←back to thread

600 points codetrotter | 9 comments | | HN request time: 1.291s | source | bottom
Show context
subsubzero ◴[] No.35461974[source]
Congrats Dang, you have done a wonderful job so far and moderate one of the most fantastic online communities out there. I am sure most of the job feels somewhat thankless but I want to let you know I(and many many other users on this site) appreciate your hard work and dedication.
replies(3): >>35462601 #>>35462773 #>>35463700 #
codeddesign ◴[] No.35462773[source]
If by “finest” you mean a Reddit mob mentality for tech, then yes I completely agree with this statement.
replies(6): >>35462836 #>>35463131 #>>35463193 #>>35463875 #>>35464427 #>>35464999 #
dang ◴[] No.35463131[source]
What do you think we could do differently? Serious question.

I don't like the mob thing either but it's how large group dynamics on the internet work (by default). We try to mitigate it where we can but there's not a lot of knowledge about how to do that.

replies(24): >>35463179 #>>35463213 #>>35463257 #>>35463371 #>>35463548 #>>35463713 #>>35463749 #>>35464099 #>>35464410 #>>35464467 #>>35464570 #>>35464688 #>>35464754 #>>35465446 #>>35465523 #>>35465648 #>>35465794 #>>35466615 #>>35466946 #>>35467134 #>>35468675 #>>35469283 #>>35476621 #>>35488228 #
1. harry8 ◴[] No.35463713[source]
How do you attempt to counter groups who deliberately game things?

Newer, widely purposed tech that hasn't taken hold yet is the one I see most often. Where people have invested non-trivial time. Eg the rust crew which I choose as an example because I quite like rust so I'm not bashing the tech here.

There are bogus "here's a cool library/app/thing" articles which get to the front page where I scratch my head and then discover "oh, it's boring as hell but in rust." I see people expressing legitimate points of view (which I frequently disagree with) about cases where C might be a better choice or where a rust re-write doesn't buy the user much that are immediately massively downvoted. It's mob mentality, is it organised or self-organising? Doesn't matter either way. Makes rustaceans look pretty stupid though. And yes you can do this for non-rust stuff and it happens. And it's easy to see why. You invest a ton of time into a tech you really want it to succeed to maximise the payoff for doing so.

What do you do to counter it? Anything at all?

HN is gamified on karma - an idea taken from slashdot with a mildly different implementation but a really good idea that has been under-applied accross the web. In this game popular in the zeitgeist massively trumps interesting, well written, thought provoking and well supported. Sure by the time you've got 50 points you probably should stop caring and you've got karma to burn to be thoughtfully unpopular, but if you take the time you probably want that seen and it's the stuff you want to see - which is kind of the point.

The other question I have is why has there been next to nothing here (that I've noticed - maybe I missed it?) About the twitter files revelations and their importance or lack thereof? Intervention? If so please would you share the thinking behind it?

The point being, to suggest what you could do differently requires a clue on what you actually do now, which I'm not at all sure I have.

replies(1): >>35463892 #
2. dang ◴[] No.35463892[source]
We've worked a ton on trying to prevent that sort of gaming. I would never say that we're catching all of it. If you see a case that you think has broken through our defences, the best thing to do is to email hn@ycombinator.com so we can take a look.

Re the Twitter stuff: there have been quite a few major threads. They also attract a ton of user flags. I'm personally open to the topic but the chalice is so poisoned that I'm not sure HN can discuss it in an interesting way, and interestingness is what we're going for here.

replies(3): >>35464673 #>>35464850 #>>35466044 #
3. p-e-w ◴[] No.35464673[source]
> I'm not sure HN can discuss it in an interesting way, and interestingness is what we're going for here.

I ask this with the greatest amount of respect: Is it possible that you're taking that "mission statement" a little too seriously here?

I'm a casual HN user and I open the front page 3-4 times per day. Roughly 50% of the topics tend to be well-written but ultimately standard blog posts from random technologists on fairly standard technology topics.

Some of it is indeed interesting, and what's interesting of course depends on the reader, but I think it's safe to say that most people won't open the front page and be utterly blown away by how incredibly interesting every single post is.

Considering this, silencing a lively discussion because it might not meet certain, ultimately subjective, criteria of "interestingness" seems excessive. To be clear, I don't want to see the front page dominated by a single topic every day, but it pains me when discussions containing (among other things) thoughtful comments are effectively hidden from view in a matter of minutes because an algorithm or a moderator thinks we'd be better off reading about how someone has connected their dog's heart rate monitor to their car's entertainment system using a Raspberry Pi.

replies(2): >>35464741 #>>35465514 #
4. dang ◴[] No.35464741{3}[source]
Your comment makes me think that we should take it even more seriously. The whole idea of HN is intellectual curiosity. If the site isn't intellectually interesting, it fails—full stop.

Don't get me wrong: I'm not saying it is interesting. There are all too many ways in which it is not. I just mean that that's what what want it to be, and try to help it be. Even though we fall short.

Connecting a heart rate monitor to a car does sound kind of interesting to me though! but not with a dog - that sounds too close to abusive. But I could imagine someone making a biofeedback system where their heart drives their car or something. That would ba a great HN post.

replies(1): >>35464902 #
5. harry8 ◴[] No.35464850[source]
Pick any thread on C++ where someone suggests C has its place with reasoning. Any thread on clojure. Any thread on rust. Zero effort with popular with thread zeitgeist "wins". Reasoned supported dissent not so much.

Well I'm astounded that you let people "user flag" the story showing the evidence of government intervention in social media "moderation" which until it came out was a rank conspiracy theory. It's deeply controversial and upsetting for large numbers of people who would rather it were not true and I'm one of them. I would definitely rather it if it wasn't true about the suppression of a true story in the lead up to the election. I would definitely rather it wasn't true that the FBI, CIA, NSA etc were getting involved where they really should not. But it is true no matter how much I dislike that it is. It should be discussed widely and nowhere more so than here - it's our industry. How bad is it and how concerned should we be about it is a vital discussion, not be "user flagged" to be out of bounds. I completely understand that a noisy and vocal minority used to think Elon could do no wrong and his farts smelled like lavender and yet we had reasoned, sensible discussion here about his efforts in the Thai cave rescue. Nowadays a noisy and vocal group think Elon is in league with everything that is evil and shout it to the rooftops. Still we are likely to manage.

You didn't say it outright but I would take you at your word you, as in HN didn't moderate the story away and I would like you to confirm in the face of what we now know about pressuring moderators.

If it's "user flags" that did it are you being gamed on that? How do you know?

Would that be happening if it made Trump look like more of a crook than he is which plays to my prejudices just like 90+-ish % of those here? Zeitgeist. Major story that affects us as people, our community, many of our startups and specifically HN, which seems like it would be subject to similar pressures?

The major threads? I never saw them and I come here too often! ;-)

replies(1): >>35465557 #
6. p-e-w ◴[] No.35464902{4}[source]
I think the site is interesting overall. That doesn't mean every single post is, or has to be, for everyone.

And, ironically, even interestingness can get repetitive. When I see a post titled "How I made Netflix' video decoding on Android 25% faster", I know that I'm going to find a war story where some silly hardware bug was preventing proper cache management or something. It's interesting in a way, and if I read the full post I'm going to learn plenty of new things – but at the end of the day, I've seen (and done) it all before. Not this particular bug of course, but this type of story.

What I'm really looking for is ideas and thoughts that are completely new to me. Not just in their particulars, but in their general direction. That's very difficult to find. And "dissenting", "controversial", and "offensive" opinions are an absolutely crucial part of finding it, in my experience. A "sort by controversial" feature like Reddit has would be a godsend for HN.

7. concordDance ◴[] No.35465514{3}[source]
In my experience the lively ones aren't that productive. People start commenting mostly for the audience, with a heavier reliance on rhetoric and less on understanding others.
8. concordDance ◴[] No.35465557{3}[source]
> How bad is it and how concerned should we be about it is a vital discussion, not be "user flagged" to be out of bounds

It's also really hard to discuss because the ability of the average user to comprehend nuance seems to have gone down, instead pattern matching things to the nearest cliche. I've seen time and time again what should be a nuanced discussion having users mentally replace nuanced statements and facts with more readily accessible clichés.

I have no simple solutions to that issue, inspiring that kind of nuanced discussion would likely need it to explicitly gamified (e.g. having a specific award/karma that people can give out for someone else understanding the nuance of something someone else said and an equivalent down vote for rounding something off to a cliche) and even that probably wouldnt work.

9. revelio ◴[] No.35466044[source]
I think that touches on one of the things that could be easiest to improve. HN has upvotes and downvotes without any guidance on what people are meant to be voting about (agreement? post quality?). So people just downvote stuff they disagree with in the hope that other people won't see it.

Because flagging also affects visibility it's clear that many users simply use flagging as a kind of "super downvote" which it presumably isn't meant to be. Their goal is to suppress interesting discussion of interesting things uncomfortable for their world view, not to clean up spam. I always read with showdead on because the sheer quantity of interesting, useful posts that get flagged is well beyond the value of the flagging mechanism.

HN seems to be stuck in a form of circular reasoning in which flags are taken as a sign that some people are (or claiming to be) upset, therefore the discussion won't be "good", so it is OK to suppress it, which then encourages people to flag things. But this just empowers aggressive minorities who weaponize their own feelings to shut down interesting debates for everyone else. It seems counter to the mission.

A simple fix: put the ability to flag behind a very high karma threshold, write out a clear policy for how it's meant to be used i.e. what is considered rule breaking and what isn't, then take away flagging privs for people who consistently flag things that don't meet the policy.