PG: The biggest source of stress for me at YC was running HN

I honestly think the only solution is for individuals to recuse themselves from those networks (I say on one of those networks), lower the trust they place in digital information, etc. It's become clear that the downward spiral is intrinsic to the medium itself (or possibly just the scale). I don't believe that any amount of technology, or product-rethinking, or UX will change that. We just weren't meant to interact this way. My only hope is that people eventually get disenchanted or burned-out enough that they simply stop engaging.

I replied to the original tweet too ("what would you do if you were Jack Dorsey?"). I said I'd shut the whole thing down.

Sadly, the level headed people recuse themselves which only adds to the toxicity.

Actually what happens is the level headed people on one side of an issue divide recuse themselves, leaving a "seemingly level-headed consensus echo chamber" behind. IMHO, that's worse. This account exists largely to counter exactly that trend. It's important (to me) that newcomers to the site don't get the idea that "hackers" are all fringe libertarians on every non-technical subject.

This site may feel like a "consensus echo chamber" but in reality it is nothing remotely close to that. I think you may be running into the notice-dislike bias: https://hn.algolia.com/?dateRange=all&page=0&prefix=true&que.... Since you report noticing fringe libertarians, we can be sure that you dislike fringe libertarianism. We can also be sure that they have just the opposite picture of HN, since everyone crafts their picture in the image of what they dislike, without realizing that they're doing that. It just feels like an objective picture. I can list dozens of examples of this, but I'll restrain myself for once and spare you.

Unfortunately, these extremely contradictory subjective images of HN seem to be a consequence of its structure, being non-siloed: https://hn.algolia.com/?dateRange=all&page=0&prefix=true&que.... This creates a paradox where precisely because the site is less divisive it feels more divisive—in the sense that it feels to people like it is dominated by their enemies, whoever their enemies may be. That's extremely bad for community, and I don't know what to do about it, other than post a version of this comment every time it comes up.

Thanks for caring about level-headeness, in any case.

The sensitivity with which you replied just tells me I'm probably right about this. I assure you that, having dealt with HN readers in real life contexts on both sides of that divide, that the perception absolutely isn't symmetric. HN is seen as a "safe space" for some demographics and definitely as hostile by others. (Edit: I'll just say it. I've had multiple conversations with real life women where I have to make excuses for the perspective of posters here and explain why it's still a valuable forum anyway.)

I mean, I agree with you that we all have biases and blind spots in our perception. Which means... so do the mods. I comment because I want HN to continue to be a site that people like me want to comment on. The site that "people whose comments dang likes" want to comment on surely looks different.

I totally empathize with what it's like to try to defend HN as a worthy place to participate when you're talking with someone who has extremely strong feelings about how awful it is, and the fact that you're willing to do that makes me feel much more sympathy and common ground with you than any disagreement we may have on other points.

But I think your explanation of why this is is much too simplistic. The difference seems to be that you aren't being bombarded every day with utterly contradictory extremely strong feelings about how awful it is. If you were, you wouldn't be able to write what you just posted. Your judgment that the perception "isn't symmetric" is wildly out of line with what I encounter here, so one of us must be dealing with an extremely skewed sample. Perhaps you read more HN posts and talk to a wider variety of people about HN than I do. From my perspective, the links below are typical—and there are countless more where these came from. Of course, there are also countless links claiming exactly the opposite, but since you already believe that, they aren't the medicine in this case. I sample that list when responding to commenters who see things this way:

https://news.ycombinator.com/item?id=23729568

https://news.ycombinator.com/item?id=17197581

https://news.ycombinator.com/item?id=23429442

https://news.ycombinator.com/item?id=20438487

https://news.ycombinator.com/item?id=15032682

https://news.ycombinator.com/item?id=19471335

https://news.ycombinator.com/item?id=15937781

https://news.ycombinator.com/item?id=21627676

https://news.ycombinator.com/item?id=15388778

https://news.ycombinator.com/item?id=20956287

https://news.ycombinator.com/item?id=15585780

https://news.ycombinator.com/user?id=BetterThanSlave

https://news.ycombinator.com/user?id=slamdance

https://news.ycombinator.com/item?id=15307915

A sample email, for a change of pace: "It's clear ycombinator is clearly culling right-wing opinions and thoughts. The only opinions allowed to remain on the site are left wing [...] What a fucking joke your site has become."

https://news.ycombinator.com/item?id=20202305

https://news.ycombinator.com/item?id=18664482

https://news.ycombinator.com/item?id=16397133

https://news.ycombinator.com/item?id=15546533

https://news.ycombinator.com/item?id=15752730

https://news.ycombinator.com/item?id=20645202

https://news.ycombinator.com/item?id=21325122

https://news.ycombinator.com/item?id=23719343

I think there is a solution to this problem. If moderator decisions are made and recorded publicly then the data can at least be analyzed objectively. If there is indeed a bias then someone should be able to sit down and do the statistical analysis and show that "Yes, X type of stories / comments are more consistently flagged / removed / downvoted / etc." or "No, there is actually no bias in this instance".

I think there is contention right now because moderator decisions are opaque so people come up with their own narratives. Without actual data there is no way to tell what type of bias exists and why so it's easy to make up a personal narrative that is not backed with any actual data.

User flagging is also currently opaque and a similar argument applies. If I have to provide a reason for why I flagged something and will know that my name will be publicly associated with which items I've flagged then I will be much more careful. Right now, flagging anything is consequence free because it is opaque.

I completely understand, believe me I get it—but based on everything I've seen, it's a hopelessly romantic view. If I've learned one thing, it's that people are going to "come up with their own narratives", as you aptly put it, no matter what we do. Adding energy into that would only create more pressure and demand on a system which is maxed out already.

Making this mistake would lead to more argument, not less—the opposite of what was intended. It would simply reproduce the same old arguments at a meta level, giving the fire a whole new dimension of fuel to burn. Worse, it would skew more of HN into flamewars and meta fixation on the site itself, which are the two biggest counterfactors to its intended use.

Such lists would be most attractive to the litigious and bureaucratic sort of user, the kind that produces two or more new objections to every answer you give [1]. That's a kind of DoS attack on moderation resources. Since there are always more of them than of us, it's a thundering herd problem too.

This would turn moderation into even more of a double bind [2] and might even make it impossible, since we function on the edge of the impossible already. Worst of all, it would starve HN of time and energy for making the site better—something that unfortunately is happening already. This is a well-known hard problem with systems like this: a minority of the community consumes a majority of the resources. Really we should be spending those making the site better for its intended use by the majority of its users.

So forgive me, but I think publishing a full moderation log would be a mistake. I'll probably be having nightmares about it tonight.

[1] https://hn.algolia.com/?dateRange=all&page=0&prefix=true&que...

[2] https://news.ycombinator.com/item?id=23656311

I merely outlined what I would want if I was a moderator. I would rather receive email with statistical analysis than be compared to Hitler and Stalin without any data to back it up. It would be way funnier if someone proved statistically that I was Hitler and Stalin at the same time. They'd have to go through a lot of trouble to actually do that and if they managed to do so then that would be some high art.

Any complaint without data to back it up would be thrown in the trash pile.

In any case. It's a worthwhile experiment to try because it can't make your life worse. I can't really imagine anything worse than being compared to Hitler and Stalin especially if all that person is doing is just venting their anger. I'd want to avoid being the target of that anger and I would require mathematical analysis from anyone that claimed to be justifiably angry to show the actual justification for their anger. Without data you will continue to get hate mail that's nothing more than people making up a story to justify their own anger. And you have already noticed the personal narrative angle so I'm not telling you anything new here. The data takes away the "personal" part of the narrative which I think is an improvement.

Alas, there's no way to avoid being the target of that anger. It's inevitable in the system. You're right that one has to develop strategies for managing one's reaction to it. I don't think requiring a mathematical analysis would work in my case. It might not work in any case; I expect most angry people would probably get angrier if told that their anger is invalid because not backed up mathematically, and the dynamics of escalating mass anger could end up destroying the whole system.

There's a deeper issue though. Such an analysis would depend on labeling the data accurately in the first place, and opposing sides would never agree on how to label it. Indeed, they would adjust the labels until the analysis produced what they already 'know' to be the right answer—not because of conscious fraud but simply because the situation seems so obvious to them to begin with. As I said above, the only people motivated enough to work on this would be ones who would never accept any result that didn't reproduce what they already know, or feel they know.

Hey dang, first, I appreciate all of your comments in this post and your deep commitment to both HN and the state of online discussion. I'm learning a lot from reading your comments.

I'm curious if given all that you've shared, you think it's even _possible_ to scale a "healthy" discussion site any larger than HN currently is? It's clear that HN's success is in no small part due to the commitment, passion, and active participation of the few moderators. Contrast that with some of the top comments, which describe how toxic Twitter is, and I wonder if there's some sort of limit to effective moderation, or if we just haven't found more scalable solutions to manage millions of humans talking openly online sans toxicity? cheers

I'm not dang, obviously, but I think it's a testament to his hard work that HN functions even as well as it does.

Most sites its size are far, far worse, I think.

I personally believe that is due to human nature.

I think that is what dang has observed and is trying to articulate - no matter how smart or rigorous or mathematical you are, you still are human and thus subject to the human condition.

One way that manifests is this persuasion that the Other is winning the war (and that there is a war, for that matter).

I take it as almost axiomatic that a site with Twitter's volume cannot be anything but the cesspool it is.

It's too big for a single person to even begin to read a statistically-significant fraction of the content.

That means moderation is a hilariously-stupid concept at that scale. Any team of moderators large enough to do the job will itself suffer the fragmentation and conflicts that online forums do, and find itself unable to agree on what the policies should be, let alone how they should be adapted in contentious cases (and by definition, you only need moderation in contentious cases).

I agree with a lot of this, but the things you're talking about already apply to HN, so the argument can't really be used to show that a larger site would necessarily fail to work as well as HN (however 'well' that really is—I'm not making any grand claims here).

For example, the human nature you're talking about is by far the strongest force on HN, and the scale (though tiny compared to Twitter or Facebook or Reddit) is already beyond what one would suppose possible for a forum like this.