This post from one year ago discussed a similar problem. Suggestions for returning LessWrong to a position of centrality included:
Allowing and encouraging more links posts and the discussion of them, on topics of interest to rationalists, such as machine intelligence and transhumanism, as Hacker News does now.
Allow and encourage posts on more political topics in Discussion, but probably not Main. Dangers here could be mitigated by banning discussion of current politicians, governments, and issues, or banning discussion on specific topics. I personally think this wouldn’t work because moderation and banning would need to be strictly enforced, assuming the user base doesn’t naturally follow the ban. Considering LessWrong has a history of fatigue among moderators, doing something like this which may effectively lower the sanity waterline here (for a temporary period) might ruin it more.
Get rid of Open Threads and create a new norm that a discussion post as short as a couple sentences is acceptable.
I think creating new norms is a collective action problem. For whatever reason(s), maybe mostly fear of downvoting, thinking what would be posted isn’t “appropriate enough” for LessWrong, and indifference, no single individual(s) are incentivized to take risks in posting more and more novel content. Or something like that. Generating a new norm of encouraging others to give more upvotes to posts which are on the edge of LessWrong’s Overtown windown, or “appropriate content” criterion, may again be another collective action problem. Also, that seems risky.
I think some actions were provided in the previous threads, they just weren’t made actionable. John Maxwell made some observations, which could be turned into actions.
Users on Less Wrong could downvote less. I personally use both upvoting and downvoting sparingly on LessWrong, unless a comment or post really stands out as great or awful. This seems like a thing we can’t get a whole community to do.
Instead of merely upvoting a post or comment, leave a comment like “great post” as a comment, or whatever positive feedback, as this is more a motivator. This in turn may incentivize people to post more often over the long-term.
I bolded the last one because it seems actionable. I think another bottleneck is many suggestions to fix this sort of problem revolve around changing site mechanics, level of moderation, and encouragement from popular figures for a change in culture and/or behavior. Nobody seems to think we can fix all these things by contacting Trike Apps (who maintains and bulids LessWrong), and asking them to change the site mechanics. I don’t think that would work, anyway. I think if one want to change how LessWrong works, one needs to contact the moderators of the site, its real owners, or whatnot, and bring proposals directly to them.
Users on Less Wrong could downvote less. I personally use both upvoting and downvoting sparingly on LessWrong, unless a comment or post really stands out as great or awful. This seems like a thing we can’t get a whole community to do.
Users on Less Wrong could downvote less. I personally use both upvoting and downvoting sparingly on LessWrong, unless a comment or post really stands out as great or awful. This seems like a thing we can’t get a whole community to do.
LessWrong doesn’t have to have a uniform standard for upvoting and downvoting. For example, upvoting doesn’t differentiate between “interesting” and “correct”. As I have said before, someone might even feel compelled to downvote an interesting speculative idea for the fear that other readers might mistake the lack of downvotes for the speculative idea being thought of as mostly correct by other LWers.
To solve it, we could encourage people to use tags or informal tags, such as putting a tag inside square brackets in the title of the post, to clearly indicate, for example, how certain a poster is about their idea. For example, a post could have a title like this: “A statement [Epistemic state:possible][Topic: something]” or “A statement [Epistemic state: a speculation] [Topic: something]”. I think that it is likely that readers would treat different tags differently, for example, something might still be a curious idea, even if it is unpolished and has flaws, thus, if it was tagged properly (e.g. “unlikely” or “speculation”, “a spherical cow style model”), it would not merit a downvote, because it would be clear that that post is not going to be mistaken by other readers for a one that purports to be accurate and certain. On the other hand, tags “certain” or “highly likely” would be useful for readers who prefer not having to wade through various speculations and want to read more reliable posts. Of course, if someone tried to pass off their pet idea as a certain fact, they could be downvoted.
Tags work pretty well on reddit, and folks already use the Link: tag here. However I think that having too many tags or too complex of a tag system could also just contribute to the low volume problem.
In particular I like the idea of tags for “fiction” and possibly for “speculative”. Although if we are to be completely honest with ourselves, the sequences contains many posts that should be tagged fiction or speculative. The fiction ones are obvious, but its not always obvious which ones are speculative.
Well, a tag system doesn’t have to be strict or predetermined in advance. I think that if posters were allowed to create new tags to express their intent and their certainty about their posts, suitable and expressive tags would likely prevail and become common whereas unexpressive tags would be used only a few times and then fall out of use. People would pick up usage of various tags from observation.
A similar thing happened on Imgur when tags were introduced there; it seems to have worked fine on their end, though obviously it’s a much larger community. I am not certain how the differences there, much less the differences in culture, would affect the adoption of tags as a means of identifying the nature in which a post is intended.
Additionally, it’s worth considering the use of the same system for commenting on posts; there are comment threads out there that I could see having used such a tagging system. The question, then, is whether that would create too much clutter.
This post from one year ago discussed a similar problem. Suggestions for returning LessWrong to a position of centrality included:
Allowing and encouraging more links posts and the discussion of them, on topics of interest to rationalists, such as machine intelligence and transhumanism, as Hacker News does now.
Allow and encourage posts on more political topics in Discussion, but probably not Main. Dangers here could be mitigated by banning discussion of current politicians, governments, and issues, or banning discussion on specific topics. I personally think this wouldn’t work because moderation and banning would need to be strictly enforced, assuming the user base doesn’t naturally follow the ban. Considering LessWrong has a history of fatigue among moderators, doing something like this which may effectively lower the sanity waterline here (for a temporary period) might ruin it more.
Get rid of Open Threads and create a new norm that a discussion post as short as a couple sentences is acceptable.
I think creating new norms is a collective action problem. For whatever reason(s), maybe mostly fear of downvoting, thinking what would be posted isn’t “appropriate enough” for LessWrong, and indifference, no single individual(s) are incentivized to take risks in posting more and more novel content. Or something like that. Generating a new norm of encouraging others to give more upvotes to posts which are on the edge of LessWrong’s Overtown windown, or “appropriate content” criterion, may again be another collective action problem. Also, that seems risky.
I think some actions were provided in the previous threads, they just weren’t made actionable. John Maxwell made some observations, which could be turned into actions.
Users on Less Wrong could downvote less. I personally use both upvoting and downvoting sparingly on LessWrong, unless a comment or post really stands out as great or awful. This seems like a thing we can’t get a whole community to do.
Instead of merely upvoting a post or comment, leave a comment like “great post” as a comment, or whatever positive feedback, as this is more a motivator. This in turn may incentivize people to post more often over the long-term.
I bolded the last one because it seems actionable. I think another bottleneck is many suggestions to fix this sort of problem revolve around changing site mechanics, level of moderation, and encouragement from popular figures for a change in culture and/or behavior. Nobody seems to think we can fix all these things by contacting Trike Apps (who maintains and bulids LessWrong), and asking them to change the site mechanics. I don’t think that would work, anyway. I think if one want to change how LessWrong works, one needs to contact the moderators of the site, its real owners, or whatnot, and bring proposals directly to them.
LessWrong doesn’t have to have a uniform standard for upvoting and downvoting. For example, upvoting doesn’t differentiate between “interesting” and “correct”. As I have said before, someone might even feel compelled to downvote an interesting speculative idea for the fear that other readers might mistake the lack of downvotes for the speculative idea being thought of as mostly correct by other LWers.
To solve it, we could encourage people to use tags or informal tags, such as putting a tag inside square brackets in the title of the post, to clearly indicate, for example, how certain a poster is about their idea. For example, a post could have a title like this: “A statement [Epistemic state:possible][Topic: something]” or “A statement [Epistemic state: a speculation] [Topic: something]”. I think that it is likely that readers would treat different tags differently, for example, something might still be a curious idea, even if it is unpolished and has flaws, thus, if it was tagged properly (e.g. “unlikely” or “speculation”, “a spherical cow style model”), it would not merit a downvote, because it would be clear that that post is not going to be mistaken by other readers for a one that purports to be accurate and certain. On the other hand, tags “certain” or “highly likely” would be useful for readers who prefer not having to wade through various speculations and want to read more reliable posts. Of course, if someone tried to pass off their pet idea as a certain fact, they could be downvoted.
Tags work pretty well on reddit, and folks already use the Link: tag here. However I think that having too many tags or too complex of a tag system could also just contribute to the low volume problem.
In particular I like the idea of tags for “fiction” and possibly for “speculative”. Although if we are to be completely honest with ourselves, the sequences contains many posts that should be tagged fiction or speculative. The fiction ones are obvious, but its not always obvious which ones are speculative.
Well, a tag system doesn’t have to be strict or predetermined in advance. I think that if posters were allowed to create new tags to express their intent and their certainty about their posts, suitable and expressive tags would likely prevail and become common whereas unexpressive tags would be used only a few times and then fall out of use. People would pick up usage of various tags from observation.
A similar thing happened on Imgur when tags were introduced there; it seems to have worked fine on their end, though obviously it’s a much larger community. I am not certain how the differences there, much less the differences in culture, would affect the adoption of tags as a means of identifying the nature in which a post is intended.
Additionally, it’s worth considering the use of the same system for commenting on posts; there are comment threads out there that I could see having used such a tagging system. The question, then, is whether that would create too much clutter.
Good point—agreed. I will try to remember to come up with an appropriate tag for my next post.