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

RyanCarey 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: RyanCarey 27 November 2016 06:43:11AM *  12 points [-]

Thanks for addressing what I think is one of the central issues for the future of the rationalist community.

I agree that we would be in a much better situation if rationalist discussion was centralized and that we are instead in a tragedy of the commons - more people would post here if they knew that others would. However, I contend that we're further from that desired equilibrium that you acknowledge. Until we fix the following problems, our efforts to attract writers will be pushing uphill against a strong incentive gradient:

  1. Posts on LessWrong are far less aesthetically pleasing than is now possible with modern web design, such as on Medium. The design is also slightly worse than on the EA Forum and SSC.
  2. Posts on LessWrong are much less likely to get shared / go viral than posts on Medium and so have lower expected views. This is mostly because of (1). (Although posts on LW do reliably get at least a handful of comments and views)
  3. Comments on LessWrong are more critical and less polite than comments on other sites.
  4. Posts on LessWrong are held in lower regard academic communities like ML and policy than posts elsewhere, including on Medium.

The incentive that pushes in our favor is that writers can correctly perceive that by writing here, they are participating in a community that develops very well-informed and considered opinions on academic and future-oriented topics. But that it not enough.

To put this more precisely, it seems to me that the incentive gradient is currently pointing far too steeply away from LessWrong for 'I [and several friends] will try and post and comment here more often...' to be anything like a viable solution.

However, I would not go as far as to say that the whole project is necessarily doomed. I would give the following counterproposals:

  • i) Wait for Arbital to build something that serves this purpose,thereby fixing (1)-(4)
  • ii) Build a long list of bloggers who will move back (for some reasonable definition) to LessWrong, or some other such site, if >n other bloggers do. It's the "free state project" type approach where once >n people commit, you "trigger the move", thereby fixing the tragedy of the commons dynamic. Maybe one can independently patch (3) in this context by using this as a Schelling point to improve on community norms.
  • iii) Raise funds for a couple of competent developers to make a new LessWrong in order to fix (1) and (2).

I think (i) or (ii) would have some reasonable hope of working. Maybe we should wait to figure out whether (i) will occur, and if not, then proceed with (ii) with or without (iii)?

Comment author: AnnaSalamon 27 November 2016 08:17:29AM *  14 points [-]

Thoughts on RyanCarey's problems list, point by point:

Until we fix the following problems, our efforts to attract writers will be pushing uphill against a strong incentive gradient:

Not sure all of them are "problems", exactly. I agree that incentive gradients matter, though.

Comments on the specific "problems":

1 Posts on LessWrong are far less aesthetically pleasing than is now possible with modern web design, such as on Medium. The design is also slightly worse than on the EA Forum and SSC.

Insofar as 1 is true, it seems like a genuine and simple bug that is probably worth fixing. Matt Graves is I believe the person to talk to if one has ideas or $ to contribute to this. (Or the Arbital crew, insofar as they're taking suggestions.)

2 Posts on LessWrong are much less likely to get shared / go viral than posts on Medium and so have lower expected views. [snip]

The extent to which this is a bug depends on the extent to which posts are aimed at "going viral" / getting shared. If our aim is intellectual generativity, then we do want to attract the best minds of the internet to come think with us, and that does require sometimes having posts go viral. But it doesn't require optimizing the average post for that; it in fact almost benefits from having most posts exist in the relative quiet of a stable community, a community (ideally) with deep intellectual context with which to digest that particular post, such that one can often speak to that community without worrying about whether one's points will be intelligible or palatable to newcomers.

Insofar as writers expect on a visceral level that "number of shares" is the useful thing... people will be pulling against an incentive gradient when choosing LW over Facebook. Insofar as writers come to expect on a visceral level that “adding to this centralized conversational project” tracks value, and that number of shares (from parties who don’t then join the conversation, and who don’t carry on their own good intellectual work elsewhere) is mostly a distraction or blinking light… the incentive may actually come to feel different.

People do sometimes do what is hard when they perceive it to be useful.

3 Comments on LessWrong are more critical and less polite than comments on other sites.

I feel there’s an avoidable part of this, which we should avoid; and then an actually useful part of this, which we should keep (and should endeavor to develop positive affect around — when one accurately perceives the usefulness of a thing, it can sometimes come to feel better). See Sarah’s recent post: On Trying Not To Be Wrong

4 Posts on LessWrong are held in lower regard academic communities like ML and policy than posts elsewhere, including on Medium.

This seems like a bad sign, though I am not sure what to do about it. I don’t think it’s worth compromising the integrity of our conversation for the sake of outside palatability; cross-posting seems plausible; I’d also like to understand it more.

Comment author: Vaniver 27 November 2016 05:41:07PM 5 points [-]

Matt Graves is I believe the person to talk to if one has ideas or $ to contribute to this.

Yep, message me about this, either here or by email (this username at gmail).

Comment author: AnnaSalamon 27 November 2016 08:30:43AM *  8 points [-]

(ii) seems good, and worth adding more hands and voices to; it seems to me we can do it in a distributed fashion, and just start adding to LW and going for momentum, though.

sarahconstantin and some others have in fact been doing something like (ii), and was I suspect a partial cause of e.g. this post of mine, and of:

Efforts to add to (ii) would I think be extremely welcome; it is a good idea, and I may do more of it as well.

If anyone reading has a desire to revitalize LW, reading some of these or other posts and adding a substantive (or appreciative) comment is another way to encourage thoughtful posting.

Comment author: sarahconstantin 27 November 2016 10:19:12AM 11 points [-]

I also support (ii) and have been trying to recruit more good bloggers.

I'll note that good writers tend to be low on "civic virtue" -- creative work tends to cut against that as a motivation. I'm still trying to think of good ways to smooth the incentive gradient for writers.

One possibility is to get some people to spend a weekend together -- rent a place in Big Sur or something -- and brainstorm/hype up some LW-specific ideas together, which will be posted in real time.

Comment author: Vaniver 27 November 2016 05:38:45PM 5 points [-]

One possibility is to get some people to spend a weekend together -- rent a place in Big Sur or something -- and brainstorm/hype up some LW-specific ideas together, which will be posted in real time.

This sounds like an excellent idea.

Comment author: RyanCarey 27 November 2016 09:05:12AM 1 point [-]

I agree that this is great.

I meant to propose something even more specific. Using for example a Google Form, you collect a list of people who agree to post on LW if and only if that list surpasses 200 names.

Once it gets to 200, you email everybody and tell them LW is relaunching.

Do I think it'd work? Maybe.

Comment author: TheAltar 27 November 2016 08:06:52AM *  4 points [-]

A separate action that could be taken by bloggers who are interested in it (especially people just starting new blogs) is to continue posting where they do, but disable comments on their posts and link people to corresponding LW link post to comment on. This is far less ideal, but allows them to post elsewhere and to have the comments content appear here on LW.

Comment author: sarahconstantin 27 November 2016 10:21:45AM 9 points [-]

This is a nontrivial cost. I'm considering it myself, and am noticing that I'm a bit put off, given that some of my (loyal and reflective) readers/commenters are people who don't like LW, and it feels premature to drag them here until I can promise them a better environment. Plus, it adds an extra barrier (creating an account) to commenting, which might frequently lead to no outside comments at all.

A lighter-weight version of this (for now), might be just linking to discussion on LW, without disabling blog comments.

Comment author: FeepingCreature 29 November 2016 12:28:47PM 1 point [-]

Would you use the LW comments section if it was embeddable, like Disqus is?

Comment author: casebash 27 November 2016 02:23:53PM 0 points [-]

I thought ii) had been discussed in the past and was supposedly happening, but nothing ever came from it.