This article is something that has been in my head for a while. I hadn't planned on doing a write-up so soon. I wanted to take the time to a) refine my ideas and b) figure out how to express them clearly before posting. But the recent post Less Wrong lacks direction made me change my mind. My thinking now is that I overestimated the downside (wasting peoples time with a less than fully thought out post) and that there's enough value to justify posting a rough draft now.


LessWrong has been one of the most amazing things I've experienced in my life.

  1. I have learned a ton, and have "leveled up" quite a bit.
  2. Knowing that there are this many other relatively rational people in the world and being able to interact with them is a truly amazing thing.
But I see so much opportunity for LW to do more. Below are some thoughts.

Easy

  • A way to discuss ideas for the site, vote on them, and incentivize the generation of good ideas. I sense that having this would be huge. a) I sense that there are a lot of good ideas out there in people's heads but that they haven't shared. b) I sense that by discussing things, there could be a lot of refinement of current ideas, and a lot of generation of new ideas.
  • More generally, my impression is that it'd be a good idea to subdivide sections for posts. Right now it's pretty much Main, Discussion or Open Thread. Ex. someone who has an idea to improve LW might not think it's "Discussion worthy" (or even "Open Thread worthy"), but I sense that if there were a section explicitly for "LW Ideas", they'd be a lot less reluctant to post. More generally, it'd justify more "bite sized posts" rather than requiring a full write-up.
    • One example of a subsection that I think would be cool is a Personal Advice section. The ability to post anonymously seems like it'd be a useful feature here. Other ideas for subsections: AMA!, Brainstorming/Unrefined Thoughts, I Don't Understand X, Contrarian Thoughts.
  • Social coordination:
    • Apartments/living together.
    • "What are you currently learning? What do you want to learn?". so8res recommends pairing up, and I agree.
    • Geographical map of users to facilitate friendships and/or dating. (This already exists. But it seems that a low proportion of LW users added a pin on the map. My impression is that because of network effects, the usefulness of this is very much a function of how many users there are. Also, I sense that there'd need to be some sort of a different UI that with some sort of organization.)
    • Online chat. Like Facebook. I think it'd be a) cool and b) sometimes useful.
  • Project ideas. There are a lot of smart, skilled and ambitious people here who want to do good things. If LW made it easier to coordinate and work with people, I could see it having a huge impact.


Harder

  • Crowdsource the refinement of posts. 
    • Maybe have an answer wiki for each article that summarizes the main points.
    • Maybe let the author award karma to people who submit a diagram of something explained in the article (I'm a big fan of explaining things visually). Along similar lines, maybe do the same thing for people who submit relevant YouTube videos. Ex. I think that this would be a relevant clip to add to an article about expected value (beware: cringeworthy). (Again, I'm really not a big fan of writing as a medium)
    • Maybe allow collaboration on drafts. And allow the author to award karma to collaborators.
  • Side comments. I really think that for a lot of scenarios, this is a much better UI. But I also think there are use cases for the traditional comments at the bottom of the page, and so there should be both.
  • Make use of some sort of debate tool. I think there are a lot of improvements that could be made to the current approach of having nested comments. It might be sufficient for the level of conversation elsewhere on the internet, but not here.
    • I should emphasize that this seems like it'd be a large and difficult undertaking.
    • But I should also emphasize how important I think it is. Media For Thinking The Unthinkable largely expresses my views here. I think that the mediums we use to write, think and communicate play a large and very underrated role in determining how well we could think. As a society, we don't seem to really recognize this, and we don't seem to have made much progress as far as inventing such tools goes. The importance of such tools goes way beyond LessWrong, but I guess I'm just noting here that LW would benefit greatly from it. I don't think that there are many legitimately deep conversations on LW, and I think that the limitations of nested comments are a big part of the reason why.
    • Along similar lines, I think it's pretty important for there to be a way to highlight and take notes on articles (currently, most people don't). I've been using scrible. Come to think of it, scrible actually isn't that bad of a solution, but I think it'd be awesome if there were a better way to do this built in to the site. (This is another thing I'd like to see across the internet, but I digress...)


There are also a bunch of ideas I have on how the design of the site could be better, but those don't seem relevant to post now (I think we all agree that there's a lot to improve on).

(As far as actually implementing change, things in the "Easy" section seem rather... easy to implement. One option is to just "hack it together". Have "thread" posts, like "Stupid Questions Thread" and use some sort of community managed Google Doc for things like project ideas and social coordination. Idk, I actually haven't thought through what a hacky/MVP implementation would look like. But as far as making changes to the actual site, the code wouldn't be that hard.)

Edit: Some concrete steps to be taken for "version 1s" of the easy ideas. Not well thought out enough; needs vetting.
  • Have a "LessWrong Ideas" thread. Preferably link to it on the sidebar so we don't get too many repeat ideas, so the good ideas have enough time to be voted up, and so ideas don't get "lost in time".
  • Have a "Project Ideas" thread. Preferably link to it on the sidebar, for similar reasons.
  • Have a Project Ideas Google Doc. This would be a list of the more serious and vetted ideas, with brief summaries, skills required, and you could add your name to the list of people interested in working on it.
  • Link to the map of community members on the sidebar. Give, say 50 karma points for adding your data point. I'm not sure how the data would be used for social coordination though. It'd be incredible if there was an API.
    • Actually, maybe it'd be a better idea to create a site that allows users to input their data point on the map, and it'd create the API for us. And you could add things like contact info, interested in finding a roommate? dating? friends/fun activities? On second thought, maybe this is getting too far from the idea of a version 1.
  • Discuss a) the idea of having subsections (ex. Personal Advice, Unrefined Thoughts), and b) which ones we'd like to see. Then create and manage threads based on interest and traction.
  • Have a Google Doc to help people learning the same things pair up. Potential information to include: what you want to learn, how much time per week you want to spend, how many months you'd like to spend learning, fields in which you're knowledgeable (ex. math, psychology, genetics...).
Edit: I didn't realize how reasonable the first steps would be. If after a discussion and vetting people agree on some of these things (or other ideas), I'd be willing to manage the Google Docs and/or threads. I'm also a web developer and would be willing to work with any code for the site. I only have about a year of experience though, so I wouldn't be able to lead any efforts.

Things in the "Harder" section seem, well... harder. It seems that it'd take a non-trivial amount of coding. Especially the debate tools stuff - that'd take an incredible amount of time, expertise and iteration. I think that they're absolutely worth the effort though.


Anyway, at this point I don't even think I'm even proposing anything (I haven't thought it through nearly enough). But I do sense that the ideas are thought out enough to start a conversation. Thoughts?
New Comment
30 comments, sorted by Click to highlight new comments since:
[-][anonymous]170

One problem with a lot of these META: discussions is that people tend to come up with suggestions for things other people ought to do to make LW better.

Why is that a problem? To me it seems suboptimal, but still a net positive.

In particular, I sense that if there was a enough consensus behind an idea, it'd really help motivate volunteers/whoever to get it done.

In particular, I sense that if there was a enough consensus behind an idea, it'd really help motivate volunteers/whoever to get it done.

Historically, this has not been the case, or at least the people whose 'consensus' would be powerful don't participate in the discussion. (As the troll toll change demonstrated, there's really only one person who needs to approve of a change for a change to be made.)

I am curious about the use of 'sense' instead of 'suspect.' My impression is that the trouble is not motivating volunteers to "make the change," the trouble is motivating volunteers to overcome the hurdles to get to the point where they are able to make the change.

Historically, this has not been the case

Hmm, that's unfortunate and decreases my confidence that this could work out.

I am curious about the use of 'sense' instead of 'suspect.'

Just trying to qualitatively communicate my level of confidence. Suspect might mean ~60% confident where sense might mean ~75%.

My impression is that the trouble is not motivating volunteers to "make the change," the trouble is motivating volunteers to overcome the hurdles to get to the point where they are able to make the change.

My view is actually more that there are certain "gates to be opened". If you open 4/5 gates but the fifth gate is still closed, it won't be enough. Something along those lines. As to your points, they both seem like "gates" to me.

A way to discuss ideas for the site, vote on them, and incentivize the generation of good ideas. I sense that having this would be huge. a) I sense that there are a lot of good ideas out there in people's heads but that they haven't shared. b) I sense that by discussing things, there could be a lot of refinement of current ideas, and a lot of generation of new ideas.

By ideas for the site, do you mean changes for the site code? And if you mean changes to the code, do you mean the change ideas, or the change implementations?

The backend is open source, and posts were made in 2011 and 2012 on how to make modifications to the code so that users could submit improvements. I don't pay close attention, but my impression is that very few, if any, improvements have been submitted by users.

Issues can be reported (i.e. requests made) at the Google Code page. But resources to implement those changes are minimal, as I understand it.


It seems to me that the primary piece missing is the effort. Either volunteers need to familiarize themselves with the codebase and then make changes, or volunteers need to put up enough cash to hire devs to make changes. (Presumably the folks at Trike Apps, who would need to approve those changes anyway.)

A way for people to assign monetary value to requests ("I would pay $20 for five separate subreddits") and aggregate those pledges into prizes would perhaps be interesting and solve some of the core problems (while causing other, hopefully more minor, problems). Either it's clear that something is worth doing, it gets done, and the doer is rewarded, or it's clear that something is not worth doing, and it remains deliberately undone.

If we have a way to actually implement improvements to the site, I'd be interested in learning how to do so. I have some development experience. Monetary rewards could certainly motivate me to do so (get me to do it sooner), but I'll probably start researching and working anyway.

As adamzerner asked in another comment, will my contributions actually make it to the site? I need to do more research.

By ideas for the site, do you mean changes for the site code? And if you mean changes to the code, do you mean the change ideas, or the change implementations?

I mean features.

Issues can be reported (i.e. requests made) at the Google Code page. But resources to implement those changes are minimal, as I understand it.

My impression is that because of trivial inconveniences, these are actually huge problems. I'm not sure though. I'm an inexperience developer, and in looking at it I feel like I don't know where to start. Plus, it has the feeling of a dead project (because it is), and this is very demotivating. "Will anyone respond to the Issue I create? The pull request I issue? Is this even the right place for me to do that stuff?"

It seems to me that the primary piece missing is the effort.

Right. I should have talked about this in my post, but my thoughts are still very unrefined. Anyway, here's what I'm thinking:

  • My overarching point in this post is that improving LW is a high level action that could be very very impactful. I don't know how to argue this well yet. Some outside view evidence might be that Eliezer judged it worth it to invest his time in the Sequences and in HPMOR to develop the rationality community. I think that there's still a ton of room for improvement, and that it wouldn't take that much effort.
  • I think that others could be convinced of this (although I definitely don't expect this post to do it).
  • I think that if others were convinced of this, they'd want to contribute. Well, enough people anyway. There seems to be a lot of people who are capable of contributing, and we'd only need a handful. (I wish I could give a good estimate on how many man-months it'd take.)
  • I think that trivial inconveniences are a big reason why people don't contribute (sort of). I sense that if a reasonably organized plan was laid out, it seemed like it'd be impactful, and contributing was low friction, we'd get more than enough volunteers.

I think trivial inconveniences are a huge part of it. I once was part of this group that created a website, and the owner put an edit button on the main page. If you clicked it you could see the source code of the entire website and modify any part of it.

I only have basic knowledge of HTML and javascript, but I ended up spending a bunch of time improving the site. I mean it wasn't a lot, I think I just did some minor editing and added a few pages that embedded an IRC client and an embedded version of our subreddit.

But to my memory, it's the only time I've contributed anything to any kind of open source project. And just because it was so trivial to jump in and get started.

I think having more "subreddits" would be worthwhile - many times I've had something I'd like to discuss with LWers but not posted it because it was too off-topic. However, as others have pointed out, this has been suggested many times but not done.

Perhaps if we can find out how much money it would take for Trike Apps to be willing to make this (or any other popular) change, and it was a reasonable amount, we could arrange a Kickstarter or similar to see if we could get enough funds.

The volume of quality content getting posted to LW is not as high as it could be. So one idea is code modifications that nudge users to publish more high-quality content.

When you write:

This article is something that has been in my head for a while. I hadn't planned on doing a write-up so soon. I wanted to take the time to a) refine my ideas and b) figure out how to express them clearly before posting. But the recent post Less Wrong lacks direction made me change my mind. My thinking now is that I overestimated the downside (wasting peoples time with a less than fully thought out post) and that there's enough value to justify posting a rough draft now.

I have a feeling that there are many articles like this that people have half-baked that they haven't gotten around to refining. I have lots of half-baked articles, for example. In my case, I endorse not working on those articles because I feel like I've got higher-priority stuff to do. Commenting on your article is more relaxing and something I can do in my spare time. But maybe that's the problem. Maybe the problem is that because posting to LW is framed as "work" for almost everyone, we get very few posts. (How many successful online forums do you think make posting feel like work? E.g. think of the GameFAQs forum, the XKCD forum, etc. Maybe the reason Q&A sites like Quora do so well is that writing a high quality response to a question doesn't feel like work. So, more LW Q&A threads?)

Or maybe it's fine that people see writing quality LW posts as work. Maybe the real issue is that they don't think it's valuable work. They notice that most comments even on highly-upvoted LW posts are critical, and subconsciously infer (with their primate brains) that making posts is not a socially sanctioned activity. Maybe what we need is a "blogging carnival" (like the EA community was doing for a while) where someone chooses a prompt at the beginning of the month and challenges people to write a post on that topic before the month is over.

[-][anonymous]50

I think half-baked articles should be entirely okay in Discussion although not in Main, because you often need help / brainstorming to work it out or test some aspects, and Discussion is just the ideal place for it. IMHO Discussion should be seen as a community bakery baking each others half-baked ideas then promoting to Main when done.

I somewhat strongly agree. Regardless of what the "official" policy is, I think it should be made much more explicit. I sense that a lot of people are reluctant to post because they're holding themselves to too high a standard.

Well, it's very hard to get a post that gets lots of upvotes because everyone has their own ideas of how things should be done.

Project ideas. There are a lot of smart, skilled and ambitious people here who want to do good things. If LW made it easier to coordinate and work with people, I could see it having a huge impact.

I see no obstacle to open up threads in discussion for any project. There's also https://www.facebook.com/groups/dotimpact/

Online chat. Like Facebook. I think it'd be a) cool and b) sometimes useful.

We have a facebook group and there's an IRC channel.

Also we have written conversation here and a chat doesn't provide in person communication.

Social coordination:

Happens via Meetups. http://wiki.lesswrong.com/wiki/Less_Wrong_meetup_groups

I see no obstacle to open up threads in discussion for any project.

My impression is just that people are much less inclined to do so than they would be if there was a section explicitly for project ideas. And that it's less organized.

There's also https://www.facebook.com/groups/dotimpact/

Very cool. I actually signed up for dotimpact a while ago, but have forgotten about it. Given the existence of dotimpact, I'm now not so sure whether it's worth having a separate place on LW for it. My first thoughts are:

1) It's worth linking to dotimpact. My thinking is that helping people on LW coordinate on projects has the potential to be a very impactful high level action.

2) It still seems worth it to have a section/thread for project discussions on LW.

We have a facebook group and there's an IRC channel.

Good point. I feel slightly stupid :) I'm not sure what my point is there or whether it's valid. I'm just envisioning it on LW and it feeling cool.

[-]Elo00

Live/Realtime chat is a definite possibility.

Tools like hangouts/slack are the next generation of social interactions and I am already in a few groups based around those systems. The trick is finding the right interface to make it work. (the two mentioned seem to work okay)

My impression is just that people are much less inclined to do so than they would be if there was a section explicitly for project ideas.

Creating additional sections increases complexity. Creating additional section for things that could be solved via a thread in discussion or a Wiki page, seem like a bad idea to me.

[-]V_V10

I think that the home page could use a redesign: right now, most of the the area on the first screen is occupied by static text. I have to scroll down or click to see new posts in Main (which is not very much active) and I have to click to see new posts in Discussion (where most of the activity occurs).

New posts should be immediately visible on the first screen of the home page. Also, there is probably no reason to keep Main and Discussion separated at the moment. They should be merged.

[-][anonymous]00

I thought the same as well - but some time ago someone mentioned the recentposts URL and now I strictly use this URL as my LW starting page. In addition to recency sort, viewing site content this way has the effect of combining Main and Discussion into a single post list.

[-]V_V00

I thought the same as well - but some time ago someone mentioned the recentposts URL

That's helpful, thanks.

What are you trying to improve on LW, and why? What is the purpose of improvements? What do you want LW to be like after you'll apply them?

Personally, I'd want LW to be an effective tool in learning how to apply rationality, discussing rationality and rationality-related topics, and developing new rationality techniques. Every instrument has its purpose; if you want to study, say, math, isn't it more effective to go to some math communities and seek help and assistance there? If you want to chat like in Facebook, why not go to Facebook? If you have a brilliant startup idea, why not go to the (potentially much broader) community of people interested in such kind of ideas? And so on.

The obvious, and very important, improvement is writing more high-quality posts on rationality-related topics, like the sequences.

The harder subsection mostly consists of technical improvements. Is it an actual bottleneck of LW, or our needs are mostly satisfied with nested comments?

[pollid:986]

Good points, thanks for bringing them up.

On second thought, I think I may have overestimated the value of things like study partners and project collaboration. Personally, I just really value being able to do these things with other LW users. Inferential distance is one big reason. I sense that I'm not alone here, but I'm not sure.

In addition to these sorts of tangential things, I also would like to see the quality of posts and conversation on LW improve, and I think some of the other points I brought up would help address that.

The harder subsection mostly consists of technical improvements. Is it an actual bottleneck of LW, or our needs are mostly satisfied with nested comments?

Hmm, I don't want to nitpick, but I'm struggling to answer this question because I don't know how to interpret words like "bottleneck" and "satisfied". I think that a more sophisticated UI could really transform how people think and communicate. As for how things are right now, I'd qualitatively describe it as "fine". Maybe "mildly satisfying".

Thanks for posting this thread. It's good to see a bunch of ideas floating around.

I would imagine that multiple discussion areas should be easy given that this code is based off Reddit. This has been discussed before, but no action has been taken on this.

Social co-ordination is handled to a degree by local Less Wrong Groups.

Re: debate tools. Some tools already exist. Have you looked into these?

Re: highlighting and taking notes across the Internet. Apparently the replacement for Internet Explorer will do something like this.

Having a system for enabling pairing up for learning also sound like a good idea.

I have more to say on these ideas. I'll message you later when I'm not overloaded with work.

Social co-ordination is handled to a degree by local Less Wrong Groups.

I've only been to one meet-up and there's none in my area, so I don't know too much about the extent to which this is true. What do you think? Do you see room for improvement?

Re: debate tools. Some tools already exist. Have you looked into these?

I've looked through the LW wiki page and wasn't impressed. I haven't looked too hard though.

Re: highlighting and taking notes across the Internet. Apparently the replacement for Internet Explorer will do something like this.

I just looked into it and it seems awesome... but I hate Internet Explorer so much and am prepared to be disappointed.

I have more to say on these ideas. I'll message you later when I'm not overloaded with work.

Cool, I look forward to it!

[-]Elo00

sounds like you need to start a meetup. PM me for help on this.

I am new here - and so do not have enough experience to make a judgement call, but I do have a question:

Why do you want to "improve" it? What are the aspects of it's current operation that you think are sub-optimal, and why?

I see a lot of interesting suggestions for changes, and a wishlist for features - but I have no inkling if they might "improve" anything at all. I tend to be of the "simpler is better" school, and from the sound of things, it seems things are already pretty good, or at least pretty non-bad?

STORYTIME!

I used to play a lot of World of Warcraft. I mean - a lot. I had always been a big fan of Blizzard, and when WoW came out, I participated eagerly in the beta, and played it heavily for many years. I eventually left, for a number of reasons - but the relevant one, here, is that Blizzard had been steadily "improving" WoW to the point where it was not what I wanted. In the early days, a lot of WoW was hard, and thus rewarding. You had giant questlines, 40 man raids, and it would take months, maybe a year, to complete goals. Doing so was rewarding, as it was challenging to the intellect, and demonstrated mastery to my peer group - It's fun to brag and show off, even in a video game. But - my goals were not Blizzards, and they steadily "improved" things by making it simpler - rather than a 40 man raid where everyone must be in top form, you could do 25 man, 10 man, 5 man "raids", and you could earn some things by virtue of just grinding (quantity) rather than excellence (quality). Eventually, they started simply selling the types of things that I had spent a great deal of time earning, further invalidating it in my eyes. They improved themselves out of a paying customer, and while they maybe picked up 5 in my place - for me, at least, it ruined the game.

The moral is - beware of "improving" things so much that you alter them fundamentally. I'll be blunt - very little of what you propose above can't be done in discussion threads, and the world has enough social networks. Part of the reason I joined here is the fact that I cannot ask or discuss these things on twitter or shudder facebook - well, I can, but I would get very little but the blank stares of bumpkins. I love humanity, but on a whole we are a bunch of bumpkins, sorry to say.

My thinking stems from the belief that improving this community would be a high level action that would do a lot of good. Improving the quantity and quality of conversation could 1) help spread rationality and 2) improve the experience for current members. And I think that it could lead to 3) intellectual progress and 4) progress stemming from people working together on projects.

The explanation above is incomplete, but hopefully it communicates the big picture - I see a lot of untapped potential for this community to make important progress in discovering things and achieving things.

Why do I think this? It'll take me a good amount of time to answer that properly and now isn't the time for me to do so, sorry. I plan on posting again with that answer at some point though.

Fair enough. I should mention my "Why" was more nutsy-and-boltsy than asking about motive; it would perhaps more accurately have been asked as "What do you observe about lesswrong, as it stands, that make you believe it can or should be improved". I am willing to take the desire for it as a given.

The goal of the why, fwiw, was to encourage self-examination, to help perhaps ensure that the "improvement" is just that. Fairly often, attempts to improve things are not as successful as hoped (see most of world history), and as I get older I begin to think more and more that most human attempts to "fix" complex things just tend to screw em up more.

Imagine an "improvement" where your picture was added as part of your post. There are perhaps some who would consider that an improvement - I, emphatically, would not. Not that you are suggesting that - just that the actual improvements should ideally be agreed upon (or at least tolerable to) most or all of the community, and sometimes that sort of consensus is just impossible.

One simple UI improvement for the site: add a link from comments in inbox to that comment in the context of its post; now I have to click twice to get to the post and then scroll down to the comment.

That plus 10,000 other things :)