Use this thread for discussion of Less Wrong itself and all things meta, meta-meta, meta-meta-meta, omega, etc.

This is a breakdown of Less Wrong's recent new user traffic, data sourced from the Less Wrong Google Analytics account.

67% StumbleUpon
16% Google
5.4% Reddit
3.6% Hacker News
3% Harry Potter story
0.7% Facebook
0.3% Overcoming Bias
4% "The Long Tail"

The 16% for Google is artificially high because many of those hits are users that are using Google as an address bar by searching for Less Wrong.

So we get an order of magnitude more traffic from Stumble Upon than anywhere else -- sometimes thousands of new users a day. Stumble Upon has been Less Wrong's biggest referrer of new users from the beginning of the site. That was surprising to me and I suspect it is also surprising to you. Some of our very best users, like Alicorn, came from Stumble Upon.

What can you do?

1. Sign up for Stumble Upon and start "thumbs up'ing" or likeing LW articles that you sincerely like and want to recommend to others. You could start by stumbling one of my favorite Less Wrong articles on the power of a superintelligence and the real meaning of making efficient use of sensory information.
   
   In order to get maximum Stumble power, you can't just stumble LW articles and only LW articles. You need to use Stumble Upon for a minute or two every now and then and vote up or down the random links it gives you. I know, it's annoying, but what are a few dust specks when we are talking about saving the world?
2. Help our Google traffic by linking to Less Wrong using the word rationality. Less Wrong is the best web site out there on rationality. We should rank #1 on Google for rationality, not #57. At this point in Google's metaphorical paperclipping of the web, some evidence of effort going into increased inbound links is a sign of a high-quality site.
3. When you stumble something on Less Wrong you like or post a link to Less Wrong on the greater Internet, post here. You will be rewarded with large amounts of karma and kudos. Also, cake.

Thanks to Louie for help with this post.

Why do long, uninspiring, and seemingly-childish debates sometimes emerge even in a community like LessWrong?  And what can we do about them?  The key is to recognize the potentially harsh environmental effect of an audience, and use a dying debate to fertilize a more sheltered private conversation.

Let me start by saying that LessWrong generally makes excellent use of public debate, and naming two things I don't believe are solely responsible for fruitless debates here: rationalization biases and self-preservation¹.  When your super-important debate grows into a thorny mess, the usual aversion to say various forms of "just drop it" are about signaling that:

1. you're not skilled enough to continue arguing, so you'd look bad,
2. the other person isn't worth your time, in which case they'd be publicly insulted and compelled to continue with at least one self-defense comment, extending the conflict, or
3. the other person is right, which would risk spreading what appear to be falsehoods.

"Stop the wrongness", the last concern, is in my opinion the most perisistent here simply because it is the least misguided.  It's practically the name of the site.  Many LessWrong users seem to share a sincere, often altruistic desire to share truth, abolish falsehood, and overcome conflict.  Public debate is a selection mechanism generally used very effectively here to grow and harvest good arguments.  But we can still benefit from diffusing the weed-like quibbling that sometimes shows up in the harsh environment of debate, and for that you need a response that avoids the problematic signals above.  So try this:

"I'm worried that debating this more here won't be useful to others, but I want to keep working on it with you, so I'm responding via private message.  Let's post on it again once we either agree or better organize our disagreement.  Hopefully at least one of us will learn and refine a new argument from this conversation."

Take a moment to see how this carefully avoids (1)-(3).  Then you can try changing the tone of the private message to be more collaborative than competitive; the change in medium will help mark the transition.  This way you'll each be less afraid of having been wrong and more concerned with learning to be right, so rationalization bias will also be diminished.  As well, much social drama can disintegrate without the pressure of the audience environment (I imagine this might contribute to couples fighting more after they have children, though this is just anecdotal speculation).  Despite being perhaps obvious, these effects are not to be underestimated!

But hang on...  if you're convinced someone is very wrong, is it okay to leave such a debate hanging midstream in public?  Why doesn't "stop the wrongness" trump our social concerns and compel us to flog away at our respective puddles of horsemeat?

Incident report and hosting update

In the leadup to 2010-05-11 we (Tricycle) were unhappy with repeated short downtime incidents on the Less Wrong (LW) server (serpent). The apparent cause was the paster process hanging during heavy IO. We had scripted an automatic restart of the process when this problem was detected, but each incident caused up to a minute of downtime and it was obvious that we needed a proper solution. We concluded that IO on serpent was abnormally slow, and that the physical machine at Slicehost that serpent ran on had IO problems (Slicehost was unable to confirm our diagnosis). We requested migration to a new physical machine.

Error 1: We requested this migration at the end of our working day, and didn't nurse the migration through.

After the migration LW booted properly, but was quickly unstable. Since we didn’t nurse the migration through we failed to notice ourselves. Our website monitoring system (nagios) should have notified us of the failure, but it, too failed. We have a website monitoring system monitoring system (who watches the watchers? this system does - it is itself watched by nagios).

Error 2: Our website monitoring system monitoring system (a cron job running on a separate machine) was only capable of reporting nagios failures by email. It "succeeded" in so far as it sent an email to our sysadmin notifying him that nagios was failing. It clearly failed in that it failed to actually notify a human in reasonable time (our sysadmin very reasonably doesn’t check his email during meals).

serpent continued to be unstable through our next morning as we worked on diagnosing and fixing the problem. IO performance did not improve on a new physical server.

2010-05-17 we migrated the system again to an AWS server, and saw significant speed and general stability improvements.

Error 3: The new AWS server didn’t include one of the python dependencies the signup captcha relies on. We didn’t notice. Until davidjr raised an issue in the tracker (#207), which notified us, no-one was able to sign up.

What we have achieved:

LW is now significantly faster and more responsive. It also has much more headroom on its server - even large load spikes should not reduce performance.

What has been done to prevent recurrence of errors:

Error 1: Human error. We won’t do that again. Generally “don’t do that again” isn’t a very good systems improvement… but we really should have known better.

Error 2: We improved our monitoring system monitoring system the morning after it failed to notify us so that it now attempts to restart nagios itself, and sends SMS notifications and emails to two of us if it fails.

Error 3: We’re in the process of building a manual deploy checklist to check for this failure and other failures we think plausible. We generally prefer automated testing, but development on this project is not currently active enough to justify the investment. We’ll add an active reminder to run that checklist to our deploy script (we’ll have to answer “yes, I have run the checklist” or something similar in the deploy script).

ETA 2010-06-02:

Clearly still some problems. We're working on them.

ETA 2010-06-09:

New deployment through an AWS elastic load balancer. We expect this to be substantially more stable, and after DNS propagates, faster.

"How can LessWrong maintain high post quality while obtaining new posters?  How can we encourage everyone to read everything, but not everyone to post everything?  How can we be less intimidating to newcomers?"

A lot of Meta conversation goes on here, and the longer it goes on without having a great example to learn from, the longer our discussion will be more aimless and less informed than it could be.  Consider speculating whether blue mould from bread could treat supporating eye infections before you knew it also treated supporating flesh wounds...  it would seem pretty random, and the discussion would be fairly aimless. 

But LessWrong.com is the first successful community of its kind! There is no example to learn from, right?

With the latter, I wouldn't agree: http://mathoverflow.net

[What I've already said in comments:  MathOverflow is a Q&A forum for research-level mathematicians, aimed at each other, created by a math grad student and a post-doc in September 2009.  As hoped, it expanded very quickly, involving many famous mathematicians around the world.  You can even see Fields Medalist — the math equivalent of Nobel Laurate — Terrence Tao is a regular contributor (bottom right).]

MathOverflow awards Karma for good questions and good answers, it's moderated, it's open to new users, and maintains a high standard so professionals stay interested and involved.  Sound familliar?  Well, what about these features...

The top of every page links to:

• Frequently Asked Questions
• How to write a good MathOverflow question
• meta.mathoverflow.net, a separate forum for questions/suggestions about site policy.

Have a look at those links.  If your first reaction is "Sure, precise guidelines worked for a professional mathematics Q&A site...", consider this:  they didn't start out as a professional mathematics Q&A site.  They started out wanting to be one.  They had to defend against wave after wave of undergraduate calculus students posting for homework help.  They had to defy the natural propensity of the community to become an open discussion forum for mathematicians.  I watched as these problems arose, were dealt with, and subsided.  For example: 

Less Wrong is extremely intimidating to newcomers and as pointed out by Academian something that would help is a document in FAQ form intended for newcomers. Later we can decide how to best deliver that document to new Less Wrongers, but for now we can edit the existing (narrow) FAQ to make the site less scary and the standards more evident.

Go ahead and make bold edits to the FAQ wiki page or use this post to discuss possible FAQs and answers in agonizing detail.

Followup to: Announcing the Less Wrong Sub-Reddit

After the recent discussion about the Less Wrong sub-reddit, me and Less Wrong site designer Matthew Fallshaw have been discussing possible site improvements, and ways to implement them. As far as I can tell, the general community consensus in the previous post was that a discussion section to replace the Open Thread would be a good idea, due to the many problems with Open Thread, but that it would be problematic to host it off-site. For this reason, our current proposal involves modifying the main site to include a separate "Discussion" section in the navigation bar (next to "Wiki | Sequences | About"). What are now Open Thread comments would be hosted in the Discussion section, in a more user-friendly and appropriate format (similar to Reddit's or a BBS forum's). If my impression was mistaken, please do say so. (If you think that this is a great idea, please do say so as well, to avoid Why Our Kind Can't Cooperate.)

We have also identified another potential problem with the site: the high quality standard, heavy use of neologisms, and karma penalties for being wrong might be intimidating to newcomers. To help alleviate this, after much discussion, we have come up with two different proposals. (To avoid bias, I'm not going to say which one is mine and which one is Matthew's.)

- Proposal 1: Posts submitted to Less Wrong can be tagged with a "karma coward" option. Such posts can still be voted on, but votes on them will have no effect on a user's karma total. There will be a Profile option to hide "karma coward" posts from view.

- Proposal 2: A grace period for new users. Votes on comments from new users will have no effect on that user's karma total for a certain period of time, like two weeks or a month.

- Proposal 3: Do nothing; the site remains as-is.

To see what the community consensus is, I have set up a poll here: http://www.misterpoll.com/polls/482996. Comments on our proposals, and alternative proposals, are more than welcome. (To avoid clogging the comments, please do not simply declare your vote without explaining why you voted that way.)

EDIT: Posts and comments in the discussion section would count towards a user's karma total (not withstanding the implementation of proposal 1 and proposal 2), although posts would only earn a user 1 karma per upvote instead of 10.

EDIT 2: To avoid contamination by other people's ideas, please vote before you look at the comments.

This post is a place to discuss meta-level issues regarding Less Wrong. Previous thread.

This example (and the whole method for modelling blackmail) are due to Eliezer. I have just recast them in my own words.

We join our friends, the Countess of Rectitude and Baron Chastity, in bed together. Having surmounted their recent difficulties (she paid him, by the way), they decide to relax with a good old game of prisoner's dilemma. The payoff matrix is as usual:

Were they both standard game theorists, they would both defect, and the payoff would be (1,1). But recall that the baron occupies an epistemic vantage over the countess. While the countess only gets to choose her own action, he can choose from among four more general tactics:

1. (Countess C, Countess D)→(Baron D, Baron C)   "contrarian" : do the opposite of what she does
   
2. (Countess C, Countess D)→(Baron C, Baron C)   "trusting soul" : always cooperate
   
3. (Countess C, Countess D)→(Baron D, Baron D)   "bastard" : always defect
   
4. (Countess C, Countess D)→(Baron C, Baron D)   "copycat" : do whatever she does

Recall that he counterfactually considers what the countess would do in each case, while assuming that the countess considers his decision a fixed fact about the universe. Were he to adopt the contrarian tactic, she would maximise her utility by defecting, giving a payoff of (0,5). Similarly, she would defect in both trusting soul and bastard, giving payoffs of (0,5) and (1,1) respectively. If he goes for copycat, on the other hand, she will cooperate, giving a payoff of (3,3).

Thus when one player occupies a superior epistemic vantage over the other, they can do better than standard game theorists, and manage to both cooperate.

"Isn't it wonderful," gushed the Countess, pocketing her 3 utilitons and lighting a cigarette, "how we can do such marvellously unexpected things when your position is over mine?"