Frankly, LLMs with transparent prompts, as well as user-side filters based on LLM prompts (e.g. "Don't show me comment threads talking shit about Attack on Titan") could do a better and more "fair" job than meat-based moderators now.
They won't have personal biases, don't need to sleep (ending the infamous "mods are asleep, post xxx" waves), their prompts would be visible to everyone, and there could be ways for the users themselves to update the space's rules/prompts.
But either way, I want people like dang to be the ones moderating and managing a community - call it "personal bias" if you'd like, but they have a vision for the space, and as long as I as a user think that that vision is of a community I want to be in, then it's fine. If I no longer think it is... I leave.
The evolution of governance of online communities mirrors that of the real-world.
First, everyone did what they wanted. As conflict became more common, power hierarchies started to emerge. we're now at a stage where every place needs to be governed, yet its members have no influence over who does it.
I have online communities will transition into something resembling democracy where moderators are elected from members by members.
---
While HN is fairly lenient, moderators in pretty much all online spaces are effectively dictators, they are not elected and they cannot be removed by ordinary users, no matter how many disagree.
And of course, such positions attract people who want power for its own sake and who have agendas they want to push.
HN is NOT fairly lenient. HN has a very strict set of rules (applied with infinite discretion) and absolute bunches of tiny rules and quirks that are completely hidden and no real transparency of any kind.
HN has basically an official party line for heavens sake! This is a site for disseminating information about VC things and driving engagement about things that VCs want people to talk about and think, driving traffic to Paul Graham things, and advertising YC businesses and people and ideology.
And not politics unless it's positive towards the ideology of VCs
There aren't official punishment policies or official ways to appeal anything. There's no higher power to call out to. There's a semisecret clique of users.
HN, like most places that are actually good to participate in, is a strict, tyrannical dictatorship that usually uses it's powers to shape behavior towards "discussion", but what that means is entirely up to dang and now tomhow.
The internet requires such behavior because it's just too easy to participate in a non-genuine way and entirely escape any retaliation. You cannot shun a human in an internet setting like you can in real life. The social tools humans and other animals use to shape community behavior are impossible online.
This idea that if we just let people speak absolutely free on the internet things will work better is hilariously uninformed. Humans do not pick or latch on to narratives that are correct, they pick narratives that feel the best and in the modern world, that is almost never the "correct" one. Brains hate nuance, but reality is nuanced.
It's funny, the same exact people on here who insist they can't ride the bus or walk around cities because they freak out if a homeless person accosts them seem to be blind to the concept of how other people's free expression can have a chilling effect.
Yes this is absolutely correct. I can think of more content that's disallowed on HN than content that's allowed: no politics (for the most part), no flamewars/aggressiveness/name-calling, no self-promoting links to your OnlyFans, nothing hugely offtopic, etc. And that's on top of very aggressive moderation of things other social media sites are filled with but are de facto banned here: shitty puns or jokes, one-line zingers, meaningless affirmation comments like "So much this" or "This is the way", nitpicks about submitted articles or personal swipes at the authors' politics...
HN has incredibly strict moderation, and to be clear, that's a good thing. It keeps discussion in line and useful, for the most part.
> It's funny, the same exact people on here who insist they can't ride the bus or walk around cities because they freak out if a homeless person accosts them seem to be blind to the concept of how other people's free expression can have a chilling effect.
I've seen that the term "gatekeeping" is recently starting to be reclaimed as people realize this, to emphasize that while anyone is welcome to participate, the community is not required to bend its rules or standards to accommodate new people. i.e. anyone is welcome to use the bus, but openly shooting heroin while you're on it won't be tolerated.
Oh please. Vote-based social networks are way too vulnerable to burying the truth and boosting lies.
It just takes the first 3-4 viewers to downvote you to prevent the next 10000 people from seeing what you said. There's no downside to downvoting just because you don't like what someone says, even if it's true.
And usually no amount of corrections can outshout a lie/mistake with 100+ votes.
I was thinking of discussion platforms but for general government I had a great idea: we know how to write and publish new laws but we are terrible at unpublishing outdated garbage laws. AI would be great for finding all laws that it thinks are dubious, impractical or otherwise undesirable in today's context and zeitgeist. There can be various manual deletion processes for different categories of absurdity.
Vote manipulation is a non-issue here because users require a minimum of 500 karma to vote, and because the site is so much smaller than Reddit it can take months to reach that threshold. Downvoting is also capped so that you're very unlikely to get pushed back below the 500 karma threshold unless you are consistently making comments that the community doesn't like. I post things I know won't be well-received here all the time and it's quite rare for a comment to go below -2 karma, but comparatively common for these sorts of comment to get flagged despite not breaking any rules.
4chan was great in 2015 precisely because anyone could comment, but it's a young man's website in that scrolling through a 300 comment thread to find the worthwhile parts of the discussion will require upwards of fifteen minutes, whereas on Reddit or Hacker News most of that sorting is already done. This does have censorial effects, so it isn't ideal for controversial topics like politics, but it's better for almost everything else.
What stops people from setting up and aging (or buying) sockpuppet accounts to the point where they control 10+ or even 100+ flag-capable / vote-capable HN accounts, and then using them as a network to deny or boost certain topics? This kind of behavior almost certainly goes on here.
> What stops people from setting up and aging (or buying) sockpuppet accounts to the point where they control 10+ or even 100+ flag-capable / vote-capable HN accounts, and then using them as a network to deny or boost certain topics?
It’s a single board with a full-time moderator and almost everyone on it has a background in information technology. These kinds of networks leave very obvious signatures, and the site simply isn’t a big enough place for them to hide.
> This kind of behavior almost certainly goes on here.
Of course, only the site admins would be able to show you actual examples. But this kind of stuff happens everywhere on the internet where you can post for free, so there is no reason to think it's not happening here.
> the site is so much smaller than Reddit it can take months to reach that threshold.
You can get there in days if you just spot a few bandwagons to hop on.
> I post things I know won't be well-received here all the time and it's quite rare for a comment to go below -2 karma, but comparatively common for these sorts of comment to get flagged despite not breaking any rules.
Yep, there's no downside to frivolously downvoting/flagging: It just takes a 2-3 people to hide your comment from the majority of the users as soon as it's posted, easy for a PR firm with paid people watching a topic like hawks.
Sometimes when I get insta-downvoted in a heated topic, if I delete my comment and repost later, the first few votes are positive. So it's clearly dependent on luck/time, which it shouldn't be.
I and others suggested this years ago: Maybe votes shouldn't have any effect for the first 12 or 24 hours.
Hell look at HN and literally anywhere. Everybody has their own "ideal" world.
I for instance don't want anybody talking shit about anime or video games ever.