record: I predicted that it would), and indeed it seems my prediction was borne out.
Now, recognizing that you (Zack) obviously don’t speak for the moderation team, I’d nonetheless like to ask you (and any other bystanders or—indeed—moderators who might happen to be reading this): what role do you think things like subject matter expertise ought to play in deciding whether to evict a user from an online forum?
Note 1: Despite the somewhat snide-sounding tone of the above, I do intend my question as a genuine, non-rhetorical question; I am open to the answer being something other than “none whatsoever”. I do think, however, that whatever the correct norm is here, it would benefit from being made common knowledge, even if that involves making somewhat ugly-sounding statements like “The LW moderation team will treat you differently if you are a subject matter expert compared to if you are not.”
Note 2: I also don’t mean to imply that, if Ilya were not a subject matter expert who occasionally contributes comments of real value, the comments he made here would in and of themselves be ban-worthy. This is the other part of the reason why I avoided talking about Ilya’s credentials until it was brought up by someone else: I’m entirely open to the answer to my original question (“what does it take to get a LW mod to ban somebody”) being something like, “We have a bunch of red lines, which have nothing to do with credentials, and also which Ilya’s comments entirely fail to cross, such that our decision not to ban (or take any other administrative action against) him would hold even if Ilya was not a subject matter expert.”
Note 3: Having said that, suppose it is the case that whether a commenter is ban-worthy is dependent, not purely on whether they cross some set of red lines, but on some kind of cost-benefit calculation. Then to what degree do a commenter’s non-constructive (or outright destructive) comments have to outnumber their productive contributions before the scales are considered to have tipped? Looking at Ilya’s recent comment history, the ratio of “useless” comments to “useful” comments seems quite heavily skewed in favor of “useless”, and that’s without counting the slew of comments he’s left on this post. Is the argument here that “useless” comments are in some sense “okay”, because they can all be “downvoted into invisibility”, such that the implied ratio is actually infinite, i.e. someone can make as many terrible comments as they want, as long as they’ve made at least one positive contribution in the past? Or is it merely that the ratio is some really large number? Or something else entirely?
Note 4: Perhaps the ratio isn’t the right way to think about it at all. Perhaps the idea is simply that banning a user from LW is a really serious thing to do (which somewhat lines up with Zack calling it an “escalation”), and each instance of a ban requires an in-depth discussion (cf. the decision to ban Brent’s account), such that the effort involved isn’t worth it unless the harms are really huge and obviously visible?
It’s not clear to me what the right way is to think about this. What I do know is that the impulse which triggered my initial comment was a thought along the lines of “If this was my personal blog or Facebook wall, I would consider multiple comments as bad as Ilya’s to be a ban-worthy offense.” To the extent that LW moderation norms differ from those of a personal blog or Facebook wall (and again, I am entirely open to the idea that they do differ, for sensible, important reasons!), I think it’s useful to have an open, transparent discussion of how, where, and why.
I do a cost-benefit calculation.
If someone’s producing lots of great ideas and posts on the site, but they’re sometimes aggressive or rude or spiky, then I will put in more effort to give them feedback and give them a lot more rope than if (on the other end of the spectrum) they’re a first time poster. If Ilya’s comment was an account’s first comment, I’d ban the account and delete the comment. That sort of new user growth is bad for the site.
Responding to this situation in particular: I had the perception that Ilya had in the past contributed substantially to the site (in large part on the topic of causal modeling), and have (in my head) been giving him leeway for that. Also I met him once at LessWrong thing in Cambridge UK when I was 16 and he was friendly, and that gave me a sense he would be open to conversation and feedback if it came. That said, looking over his past comments was much more heat to light than I expected (lots more random unpleasant and rude comments and way fewer substantive contributions), so I am a bit surprised.
I’ve now given Ilya a warning upthread.
Thanks for replying; strong-upvote for displaying transparency.
Personally, I lean laissez faire on moderation: I consider banning a non-spam user from the whole website to be quite serious, and that the karma system makes a decently large (but definitely not infinite!) ratio of useless-to-useful comments acceptable. Separately from that, I admit that applying different rules to celebrities would be pretty unprincipled, but I fear that my gut feeling actually is leaning that way.