Threatening violence is against Twitter's rules and yet world leaders have threatened war on Twitter. Those tweets deserve to stay up because they have public value, but Twitter needs some concrete policy on why a world leader saying they will "obliterate Iran" is different than a random person saying they will assault another random person. I am not sure this policy is a good answer to that, but they needed to do something.
> a world leader saying they will "obliterate Iran"
It's also worth noting that saying they would "obliterate Iran" under a specific set of circumstances is completely different from saying they will "obliterate Iran".
One is an (admittedly blunt) form of diplomatic and political threat, the other is a statement deliberately taken out of context to make someone look worse than they already do.
I disagree. One of the primary aspects of many threats is defining the specific criteria that would lead to violence. The point of a threat is usually not to actually inflict the violence but instead to use that possibility of violence as a motivation for changing behavior. It is roughly equivalent to preemptive terrorism. I therefore don't see how putting qualifications on the threat of violence should be enough to circumvent the rules. I would still expect to get in trouble for saying "I will kill you if you ever state that political opinion again" even though the threatened party could easily avoid any fear of violence by simply refraining from stating that opinion.
I suspect the answer is less “deserves to be kept up” and more “mustn't be pulled down lest we risk the wrath of or lose the attention caused by having the author”. Can you imagine if Twitter suspended the current President of the United States? They'd have the Republican political machine at their throats and lose a ton of attention and thus ad revenue that they currently get from his tweets.
Critical. I applaud them taking leadership here. I feel like the climate of the internet has us in a nasty quagmire where these companies are find themselves in a “damned if they don’t but perhaps double-damned if they do” kind of impasse. So kudos to twitter for actually trying.
I really hope they create a system whereby the public can audit their decisions and deliberations.
So what is the alternative here? Allow everything from copyright infringement to threats of violence to child porn?
That is what I don't get about the complaints of these companies censoring users. Every mature government has already agreed that completely unlimited free speech is not a good thing as speech can infringe on the rights of others. We are just debating about where exactly the line is now. You are free to complain about specific instances of Twitter censoring users, but the idea that they shouldn't be able to censor anyone is a lost cause.
If a world leader with 100k+ followers tweets out child porn and Twitter just tags it with a warning instead of removing it I agree that would be extremely controversial!
Twitter reserves the right to remove any Tweet. Those three examples were specifically used as a progression of severity and I would imagine we would all approve of Twitter removing the last example.
However the change to the policy that Twitter is detailing here isn't really about deleting tweets. This policy creates a middle ground for tweets that break the rules but shouldn't be deleted. Someone could even make the argument that middle ground gives Twitter the flexibility to censor less.
Yes, but "selectively enforcing" the rules is a subjective thing, not an objective thing. There will always be some people who say that ANY removal of ANY post is censorship.
So by definition there is no objectively "pure enforcement" possibility. And the other end of the spectrum, no enforcement whatsoever makes the site suck for lots of people so that's out of the question.
So there's going to be "selective enforcement," as you call it. And if you were in charge of enforcement, I bet you'd think you were being 100% objective, not selective/subjective, but I bet more that somebody would disagree.
I do think the impact of this will be overstated, in the big picture. The vast majority of Americans (and people) don't use Twitter. An editorial tag from twitter isn't really going to change the impact of a tweet in my opinion, other than limiting its exposure on Twitter itself.
But it would be hilarious if news orgs use this as a cue to give trigger warnings when they quote a tagged tweet. "Warning: the tweet we are about to show you violates Twitter's trust and safety rules; hide your children before we reveal what [prominent public figure] tweeted!"
I do agree with you on this. Especially this rule.
Abuse/harassment: You may not engage in the targeted harassment of someone, or incite other people to do so. This includes wishing or hoping that someone experiences physical harm.
Obviously there is a gray area where people use 'groups' rather than individuals to claim they are not targeting people.
I.E. I wish people would kill John Smith vs I wish all white people would die.
If I understand the new policy correctly, what Twitter is doing is taking an official step back from simply deferring to the Trust and Safety Council, by allowing public figures of sufficient stature to avoid having their posts removed, even if they do violate Twitter's rules.
Do you have a better solution?
I am honestly thinking on how would I solve the problem where I own a popular social network or forum/subreddit, I want to prevent it to become a shit place, so I would like to put some rules like HN and other communities have , how can I prevent the situation where people will call me as censoring them on my own community? My only solution would be to ban political discussion completely and implement some hard punishment for people that don't respect the community rules of conduct even banning the bad actors.
I actually think it's a great compromise, but I do wish users could see which Trust and Safety partner determined each violation on a tweet by tweet basis.
I.e. Twitter in partnership with [GLAAD; ADL; etc] determined this tweet contained [violation]. Then we could quantify, drill down and get some great insights.
I think your comment that politicians "better conform to Twitter's code of conduct!" if they want to use Twitter is a perfect summary of the underlying philosophy here. A system where you have bored teens posting alongside the POTUS, desirability aside, is very difficult to regulate. All this move amounts to is a statement that if you can't make your point without breaking the Twitter rules, it belongs somewhere other than Twitter.
In my own opinion, it is hugely irresponsible for a business like Twitter to give politicians a platform in the first place. If this is a move away (*s/toward/away) from that, by hindering usefulness of Twitter as a campaign tool, I'm completely in favor of that.
ps: TASC don't have any influence on which posts are considered rule-breaking, only on which are considered noteworthy enough to stay up. So they have no influence on "throttling" any tweets, only on saving ones which would otherwise be deleted.
1 point by creaghpatr 1 minute ago | parent | edit | delete [-] | on: Twitter adds labels for tweets that break its rule...
>When a Tweet has this notice placed on it, it will feature less prominently on Twitter, and not appear in:
>Safe search >Timeline when switched to Top Tweets >Live events pages >Recommended Tweet push notifications >Notifications tab >Explore
Getting out in front of the 2020 election, I see. Better conform to Twitter's code of conduct if you want to maximize reach!
Why does this decision change anything? They've been allowing rule-breaking material from politicians for a while, they just now let the user know and prevent the rule-breaking content from spreading.
No company should be attempting to define what's in public interest. Especially not Twitter. There's huge information asymmetry not to mention that they've already demonstrated themselves as being not even remotely politically neutral.
Elon already has the SEC getting on him for his tweets, I imagine they would go after Bezos too. Both would face shareholder lawsuits for anything sufficiently controversial, so it probably doesn't need to be as much of a priority for Twitter.
>We're not going to delete it but we're going to algorithmically prevent discovery
Tech monoliths like facebook, google, and twitter are starting to show very obvious political biases.
If we push them to be equitable to differing opinions, they basically get to continue enforcing their own political biases because we can only ever do blackbox testing. God knows what the backend is doing.
If we push them to own their biases try to get more platforms representing differing political opinions, we create admitted echo chambers and further polarize a political climate (in the USA) where we're already arguing whether communists or nazis are trying to destroy the country.
This is a complex problem without an easy solution but it's not unreasonable to bet twitter is going to abuse this new feature to either allow people they like to blatantly violate the rules (without actually algorithmically demoting the tweets, people will ignore the notice) or effectively shadowban people they don't like without explaining why a tweet doesn't follow the ToS until the subject matter is out of the news cycle.
> Tech monoliths like facebook, google, and twitter are starting to show very obvious political biases.
How is this policy political bias? If one political group makes more calls to violence or for discrimination than another perhaps the problem is with the group, not Twitter.
If not political advocacy groups, who would be a better consultant choice? Isn't the entire purpose of advocacy groups to encourage the safety of certain groups?
I think an apolitical philosophical framework would be better than _biased_ (however good their intentions) groups who have the mostly the well being of their interest group in mind. An advocacy group isn't a group which will take a balanced approach. Naturally, if they are advocates, they would represent their position _fully_ and uncompromised.
An apolitical (not advocating anything outside a framework) would be better and more neutral way to manage these differing opinions.
I disagree. Society is highly political, so why should Twitter be "apolitical" with their moderation? (if such a thing is even possible. I doubt the feasibility of apolitical moderation in general)
Nobody is saying this policy is political bias. The problem is with who gets to define things like discrimination.
As an example, there are those who believe any form of border security, or any enforcement of existing immigration law is discrimination and immoral. Which side defines the terms used is absolutely a form of bias.
If your political stance violates the TOS that is enforced on non-public offical users, is it really "political bias"? Maybe you should rethink your politics if bullying, inciting violence, and inciting hate speech is part of your position.
The problem will arise when twitter selectively enforces these rules on people with only one political stance.
Hypocrisy is a common human trait, and one that should be accounted for in such a process - it's hard to look at Twitter's past actions and assume they'll do this one perfectly.
Even if that's true, people on the right seem to be more likely to actually follow threats with physical violence. At least this is the case in Germany where the deaths caused by right-wing extremism since the beginning of the 90s number more than a hundred, while there have been only two deaths since 2001 caused by left-wing radicals. Unfortunately the article I read on this didn't have numbers for left-wing violence in the 90s, but I doubt they would make up for this significant difference.
Since we can apparently post any anecdata we want here now, I've seen the exact opposite.
I'm curious, what are the leftist counterparts to the Proud Boys, "Day of the Rope" posts, "Free Helicopter Rides" T-shirts, widespread embrace of Nazi imagery, and armed Pizzagate conspiracy crusaders?
Whether or not it's political bias depends on what's in the TOS and who gets to define "bullying", "inciting violence", and "hate speech".
I can definitely think of ways any of those could be defined by Trumpists. Suffice to say those definitions are likely to differ substantially from those that a typical Pride attendee might offer.
We've banned this account for taking HN threads into ideological flamewar. If you don't want to be banned, you're welcome to email hn@ycombinator.com and give us reason to believe that you'll follow the rules when posting here.
There is a great example in the thread below that might as well be surfaced here. When blue-checks on twitter post 'kill white people' it is magically not bullying, not inciting violence and not hate speech.
What is worse is the secondary effects when people point to twitter not taking action as a justification that posting 'kill <skin colour>' is somehow acceptable if <skin colour> == white. After all, twitter allows it so it must be OK.
- Teddy Wilson was mocking the "white genocide" conspiracy theory. He's also white.
- Dadito Calderone was making a joke about people who pretend to be concerned about crime in black neighborhoods when really they're trying to demonize black people as criminals.
- ka5sh, and possibly Brian Clevinger, is quoting the TV series Roots.
- Bobbie Oliver was talking about a movie in which a black man escapes slavery by killing his captors.
- Shay Stweart Bouley was sharing a story with an intentionally provocative title
- Tracy Bommeisha-Ann Clayton was ironically using a painfully exaggerated black accent.
- Yasser Lester is saying something similar to Nick Hanauer telling plutocrats to beware of pitchforks and torches.
None of these people were actually advocating for killing white people, nor or they making these jokes in a targeted way that would constitute bullying. Furthermore, many of these tweets are many years old, which indicates that they were found by scraping the bottom of the barrel.
I am pretty sure that if one would take the approaches you have outlined - e.g. quoting some TV show, making a 'provocative title', speakig in an exagerrated accent -- but with the people <colour> swapped, they would be flagged in a relatively short order. No matter how bening the motive, the message is not OK. The fact that you think that it is OK and make excuses for it is the illustration of my GP post's last point.
I disagree. If you go looking, you can find verified accounts suggesting killing Hispanics [1], putting Jews in ovens [2], killing Indians [3], and otherwise committing genocide [4].
>We're not going to delete it but we're going to algorithmically prevent discovery
Tech monoliths like facebook, google, and twitter are starting to show very obvious political biases.
If we push them to be equitable to differing opinions, they basically get to continue enforcing their own political biases because we can only ever do blackbox testing. God knows what the backend is doing.
If we push them to own their biases try to get more platforms representing differing political opinions, we create admitted echo chambers and further polarize a political climate (in the USA) where we're already arguing whether communists or nazis are trying to destroy the country.
This is a complex problem without an easy solution but it's not unreasonable to bet twitter is going to abuse this new feature to either allow people they like to blatantly violate the rules (without actually algorithmically demoting the tweets, people will ignore the notice) or effectively shadowban people they don't like without explaining why a tweet doesn't follow the ToS until the subject matter is out of the news cycle.
79 comments
[ 0.30 ms ] story [ 158 ms ] threadIt's also worth noting that saying they would "obliterate Iran" under a specific set of circumstances is completely different from saying they will "obliterate Iran".
One is an (admittedly blunt) form of diplomatic and political threat, the other is a statement deliberately taken out of context to make someone look worse than they already do.
I really hope they create a system whereby the public can audit their decisions and deliberations.
>Safe search >Timeline when switched to Top Tweets >Live events pages >Recommended Tweet push notifications >Notifications tab >Explore
Getting out in front of the 2020 election, I see. Better conform to Twitter's code of conduct if you want to maximize reach!
Edit: Good time to get to know who will be making the decisions on throttling exposure- do not cross these groups! https://about.twitter.com/en_us/safety/safety-partners.html
That is what I don't get about the complaints of these companies censoring users. Every mature government has already agreed that completely unlimited free speech is not a good thing as speech can infringe on the rights of others. We are just debating about where exactly the line is now. You are free to complain about specific instances of Twitter censoring users, but the idea that they shouldn't be able to censor anyone is a lost cause.
However the change to the policy that Twitter is detailing here isn't really about deleting tweets. This policy creates a middle ground for tweets that break the rules but shouldn't be deleted. Someone could even make the argument that middle ground gives Twitter the flexibility to censor less.
So by definition there is no objectively "pure enforcement" possibility. And the other end of the spectrum, no enforcement whatsoever makes the site suck for lots of people so that's out of the question.
So there's going to be "selective enforcement," as you call it. And if you were in charge of enforcement, I bet you'd think you were being 100% objective, not selective/subjective, but I bet more that somebody would disagree.
But it would be hilarious if news orgs use this as a cue to give trigger warnings when they quote a tagged tweet. "Warning: the tweet we are about to show you violates Twitter's trust and safety rules; hide your children before we reveal what [prominent public figure] tweeted!"
The vast majority of Americans hold little of the country's wealth, so the actions of rich people have little effect.
Abuse/harassment: You may not engage in the targeted harassment of someone, or incite other people to do so. This includes wishing or hoping that someone experiences physical harm.
Obviously there is a gray area where people use 'groups' rather than individuals to claim they are not targeting people.
I.E. I wish people would kill John Smith vs I wish all white people would die.
Twitter Verified Hate - https://www.label56.com/wp-content/uploads/verified-hate-1.j...
I.e. Twitter in partnership with [GLAAD; ADL; etc] determined this tweet contained [violation]. Then we could quantify, drill down and get some great insights.
In my own opinion, it is hugely irresponsible for a business like Twitter to give politicians a platform in the first place. If this is a move away (*s/toward/away) from that, by hindering usefulness of Twitter as a campaign tool, I'm completely in favor of that.
ps: TASC don't have any influence on which posts are considered rule-breaking, only on which are considered noteworthy enough to stay up. So they have no influence on "throttling" any tweets, only on saving ones which would otherwise be deleted.
>When a Tweet has this notice placed on it, it will feature less prominently on Twitter, and not appear in: >Safe search >Timeline when switched to Top Tweets >Live events pages >Recommended Tweet push notifications >Notifications tab >Explore
Getting out in front of the 2020 election, I see. Better conform to Twitter's code of conduct if you want to maximize reach!
Good time to get to know who will be making the decisions on throttling exposure- do not cross these groups! https://about.twitter.com/en_us/safety/safety-partners.html
This political cartoon sums it up nicely. https://i.imgur.com/kLkMthu.png
This isn't as disturbing as Vimeo's takedown of Project Veritas, or the bias exposed in their video, but the trend is seriously alarming.
I totally agree with the second part of your statement though.
However if they were to always publish things like this which are factually incorrect and untrue they would be sued for libel.
Oh look they have been sued for libel many many times. Try a google search for "project veritas libel suit" and see how many convictions you find.
Tech monoliths like facebook, google, and twitter are starting to show very obvious political biases.
If we push them to be equitable to differing opinions, they basically get to continue enforcing their own political biases because we can only ever do blackbox testing. God knows what the backend is doing.
If we push them to own their biases try to get more platforms representing differing political opinions, we create admitted echo chambers and further polarize a political climate (in the USA) where we're already arguing whether communists or nazis are trying to destroy the country.
This is a complex problem without an easy solution but it's not unreasonable to bet twitter is going to abuse this new feature to either allow people they like to blatantly violate the rules (without actually algorithmically demoting the tweets, people will ignore the notice) or effectively shadowban people they don't like without explaining why a tweet doesn't follow the ToS until the subject matter is out of the news cycle.
How is this policy political bias? If one political group makes more calls to violence or for discrimination than another perhaps the problem is with the group, not Twitter.
An apolitical (not advocating anything outside a framework) would be better and more neutral way to manage these differing opinions.
As an example, there are those who believe any form of border security, or any enforcement of existing immigration law is discrimination and immoral. Which side defines the terms used is absolutely a form of bias.
If your political stance violates the TOS that is enforced on non-public offical users, is it really "political bias"? Maybe you should rethink your politics if bullying, inciting violence, and inciting hate speech is part of your position.
Hypocrisy is a common human trait, and one that should be accounted for in such a process - it's hard to look at Twitter's past actions and assume they'll do this one perfectly.
I'm curious, what are the leftist counterparts to the Proud Boys, "Day of the Rope" posts, "Free Helicopter Rides" T-shirts, widespread embrace of Nazi imagery, and armed Pizzagate conspiracy crusaders?
I can definitely think of ways any of those could be defined by Trumpists. Suffice to say those definitions are likely to differ substantially from those that a typical Pride attendee might offer.
Any of the 3 is cool.
https://news.ycombinator.com/newsguidelines.html
https://www.label56.com/wp-content/uploads/verified-hate-1.j...
What is worse is the secondary effects when people point to twitter not taking action as a justification that posting 'kill <skin colour>' is somehow acceptable if <skin colour> == white. After all, twitter allows it so it must be OK.
- Dadito Calderone was making a joke about people who pretend to be concerned about crime in black neighborhoods when really they're trying to demonize black people as criminals.
- ka5sh, and possibly Brian Clevinger, is quoting the TV series Roots.
- Bobbie Oliver was talking about a movie in which a black man escapes slavery by killing his captors.
- Shay Stweart Bouley was sharing a story with an intentionally provocative title
- Tracy Bommeisha-Ann Clayton was ironically using a painfully exaggerated black accent.
- Yasser Lester is saying something similar to Nick Hanauer telling plutocrats to beware of pitchforks and torches.
None of these people were actually advocating for killing white people, nor or they making these jokes in a targeted way that would constitute bullying. Furthermore, many of these tweets are many years old, which indicates that they were found by scraping the bottom of the barrel.
[1] https://twitter.com/KillerMike/status/612315197663940608.
[2] https://twitter.com/TheBillyProcida/status/62535136510415257...
[3] https://twitter.com/ZaidZamanHamid/status/505972865766658048
[4] https://twitter.com/AlexisGZall/status/604026752013230080
Tech monoliths like facebook, google, and twitter are starting to show very obvious political biases.
If we push them to be equitable to differing opinions, they basically get to continue enforcing their own political biases because we can only ever do blackbox testing. God knows what the backend is doing.
If we push them to own their biases try to get more platforms representing differing political opinions, we create admitted echo chambers and further polarize a political climate (in the USA) where we're already arguing whether communists or nazis are trying to destroy the country.
This is a complex problem without an easy solution but it's not unreasonable to bet twitter is going to abuse this new feature to either allow people they like to blatantly violate the rules (without actually algorithmically demoting the tweets, people will ignore the notice) or effectively shadowban people they don't like without explaining why a tweet doesn't follow the ToS until the subject matter is out of the news cycle.
Oh good, another reason to always sort chronologically (until Twitter removes the ability to do so)
In other words, so-called public interest doesn't exist, it's just an excuse used by Titter to shape opinions and censor political opponents.
Unlike around last election, this time Twitter will lose influence and users.