Well-Kept Gardens Die By Pacifism

May I suggest that length of comment should factor significantly into the choice to up/downvote?

I once suggested that upvote means "I would take the time to read this again if the insights from it were deleted from my brain" and downvote means "I would like the time it took to read this back."

Time figures into both of these. If you read a few words and don't profit from them, well, neither have you lost much. If you read several paragraphs, reread them to ensure you've understood them (because the writing was obtuse, say), and in the end conclude that you have learned nothing, the comment has, in some sense, made a real imposition on your time, and deserves a downvote.

(Note) This mostly has to do with karma with a minor rant/point at the end. If that doesn't sound interesting, it probably won't be.

Because I really do honestly think that if you want to downvote a comment that seems low-quality... and yet you hesitate, wondering if maybe you're downvoting just because you disagree with the conclusion or dislike the author... feeling nervous that someone watching you might accuse you of groupthink or echo-chamber-ism or (gasp!) censorship... then nine times of ten, I bet, nine times out of ten at least, it is a comment that really is low-quality.

Some of the most interesting things I have registered about LessWrong thus far have to do with the karma game. I am convinced that there are huge swaths of information that can be learned if the karma data was opened for analysis.

If I had to guess at the weaknesses of the karma system I would peg two big problems. The first is that (some? most? many?) people are trying to assign an integer value to a post that is something outside of the range [-1,1] and then adjust their vote to affect a post's score toward their chosen value. This seems to have the effect that everything is drawn toward 0 unless it is an absolutely stellar post. Then it just drifts up. I think the highest comment I have seen was in the high teens. I know there are more than twenty people visiting the site. Do they not read comments? Do they not vote on them?

The second problem spot is that I find it hard to actually use the feedback of karma. I have no way of knowing how well I am doing other a number. I have noticed that my karma has jumped lately and this leads me to believe I have made a change for the better. Unfortunately, I have no easy way of seeing which comments did well and which did poorly. Was it my tone? Did I get wiser? Are my comments more useful? Since I am new, my comment level is low and I can dig through what is there and learn, but this will simply get harder as time goes on. The karma system seems to work well on a comment by comment basis but not so much as a teaching tool. I see this as a problem because this is exactly what I need and I feel like I am squeezing a square peg into a round hole. It makes me think I am not using it correctly.

I find both of the above problems frustrating to me personally. I see a comment get voted down and think, "Okay, that was bad." If I ask for clarification, it goes back up, which just makes it confusing. "Uh, so was it bad or not bad?" The difference between the highest rated comment of mine and the lowest is less than 10. I think the highest is 5 and the lowest was at -2 before I deleted it.

Now, don't get me wrong, I am not complaining that my super-great-excellent posts are not voted to 20 karma in a single weekend. I am complaining that my crappy posts are all sitting at 0 and -1. I just started posting here and already have over 50 karma and the dark secret is that I am a complete poser. I barely even know the terms you guys use. I have not read much of Overcoming Bias and if you gave me a test on key points of rationality I would probably limp through the guessable stuff and start failing once the questions got hard. I can pick apart the logic within a given post, but the only real contributions I have made are exposing flaws in other comments. How in the world am I succeeding? I do not know.

To put this back into the original point, if people are shy about telling me my posts are low quality I can (a) never learn the difference between "mediocre" and "bad" and (b) any fool can limp by with comments that just repeat basic logic and use key terms in the right order. The chances of that being fun are low. One of my great paranoias is that I am the fool and no one pointed it out; I am the elephant in the room but have no mirror. I don't want to trample on your garden and smush the roses. I want to partake in what seems to be a really awesome, fun community. If I don't fit, kick me out.

(To be a little less harsh on myself, I do not consider myself a fool nor am I trying to play the role of a fool. If I am one, please let me know because I apparently have not figured it out yet.)

I'd like to weigh in with a meta-comment on this meta-discussion: y'all are over-thinking this, seriously.

In the vein of Eliezer's Tsuyoku Naritai!, I'd like to propose a little quasi-anime (borrowed from the Japanese Shinsengumi by way of Rurouni Kenshin) mantra of my own:

Aku soku zan! ("Slay evil instantly!")

Don't obssess over what fractional vote a read-but-not-downvoted comment should earn, don't try to juggle length with quality with low-brow/high-brow distinctions (as Wittgenstein said, a good philosophy book could be written using nothing but jokes), don't ponder whether the poster is a female and a downvote would drive her away, or consider whether you have a duty to explain your downvote - just vote.

Is it a bad comment? (You know deep down that this is an easy question.) Aku soku zan! Downvote evil instantly! Is it a useless comment? Aku soku zan!

(And if anyone replies to this with a comment like 'I was going to upvote/downvote your comment, but then I decided deep down to downvote/upvote' - aku soku zan!)

For nearly 10 years I have referenced this thread in various forums I've moderated. While I never entirely agreed with every aspect, it has mostly held up well as a lodestar over the years.

Until recently.

And now, with the benefit of enough sequential observation over time, I am comfortable describing what I believe is a major hidden assumption, and thereby weakness, in this entire argument:

For the concept of "walled gardens" relating to online communities to succeed and thrive, there must exist an overlay of credibly alternative platforms. Or, more directly, there must exist fair and healthy competition among and within the media upon which the discussion are taking place.

This argument was created largely before the "sciences" of social media were refined. Today, we are living with the consequences of intersecting sciences of human psychology, sociology, computer science, statistics and mathematics. Dense, voluminous texts have been penned which describe precise models for determining how to create, manage and extract-value from "online communities". Some of those models even go so far as to involve manipulation of human physiological responses -- i.e., intentional mechanisms within the platforms designed specifically to trigger release of chemicals in the human brain.

There exists an analogue for this maturity curve within television advertising. As both the medium and techniques matured, the need to evaluate how we, as a society, managed its impact fundamentally changed.

Today, in late 2018, there effectively exists no credible "public town square" whereby free speech exists as it was intended (within the intention of the US Constitution). What exists in its place is a de facto oligopoly of media companies posing as tech companies who have divided up the horizontal market and who exercise overwhelming "market power" (as in HSR power) over would-be competitors. Those competitors are then relegated to competing as "free speech purists", which leads to the traps outlined by the original argument: a cesspool of fools and insults.

This situation allows the dominant players to then, in turn, point to the worst aspects of their would-be competitors whenever they feel threatened by them -- or are otherwise politically or economically motivated. Using catch phrases like, "hate speech" or whatever "ism" catches the gestalt, the oligopolists then pressure the supply-chain of would-be competitors, forcing them out of business. They eliminate their ability to process payments. They cut off their upstream bandwidth providers. They remove their ability to be routed or resolved. They eliminate all possibility of collecting advertising revenues.

And they do all this with the virtuous facade that they, the incumbent giants, are safeguarding a "well kept garden". All while conveniently forgetting that they only rose to such dominance by exploiting the very freedom of speech -- including an early tolerance for the opinions they now so self-righteously claim to oppose.

There are various academic ways to describe this type of situation. But the solution is the same: largely unrestricted free speech and a renewed focus on one's personal responsibility in selecting herself in or out of conversations.

It may be true that well-kept gardens die from activism, but it's also the case that moderation can kill communities.

Any community that really needs to question its moderators, that really seriously has abusive moderators, is probably not worth saving. But this is more accused than realized, so far as I can see.

There speaks the voice of limited experience. Or perhaps LiveJournal, Reddit, Google+ and Facebook really are not worth saving?

I've seen enough discussion forums killed by abusive moderators that I look carefully before signing up for anything these days. When I write a lengthy response, like this, I post it on my own site rather than face the possibility that it will be silently deleted for disagreeing with a moderator.

However, I've also been a moderator, and I've seen situations where moderation was desperately needed. In my experience on both sides of the issue, there are some basic criteria for moderation that need to be met to avoid abuse:

• Moderation needs to be visible. Comments that are removed should be replaced with a placeholder saying so, and not simply deleted. Otherwise there will be accusations of repeated unfair deletion, and any act of moderation will quickly snowball into an argument about how much censorship is occurring, and then an argument about whether that argument is being censored, and so on until everyone leaves the site.
• Moderation needs to be accountable. Moderators must have individual accounts, and moderation actions need to be associated with individual accounts. Without this, it's pretty much impossible to identify an abusive moderator. I recently got banned from a subreddit for asking which rule I had broken with a previous posting, and there was no way to find out who had banned me.
• Moderation needs to be consistent. There needs to be a description of what the criteria for moderation actually are. It doesn't need to be legalistic and all-encompassing, and it may be subject to change, but it needs to exist. Some people feel that actually writing down the criteria encourages people to argue about them. The alternative, though, is that person A gets banned or censored for doing something that person B does all the time; that leads to much worse ill-will and ultimately is worse for the community.
• Moderation rules need to apply to the moderators. A special case of the above, but it deserves highlighting. Few things are more infuriating than being banned by a moderator for doing something that the person doing the banning does all the time. Once this kind of moderation starts happening (e.g. Gizmodo), the atmosphere becomes extremely toxic.
• Moderation needs an appeals process. There are abusive power-tripping assholes out there, and they love to find their way onto forums and become moderators. You need a mechanism for identifying any who find their way into your forum. Having some sort of appeals process is that mechanism. Ideally appeals should be resolved by someone who isn't part of the moderation team. Failing that, they should be resolved by someone other than the person being complained about, obviously.

It also helps if the moderation activity can be openly discussed in a partitioned area of the site. There will be desire to discuss moderation policy, so plan ahead and have a space where people can do so without derailing other threads. That way, you can also redirect meta-discussion into the moderation discussion area to avoid thread derailment, without making the problem worse.

(Also posted at my web site)

Maybe I understand on a gut level that the opposite of censorship is not academia but 4chan (which probably still has mechanisms to prevent spam).

A quick factual note: 4chan unconditionally bans child pornography and blocks (in a Wikipedia sense) the IPs, as I found out myself back when I was browsing through Tor. They'll also moderate off-topic posts or posts in the wrong section. They actually have a surprisingly lengthy set of rules for a place with such an anarchistic reputation.

Update: new 'feature' - apparently, you can now only downvote if you've done less downvoting than your karma. Example from my screen:

Your total down votes (2538) must be less than your karma (528)

I can see myself linking to this more than anything else you've ever written. Sing it, brother!

Note that the voting system we have here is trivially open to abuse through mass account creation. We're not in a position to do anything about that, so I hope that you, the site admins, are defending against it.

Wikipedia is an instructive example. People think it's some kind of democracy. It is not a democracy: Jimbo is absolute ruler of Wikipedia. He temporarily delegates some of his authority to the bureaucrats, who delegate to the admins, who moderate the ordinary users. All those with power are interested to get ideas from lots of people before making decisions, but it's very explicit policy that the purpose is to make the best encyclopaedia possible, and democracy doesn't enter into it. It is heavily policed, and of course that's the only way it could possibly work.

There is no strong reason that reasonable, informative discourse should be an attractor for online communities. Measures like karma or censorship are designed to address particular problems that people have observed; they aren't even intended to be a real solution to the general issue. If you happen to end up with a community where most conversation is intelligent, then I think the best you can say is that you were lucky for a while.

The question is, do people think that this is the nature of community? There is a possible universe (possible with respect to my current logical uncertainty) in which communities are necessarily reliant on vigilance to survive. There is also a possible universe where there are fundamentally stable solutions to this problem. In such a universe, a community can survive the introduction of many malicious or misguided users because its dynamics are good rather than because its moderator is vigilant. I strongly, strongly suspect that we live in the second universe. If we do, I think trying to solve this problem is important (fostering intelligent discourse is more important than the sum of all existing online communities). I don't mean saying "lets try and change karma in this way and see what happens;" I mean saying, "lets try and describe some properties that would be desirable for the dynamics of the community to satisfy and then try and implement a system which provably satisfies them."

I think in general that people too often say "look at this bad thing that happened; I wish people were better" instead of "look at this bad thing that happened; I wish the system required less of people." I guess the real question is whether there are many cases where fundamental improvements to the system are possible and tractable. I suspect there are, and that in particular moderating online discussion is such a case.

Eliezer,

I used to be not so sure how I felt about this subject, but now I appreciate the wonderful community you and others have gardened, here.

I think I fundamentally disagree with your premise. I concede, I have seen communities where this happened . . . but by and large, they have been the exception rather than the rule.

The fundamental standard I have seen in communities that survived such things, versus those that didn't fall under two broad patterns.

A) Communities that survived were those where politeness was expected - a minimal standard that dropping below simply meant people had no desire to be seen with you.

B) Communities where the cultural context was that of (And I've never quite worded this correctly in my own mind) acknowledging that you were, in effect, not at home but at a friendly party at a friends house, and had no desire to embarrass yourself or your host by getting drunk and passing out on the porch - {G}.

Either of these attitude seems to be very nearly sufficient to prevent the entire issue (and seem to hasten recovery even on the occasion when it fails), combined they (in my experience) act as a near invulnerable bulwark against party crashers.

Now exactly how these attitudes are nurtured and maintained, I have never quite explained to my own satisfaction - it's definitely an "I know it when I see it" phenomena, however unsatisfying that may be.

But given an expectation of politeness and a sense of being in a friendly venue, but one where there will be a group memory among people whose opinions have some meaning to you, the rest of this problem seems to be self-limiting.

Again, at least in my experience - {G}. Jonnan

One problem I have with hesitation to downvote is that some mediocre comments are necessary. Healthy discussion should have the right ratio of good comments to mediocre comments, so that people may feel relaxed, and make simple observations, increasing rate of communication. And current downvote seems too harsh for this role. On the other hand, people who only make tedious comments shouldn't feel welcome. This is a tricky balance problem to solve with comment-to-comment voting.

I would downvote more, if we had a separate button, saying "mediocre", that would downvote the comment, say, by 0.3 points (or less, it needs calibration). The semantics of this button is basically that I acknowledge that I have read the comment, but wasn't impressed either way. From the interface standpoint, it should be a very commonly used button, so it should be very easy to use. Bringing this to a more standard setting, this is basically graded voting, --, - and ++ (not soft/hard voting as I suggested before though).

An average mediocre comment should have (a bit of) negative Karma. This way, people may think of good comments they make as currency for buying the right to post some mediocre ones. In this situation, being afraid to post any mediocre comments corresponds to excessive frugality, an error of judgment.

Also, this kind of economy calls for separation of comment Karma and article Karma, since the nature of contributions and their valuation are too different between these venues.

This post makes me think of SL4:

The most active place on the internet for discussing Friendly AI is the SL4 email list. Ironically, it must be one of the most hostile email lists on the internet with frequent flame wars and people being banned from the list. The moderation system consists of so-called “list snipers” whose job it is to ban discussions that they don’t like. If these people are experts in friendliness… lord help us.

• http://www.vetta.org/2006/09/friendly-ai-is-bunk/

Good thing this community died for entirely unrelated reasons, then!

Different people will have different ideas of where on the 4chan - colloquium continuum a discussion should be, so here's a feature suggestion: post authors should be able to set a karma requirement to comment to the post. Beginner-level posts would draw questions about the basics, and other posts could have a karma requirement high enough to filter them out.

There could even be a karma requirement to see certain posts, for hiding Beisutsukai secrets from the general public.

4chan is actually pretty popular, I don't know if you are aware. Somehow their lack of censorship hasn't kept them from being "fun" for millions of people.

The balance for a moderator is between too much craziness and too much groupthink.

Moderation easily becomes enforcement of a dogma. In English literary theory today, you're required to be a cultural relativist. You only get to choose one of three kinds of cultural relativist to be: Marxist, feminist, or post-modernist. Anyone else is dismissed as irrelevant to the discourse. This is the result of "moderation," which I place in quotes because it is anything but moderate.

It is especially problematic when the moderator is a key contributor. A moderator should, ideally, be a neutral referee.

Revisiting this post in 2017, I'm calling it wrong in retrospect. It seems to me that LessWrong is less vibrant than it used to be, and this is not because of too little moderation, but may be partly because of too much, both from above (post promotion, comments from EY, and harassment of divergent views from moderators) and from below (karma voting). LW has become a place of groupthink on many issues. Karma did not prevent that, and may have accelerated it.

EY encouraged this. He refused to engage with criticism of his ideas other than with rudeness or silence. He chased away Richard Loosemore, one of the only people on LW who was qualified to talk about AI and willing to disagree with EY's ideas. EY's take on him was:

Warning: Richard Loosemore is a known permanent idiot,

(And, looking at that thread, how exactly did timtyler, one of the other stars of LW, get banned?)

Given a finite amount of time in a day, I have to decide how to use it. While I can afford to take a quick look at each comment when there are only few of them, I have no choice but to ignore some when there are pages of them (and other top-level posts to read). One nice thing with the karma system is the "best to worst" comments order: I can read the first ones and stop reading when encountering too many "boring" ones in a row (but maybe not "boring" enough to merit a downvote).

However, if many people use a similar algorithm to mine, the "bad" comments won't be read often and thus won't get further downvotes. Worst: the "good but new" comments (starting at 0) can get stuck in that pool of unread comments.

Vladimir_Nesov suggested to add a "mediocre" voting option affecting karma by -0.3 (instead of the -1 or +1). I would instead suggest a "I read this" button, worth 0 karma, together with some counter indicating the total amounts of votes irrespective of them being -1, 0 or +1. When you read a post/comment, you always vote: -1 if you judge it bad, +1 if you judge it good and 0 if you are not ready to do any of the previous.

With such a device, people could once in a while "sacrifice" some of their time reading low karma comments with few total reading count. Moreover, the current -4 threshold for hiding a post could become a function of this total count (some kind of variance).

My secret garden or why my garden is not well-kept

I talk of a real garden here - my garden, a place for me to rest, to see flowers, birds and butterflies, to read in.

Years ago, before my time, it was a well-kept garden.

There was a lot of grass and some bushes, and that was it, easy to maintain, well kept.

Because I like flowers, i tried this and that, some things thrived and others died and it was never the garden I planned in the beginning.

I never planted the clover and the moss that creeped into the grass. In the beginning, I tried to fight it, but suddenly I began to understand that the garden won life trough it. There suddenly were bees and little flowers in the grass.

I despised roses in the beginning. They seemed like plastic flowers to me. But then I discovered the old roses, roses who did the one thing a rose to my mind should do : they smell. All the roses in my garden have different smells.

I have plants I never planted, but I like them. I even like dandelion, even if I have to fight it every year.

There will be dandelion anyway. A bit of fighting is necessary, because else the garden would have only dandelion.

I took my gardening philosophy a bit from "the secret garden" from Burnett and a bit from chinese and japanese gardens, including new plants, if they take root, only rooting out the ones that try to overtake the garden or that I know are poisenous.

This year, because of clima change, a lot - even the big tree I love so - has to go. I don´t know which flowers will thrive next year. I wait and see.

I know this: in the oh so well kept gardens in my childhood there were no sweet smells and nothing was allowed to grow, if it was not planted in that exact place.

Let the new seeds, that fly in, place to grow. Don´t root them out, because they are not what you expected.

A garden can be too well kept to be a living, thriving thing.

"Don't believe in yourself! Believe that I believe in you!"

If you're trying to quote Gurren-Lagann here, I believe you botched the quote. "Believe in me who believes in you!" But maybe it was dubbed differently. In any case, I do find some amusement in your approvingly quoting a show which was more or less premised on a rejection of rationality. "Throw away your logic and kick reason to the curb!" I'll have to remember that for the next anti-rationalism quotes thread.

But anyways, I did like this post, although as you implicitly concede it's just one narrative of community development among many. I'm sure that there have been as many communities to have fallen due to despotic moderation or impoverished by rigid ideological guidelines as there have been ruined in the ways described in the OP. Oftentimes the "idiots" who "ruin" the comm are actually the lonely voices of reason. It's a fine line to walk, and I look forward to someday seeing a modern-day Machiavelli write a tract on "The Internet Community Moderator". Because it really is that tricky.

the opposite of censorship is [...] 4chan

I intend to steal this quip frequently.

In my opinion, if handled improperly (E.g. Reddit) up/down vote system enable predatory herd dynamics since the accumulation of karma points trigger a competition between members and an excessive sense of self-worth.

When you don't have a specific karma count for every single community you are in, what can happen is people starting acting on purpose to get karma (that is also dopamine), adapting their behaviour to the general trend of every community they join, for the mere sake of acquiring what is perceived as "credibility" and "respect".

Gaining karma throwing toxic, very low-effort, provocative comments becomes the praxis, and it's very effective since people always struggle to prove their superiority, be it charismatic or intellectual.
People also tends to validate who keeps the game on the same field they feel confident, leaving little to no room for a real, healthy exchange of different perspectives.
 

Lots of interesting and good things come out of 4chan. The signal/noise is low, but there is still lots of signal, and it had no high ideals to start with at all.

I wonder if an explicitly rationalist site without standards would devolve into something that was still interesting and powerful. I think I would trade LW/OB for a site where a thousand 13 year old bayesians insulted each others' moms and sometimes built up rationality. In the long run it's probably worth more.

Also, I have a higher quality comment which my posting time is too small to contain.

Just like it is easy to be naive about the universal virtue of unconditional nonviolent pacifism, when your country already has armed soldiers on the borders, and your city already has police.

Kudos! A lot of my friends don't understand why I practice martial arts, they just don't understand how priviledged they are in never having needed it.

Typo: at the bottom of the post, where the previous post is referred. Underconfidence has an extra 'e'

I see that Robert Scoble has recently posted on a good way of creating a responsible online community:

http://scobleizer.com/2009/04/20/decentralized-moderation-is-the-chat-room-savior/

Um, so, who is the fool you're talking about, or are you just speaking hypothetically?

You seem to be implying that communities such as 4chan are "bad". You did not say that explicitly, but you heavily implied it. Why do you think free-speech communities are bad? Your choice of the word "fool" to characterize someone with whom you disagree is questionable at best.

Maybe I understand on a gut level that the opposite of censorship is not academia but 4chan (which probably still has mechanisms to prevent spam).

A quick factual note: 4chan unconditionally bans child pornography and blocks (in a Wikipedia sense) the IPs, as I found out myself back when I was browsing through Tor. They'll also moderate off-topic posts or posts in the wrong section. They actually have a surprisingly lengthy set of rules for a place with such an anarchistic reputation.

One: I support the above post. I've seen quite a few communities die for that very reason.

Two: Gurren Lagann? (pause) Gurren Lagann? Who the h*ll do you think I am?

Barry Kort posted notice of this article on a thread at The Wikipedia Review.

We have there, over the years, often considered this problem, with Wikipedia being the sublimely ridiculous example.

Here are a couple of threads that come to mind:

• Wiki Induced Knowledge Impairment

• The Web Is Making People Stupid

Open to discussion here or there …

Jon Awbrey

What communities actually die in that way? If they don't actually end but continue differently then it's like saying science fiction died because new authors with their newfangled take on the genre changed things (disclaimer: I don't really know anything about science fiction).

In the case of spam there is a problem of high volume (raivo pommer estee is a good counter-example, as there's generally no more than 1 per thread and it's short) but otherwise I don't really see the harm in idiots posting. Anybody is free to skip past stuff they don't care about (I do it all the time) and people get value out of even reading stupid comments, so I don't see what's so terrible that it outweighs it. I'm with Hopefully Anonymous on how I rate blogs by their comment policies.

For the record, I am also having a hard time deciding how to vote on comments to this post. Is it too early to turn up the standards and vote everyone down?

awesome post, eliezer. you sound like quirrel.

Reading this, I found myself wondering what Eliezer felt he needed to justify censoring, and why he felt the need to talk us into approving his doing so.

Then I took a look at the "Top Contributors", and noticed that despite having a karma score higher than several people on the list, my account wasn't listed.

Hm.

PRACTICE DOWNVOTING ON THIS COMMENT FAGS LOLOLOLOL