I think in some sense both making top-level posts and criticism are thankless jobs. What is your motivation to spend a lot of effort to write up your arguments in top-level post form in the first place? I feel like all the things you list as making things unrewarding apply to top-level posts just as much as writing critical comments (especially in as much as you are writing on a topic, or on a forum, where people treat any reasoning error or mistake with grave disdain and threats of social punishment).
If you get rid of people like Said or otherwise discourage low-effort criticism, you’ll just get less criticism not better criticism.
I don’t buy this. I am much more likely to want to comment on LessWrong (and other forums) if I don’t end up needing to deal with comment-sections that follow the patterns outlined in the OP, and I am generally someone who does lots of criticism and writes lots of critical comments. Many other commenters who I think write plenty of critique have reported similar.
Much of LessWrong has a pretty great reward-landscape for critique. I know that if I comment on a post by Steven Byrnes, or Buck, or Ryan Greenblatt or you, or Scott Alexander or many others, with pretty intense critique, that I will overall end up probably learning something, while also having a pretty good shot at correcting the public record on some important mistakes, and also ending up with real status and credibility within our extended ecosystem. Indeed, you personally come to mind as someone who I have come to respect largely as a result of writing good critiques and comments.
This is generally not the case with Said in my experience. It is very rare that I have a good time responding to any of his critiques, or reading the resulting comment threads. It burns an enormous amount of motivation, and by my best judgement of the situation the critiques do not end up particularly important or relevant when I try to evaluate the work many years later with more distance from the local discussion. They aren’t always wrong, but rarely have the structure of making my understanding actually much deeper.
By far the most likely way you end up with less critique is to make commenting on LessWrong feel generally unrewarding, drag everything into matches of attrition, and create an overall highly negative reward landscape for almost any kind of real or detailed contribution (whether top-level post or critique). If you want more critique, I think the goal is not to never punish critique, but to reward good critique. I am pretty happy with a bunch of things we’ve done for that over the years (like the annual review), and I would like to do more.
Top-level posts are not self-limiting (from a status perspective) in the way I described for a critical comment. If you come up with a great new idea, it can become a popular post read and reread by many over the years and you can become known for being its author. But if you come up with a great critical comment that debunks a post, the post will be downvoted and forgotten, and very few people will remember your role in debunking it.
I agree this is largely true for comments (largely by necessity of how comment visibility works)[1]. Indeed one thing I frequently encourage good commenters to do is to try to generalize their comments more and post them as top-level posts.
Dialogues were also another attempt at making it so that critique is less self-limiting, by making it so that a more conversation can happen at the same level as a post. I don’t think that plan succeeded amazingly well (largely because dialogues ended up hard to read, and hard to coordinate between authors), but it is a thing I care a lot about and expect to do more work on.
The popular comments section on the frontpage has also changed this situation a non-trivial amount. It is now the case that if you write a very good critique that causes a post to be downvoted, that this will still result in your comment getting a lot of visibility on the frontpage. Indeed, just this very moment we have a critique by sunwillrise with a bunch more karma than the post it is replying to prominent on the frontpage:
I disagree. Posts seem to have an outsized effect and will often be read a bunch before any solid criticisms appear. Then are spread even given high quality rebuttals… if those ever materialize.
I also think you’re referring to a group of people who write high quality posts typically and handle criticism well, while others don’t handle criticism well. Despite liking many of his posts, Duncan is an example of this.
As for Said specifically, I’ve been annoyed at reading his argumentation a few times, but then also find him saying something obvious and insightful that no one else pointed out anywhere in the comments. Losing that is unfortunate. I don’t think there’s enough “this seems wrong or questionable, why do you believe this?”
Said is definitely more rough than I’d like, but I also do think there’s a hole there that people are hesitant to fill.
So I do agree with Wei that you’ll just get less criticism, especially since I do feel like LessWrong has been growing implicitly less favorable towards quality critiques and more favorable towards vibey critiques.
That is, another dangerous attractor is the Twitter/X attractor, wherein arguments do exist but they matter to the overall discourse less than whether or not someone puts out something that directionally ‘sounds good’. I think this is much more likely than the sneer attractor or the linkedin attractor.
I also think that while the frontpage comments section has been good for surfacing critique, it encourages the “this sounds like the right vibe” substantially. As well as a mentality of reading the comments before the post, encouraging faction mentality.
I think in some sense both making top-level posts and criticism are thankless jobs. What is your motivation to spend a lot of effort to write up your arguments in top-level post form in the first place? I feel like all the things you list as making things unrewarding apply to top-level posts just as much as writing critical comments (especially in as much as you are writing on a topic, or on a forum, where people treat any reasoning error or mistake with grave disdain and threats of social punishment).
I don’t buy this. I am much more likely to want to comment on LessWrong (and other forums) if I don’t end up needing to deal with comment-sections that follow the patterns outlined in the OP, and I am generally someone who does lots of criticism and writes lots of critical comments. Many other commenters who I think write plenty of critique have reported similar.
Much of LessWrong has a pretty great reward-landscape for critique. I know that if I comment on a post by Steven Byrnes, or Buck, or Ryan Greenblatt or you, or Scott Alexander or many others, with pretty intense critique, that I will overall end up probably learning something, while also having a pretty good shot at correcting the public record on some important mistakes, and also ending up with real status and credibility within our extended ecosystem. Indeed, you personally come to mind as someone who I have come to respect largely as a result of writing good critiques and comments.
This is generally not the case with Said in my experience. It is very rare that I have a good time responding to any of his critiques, or reading the resulting comment threads. It burns an enormous amount of motivation, and by my best judgement of the situation the critiques do not end up particularly important or relevant when I try to evaluate the work many years later with more distance from the local discussion. They aren’t always wrong, but rarely have the structure of making my understanding actually much deeper.
By far the most likely way you end up with less critique is to make commenting on LessWrong feel generally unrewarding, drag everything into matches of attrition, and create an overall highly negative reward landscape for almost any kind of real or detailed contribution (whether top-level post or critique). If you want more critique, I think the goal is not to never punish critique, but to reward good critique. I am pretty happy with a bunch of things we’ve done for that over the years (like the annual review), and I would like to do more.
Top-level posts are not self-limiting (from a status perspective) in the way I described for a critical comment. If you come up with a great new idea, it can become a popular post read and reread by many over the years and you can become known for being its author. But if you come up with a great critical comment that debunks a post, the post will be downvoted and forgotten, and very few people will remember your role in debunking it.
I agree this is largely true for comments (largely by necessity of how comment visibility works)[1]. Indeed one thing I frequently encourage good commenters to do is to try to generalize their comments more and post them as top-level posts.
And as far as I can tell this is an enormously successful mechanism for getting highly-upvoted posts on LessWrong. Indeed, I would classify the current second most-upvoted post of all time on LessWrong as a post of this kind: https://www.lesswrong.com/posts/CoZhXrhpQxpy9xw9y/where-i-agree-and-disagree-with-eliezer
Dialogues were also another attempt at making it so that critique is less self-limiting, by making it so that a more conversation can happen at the same level as a post. I don’t think that plan succeeded amazingly well (largely because dialogues ended up hard to read, and hard to coordinate between authors), but it is a thing I care a lot about and expect to do more work on.
The popular comments section on the frontpage has also changed this situation a non-trivial amount. It is now the case that if you write a very good critique that causes a post to be downvoted, that this will still result in your comment getting a lot of visibility on the frontpage. Indeed, just this very moment we have a critique by sunwillrise with a bunch more karma than the post it is replying to prominent on the frontpage:
Though I do think it’s been changing and we’ve made some improvements on this dimension, see my last paragraph
I disagree. Posts seem to have an outsized effect and will often be read a bunch before any solid criticisms appear. Then are spread even given high quality rebuttals… if those ever materialize.
I also think you’re referring to a group of people who write high quality posts typically and handle criticism well, while others don’t handle criticism well. Despite liking many of his posts, Duncan is an example of this.
As for Said specifically, I’ve been annoyed at reading his argumentation a few times, but then also find him saying something obvious and insightful that no one else pointed out anywhere in the comments. Losing that is unfortunate. I don’t think there’s enough “this seems wrong or questionable, why do you believe this?”
Said is definitely more rough than I’d like, but I also do think there’s a hole there that people are hesitant to fill.
So I do agree with Wei that you’ll just get less criticism, especially since I do feel like LessWrong has been growing implicitly less favorable towards quality critiques and more favorable towards vibey critiques. That is, another dangerous attractor is the Twitter/X attractor, wherein arguments do exist but they matter to the overall discourse less than whether or not someone puts out something that directionally ‘sounds good’. I think this is much more likely than the sneer attractor or the linkedin attractor.
I also think that while the frontpage comments section has been good for surfacing critique, it encourages the “this sounds like the right vibe” substantially. As well as a mentality of reading the comments before the post, encouraging faction mentality.