Uighur treatment by China amounts to 'Genocide' says formal legal text - Page 25 - Politics Forum.org | PoFo

Wandering the information superhighway, he came upon the last refuge of civilization, PoFo, the only forum on the internet ...

Talk about what you've seen in the news today.

Moderator: PoFo Today's News Mods

#15166610
Fasces wrote:This is so disingenuous I don't even know where to begin.


Your Chinese apologies are very disingenuous indeed.

Do we start with the European conception of ethnicity that isn't found in the mindsets of the average East Asian? Mind you, 'Han' is a very inclusive and very vague label that doesn't really mean anything.

Do we start with the fact that historical immigration to the West is the product of colonialism, and mass economic migration in the present - the former of which cannot be applied to China, and the latter of which is a current trend?

The complete lack of evidence for any of these claims, except in some nebulous vague cloud of "common sense and commonly known" that leads all the commentators on PoFo with no understanding of Chinese, no close connection to China, and no experience with Chinese culture to authoritatively state absurd 'facts' about how Chinese think, and conceive of themselves?


Do we start with the fact that China is

1) a totalitarian dictatorship.
2) "re-educating" Uyghurs to turn them into proper Han.
3) sterilising Uyghur women en mass.
4) suppressing the Tibetans.
5) suppressing the Cantonese people of Hong Kong.
6) suppressing the Mongolians.
7) controls a territory 2.5 times the size of the EU, inhabited by 1.4 billion people that 95% happened to be Han Chinese as if by immaculate conception.

China is the necromonger imperium from the Chronicles of Riddick in real life.

#15170784
@Igor Antunov

Igor Antunov wrote:Do you guys have technology institutes and trade schools over there? I was in one government run tertiary education institution (not college or uni) three weeks ago for a speed keyboarding test. It was dilapidated, crumbling, dirty and looked like a welfare version of a crappy college. I bet the BBC could do an even more dire documentary on this one.

Well these schools in xinjiang look damn nice compared to Australia's 'student internment camps'. And yes here some have accommodation and you can stay at them while studying anything from knitting to positive thinking and other life skills shit from cert 1-4. It's usually full of young single mothers and binge drinking young men that are trying to half ass a trade apprenticeship.

AND YES, young paroles (many of them indigenous) are forced to study bridging courses for work or go back to jail. Australia is committing genocide against aborigines. It's not like their culture had organized education and basic geometry. It is now being wiped out with education and carpentry job skills training. :*( /anti-china logic.

I guess this all all alien and scary to you chaps. I like how even those who have the self awareness to question these flimsy accusations try to qualify and conform to the manufactured consent with 'I know china does some terrible shit but....this adrian guy is a nutjob etc etc'. No, all actual measurable statistics and the situation on the ground shows china does some amazing stuff for minorities and the poor in isolated regions of the country and has done so for decades. All dumb accusations suggest otherwise. No need to qualify dumb accusations by some German employed by the CIA just so you have permission in western discourse to bring up the achievements of china's anti-poverty, healthcare and education.


My understanding is that the Australians haven't treated the aborigines very well. That being said, I don't know what the conditions of those camps are in Australia. I do know that these Muslims in China haven't been treated too well judging by the BBC video. I would think all of this would be obvious to you given the videos. The "re-education camp" seems to be a favorite of communist regimes around the world to be used as what are in reality internment camps that have torture and abuse and/or concentration camps with hard labor designed to literally work people to death as part of "teaching the value of work." That is the case here with China and it's communist regime that seeks to keep power and control over others at any cost.
#15174233
It is now reported that China is implementing emotion detection via their cameras + ML processing engines. I'm sure this will be a test bed for a broader roll out across the entire country. That place will go above and beyond 1984 faaaaast. Still don't see how some of our pofoers here can defend this regime. :lol: Crazier than MTG!

China is becoming what NK would be if it had money.

I guess they are going to try the "Beatings/sterilizations/reeducation will continue until morale improves" approach with that AI deployment. :lol: What is going on with that government. It's a complete joke. Bonkers... absolutely bonkers.
#15174324
Red_Army wrote:Spotify does that! I'm not a China apologist, but it's hard to look at wiki leaks and the corporate and intelligence surveillance in our country and throw many stones about 1984.


Are you saying we cannot criticize the Chinese mega Nazi dictatorship without first taking 20 self-administered lashes?

Whataboutism is something that only apologists do.

I have never ever seen you or any other leftist in here mentioning China in all the anti-American threads. I have never seen you attempting to derail these plentiful threads with anti-Chinese whataboutism. I have seen you though using anti-US whataboutism in the very few threads that criticize China or Russia. How do you reconcile that enormous double-standard?

You have seen me and Rancid criticizing everybody(the US first and foremost) in here, so Rancid can throw as many stones as his heart desires, you on the other hand only throw them selectively. So no, you cannot say that to Rancid especially when you clearly fail to abide by your own logic.

I'm going to say that as plain as possible:

Russian and Chinese apologists in here are directly threatening our and our family's personal well-being, literally threatening our lives just like Protasevich. We have reached a point where EU bloggers get snatched inside the EU for criticizing the Kremlin.

It is high-time people start being responsible for the consequences of their words and arguments instead of childishly behaving as if it's just words in a forum. This is not a PS4 game.

Words like actions have consequences, of course if people did not believe that they would not even be posting in here in the first place.
#15174337
Red_Army wrote:Spotify does that! I'm not a China apologist, but it's hard to look at wiki leaks and the corporate and intelligence surveillance in our country and throw many stones about 1984.


This defelction/what about and false equivalence is so off, I'm not even sure you are being serious or sarcastic. What the fuck is this? :lol: And whooever liked your statement is a fucking jackass.

I have the choice to not use Spotify (I don't use it actually). Most people in China don't have the choice to not live in China. Spotify isn't a government that tells me what I can and can't do in my life.

Serious what kind of dumb shit "argument" is this? You a fan of authoritarianism or something?

Further, are you saying this is ok? You want the US government doing that shit to you? If you want to talk about corporations being evil, you can start another thread about that, and i will happily join you in talking about how evil corporations are (they absolutely are evil, so we will be in alignment in that thread).

Side note: Spotify doesn't use facial emotion AI engines.
#15174355
From what I can tell, this software has issues when it comes to analysing people who are not white and western, simply because computers have trouble reading non-white faces, and emotional cues are sometimes culturally specific.

Consequently, the west uses it in very specific ways, like checking if drivers in self-driving cars are getting sleepy. I assume this works because the vast majority of drivers of self driving cars are white professionals.

So, if the AIs have been calibrated for Han features and culture, and the Uighur cultures and look are different enough, these would not give accurate readings anyway.
#15174443
Pants-of-Dog wrote:the west uses it in very specific ways


"The west" is not an organized entity to use anything. "The west" only exists as an idea and even that with no consensus on what it actually is. Your statement is totally false.

Private individuals and companies in the west do research and development as they do all around the world, not "the west".

China uses this technology to maintain its Nazi dictatorship.
#15174453
Pants-of-dog wrote:From what I can tell, this software has issues when it comes to analysing people who are not white and western, simply because computers have trouble reading non-white faces, and emotional cues are sometimes culturally specific.

Consequently, the west uses it in very specific ways, like checking if drivers in self-driving cars are getting sleepy. I assume this works because the vast majority of drivers of self driving cars are white professionals.

So, if the AIs have been calibrated for Han features and culture, and the Uighur cultures and look are different enough, these would not give accurate readings anyway.


Very clearly, you do not understand how ML training and inference works.

Three pre-points before my real point:
First:
You are confounding different issues/problems with AI. The issue of bias in ML engines isn't with the tech itself. The issue of bias comes from the data sets on which these engines are trained on. The data sets are selected by the builders of these systems, who tend to be of a specific demographic. In development, they tended to work with relatively small data sets they selected. Naturally, their personal and cultural biases get worked into those small data sets. Thus, an unintentional bias can get built in (there isn't always a bias and the industry is getting better at eliminating clear biases). This is one of the growing pains of the industry; the realization of biases. The industry worldwide (except in China of course) is trying to develop standards and methodologies to prevent these biases, as well as come up with ethical guidelines around the use and deployment of such systems. China doesn't give a shit about AI ethics. They state they care about this, but these are just statements because when you look at how they are deploying AI (in concentration camps and Uighur cities for example ), clearly, they don't give a shit.

Second:
What a shit deflection. "The tech isn't good, so it's ok."

Third:
The second statement above is false. The tech is very good, and it's very possible to eliminate bias, or intentionally build it in (which is very easy to do). It is all about how you define bias, and what your intent is with the inference engine you are training. At its core, it's all about the data set the models are trained on, which is largely selected by people (with Automation assistance which brings another layer of complexity of course).

All of that said, it is not difficult at all for China to train their models against Uighurs. That is to say, intentionally built in bias, which is what they want. In fact, it's very easy to build in this bias, the harder challenge is to eliminate it (what people in free nations are trying to do). China, wants to track these people specifically; and it can be done easily. China, being an authoritarian regime with no interest in protecting privacy or addressing the ethical problems that come with this tech, are very well positioned to create the data sets needed to deploy VERY well refined models that can be used against these people.

Training models is something that be put on nightly regressions to refine/improve it as well as the data sets grow. It's a continuous cycle (in the industry it's called continuous integration and continuous delivery or CI/CD)

In fact, that is why ML deployments are the most effective in China compared to everywhere else. They simply don't give a shit about privacy, bias, or ethics. They have more data on people than anywhere else. Additionally, it is easier for them to say fuck ethics and laws abuse the data.

Why do you people keep trying to defend this garbage from the CCP? What is wrong with you people?
Last edited by Rancid on 28 May 2021 16:37, edited 2 times in total.
#15174455
Rancid wrote:Very clearly, you do not understand how ML training and inference works.

Three pre-points before my real point:
First:
You are confounding different issues/problems with AI. The issue of bias in ML engines isn't with the tech itself. The issue of bias comes from the data sets on which these engines are trained on. The data sets are selected by the builders of these systems, who tend to be of a specific demographic. Thus, an unintentional bias is built in. The industry worldwide (except in China of course) is trying to develop standards and methodologies to prevent these biases, as well as come up with ethical guidelines around the use and deployment of such systems.


Yes, that is what I meant when I said that if it was calibrated for Han people, it might give false readings for Uighurs anyway.

Second:
What a shit deflection. "The tech isn't good, so it's ok."


Actually, the implication is the opposite: since the tech does not work accurately, this could provide an opportunity for China to make whatever claims they want about the Uighur and say the tech backs them.

Third:
The second statement above is false. The tech is very good, and it's very possible to eliminate bias, or intentionally build it in (which is very easy to do). It is all about how you define bias, and what your intent is with the inference engine you are training. At its core, it's all about the data set the models are trained on, which is largely selected by people (with Automation assistance which brings another layer of complexity of course).

All of that said, it is not difficult at all for China to train their models against Uighurs. That is to say, intentionally built in bias, which is what they want. In fact, it's very easy to build in this bias, the harder challenge to eliminate it. They want to track these people specifically. China, being an authoritarian regime with no interest in protecting privacy or address the ethical problems that come with this tech, are very well positions to create the data sets needed to deploy VERY well refined models that can be used against these people.

In fact, that is why ML deployments are the most effective in China compared to everywhere else. They simply don't give a shit about privacy, bias, or ethics.

Why do you people keep trying to defend this garbage from the CCP? What is wrong with you people?


I had no idea about intentionally adding bias. I assumed that China would simply use the existing tech problems to justify oppression. It is possible that the articles I read were a few years old.
#15174456
Pants-of-dog wrote:Yes, that is what I meant when I said that if it was calibrated for Han people, it might give false readings for Uighurs anyway.


What I'm saying is that it's very easy to retrain on Uighur data (i.e. intentionally force in bias). You can use the exact same model you used to train a Han biased engine, and then just train up a Uighur biased engine. You can do this literally overnight. In fact,that's how it's usually done, overnight when computers and networks are more idle.

This is way easier to do than trying to train for an unbiased model.

Pants-of-dog wrote:Actually, the implication is the opposite: since the tech does not work accurately, this could provide an opportunity for China to make whatever claims they want about the Uighur and say the tech backs them.

The tech does work, and China will make up whatever the fuck they want to say whether it works or not. However, it does work very well.

Pants-of-dog wrote:I had no idea about intentionally adding bias. I assumed that China would simply use the existing tech problems to justify oppression. It is possible that the articles I read were a few years old.

Yea, ML is very malleable.This isn't like manufacturing chips which can take months to years to produce a new upgraded chip. THis is one of the things that makes so many people worried about this tech (including myself). As long as you have the data, and CHina is known to be the worlds largest holder of data on people in the planet. It's not google or facebook.
Last edited by Rancid on 28 May 2021 16:49, edited 1 time in total.
#15174459
@Rancid

(getting off the specific Uighur topic for a second)

Yes, the implications of this tech is pretty scary. I assume governments (including western ones) will use this, or already do and do not tell us. Much like the different tracking and sensing software on cell phones.
#15174460
Pants-of-dog wrote:@Rancid

(getting off the specific Uighur topic for a second)

Yes, the implications of this tech is pretty scary. I assume governments (including western ones) will use this, or already do and do not tell us. Much like the different tracking and sensing software on cell phones.



Yea, I'm sure many governments are using them too. One application I'm aware of in the west is for an ML engine to scan financial transactions to look for signs of criminal activity. I don't believe this is in production deployments just yet, but I know this sort of thing is being worked on. I'm sure there are more nefarious uses that governments are considering. This is something that needs to be pushed against on a global scale. That is, we need to push for governments to not do dumb shit that infringes on rights/freedom/etc. East or west it all has to be called out.
#15174533
@Rancid @noemon I have read about 3 posts in this thread because I don't see the value in arguing with people about this. I don't trust western media reporting about China. I don't trust Chinese media reporting about China. I have no idea what's going on and don't see the utility of this or any argument on this forum. I just stop by when I'm bored to interject occasionally. You can both rest happily in having the correct opinion about this and none of us will have accomplished anything. I may occasionally return in a dozen pages to comment on the latest comment, but maybe not. If we ever meet in person I will happily explain my feelings to you.
  • 1
  • 21
  • 22
  • 23
  • 24
  • 25
Russia-Ukraine War 2022

I don't know who are you are referring to, but th[…]

PoFo would be a strange place for them to focus o[…]

In my opinion, masculinity has declined for all o[…]

@ingliz good to know, so why have double standar[…]