I have been thinking more about this comment by clone of saturn, which seems to be even more insightful and relevant than I noticed at first:
I don’t think high-effort posts are more likely to contain muddled thinking, but I do think readers are less likely to notice muddled thinking when it appears in high-effort posts, so suppressing criticism of high-effort posts is especially dangerous.
I have long noticed (both on Less Wrong and elsewhere) that people often can’t tell the difference between something that looks very impressive (or high-effort, etc.), and something that is actually good.
This is a huge problem.
As has been alluded to elsethread, people are liable to conflate “high-effort” with “high-value”. But that’s not the whole of the problem. In fact all of the following things often get conflated—despite the fact that they’re very different:
A post that is correct and/or valuable and/or good, etc.
A post that took a great deal of effort to produce.
A post that looks like it took a great deal of effort to produce.
A post that is long.
A post that contains a great deal of information.
A post which contains a very high density of information.
That Dijkstra quote I cited in another comment—it applies to posts at least as much as to comments. It is a grave mistake, to reward mere visible quantity of effort. If a post is long, and has many diagrams, and other clear markers of work put in—what use is that?
I have made this longer than usual because I have not had time to make it shorter.
Consider a very short post (or comment), which—briefly, elegantly, with a minimum of words—expresses some transformative idea, or makes some stunningly incisive point. Forget, for now, the question of its quality, and consider instead: how much effort went into writing it? Do you tally up only the keystrokes? Or do you count also the years of thought and experience and work that allowed the writer to come up with this idea, and this sequence of words to express it? Do you count the knowledge of a lifetime?
If I write six thousand words on some point, and it is rambling, verbose, meticulously cited, illustrated, etc., but it is all nonsense, and this fact is obscured by the post’s sheer weight… why is it a good thing, that I should write this, and post it? Especially given that readers are less likely to notice mistakes, confusions, etc., than they would in a shorter post? (And those mistakes and confusions take considerably more effort to unravel—which often means they simply aren’t unravelled.)
If Hegel and Gettier both posted their work on Less Wrong, which of them should we reward more?
I agree that length should be considered a cost not a benefit. (My own posts seem to average around 2 minutes of read time.) When the situation calls for it, I prefer to signal effort (which I think is often useful for passing people’s initial filters and getting their attention) by doing things like being meticulous about syntax/spelling, using more formal/academic language, and not by increasing length.
I’m curious if you see a serious problem with people considering length as a benefit on LW, or confusing effort with quality (beyond using effort as an initial filter). For example, can you cite a post that is not very good but has apparently been highly upvoted because it’s long or otherwise seems high-effort?
Yes, I could cite many such posts. But I’m not entirely sure what the purpose would be, except to publicly antagonize popular contributors to Less Wrong. And the incentives for others to publicly disagree with my evaluation would obviously be great (while the incentives to agree, essentially nonexistent).
That said, your request for examples is, obviously, quite reasonable in the general sense (I would be quite hypocritical to claim otherwise). I’m willing to cite examples via private message, if you like. A question first, if I may: do you, yourself, think that examples of this sort of thing exist on Less Wrong? Or, do you think there are no such examples (or perhaps that they are very rare, at best)?
while the incentives to agree, essentially nonexistent
To me, an obvious incentive is to reach accurate consensus about the problem so that (if real) it can be solved and so upvotes can better align with quality.
I’m willing to cite examples via private message, if you like.
Sure, please do. I think if the problem is real then eventually we’ll probably have to discuss such examples in public but we can certainly start in private if you prefer.
A question first, if I may: do you, yourself, think that examples of this sort of thing exist on Less Wrong? Or, do you think there are no such examples (or perhaps that they are very rare, at best)?
No salient examples come readily to mind, but I think there could be any number of explanations for that aside from such examples being very rare or nonexistent.
ETA: I have not yet received any examples via PM, as of Aug 6.
I have been thinking more about this comment by clone of saturn, which seems to be even more insightful and relevant than I noticed at first:
I have long noticed (both on Less Wrong and elsewhere) that people often can’t tell the difference between something that looks very impressive (or high-effort, etc.), and something that is actually good.
This is a huge problem.
As has been alluded to elsethread, people are liable to conflate “high-effort” with “high-value”. But that’s not the whole of the problem. In fact all of the following things often get conflated—despite the fact that they’re very different:
A post that is correct and/or valuable and/or good, etc.
A post that took a great deal of effort to produce.
A post that looks like it took a great deal of effort to produce.
A post that is long.
A post that contains a great deal of information.
A post which contains a very high density of information.
That Dijkstra quote I cited in another comment—it applies to posts at least as much as to comments. It is a grave mistake, to reward mere visible quantity of effort. If a post is long, and has many diagrams, and other clear markers of work put in—what use is that?
Pascal once wrote:
Consider a very short post (or comment), which—briefly, elegantly, with a minimum of words—expresses some transformative idea, or makes some stunningly incisive point. Forget, for now, the question of its quality, and consider instead: how much effort went into writing it? Do you tally up only the keystrokes? Or do you count also the years of thought and experience and work that allowed the writer to come up with this idea, and this sequence of words to express it? Do you count the knowledge of a lifetime?
If I write six thousand words on some point, and it is rambling, verbose, meticulously cited, illustrated, etc., but it is all nonsense, and this fact is obscured by the post’s sheer weight… why is it a good thing, that I should write this, and post it? Especially given that readers are less likely to notice mistakes, confusions, etc., than they would in a shorter post? (And those mistakes and confusions take considerably more effort to unravel—which often means they simply aren’t unravelled.)
If Hegel and Gettier both posted their work on Less Wrong, which of them should we reward more?
I agree that length should be considered a cost not a benefit. (My own posts seem to average around 2 minutes of read time.) When the situation calls for it, I prefer to signal effort (which I think is often useful for passing people’s initial filters and getting their attention) by doing things like being meticulous about syntax/spelling, using more formal/academic language, and not by increasing length.
I’m curious if you see a serious problem with people considering length as a benefit on LW, or confusing effort with quality (beyond using effort as an initial filter). For example, can you cite a post that is not very good but has apparently been highly upvoted because it’s long or otherwise seems high-effort?
Yes, I could cite many such posts. But I’m not entirely sure what the purpose would be, except to publicly antagonize popular contributors to Less Wrong. And the incentives for others to publicly disagree with my evaluation would obviously be great (while the incentives to agree, essentially nonexistent).
That said, your request for examples is, obviously, quite reasonable in the general sense (I would be quite hypocritical to claim otherwise). I’m willing to cite examples via private message, if you like. A question first, if I may: do you, yourself, think that examples of this sort of thing exist on Less Wrong? Or, do you think there are no such examples (or perhaps that they are very rare, at best)?
To me, an obvious incentive is to reach accurate consensus about the problem so that (if real) it can be solved and so upvotes can better align with quality.
Sure, please do. I think if the problem is real then eventually we’ll probably have to discuss such examples in public but we can certainly start in private if you prefer.
No salient examples come readily to mind, but I think there could be any number of explanations for that aside from such examples being very rare or nonexistent.
ETA: I have not yet received any examples via PM, as of Aug 6.