Content moderation is broken (2019)
eff.org
eff.org
I'm kind of disappointed that the EFF never issued a correction or retraction for this claim.
> users being censored for engaging in counterspeech or for using reclaimed terms
A few paragraphs up, weren't they saying it was a bad thing that different rules apply to different people? Now they're saying it's a bad thing if the same rules apply to everyone?
What kind of correction or retraction would you be looking for? That is explicitly Twitter policy[1].
Trump may have eventually crossed a line. That doesn't mean his line was in the same place as other people's.
[1] https://help.twitter.com/en/rules-and-policies/public-intere...
Trump isn't the only one who had a different line, though his was probably looser than anyone else got. Any celebrity or public figure is going to have more leeway in what he or she is allowed to say.
Or they're saying that the rules themselves are problematic, and more sophistication is needed than just "this word is not allowed."
No, the rule is you can’t use reclaimed slurs unless you’re part of the group targeted by them. So yeah, you don’t get to call someone who’s gay a fa__ot unless you are gay yourself. That’s like the whole point of reclaiming them.
The issue is that a dumb keyword filter can’t properly enforce this.
Are we going to have the content moderation team guess the sexual preferences of random twitter accounts?
No, he was still President when he got banned.
For example:
- Blacklist using censor X for nsfw content with anything above level 3.
- Blacklist using censor Y for political content to filter any ad hominem
- Blacklist using censor Z on scientific studies and information to filter anything with small sample size or that’s politicized
Obviously rather complicated, but this is what I want. A censor doesn’t have to be an org, it could be a person or a network of people you trust / follow.
Yes bubbles will form. That’s fine, that’s our right to join and filter on a bubble. Each bubble will exhibit different behaviors, and much like nature the fittest will rise and “survive”.
The amount of work required just goes way beyond checking some posts here and then towards the day, and distributing the load towards a huge community of moderators would give other issues on consistency, internal battles and keeping up to date on changes.
To note the people getting paid don’t need to be employed by the platform, I’d totally see a community crowdfund a moderator to tag for them specific content.
If there were 5 Twitters all competing against one another, they would have to have customer service since you could move to another company. The companies would actually allocate the necessary resources and that would dampen the worst excesses.
Until somebody (US, EU, etc.) gets serious and starts busting up these companies that rely on "network effect", this will all continue.
Your idea would work wonderfully for the first point, assuming you had enough censors devoting enough time to the social networks you care about. If this scheme was adopted, people would be able to craft the perfect echo chamber, seeing only what they want to see, being exposed only to the ideas they want to think about. It would be the current algorithms on steroids.
But it would totally fail for the second point.
When $social_media_empire deplatforms someone calling for violence against the Jews, it's not (entirely) out of altruism, it's because their investors don't want to see headlines declaring $social_media_empire as the home of violent, holocaust-denying sociopaths.
It's not enough to hide $antisocial_behvior from people that don't want to encounter it, it's important to the platform that the platform not be seen as a home for that type of content. So, at some point, it becomes necessary to hide content even from the people who want to see it.
Inciting violence or holocaust denying is outside the spectrum of what can be accepted on any moderation platform (outside of the sheer legal issues, also for the reasons you are citing). Some people want those kind of speeches represented as well, but it just won’t happen, at least not on any platform with public and international exposure.
With anything illegal off the table, sthe debate is more on how to deal with things ranging from flat earthers, hate on fat, furries, sex, anti-capitalism, hyper-capitalism, bigotry, grifting, crypto bros etc
In that “somewhat legal but possibily be deeply disturbing or unwelcoming from some” fringe, advertisers will also gain from being allowed to filter on what moderation strategies people are choosing. Basically it gives them ways to dictate specific rules for their content (e.g a diet product could be fine being advertised next to sex and fat-hate but want nothing to do with crypto-bros, etc.)
For example, I've seen many ad hominem attacks on one group of people but do the exact same thing to another more "popular" and "in" group and you get ban-hammered.
You can't really trust anyone. Unless you want biased content - if that is the case, then censor away and be happy in the bubble.
While current large social networks are not a real monopoly, they're not far away... if you want profiles, photos etc., you're pretty much stuck with facebook (this also counts instagram), forums are replaced by reddit and chat is concentrated on a few major platforms (whatsapp... so facebook again, whatever the current google chat is, viber, telegram and maybe signal).
While saying "fuck you guys, i'm building my own reddit, but with hookers, and blackjack!" is simple, and setting up some reddit-like software is simple, moving users there is hard... that goes double for stuff like twitter or facebook, where in same cases, you need to use those to even get official government info.
And even if you do succeed with making your own facebook, due to hookers and blackjack, even if perfectly legal (free spech and everything), hosting providers and domain registrars might turn off your service, because you're doing something perfectly legal, but they still don't like it.
Proving a monopoly would probably be impossible, but avoiding one of the big social networks is becoming harder and harder... and who knows what new service someone like facebook or google might buy... could be telegram, could be viber,.. you never know.
If you look further (eg china), it becomes ever harder to avoid those, since many real-life services (eg restaurants) accept payment only via WeChat ("ap pfor everything"), and if you get banned from there in china, you're even more fucked than if you lose your google account over here.
What seems inevitable is the US government starts hosting public forums to talk; but being bound by free speech. Virtually anything goes unless it's an actual crime. What happens after this?
How can you give people a platform, but at the same time, tame the conversation?
Everyone gets tired of 19 out of every 20 posts being about why they need Viagra, and goes back to MySpace.
So "just don't follow" is not only meaningless advice but it's painfully naive.
If you use Twitter as is, there are a lot of things promoted for you but it's not any sort of public feed, it's (theoretically) based on who people you follow follow, who they like tweets from consistently, that sort of thing. No Viagra ads are going to be showing up in that.
> Please respond to the strongest plausible interpretation of what someone says, not a weaker one that's easier to criticize. Assume good faith.
> Please don't post shallow dismissals
Perhaps you should give these a read and edit your comment accordingly: https://news.ycombinator.com/newsguidelines.html
We better tell every big tech company that what they’ve been doing forever is impossible to do if they don’t also de platform those deplorables.
So lets say advertising is considered free speech and you may freely do it. In essence it would be a tragedy of the commons. The out of control advertising would eliminate the usefulness of the forum.
We know exactly what happens because it's been attempted. It fills up with spam and screaming bigots and nobody uses it beside spammers and screaming bigots. Which is why the the only 'inevitable' thing about such a government-run forum is that it won't happen.
When have we ever had a government-run social network before? And why aren't physical, offline public squares full of screaming bigots?
We've had many failed attempts at 'anything as long as it's not breaking a law' online spaces.
why aren't physical, offline public squares full of screaming bigots?
Because we impose severe social costs on people for that sort of thing. The screaming bigots are only bold and screaming when they feel they can avoid them. An 'anything goes' pseudonymous online forum is effectively a gift to them and mostly them - most people prioritize other things in their online social spaces far above 'anything goes'.
They are, but only now and then.
The reason is that, for most (not all) communities, there are only a few screaming bigots within easy reach of the town square, so their demonstrations often look like this[0].
With teh Internets tubes, you can gather together all those little groups of nutters, into one big nutball, so it looks like there's a lot more of them, per capita, than there actually are.
Because the security guards responsible for the event probably throw them out. If you behave like a crazy person at a public debate or gathering you'll probably get removed from the venue or delegated to the protest area with the other crazy people and their signs.
If you're attenting a townhall, a community meeting, a church gathering or what have you you've always been expected to behave like a civilized human being, moderation is omnipresent in the good old offline world. This anthing goes mentality on the internet has no precedent.
https://en.wikipedia.org/wiki/Forum_(legal)
https://en.wikipedia.org/wiki/Freedom_of_speech_in_the_Unite...
4chan
Then you would, forgo all the arguments, the arm chair lawyering, all the whataboutism, and trolls wouldn't have rules to push up against. No one could ever "technically not break any rules" by being an ass.
When companies try to act like a law institution, they get all the politics of making and enforcing laws dropped in their lap.
There is also censorship when a comment is flagged even without actual reason.
Another particularly annoying form of censorship is being unable to defend a position when providing something controversial which limits your submission rate.
HN often comes out with very tone deaf responses and takes, so I don’t think that it is the best argument. HN is also a rather niche community that caters to a very specific audience.
That is, essentially, what the terms of service for every social media platform (that no one ever bothers to read) have always said.