Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

You either die an MVP or live long enough to build content moderation

mux.com

Post reply on HN
Print view View original

381–390 of 429 posts

Re: You either die an MVP or live long enough to build content moderation

#381

Post by specialist »

armchairhacker wrote: post #245
Animats wrote: post #216

Roblox is working on a "moderation" system that can ban a user within 100ms after saying a bad word in voice. But their average user is 13 years old. Interestingly, Second Life, the virtual world built of user-created content, does not have this problem. Second Life has real estate with strong property rights. Property owners can eject or ban people from their own property. So moderation, such as it is, is the respon…

The biggest reason IMO for moderation in the first place is because if you don't block/censor some people, they will block/censor others. Either by spamming, making others feel intimidated or unwelcome, making others upset, creating "bad vibes" or a boring atmosphere, etc. So in theory, passing on moderation to the users seems natural. The users form groups where they decide what's ok and what's banned, and people jo…

> ...if you don't block/censor some people, they will block/censor others.

Terrific rephrasing of the paradox of tolerance. Thank you.

Re: You either die an MVP or live long enough to build content moderation

#382

Post by vanderZwan »

LuisMondragon wrote: post #252
Alir3z4 wrote: post #17

Around 1 year ago we got hit badly on our [blogging platform][0] by people/groups submitting fake customer support description of other big companies, either being Microsoft, Facebook, Comcast etc. Rolled out a machine learning model and trained it on the database. 99% of them vanished. Next day, the machine didn't work and success rate was around 5%. Found out, they have learned the trick and now using symbols from…

> Found out, they have learned the trick and now using symbols from different languages to make it look like English. I wonder if you can train a ML model using text as images . For example, taken as strings, "porn" and "p0rn" are not very similar, but visually they are.

Are you familiar with the Xerox copying flaw? Not exactly the same problem and unlikely to apply here but still a fun tangent:

https://www.theregister.com/2013/08/06/xerox_copier_flaw_mea...

Re: You either die an MVP or live long enough to build content moderation

#383

Post by morenoh149 »

nebula8804 wrote: post #307
Alir3z4 wrote: post #94

Earlier quoted context omitted.

Yep, shadow ban was in place as well. The thing is, the people that weren't the ones posting the content. It appeared their computer was affected by some type of malicious file to be part of a bigger network. (botnet?) I could see from the thousands of different IPs in different countries around the globe, that it could be affected personal computers. Very few of them were computers from hosting companies, the rest w…

This seems like a lot of effort? What do you think was the motivation for something like this on your site?

The cynical answer is backlinks haha

Re: You either die an MVP or live long enough to build content moderation

#385

Post by spurgu »

woko wrote: post #363
michaelpb wrote: post #73

Earlier quoted context omitted.

Yeah, there's an entire category of "idea guys" who don't get this. They repeatedly try to crack the code on a truly moderation-free or purely crowd-moderated platform, and it never, ever, ever works. It almost always boils down to a poor understanding of how humans work (usually some sort of "homo economicus") or how computers work (usually some sort of "AI magic wand").

It depends on the scale. I can vouch that crowd-moderation works fine for a small forum (~ 1000 members) which I am part of. And there is no karma system. You get to report posts (3 reports mean that the post is deleted), and warn users (24-hour ban after 3 "active" warnings, and then it scales up to a permanent ban after 15 "active" warnings). Warnings become "inactive" after a month. It also depends on the threat m…

> You get to report posts (3 reports mean that the post is deleted), and warn users (24-hour ban after 3 "active" warnings, and then it scales up to a permanent ban after 15 "active" warnings). Warnings become "inactive" after a month.

Sounds like I could do some damage there by signing up with three accounts?

Re: You either die an MVP or live long enough to build content moderation

#386

Post by notacoward »

hnbad wrote: post #355
Goronmon wrote: post #100

Earlier quoted context omitted.

I'll never forget having to be a moderator for a somewhat popular forum back in the day and oh man did I learn how a few people can make your life hell. I was also a mod for a popular gaming forum way back in the day. It was pretty miserable looking back. Personally, for me, the extreme/shocking content wasn't the biggest issue. That stuff was quick and easy to deal with. If you saw that type of content you just imme…

> What was a lot harder were the toxic users that just stuck around. Not doing anything bad enough to necessarily warrant a permanent ban, but just a constant stream of shitty behavior. Especially sometimes when the most toxic users were also some of the most popular users. This is the problem with community guidelines being the be-all and end-all. Hard rules are great for catching insults or slurs. They're not so gr…

> Hard rules are great for catching insults or slurs.

Bingo. Hard rules encourage brinksmanship. There's always a class of "picador" users who will poke and prod and provoke just up to the line where the rules are, then flag the response. A moderator too wrapped up in rule by technicality (or too lazy to look at context) will then come down as harshly as they can on the author of the flagged comment, and give the picador a total pass. Problem is, the picador does this again and again and again, never making a positive contribution, while their targets are often chosen precisely for their prominence. Guess which one is encouraged to continue their behavior, and which one is encouraged to go away. Has the "moderator" helped to improve discourse on the site, or helped to ruin it?

Re: You either die an MVP or live long enough to build content moderation

#387

Post by hnbad »

tomjen3 wrote: post #72
dylanjha wrote: post #2

Hey there :). Author here. It was a fun experiment to play around with some different strategies for adding content moderation to https://stream.new Hive ended up being the one I landed on after trying Google Vision first ( https://cloud.google.com/vision ). The other one I was looking at is Clarity.ai but I didn't get a chance to try that one yet.

One thing that wasn't obvious to me is, why did you care about uploads of NSFW? As I understood it, you want to become Imgur of video. Imgur only became so big because they allowed NSFW stuff.

In addition to what others have said:

* Porn is almost always posted in violation of copyright.

* Hosting porn opens you up to legal issues if you can't verify that everyone involved is adult and consenting.

* Payment processors, hosting companies and other service providers you rely on usually have strict policies excluding porn.

And that's just addressing legal pornography, not other "NSFW" content like child abuse, animal abuse, general violence or gore. If you run a large enough public user generated content service people will use it to distribute illegal content or flood it with jihadist execution videos to ruin someone's day.

Re: You either die an MVP or live long enough to build content moderation

#388

Post by derekzhouzhen »

unraveller wrote: post #361
derekzhouzhen wrote: post #83

There is no correct way to do the content moderation, so on my website, https://roastidio.us I don't plan to have any. My rules are simple: * you can say whatever you want to say, but only one person, the one you replied to, get to see it other than youself * the one you replied to get to approve whether it can be shown to other people. No further reply is allowed until this one is approved. * Even the one you replie…

The opt-in to see comments and user based approval method has been tried before. It's just pre-banning and always fails to gain steam on websites because it achieves what banning aims to achieve, it places a long artificial delay to time-sensitive posts. Many better places with more rewarding interactions exist to achieve the same end. Something tells me that these roast-posts (or "this shouldn't be in anyone's feed"…

Agreed. I don't want to gain steam though. I want to facilitate meaningful, one-on-one, anonymous discussion, through minimal interactions with my website. Showing the resulting discussions to the general public only comes second.

Re: You either die an MVP or live long enough to build content moderation

#389

Post by NineStarPoint »

pier25 wrote: post #195
Alir3z4 wrote: post #128

Earlier quoted context omitted.

Oh yes. Never ending. We still keep the free plans even through there are abusers, but that won't be a reason to retire it. So many people using for legitimate reasons and keeping their personal writing there. It's unfortunate, but well, it happens.

Why do you think they were using your platform to do this? Is it because these scammers were non technical? Or maybe the anonymity provided by your platform?

If I was this sort of scammer, I’d have the text being posted and progression of how hard it is to figure out pretty well ironed down. Then you just try it on lots of platforms to find the ones that don’t figure out how to solve the issue. So my guess would be they tried because they might as well, and then they hit the limit of how much effort they had put into their own generation systems and moved on to the next mark.

Re: You either die an MVP or live long enough to build content moderation

#390

Post by patall »

raxxorrax wrote: post #369
patall wrote: post #170

One interesting approach I heard about in this domain is the Trolldrossel (German for troll throttle) by Linus Neumann from CCC. He implemented a captcha test for a comment server that would fail with a certain percentage when encountering certain key-words in the comment, even when the captcha was solved correctly. While I have no notes about the effects and the corresponding talk seems to have vanished from the int…

I think this is a really bad solution that only reinforces bad practices of hiding content. Even worse than direct censorship, it generates more problems than it solves.

When everyone knows it, I agree. As long as it is not mentioned anywhere, and people do not write hundreds of comments, its just a failed captcha that appears more often when someone writes something in range, possibly forcing them to rethink their argument. Its no solution for all, more an interesting thought.
Post reply on HN