The evolution of expert communities
lcamtuf.substack.com
lcamtuf.substack.com
I can compare with other communities that I participate. For example, some subreddits were incredible but "suddenly" something happened that changed the rules of the game and you find yourself as an alien when before you felt at home.
Most probably dang, the moderator, [1] adds a lot of value here?
[1] https://news.ycombinator.com/item?id=34920400 | https://archive.is/czYVG
That's why I recommend upvoting valuable comment "TIL", and actively sending down the yolo/kneejerks/meh/me2 comments that don't add value/different perspective, to keep the comment section valuable for the next gal/guy. If I don't agree, I tend to leave it alone... I value someone not agreeing if they have a point and bring a different perspective.
It's ok to disagree, but always attack the argument, not the person! That's a healthy debate.
I think this makes it immune to the patterns described in the article, although there are other ways for such communities to die. The decline of Slashdot is a cautionary tale.
1) Well-intentioned violators are more likely to respond positively (since the audience effect is not present) and modify their behaviour.
2) For new visitors, all members of the forum appear to be mature and self-regulating (since the moderation is invisible), and they tend to hold themselves to that high standard.
3) It is easy to detect bad-faith violators -- they are the ones who respond negatively to private moderator messages or continue to offend.
But in an online message board you have the problem that jokes are understood by everyone, whereas a deep insightful comment isn't understood by everyone, so jokes tend to go to the top, and people tend to cluster jokes on jokes in joke trees and suddenly what started as "a joke here and there lightens the mood up" turns into people trying to one-up each other on jokes instead of reading the damn article and sharing some actual insight. It's hard to control "how much" online.
HN actually has lots of jokes in absolute numbers, usually only the best ones are tolerated, it's just that on a relative basis you won't enter a thread with 20 comments where 18 are jokes, which is what I would hate.
So the lack of jokes on HN is what makes debates here possible, if we had more jokes we would no longer have debates.
I think there is a correlation between joking and being difficult to reason with, but the causality between them might not be the one you assume.
It appears to me that one of the reasons why people laugh is to release some internal tension. The tension comes from the cognitive dissonance with conflicting thoughts that they have.
The alternative would be to not release the tension but to seeing that you hold these contradictory thoughts and enduring the pain that you must be wrong, because you hold contradictory thoughts, in other words, admitting there is something for you to learn.
But that would require you to become more conscious about your conflicts. The joking is the less conscious, automatic option. And when you are not being conscious, but you automatically take the route of joking, you do not have the ability to choose between being more conscious or less conscious.
Creating original jokes is very deliberate and requires lateral thinking, familiarity with the topic, and awareness of your audience. It's very conscious!
Deploying unoriginal or clichéd jokes to a positive reception calls for some of the aforementioned.
Comedy is widely acknowledged as a means of commenting or contemplating on topics that are controversial or uncomfortable.
Some types of humor are contradictory. This is bullshitting and sarcasm. Plenty are not a fan but it could also be argued that it's useful for approaching truth or good ideas by creatively exploring the space of falsehoods and bad ideas. Conversational TDD, if you will. Bullshitting specifically can be an exercise in role-playing, which may be useful practice for situations that may actually arise.
Some jokes betray no particular belief at all but still require that you be witty or clever. These are your puns, double entendres, innuendos, and so on.
Thinking of humor as lazy or useless or some vaguely compulsive behavior is some kind of thing.
Or, to borrow a clichéd bit of sarcasm, you sound like you're a lot of fun at parties. ;)
There seems to be a threshold of consciousness where we are self-aware, but another even higher level where we are more conscious about what happens within us, for example our emotional reactions. With diligent practice such as meditation, it is possible to increase the average level of consciousness.
I am now going on a limb and trying to explain you what I mean by using your own text above as an example. I am not doing it to insult you, but help you understand what I mean.
In your first sentence you talk about "characterizing ALL jokes". But in my comment I said "ONE OF THE reasons why people laugh".
So for some reason, you misread me claiming something more general than I did.
That kind of misunderstandings are very common, and by itself does not say anything about your intelligence etc. But it is very helpful to turn such misunderstandings around, and ask yourself, why did it happen?
Our mind does tricks on us, and it is up to us to try to put our mind in order. In this case, one possible explanation is that my idea ("one of the reasons people laugh") is for some reason not psychologically acceptable to you (it contradicts something you strongly believe), and this is the reason why you misread me saying it is the ONLY reason why people laugh. This allows your mind you "prove" that my idea is wrong, by providing a counterexample.
So to take the first sentence from my comment, your mind decides to "fight" my idea, instead of laughing it off. Both are defence mechanisms of the mind, to allow you to keep your world view constant.
If you would have been a little bit more conscious the moment you read my comment, you would have noticed that there was an emotional reaction in you from reading my comment. And you could have used that emotional reaction as a signal for yourself to look into what that emotional reaction is about.
The emotional reaction happened regardless of whether my original idea was right or not. My idea just conflicts with something you believe. So it signals less about my idea and more about the ideas you have in your own head, and your emotional valence associated with those ideas.
So your reaction does not prove anything. But it is less likely that I can find good critique of my idea from a comment that shows signs that it is based on less than fully consciously explored emotional reaction. It is possible, but less likely.
The theory is not really mine. You can get an introduction to science of laughter by reading the book called Laughter: Scientigic Investigation by Robert Provine.
The researchers found that people very commonly laugh in discussions at things that nobody thinks are funny, often at the the start of a sentence. We are just not very conscious about it. How it relates to consciousness is another story, but there is a large body of "research" done about the relationship between meditation and consciousness.
One of the core advantages of HN, imo, is the absence of any of the common monetization models. The site itself and the moderation are run by a company making money in another field. Of course HN serves a function for that company, at least marketing, recruiting and community building. But, HN is not optimized to keep me "engaged", because engagement leads to more ad impressions. Things like the noprocrast feature would never make it to Reddit etc.
Another important difference is moderation is paid for and not based on volunteers. This means there is no regular way for someone to become a moderator and getting their own little fiefdom. For as long as I am following this site, dang has been the moderator here and with this consistency also comes consistency in decisions. Sure, dang does not set the tone directly, but at least we don't get "hostile takeovers" followed by a purge of supporters of the old moderation team.
Maybe it is also the topics of HN. They are at the same time really niche, but also very broad. You can find deep dives on esoteric programming languages, some social/political issues, biology research papers, history and even some gaming here. Sure, all of these topics have dedicated subreddits, Discord servers or boards, but here we get all of these topics more or less in 30 links on a single page. Maybe having a space to interact with many different topics and thus people is what keeps this interesting.
I don't know whether these things are due to something like a master plan or are just happy little accidents. No matter, it seems to work.
P.S. some comments mention the lack/low amount of humor. I agree that meme spamming is not great, but I like that we occasionally get a bit of snark or sarcasm mixed in. And without any shiny medals like Reddit gold or a point score attached to such a comment, it is easy to just skip over.
Such communities will eventually die due to poor moderation (because moderators aren't compensated, so don't bother moderating at all or are driven by ulterior motives such as ego) or due to attempts to monetise the platform which kill the user experience.
In that context HN is different in that it has a lot of resources dedicated to it, including a very diligent moderator, without any apparent need for the site to break even or turn a profit. I'm not sure how YC views HN in the context of their business model, but it seems that HN is effectively subsidised by its broader business which IMO is why it has managed to thrive for so long.
Yeah... Old times... We did just that on our forum. But it was worse than that. There were several subforums and each one had its own guideline and a couple of FAQs. When I think of it now I can't believe how stupid we were.
We scrapped them eventually (after five years or so), but it was too late. Though I do not think it was the reason why forum died. Just world had changed.
> Schisms and purges within the administrator community commonly happen at this point, too.
It was more fun for us. We had a very agnostic administrator. So being moderators we fought wars, banned each other, and then appealed to the admin. We lived through three "wars of moderators". I was banned for month in one of them. And at one point I banned another moderator.
It was so much fun. I can't find such a place today. People are too serious about their precious communities.
i.e. I follow woodworking but the communities are stale because there is rarely anything new. Woodworking is about old as a human topic can get. It’s just new memes now.
Hacker News is lively because the computer industry is brand spanking new plus HN also covers non-computer topics that are constantly changing.
I follow skateboarding and there’s always something new and always something to chat about.
wait what? have you guys already discuss all thousands type of woods, each with its own characteristics, application and unique treatments?
you also discussed in deep all the exotic tools the far east has been using for thousand of years?
and all the wood joinery used by Chinese, Japanese and Vietnamese people? there are like hundreds of those which with different usage.
are you saying that your community has already discussed all of that already? that's really awesome!
Communities are for general discussion. If you desire a thorough look of all sorts of Asian (or Western) joinery, you’d specifically look for a book or website, maybe asking a forum for pointers, but the forum itself ain’t the place.
Plus at some point, because none of what you mentioned is new, it will eventually be discussed. Then death.
Let's say someone wants to build an entire house out of solid ebony, that could generate hundreds of hours of interesting discussion.
Although I think communities can go in a few different directions after stage 3. Another common outcome seems to be: arriving at an imperfect but reasonable compromise between the needs of new users and core contributors, which keeps the community relatively stable but maybe not quite as fun as it used to be.
A lot of this does not apply to non-reddit - e.g. IRC or even here on HN or he'll the internet in general.
It is a shame Reddit became the cesspit of the internet - it was great before subreddits were a thing. As soon as you give random people power in the form of moderation/admin rights (e.g. subreddits, or Wikipedia), 95% of the time they go on a power trip, the drama/politics start, and things start to go to shit.
Ultimately the communities die when the community wants to talk about X but the vitriolic power-crazed moderators are screaming at their computer screen about how their community is only for Y and are banning people here and deleting posts there. Turns out if you shutdown conversation, there is no more conversation. Who knew!?
- Entertainment - more people watching, it means incentive to appeal to a wider audience, it often means it has to be more simple, but this is a contrast to what passionates expect,
- User interface - more people using an app means UI needs to be streamlined and be less complicated (can't use keyboard anymore, we need to use touch UI now), but this is a contrast to what power users expect,
- "Expert" communities - more people in a forum, this means less complicated subjects, so "experts" can't grow, they are only expected to give all the time.
Maybe I didn't understand the blog post, but I see it as "the water is wet" content.
When those small groups blow up the stable platform fragments and you get pointless project forks (one of which stagnates); and so on. Many of the participants in those groups take them far too seriously and are overzealous about their opinions (thats what led them to make the sacrifices that put them in that group to begin with).
There's strong echos in the shapes of the histories of cults, guilds and specialized military units.
It’s common to see professional conferences where newer people go to lots of panel sessions and more experienced people hang out in the hallways (and bars).
University and job training programs are kind of all about this process. When you connect with your old study group you’re probably sharing memes and memories, not expertise. The transition is just more explicit because usually at some point you graduate or otherwise move on.
Also the Explore - Exploit tradeoff applies here. Depending on the type of Problems the group is solving different members are in explore or exploit mode. Some problems you need a whole lot of explorers and for others you don't.
Welcoming anyone and everyone into a group without awareness of the current groups composition, what mode individuals are in, and what the real problems being solved are, is like watching what happens to a theatre company after they have finished a long run of a particular play.
The next step is to devise mechanisms to counteract the decay. Does anyone know any literature on this subject? It could be about physical expert communities.
For "helping them find your community", that means that you need to not be precious about the platform that you're currently on; you need to be where young people are. That means somehow convincing the old guard to be willing to change platforms to accommodate the possibility of new members. In practice this probably just means forking your community onto every new platform and convincing a sustainable core of experienced users to populate the new venue, leaving the rest behind; but forking small communities often means death as they lose the critical mass of users necessary for long-term appeal.
For "giving them a reason to stay", that means that your community has to be welcoming to new users. But also, since new users will probably be young-ish, the topic of the community itself needs to be something that young people have a reason to care about. If you're a community of Victrola and vintage phonograph enthusiasts, I'm afraid you're not going to appeal to the youth no matter what you do; if your topic isn't perennially cool, then you might have already lost the battle for longevity.
For "having a process to pass the torch", this means you need at least a little bit of a formal structure. If you have a forum, then who runs it, who sets the rules, who enforces the rules? If the answer is "one random guy", and then that random guy gets run over by a zeppelin, then you've lost.
Basically, creating sustainable, long-lasting, intergenerational communities is extremely difficult. Anything that you come up with to solve the problem is going to start looking like either a religion or a bureaucracy.
Most communities do not have the luxury Wikipedia has. For most of them that approach is unsustainable.
I call... baloney? Which policy, may I ask?
ArbComm and Administrators are by design incapable of blocking users based on disputes over page content. This is fine on paper, but it has resulted in the side effect that if an argument arises, the main thing that Wikipedia ends up policing is less the accuracy of what's being edited (because no authority on Wikipedia may comment on that) but rather the civility of the editors involved.
In practical terms, it's basically impossible to edit or comment on the site without breaking a policy somewhere - Wikipedia's massive list of policies and essays are self-contradicting, meaning that if you complain about a policy violation, a different editor can complain about you breaking another policy.
This means that should a discussion get to the point where administrators (let alone ArbComm) has to be involved, the default outcome is basically that both involved parties will get sanctioned; ArbComm and ANI closure notes have openly stated in the past that they've just given a page/topic ban (or worse) by community consensus to all editors involved because "it might put fresh editors on the page instead".
It's an issue further amplified by the Randy in Boise problem (aka the "fuck experts" policy), where a "sock" user can just behave in extremely polite and civil ways, yet says things that hold absolutely zero ground to the frustration of all involved. All with the goal of eventually escalating it to AN/I or ArbComm so that when their sock gets banned, at least the article is down an expert editor.
Wikipedia largely gets away with this policy on the basis that they'll accrue new editors with time anyway.
OpenAI recently announced the ability to create custom versions of ChatGPT that combine instructions, extra knowledge, and skills.
https://openai.com/blog/introducing-gpts
I could see some of the mentioned expert communities start to build on top of that to offer newcomers a customized chat with domain knowledge.
As LLMs can hallucinate bad and false answers that often skip nuance, this would not be helpful in many cases, and possibly actively harmful.
You're solving for experts not being bothered, but not for new folks being helped.
A big part of learning is learning how to learn. Trying to skip that step with lying "AI" is fools gold.
But nobody bats an eye as we throw clouds of GPUs to provide a chatbot experience.
In one real-world example of 1->5, I thought that the transition to stage 5 was because growth was limited too much: that stage 5 happens when insufficient new people keep joining.
But I haven't seen enough examples to generalize that much.
https://meaningness.com/geeks-mops-sociopaths
It's a bit bleak and gatekeeping but it raises some interesting questions on how subculture eventually turns into a capital-making machine.