Less Wrong is a community blog devoted to refining the art of human rationality. Please visit our About page for more information.

Morendil comments on On the importance of Less Wrong, or another single conversational locus - Less Wrong

84 Post author: AnnaSalamon 27 November 2016 05:13PM

You are viewing a comment permalink. View the original post to see all comments and the full post content.

Comments (362)

You are viewing a single comment's thread.

Comment author: Morendil 27 November 2016 09:49:58AM 3 points [-]

We have lately ceased to have a "single conversation" in this way.

Can we hope to address this without understanding why it happened?

What are y'all's theories of why it happened?

Comment author: sarahconstantin 27 November 2016 10:27:46AM 18 points [-]

1: the general move of the internet away from blogs and forums and towards social media.

In particular, there seems to be a mental move that people make, that I've seen people write about quite frequently, of wanting to avoid the more "official"-seeming forms of online discussion, and towards more informal places. From blogging to FB, from FB to Tumblr and Twitter, and thence to Snapchat and other stuff I'm too old for. Basically, people say that they're intimidated to talk on the more official, public channels. I get a sense of people feeling hassled by unfriendly commenters, and also a sense of something like "kids wanting to hang out where the grownups aren't", except that the "kids" here are often adults themselves. A sense that you'll be judged if you do your honest best to write what you actually believe, in front of people who might critique it, and so that it's safer to do something that leaves you less exposed, like sharing memes.

I think the "hide, go in the darkness, do things that you can't do by daylight" Dionysian kind of impulse is not totally irrational (a lot of people do have judgmental employers or families) but it's really counterproductive to discourse, which is inherently an Apollonian, daylight kind of activity.

Comment author: steven0461 27 November 2016 09:31:15PM 3 points [-]

To me, the major advantage of social media is they make it easy to choose whose content to read. A version of LW where only my 25 favorite posters were visible would be exciting where the current version is boring. (I don't think that's a feasible change, but maybe it's another data point that helps people understand the problem.)

Comment author: Evan_Gaensbauer 28 November 2016 06:10:03AM 7 points [-]

You can already do this. If you click on a user's profile, there will be a little box in the top right corner. Click on the button that says "add to friends" there. When you "friend" someone on LessWrong, it just means you follow them. If you go to www.lesswrong.com/r/friends, there's a feed with submissions from only the other users you're following.

Comment author: steven0461 28 November 2016 06:17:38AM 3 points [-]

Cool, thanks, but it looks like that's posts only, not comments.

Comment author: AnnaSalamon 27 November 2016 10:23:15PM 2 points [-]

Ignoring the feasibility question for a minute, I'm confused about whether it would be desirable (if feasible). There are some obvious advantages to making it easy for people to choose what to read. And as a general heuristic, making it easy for people to do things they want to do seems usually good/cooperative. But there are also strong advantages to having common knowledge of particular content/arguments (a cannon; a single thread of assumed "yes that's okay to assume and build on"); and making user displays individual (as e.g. Facebook does) cuts heavily against that.

(I realize you weren't talking about what was all-things-considered desirable, only about what feels exciting/boring.)

Comment author: steven0461 27 November 2016 10:39:38PM *  1 point [-]

That seems an important set of concerns, but also I'm not sure how much people are letting lack of canonicity bother them in choosing what to cite and reply to, and popular content will become canon through other mechanisms than the front page, and the more canon there exists, the harder it will be to take it as common knowledge. User-picked content is to some extent also compatible with canon, e.g. through social pressure to read a general "best of" feed. (Just to be clear, though, I don't think this is probably the way we should go / the best use of resources.)

Comment author: Morendil 27 November 2016 10:38:52AM 2 points [-]

Yes, and this would be a general trend - affecting all community blogs to some extent. I was looking for an explanation for the downfall of LessWrong specifically, but I suppose it's also interesting to consider general trends.

Would you say that LessWrong is particularly prone to this effect, and if so because of what properties?

Comment author: sarahconstantin 27 November 2016 10:52:41AM 25 points [-]

Specifically, I think that LW declined from its peak by losing its top bloggers to new projects. Eliezer went to do AI research full-time at MIRI, Anna started running CFAR, various others started to work on those two organizations or others (I went to work at MetaMed). There was a sudden exodus of talent, which reduced posting frequency, and took the wind out of the sails.

One trend I dislike is that highly competent people invariably stop hanging out with the less-high-status, less-accomplished, often younger, members of their group. VIPs have a strong temptation to retreat to a "VIP island" -- which leaves everyone else short of role models and stars, and ultimately kills communities. (I'm genuinely not accusing anybody of nefarious behavior, I'm just noting a normal human pattern.) Like -- obviously it's not fair to reward competence with extra burdens, I'm not that much of a collectivist. But I think that potentially human group dynamics won't work without something like "community-spiritedness" -- there are benefits to having a community of hundreds or thousands, for instance, that you cannot accrue if you only give your time and attention to your ten best friends.

Comment author: Vaniver 27 November 2016 05:31:57PM 14 points [-]

But I think that potentially human group dynamics won't work without something like "community-spiritedness" -- there are benefits to having a community of hundreds or thousands, for instance, that you cannot accrue if you only give your time and attention to your ten best friends.

As for why this is a problem for LW specifically, I would probably point at age. The full explanation is too long for this comment, and so may become a post, but the basic idea is that 'career consolidation' is a developmental task that comes before 'generativity', or focusing mostly on shepherding the next generation, which comes before 'guardianship', or focusing mostly on preserving the important pieces of the past.

The community seems to have mostly retracted because people took the correct step of focusing on the next stage of their development, but because there hadn't been enough people who had finished previous stages of their development, we didn't have enough guardians. We may be able to build more directly, but it might only work the long way.

Comment author: Alexei 27 November 2016 06:24:09PM 7 points [-]

Sounds interesting. I'd love to read the post.

Comment author: kechpaja 27 November 2016 11:48:15AM 8 points [-]

To expand on what sarahconstantin said, there's a lot more this community could be doing to neutralize status differences. I personally find it extremely intimidating and alienating that some community members are elevated to near godlike status (to the point where, at times, I simply cannot read i.e. SSC or anything by Eliezer — I'm very, very celebrity-averse).

I've often fantasized about a LW-like community blog that was entirely anonymous (or nearly so), so that ideas could be considered without being influenced by people's perceptions of their originators (if we could solve the moderation/trolling problem, that is, to prevent it from becoming just another 4chan). A step in the right direction that might be a bit easier to implement would be to revamp the karma system so that the number of points conferred by each up or down vote was inversely proportional to the number of points that the author of the post/comment in question had already accrued.

The thing is, in the absence of something like what I just described, I'm skeptical that it would be possible to prevent the conversation from quickly becoming centered around a few VIPs, with everyone else limited to commenting on those individuals' posts or interacting with their own small circles of friends.

Comment author: Morendil 27 November 2016 11:11:26AM 5 points [-]

There was a sudden exodus of talent, which reduced posting frequency, and took the wind out of the sails.

I'd be wary of post hoc ergo propter hoc in this context. You might also have expected that by leaving for other projects these posters would create a vacuum for others to fill. It could be worth looking at why that didn't happen.

Comment author: Kaj_Sotala 27 November 2016 05:50:24PM 12 points [-]

One interesting thing is that at one point post-Eliezer, there were two "rising stars" on LW who were regularly producing lots of fascinating content: lukeprog and So8res. Both stopped regularly posting here some time after they were recruited by MIRI and their priorities shifted.

Comment author: gworley 27 November 2016 09:49:13PM 1 point [-]

This is why I very much like Medium. I think of it as Twitter for people who want to write/read long things rather than short things. It's also much nicer than Twitter in my experience.

Comment author: SatvikBeri 27 November 2016 09:59:52AM 4 points [-]

My theory is that the main things that matter are content and enforcement of strong intellectual norms, and both degraded around the time a few major high-status members of the community mostly stopped posting (e.g. Eliezer and Yvain.)

The problem with lack of content is obvious, the problem with lack of enforcement is that most discussions are not very good, and it takes a significant amount of feedback to make them better. But it's hard for people to get away with giving subtle criticism unless they're already a high-status member of a community, and upvotes/downvotes are just not sufficiently granular.

Comment author: Morendil 27 November 2016 10:33:22AM 8 points [-]

This feels like a good start but one that needs significant improvement too.

For instance, I'm wondering how much of the situation Anna laments is a result of LW lacking an explicit editorial policy. I for one never quite felt sure what was or wasn't relevant for LW - what had a shot at being promoted - and the few posts I wrote here had a tentative aspect to them because of this. I can't yet articulate why I stopped posting, but it may have had something to do with my writing a bunch of substantive posts that were never promoted to Main.

If you look at the home page only (recent articles in Main) you could draw the inference that the main topics on LessWrong are MIRI, CFAR, FHI, "the LessWrong community", with a side dish of AI safety and startup founder psychology. This doesn't feel aligned with "refining the art of human rationality", it makes LessWrong feel like more of a corporate blog.

Comment author: John_Maxwell_IV 28 November 2016 03:39:31AM *  9 points [-]

One thought that occurs to me re: why this discussion tends to fail, and why Less Wrong has trouble getting things done in general, is the forum structure. On lots of forums, contributing to a thread will cause the thread to be "bumped", which gives it additional visibility. This means if a topic is one that many people are interested in, you can have a sustained discussion that does not need to continually be restarted from scratch. Which creates the possibility of planning out and executing a project. (I imagine the linear structure of an old school forum thread is also better for building up knowledge, because you can assume that the person reading your post has already read all the previous posts in the thread.)

A downside of the "bump" mechanic is that controversial threads which attract a lot of comments will receive more attention than they deserve. So perhaps an explicit "sticky" mechanic is better. (Has anyone ever seen a forum where users could vote on what posts to "sticky"?)