Australia puts Twitter on notice for hate speech content

[Kisai]

26 minutes ago, Sauron said:

Which might work if they had a sufficient moderation team, which they don't anymore.

The point is that any amount of porn is not illegal in most countries whereas hate speech is regulated in places like australia and the EU. Old twitter's moderation was certainly not perfect but it clearly prioritized getting rid of things that could place the platform in violation of laws and stuff that would cause significant parts of their userbase to leave, hate speech fitting both categories. Right now moderation seems to be exclusively dictated by elon musk's personal beefs and things he personally cares about having or not having on the platform, which is worse both for users and for twitter's ability to avoid legal action.

 

And let's not forget that the platform was more ready than ever to  l i t e r a l l y  censor turkish electoral discussion on request by the government of turkey, as opposed to the bullshit definition of censorship where it's just a private platform not wanting you to post the N word.

The point is, Twitter is worse under Emperor Elon than it ever was before. Elon doesn't care. Basically digital nepotism "my friends can stay, but you must go"

[wanderingfool2]

5 hours ago, Kisai said:

The point is, Twitter is worse under Emperor Elon than it ever was before. Elon doesn't care. Basically digital nepotism "my friends can stay, but you must go"

It's all going to depend on what you consider to be worse, it's all relative and evidence only gets "presented" on worst case assessments.  Even that tweet you posted goes into a hyperbole of what is said; as the "cis" term used to target someone who considers it offensive is what appears to be bannable.  Just like directing the N word towards someone repeatedly would be bannable (and in general I'm assuming tweets using those words would be torpedoed in engagements).

 

Remember what occurred under Yoel Roth (Trust and Safety):

He himself tweeted out essentially a political party was "actual Nazis"

Utilized "bans"/"verify account" to get people to enter their cell phones, which was later used for marketing (I had my account banned because I refused to verify I was a person)

Underaged pictures were posted and literal homeland security had to get involved to get it removed

Removed all posts to articles regarding a certain laptop (despite it not breaching their policies)

 

Twitter literally dragged their feet or outright ignored some of the trafficking of minors; generally speaking it depends on what you value more and it relies on data that isn't released.

 

An issue with lack of censorship in general is that people are going to get mad, and that you will always find accounts that are toeing the line...but you could have found accounts that are like that before as well.  The issue with too much censorship or censoring like what Twitter or YouTube did is you get issues like what you had happen during the pandemic.  Early on tweets/videos encouraging masks were deemed misinformation (before the official studies were reviewed); later on J&J and the TTS were labelled misinformation/removed until the final studies were accepted; the drug companies were literally allowed going on saying "100% efficacy against death" but weren't deemed misinformation.  I think cases like that highlight that it might not be that different than before.

 

  

6 hours ago, Sauron said:

The point is that any amount of porn is not illegal in most countries whereas hate speech is regulated in places like australia and the EU.

The unsolicited exposure to it is, and the giant swaths of underaged or non-consensual stuff is very much illegal (and Twitter seemingly willfully turned a blind eye to that)

 

  

6 hours ago, Sauron said:

And let's not forget that the platform was more ready than ever to  l i t e r a l l y  censor turkish electoral discussion on request by the government of turkey

The LEGAL requirement and absolute request to black accounts by Turkey

[Brooksie359]

9 hours ago, Spotty said:

That's "impressions" ie. the number of views hate speech content receives. It is good that it is lower but it doesn't tell the full picture like the actual number of removed hate speech content, the number of banned accounts, the time it takes to remove that content after it's reported, etc. It just means that the algorithm isn't promoting hate speech as much (which is good). It might mean that people scrolling their feed are seeing less hate speech content but there might be a rise in directed hate speech content that people are receiving on the platform which is getting less overall impressions.

 

The line they draw for pre-acquisition is right at the top of all of the peaks and the line they draw at the end after acquisition is below the bottom of the valleys. Hard to say just by eyeballing it but if that is meant to represent the median over that period that doesn't look right.

 

That data is from January 2022 - June 2022. It doesn't cover the period that this notice is concerning (post-Musk's takeover).

 

That's actually useful information though since it's those legal requests from governments to remove content which this notice is about. It'll be interesting to see these same stats for the January 2023 - June 2023 period to see if the number of legal requests increased or decreased.

I believe that is what Elon said his solution was. It was to suppress speech not ban it. I mean if someone posted hateful content and nobody sees it does is it an issue? Also I would imagine impressions probably matters even more as less impressions usually mean less people see it and that would effectively mean less hateful content on the platform being seen by people. 

[Kisai]

1 hour ago, Brooksie359 said:

I believe that is what Elon said his solution was. It was to suppress speech not ban it. I mean if someone posted hateful content and nobody sees it does is it an issue? Also I would imagine impressions probably matters even more as less impressions usually mean less people see it and that would effectively mean less hateful content on the platform being seen by people. 

It is if it's still searchable from dog whistles.

 

If I search twitter for "nazi slogans" or "child brides" I should see zero results. But the people posting this stuff should be under the illusion that their hateful or gross tweets aren't being straight up deleted.

 

Because that's how you keep that material from becoming proliferated. You know "warez?" all the l33t speak that script kiddies used in the late 90's? That was done intentionally to make the stuff you shouldn't know about from being found by people who were not "in" the scene. Most of that stuff isn't done now because you can generally find anything you want in google, and don't need to go looking on invite-only systems.

 

Law enforcement would prefer that the idiots keep posting their crimes on facebook and twitter like nothing is being done, so they can just grab them once they locate the producer of the illegal materials.

 

[Brooksie359]

1 minute ago, Kisai said:

It is if it's still searchable from dog whistles.

 

If I search twitter for "nazi slogans" or "child brides" I should see zero results. But the people posting this stuff should be under the illusion that their hateful or gross tweets aren't being straight up deleted.

 

Because that's how you keep that material from becoming proliferated. You know "warez?" all the l33t speak that script kiddies used in the late 90's? That was done intentionally to make the stuff you shouldn't know about from being found by people who were not "in" the scene. Most of that stuff isn't done now because you can generally find anything you want in google, and don't need to go looking on invite-only systems.

 

Law enforcement would prefer that the idiots keep posting their crimes on facebook and twitter like nothing is being done, so they can just grab them once they locate the producer of the illegal materials.

 

I guess but I am not sure why that would matter much because people who ate searching for that type of stuff are going to be able to find it even if it's not on twitter. I think the big idea is to make sure that people who aren't looking for it don't see it as much as possible because it's impossible to get the number down the 0

[Donut417]

On 6/22/2023 at 2:08 PM, Dirtyshado said:

Yes you can be criminally charged for breaking a law in a country you never been in, extradition is also an option.

Is this a criminal or a civil law or does it do both? Because generally laws that regulate business tend to have Civil penalties and nothing more. Well out side of doing creative accounting and getting fraud or tax evasion charges. I haven't read the law and I dont understand legalese, which is why I ask. 

 

Extradition only works if the other country agrees. For someone like Musk or another executive the US would be hard pressed to agree to that. 

 

On 6/22/2023 at 2:42 PM, Dirtyshado said:

always find the US citizens interpretation of the law funny, especially in international agreements, considering the US Government is a key signature in most international agreements including criminal codes, trade, communication, banking, security, human rights, extradition, taxation etc. 

Same could be said about everyone outside the US when talking about copy right laws. Because technically the US can use extradition for that purpose as well. 

[Sauron]

19 hours ago, wanderingfool2 said:

The LEGAL requirement and absolute request to black accounts by Turkey

Also known as literal government censorship, by its very definition. Musk was more than willing to immediately comply with that but seems to be unwilling to put in bare minimum moderation for hate speech, despite its requirement by other countries.

[SolarNova]

Oh no ! ..Twitter is no longer censoring things we disagree with anymore ..quick lets call 'ism this ..ism that ..ist this ..phobe that " and force them back so we dont get our feelsies hurt !

 

Seriously ...Take the L and move on.

 

[Salv8 (sam)]

On 6/22/2023 at 7:56 PM, Spotty said:

I don't believe Australia's eSafety commissioner has any actual powers to compel Twitter to make changes.

they don't, but they'll make enough of a fuss for long enough till 2025 when our election for next prime minster is due.

for those who think the Australian government is a representation of Australia as a whole, you should really look into the average Australian, then compare it to the average Australian politician.

they are so different that is my actual belief that they didn't even grow up in Australia.

this is more philosophical then tech but bear with me i'll get to the tech.

hate speech is dependent on the person you ask, take CGP Gray's video on Native American vs Indian, the further you go from an Indian reservation the less Indian is said and the more Native American is said in it's place when talking about them. what one person considers offensive/hate speech other doesn't.

while there are sentences out there that exist purely to hurt people that are categorized as hate speech. there are also sentences out there that don't but get categorized anyways because someone saw it and didn't like it/disagreed with it..

take the sentence i'll quote:

Quote

aw yea nah but the house is a bit woggy ay?

to Mediterranean's who migrated to Australia in the 70's-90's, wog is considered very offensive, but their children don't because they have co-opted the term for themselves much like Indians and African Americans.

we know the context here and can understand it.

a computer can't.

it'll see wog, and deem it hate speech. and do the relevant action for this offense.

we don't have ai's that can understand context, we are no where near that even with the recent advancements in ai. with our current tech we have two options:

1. consider all uses of a offensive term (no matter the context) inexcusable

or

2. don't moderate it and rely on your user base not to say a no no word (it's the internet so like hell that'll happen)

both are shit options and neither have good outcomes. until we get auto mods that can understand context, this is all we will have and we have to deal with the consequences of either.