December 2009 Meta Thread

This post is a place to discuss meta-level issues regarding Less Wrong. Such posts may or may not be the unique venue for such discussion in the future.

67 comments, sorted by
magical algorithm
Highlighting new comments since Today at 4:55 PM
Select new highlight date

Suggestion the first: rework the About page (possibly after a collaborative effort) to offer newbies clearer guidelines regarding what is expected in posts and in comments, what is the intended meaning of up- and downvotes, and so on.

The About Page has now been edited to (a) tell people who are repeatedly downvoted to take the hint (b) suggest that effort can be better spent than in responding to trolls (c) warn people about the downvoting limit at four times karma.

Neat ! Can I additionally suggest that the sentence:

We suggest submitting links with a short description.

be prefaced with something indicating what kind of things new posters are encouraged to submit ? Say, something like:

The typical LessWrong post is a brief essay inviting reflexion on how to think better, or reporting on relevant news; there is considerable latitude in format, topic, and tone. New posters are encouraged to use the Search box to ensure the topics they introduce are original. We suggest submitting links with a short description.

ETA: Forgot something that might be a valuable way to ease the transition from "commenter feeling their way around" to "beginning poster":

You may want to start by commenting on others' posts, and noticing when your reasoning or information is particularly appreciated by the community, as indicated by upvotes. A longer comment with many upvotes can sometimes be turned into a top-level post, if it has few dependencies on the rest of the discussion.

Perhaps we should move the About page into the wiki?

That would work, and make a collaborative effort possible on an ongoing basis, rather than a one-off before putting up the new page.

When I say "the About page", what I really mean is "whatever page the rightmost link in the top navigation bar of LessWrong points to". I notice this is already an indirect link, via an URL which redirects to an early post of Eliezer's, so I assume making it point to the Wiki instead is a matter of changing a config file somewhere.

Best compliment this site got on August 24th, 2009:

"Less Wrong. Overcoming Bias. Visiting those sites for the first time is equatable to receiving an invitation to Hogwarts."

I never would have put it that way, but.... yes.

Mormon2 has completely outlived his pretense at being anything but a pure 100% troll, so if it's all right with everyone I'm going to just start removing all his comments from this point forward. (It would be preferable to rely on readers to vote him down to -4 on everything, but we don't have that level of constant vigilance.)

Don't do it. I think it's essential for us to be able to recognize and deal with trolls (both OL and RL), and he's a perfectly good training dummy. Engaging him in his -4 threads doesn't make the signal to noise ratio worse as they're folded up. He shouldn't be allowed to make posts even if his karma rises to 20 though.

I agree, I think the system should be designed to handle trolls. Use mormon2 as a test.

If the system is not working perhaps it needs to be modified but should be modified for everyone. I like the part of the US constitution that restricts legislators from targeting individuals with legislation.

For the long run you will want a showdead function like news.arc as it allays most concerns about moderator censorship.

Mormon2 has completely outlived his pretense at being anything but a pure 100% troll, so if it's all right with everyone I'm going to just start removing all his comments from this point forward.

It would be preferable to block the user than censor comments. It is more transparent and gives less potential excuse for accusing the powers that be (you) of misconduct. Picking and choosing which comments to allow seems far more political. (I believe OvercomingBias has moved in that direction, which for better or worse has lowered my impression of the content.)

Feature request: users should be notified not only of child (direct) replies, but all descendant replies (replies to replies).

I'd like to be able to be notified of all descendants of any comment. That is, I'd like to be able to subscribe to the discussion that follows any given comment, whether or not that comment is mine.

This should be an optional preference, not a default.

I wonder if top-level posts should have a higher karma weighting than comments. A good post takes longer to write than an average comment, but a funny one-liner can get as many upvotes as a long, promoted post. If I wanted to maximize my karma total, which I admit does attract me at times, my time would be best spent with writing as many comments as possible, and no top-level posts. This doesn't feel entirely right.

Likewise for negative karma - a negative-value post is far more annoying than a negative-value comment squirreled away in a corner someplace.

I agree, and note that the difference in karma/contribution is even more noticable if you are posting someone else's one liners in the quote thread.

Do note however that the primary status boon from karma is not in being seen to have a high karma total. It is in being seen to have prominent individual posts and comments with high karma. For example, having my one-liner linked to from a context of "too much karma for comments relative to posts" may well beget downvotes for my jest. However, it means I will be seen to have made a funny one liner that got many upvotes. So the net effect is flattering.

The status that comes from writing many posts that reach the 10 karma level is significant even now. A karma modifier would just be nice for the sake of completeness.

Open Thread, Meta Thread, Rationality Quotes Thread, LW Quotes Thread... we need a forum!

Or, we could have a hack that underneath the 'Recent Posts' list showed a list of the above. That eliminates the need to create them monthly, isn't particularly intrusive and gives most of the desired functionality. That may or may not be more desirable than adding a forum to the existing wiki and blog systems.

And a newsgroup, alt.philosophy.lesswrong!

(Wait, wrong decade?)

Wait, you want to be associated with the alt.philosophy hierarchy?

*shudders*

I am not a fan of internet currency in all its forms generally because it draws attention away from the argument.

Reddit, which this is based on, went to disabling a subtractive karma rule for all submissions and comments. Submissions with down votes greater than up votes just don't go anywhere while negative comment votes get buried similar to how they do here. That seems like a good way to organize the system.

Is the reason that it was implemented in order to be signaling for other users or is it just an artifact of the reddit API? Would disabling the actual display of the "points" simultaneously disable the comment ranking? What would be the most rational way to organize the comments. The least biased way would be for it to be based on time. The current way and the way reddit works is direct democracy and that of course is the tyranny of the majority. The current way may be the most efficient if the readers have such a high vale of their time that they only have time to read the most popular comments and skip the rest. However even if that is efficient it is not necessarily optimized to elucidate the best discussion points as users typically vote up things that they agree with rather than strong arguments.

I personally do not submit more responses and posts because of the karma system. As I have seen heavily on reddit, there is karma momentum where people tend to vote similar to how others have voted (as human nature would dictate). Based on that, I know that people will reference the total points of submitters and make decisions on how to take their comments and suggestions in light of that primed information - when the arguments should be evaluated independently.

Maybe I'm missing something though.

I have no clever reply to most of your comment, but:

I personally do not submit more responses and posts because of the karma system.

In my case, it's very much a motivating factor. In fact, I do not think I would have ever been led to comment or post at all without karma. I think this is primarily because I consider it exceptionally valuable, easy-to-read instant feedback on how I'm being received, which I'm normally bad at discerning and find a very important component of any sort of interaction. I virtually never comment on other blogs at all.

The only reason I look at a commenter or poster's karma is when the post or comment seems tremendously bad, and I am trying to decide how much benefit of the doubt to give. In that case I mainly look to see if it's significantly above zero, and don't care beyond that.

I agree with the tone of your post. Voting doesn't work very effectively, in all sorts of situations (e.g., Bush x2). I do kind of like the slashdot style system where maybe you can mark a post from a list of flags, e.g., insightful, funny, intelligent, etc, and perhaps add in a few negative or critical flags like (juvenile, trivial, poorly worded, etc.)

I think these changes would encourage a more holistic evaluation of responses and would work to avoid the overall gruffness, and over-priming effect related to the current system.

Edit: Even better, you could attach a +1 signifier to positive indications (intelligent, insightful, etc.), and a -1 signifier to negative indications (poorly worded, juvenile, etc.), and then enforce that whenever someone votes, they must include a flag. Then, when displaying total points, they should take the average over all categories, so that the only truly negative posts are those which are extremely dynamically poor, i.e., poor in every category.

Do people think that "Don't Feed the Trolls" should apply to mormon2 in all his myriad identities (sharing numerous matching spelling errors, the same personal insults of Eliezer and Less Wrongers, creationism and other baiting tactics)? He has used exploits on the karma system, lied about his credentials and the links between his identities, and regularly makes rude personal attacks on people other than Eliezer. The only plausibly positive thing mormon2 has done is ask Eliezer for info about his work at SIAI, but he ignores answers and doesn't actually seem interested in answers, only in getting at Eliezer.

I propose the following:

1) Anyone other than mormon2 who has questions they think Eliezer has been inadequately answering can put them in replies to this comment, and then collectively ask Eliezer for clear straight answers.

2) Afterwards, no one responds to mormon2 or any other aliases he comes up with.

I'm new here, so i don't have anything to say on whether this mormon2 is a troll.

But, in his response where he explains his motivation, does anyone think he might have a point? Of course, if one's goal is to be received favorably, then it's best to phrase an argument in a way the audience wants to hear it. But he says that his goal was actually to see if people here could answer his criticisms despite his delivery.

At a glance, it doesn't seem that his explanation is taken seriously. But i think that, whether he's a troll or not, this is a valid question, whether we can answer a criticism even if it was phrased in a hostile manner. Although, it could be that everyone has already decided that doing so isn't worth our time or effort.

But, in his response where he explains his motivation, does anyone think he might have a point?

Quite the reverse. The initial post has a point but the response is a juvenile rationalisation.

... and quite obviously to me.

I didn't ask to become an expert in troll psychology, but I've been wasting my time on internet discussions of questionable quality for years, and this is what I have to show for it.

In the response he claims that his spelling and grammar errors were intentionally adopted as a trick, but then makes more in the response. Likewise with respect to crude and offensive language. The guy has been caught lying about numerous other things, and transparently lied in that response.

CitationNeeded, thanks for that link. I'm not sure how clear it is that these several commenters are the same person, but i can see why you'd be suspicious. Interesting that most of mormon2's early comments were upvoted, until recently he adopts this hostile tone. And rereading the "question of rationality", mormon2 remains belligerent in the response, so i'm inclined to agree with wedrifid's "juvenile rationalization" conclusion.

So, i understand dismissing mormon2 specifically, even if i think that listening-to-arguments-from-possibly-unfriendly-commenters is generally worth thinking about. I'm thinking that i may have given him too much benefit of the doubt, but stopped clock, twice a day, i suppose.

The nearby deleted comment clearly speculated that i'm mormon2 as well. I'm not, though it feels a bit silly to have to say so.

It looks like i've accidentally derailed this thread. Sorry about that. Well, as CitationNeeded originally suggested, commence with the collectively asking Eliezer for clear straight answers of inadequately answered questions.

The nearby deleted comment clearly speculated that i'm mormon2 as well. I'm not, though it feels a bit silly to have to say so.

Now I got this silly mental picture of the legacy of mormon2: from here on, every new poster will be suspected of being mormon2 for a few weeks before proving themselves. (Until it's finally revealed that mormon2 was Eliezer in disguise, testing how the community would react to such behavior.)

Hi RNO, welcome to lesswrong. Congratulations on, well, not being mormon2 and please pardon me for affirming the suggestion.

Well, as CitationNeeded originally suggested, commence with the collectively asking Eliezer for clear straight answers of inadequately answered questions.

To the extent that people are interested in how Eliezer believes his rationality work relates to his 'save the world' gig, the perfect place to ask questions is the ask Eliezer questions page. Naturally, the 'Eliezer has not been subpoenaed' caveat still applies.

Indecently, any word on when Eliezer is likely to make that presentation?

I'll repeat my suggestion that posts be able to have language tags (i.e. one tag that says "English", "Español", "Français", etc., automatically set to English for currently existing posts), and that there be an option under Preferences to filter posts by language (probably defaulting to English and, if applicable, the language of the region the user is connected from).

ETA: By "posts", I mean articles, not comments.

Why would we want to fragment the discussion to multiple different languages?

Huh, did I really use the word "posts"? I meant top-level posts, not comments.

Here's my analogous rhetorical question: why would Wikipedia want to spread its effort among multiple different languages rather than concentrating it on a single language?

Top-level posts is what I assumed you meant.

I feel there's some problem with the Wikipedia comparison, but am currently unable to formulate any satisfactory response. I'll think about it; if I don't get back to you, consider the point conceded.

Viewing info-

Originally I was viewing this blog on internet explorer, and pages with lots of comments took a really long time to load and temporarily freezed the browser. Now I have been using Firefox and the pages load almost instantly.

Can you make a subreddit where people can submit and discuss links, questions, and other self-posts that are shorter than what constitutes a typical post on the home page?

Tags now sort chronologically oldest-to-newest, which makes them much more convenient for reading through things.

I've been retagging my old posts on an ad-hoc basis. Since code changes to the LW codebase are hard to come by, there's no good way to give other people permission to do this without giving them full admin rights.

Maybe bounties could be offered for requested changes and then some elect group or you would approve acceptable funded changes and a contract a trust worthy unbiased coder.

I vote for the meta-thread convention, or for any other mechanism that keeps meta off the front page.

A means of agreeing with a post without giving the poster karma? Ideally a full blown survey feature.

The simplest solution would be to allow posters to tag their posts as "no net positive karma". Then people could 'agree' or 'vote' by voting up a post, without giving the person karma.

Ditto that -- there have been several posts and comments where people on the site have wanted to conduct a poll, but have been unable to do so in a gracefully and straightforward matter. The "you be the jury -- Amanda Knox" thread, this comment by E.Y., and Help Roko Become a Better Rationalist are all examples of this.

Since we already have a wiki running, note that votes take place on wikis all the time; such as the current round of voting on the Wiktionary logo. Maybe the standard policy could be to set up a poll on the wiki and point to it on the main site, and make templates and scripts available to make this easier for everyone. It'd also make the wiki more visible.

A model could be Stackoverflow.com which has an option to mark a post, thread or single comment as "karma neutral" (in LW terms; they call it something else). The up- and downvotes still work, they're just not applied to the author's karma.

There's some slight potential for abuse if that were transposed here, insofar as someone could take advantage of it to troll with impunity; but I don't see someone being amicable enough in the first place to amass much karma, and then turning around and becoming a troll.

I'm personally uncomfortable with endorsing another voting system that functions just like karma but with a different name. Having more than one set of numbers that can fluctuate from positive to negative and are differentiated in name alone sounds like a headache, as well as an interface that would scare me away were I not already familiar with the site.

Instead, I propose simple "I Agree" and "I Disagree" buttons. Rather than just tallying the amount of votes for either, it should simply attach one's name to the post via a pulldown tab, just like hiding and exposing child tabs in the comments section. Ideally the lists of those who agree or disagree, as well as how many votes are for either, would not be revealed until one clicked to reveal it. Or possibly make it so that one would have to vote before being able to see the results. Hopefully that would prevent people from being primed to agree with the majority (or being overly contrary).

This would reveal which discussions are actually contentious versus which are being supported by a vocal minority. It could also function as an all-purpose karma-neutral voting system.

Some possible problems: While I want to know who agrees with what, and from a Bayesian perspective this is going to help me be more accurate, I'm fearful that I would be too inclined to agree with prominent posters. I may be unable to discern whether I truly agree with something or am inclined to agree because I see that Eliezer supports an argument. Handled incorrectly it could turn into an in-group out-group situation.

edit: I forgot to include the purpose of having a list of names instead of an anonymous number. I want to know who's judgement I am relying on in matters that I don't have time to do research myself. Also, should I disagree, knowing who I am refuting would help to motivate me. Knowing that I am arguing with a practicing rationalist instead of an anonymous number helps encourage me to construct better arguments.

That's why I suggested "no net positive karma". It would be the exact same system as you just described, except that votes would be applied to an author's total karma if the cumulative votes for that comment were negative.

I'd prefer a "karma neutral" setting, going both ways: one thing that strikes me as an issue here is that an "open" thread isn't quite as open as it might be if you could declare it safe to post there without worrying about losing karma.

I am not sure where this is supposed to go (and it is now Jan) but clicking on the 'seq_mamq' link in 'Tags' causes a software error of some kind.

Karma scores >= 1000 do not fit into the green circle. That is annoying. Without changing the circle size, simply replacing 1234 with 1.2k would make this look less tacky. Eliezer's karma would be 30k and when he reaches 100k well, we can see if that fits when it happens.

But then it's harder to notice local karma changes and look up in recent comments where that happened. Total karma is an uninteresting metric, recent feedback is.

That's true. I'm also somewhat interested in the other half of the 8 on the end. Any ideas how to fit it?

A little proposal for what top-level posts should be allowed:

A post is allowed if it's on-topic, not highly objectionable, and competently done. A post is considered on-topic iff it's one of the following:

  • Something that advances the topic of rationality, where "rationality" means general ways of being less wrong in any field (or possibly a particular field), or of taking actions that are less wrong. We're relatively uninterested in being more right.
  • Information about a topic about which the Less Wrong community would probably like to learn.
  • A monthly thread originally proposed, or seconded, by someone else, which you were not the latest person to create.
  • Something exceptional that you are not deeply involved in.

Note to self: "more" is a red word and "right" is an orange word.

Most posts on here would fall under the first category. The first category was made to exclude things like general posts about quantum mechanics, which are specific to a particular field, though I also tried to include posts specifically related to rationality in quantum mechanics. The second category goes and directly includes general posts about quantum mechanics, but I've tried to make it more restrictive, covering only "information about" these topics rather than useful things that aren't information (e.g. questions and suggestions). The third category, obviously, covers monthly threads, and tries to keep them useful. The fourth category, obviously, is a catch-all, and I've made it to exclude stuff like "please come and volunteer for us; you'll like it".

So, yes, I've stuck my own opinions into it, and this is probably a pretty conservative list. Discuss.

So, yes, I've stuck my own opinions into it, and this is probably a pretty conservative list. Discuss.

I'm nonplussed. This is, as more or less described, a formalisation of your intuitive reaction to various posts. I don't share your intuitions regarding off topicness and, in particular, the 'please come and volunteer for us' example is on topic by my standards. Since formalising which posts are on topic doesn't seem particularly vital to me I reject your proposal to the extent that it seeks to embody your preferences into a general norm.

Here are the options as I see them. This list should not be seen to be exhaustive.

  1. We rely solely on the karma/voting system. The posts which are on-topic are those which are upvoted - on-topicness is thus a quality determined by the number of readers who believe a post is on-topic and of sufficient quality (either in and of itself or by virtue of the attached comment thread).

  2. We rely on the Editor's discretion. Those which are on-topic are the posts which he decides not to hide from the feed.

  3. We make the feed more dynamic - something akin to reddit's front page schema, where you can filter your front page by subreddit, hide posts from a certain user, etc.

Personally, I would love for #3 to be implemented - but LW may not be big enough for it to be effective (e.g. we have no "subreddits", only tags)

In any case, I do think we need to state our policy. As it is, I Eliezer's plan seems to be to make up policy on the fly and introduce what I think are rather ad hoc (not to mention ineffective) "vote up if you would approve of this" comments.