It isn't so much as the rest of the world having easy access. It is what the Chinese want the rest of the world to see. If you are in a South American country using a residential IP in new incognito session, doom scroll, after the initial disturbing content, you will start to notice videos of the United States government physically attacking people born in the country of the residential IP address.
The TikTok algorithm in South America. Content about Tiananmen Square and Tibet gets filtered out. Content about the United States government rolling through protesters in armored vehicles, killing people in Venezuela with bombs, and threatening Greenland, straight to top of feed.
The most brutally honest propaganda is always the most effective propaganda.
The TikTok ban is the hammer, antitrust is the anvil.
Without antitrust regulation, TikTok would have been sold to Meta, and that would be it. We'd have an even worse monopoly (which is not a good thing), but at least we wouldn't have this. With such regulations present, the US government both forced a sale and disallowed a sale to anybody who they didn't like, basically forcing TikTok to choose a government-approved partner. What did that partner do to become government approved? We'll never know.
Antitrust in the US (and GDPR in Europe) give regulators wide latitude over who to prosecute and for what. This makes it much easier to do under-the-table deals to achieve objectives that you can't or don't want to achieve by regulation, like restricting free speech.
Subjecting companies to such regulation was ok when it was about transporting cattle or selling bricks, but giving governments the ability to regulate companies that have a wide impact on speech, even if the regulations don't seem to have anything to do with speech, is just asking for trouble.
On Twitter, there's a bunch of reports that TikTok suddenly prevents people from sending the word "Epstein" in DMs [1].
I had expected an Orbanisation (aka, what happened to the media sphere in Hungary after Orban took over and his cronies bought up almost all media) of Tiktok, but not that fast, it's like less than a week after the deal [2].
Scary shit if you ask me, and it's made scarier by the fact that Tiktok has already been changing the way our youth speaks due to evading censorship (e.g. "graped" instead of "raped", "unalived" instead of kill/murder/execute/suicide).
Funny how all the people contesting other adjacent comments are noticeably absent here. It's silly in the first place, given the US government and Israeli officials were very clear in their bluntly stated aims with the tiktok take over, what we're seeing now is just the execution of clearly stated intentions, too many ideologues are in denial.
They consider free people sharing information with each other against the consent and interests of MEGAPEDOELLISON Cabal in power a "technical glitch" that they're trying hard to "patch" by slaughtering the First Amendment.
Anecdotal to myself. I shamefully sometimes use TikTok, I particularly like recipe clips and even I noticed something in the last week, most noticeably around this weekend where the algorithm for recommendations changed. It’s like they completely wiped my preferences. I try not to watch anything political so I cannot say much about censorship of content but something was noticeable in the last week.
Same experience here, and also I noticed several channels I used to be following I was no longer following after the hand offs. The feed is completely different now.
I noticed exactly the same thing. I don't recall which day it started (probably this past Sunday), but it was as if a switch flipped. My For You Page no longer has anything to do with my preferences. I'm familiar with Tiktok nudging me in different directions in the past, but I was always able to steer it back to videos I was interested in within 10-15 minutes. Three days later, and it's as if Tiktok not only has forgotten everything about my watch history, it also hasn't learned. That said, it doesn't seem to be entirely about politics. I had a mix of political/protest related content, native plant content, and woodworking videos on my For You Page. None of those are showing up for me.
It’s crazy to think that Instagram Reels, owned by Meta, is preferable to TikTok now. At least Reels now is at least competitive in terms of content - unlike two years ago when people were worried about TikTok being banned and Reels was not a good alternative.
Meta and Google (including Youtube) kowtow to the administration in what speech they promote and suppress in the exact way the administration (both parties) says China might theoretically do in the future.
Anecdotal: uploading a video of original songs with political/protest lyrics will have random background noises added to the audio track, making the songs audio seem amateurish.
That's super curious. No offense, the noise didn't make it sound more amateurish to me personally, so I wouldn't go as far as to immediately conclude that this was intentional by TikTok, let alone that it is because it didn't like the lyrics, but I'm very curious what is it anyway.
Reminds me of how someone lately was going crazy about weird video-artifacts on Youtube. It was fixed (for his videos) after contacting somebody on the technical side of Youtube, but there was never an explanation AFAIK of what actually went wrong, so I was left pondering if that could be a result of some more ambitious ML-experiments in attempt to improve compression rates or something, but never found out conclusively.
Honestly I'm surprised people don't jump ship more often with social media platforms. With TikTok this is kind of new news, but there have been related problems with it that have been pretty obvious for some time.
The same with X and, before that, Facebook.
TikTok has never worked for me though so maybe there's no real equivalent alternative. Maybe time to make one if not?
To me it says something about the public, but I'm not sure what. I'm tempted to attribute it to indifference or complacency but I'm aware of network effects and the reality of alternatives.
Sometimes I feel like education and theory about security practices needs to extend beyond micro-level phenomena like passwords, to things like administrative conflicts of interest and strength in decentralization and competition. Private monopolies and quasi-monopolies aren't just economically bad, they're bad for privacy and security, and make the public vulnerable through lack of choice. In important ways it doesn't matter if it's the government or a private company; whenever power concentrates it is easier to align and abuse.
Are you really surprised? I find the interest entirely unrelatable, but I'm not surprised. They tune these platforms for addiction. I mean, I don't even use them but I still immediately recognize their branding video-end sounds just from random exposure here or there. (I hate it)
They are in transition, so for the moment I believe them to have technical problems, because it also matches my experience. Yesterday I encountered problems with several videos, which are working today. And not all of them were political.
Going by the comments, people on TikTok seem very fast in seeing conspiracies, when many problems can be simply explained with normal problems or human failings. And it's good to be critical and aware of dangers, but I fear if they are so easy to call out problems, it will wear of fast, and people will start to ignore real problems again, like they used to be.
I am also skeptical (despite having 0 faith in the new owners). However, I am a bit confused: why would new ownership alone cause technical issues? It seems like they set new requirements that required new software. Even if the reqs are content-agnostic, I am curious what they are and how they differ from the previous tiktok.
It's kind of amazing that all the companies act in lockstep. Apple, Google, TikTok remove anti-ICE stuff, rightly or wrongly (I'll go with 'wrongly' because of freedom of speech/freedom of app choice, among other things)
it's obvious that tiktok is doing this intentionally, pretending it's a technical issue, so that people can blame the US government for forcing the sale of tiktok
I suspect these are some of those that have been banned from TikTok, and there's probably heightened moderation around this content at the moment since people are sharing AI-generated propaganda and riling others into violent confrontation with ICE.
Anecdotally my feed dramatically shifted. My politics are very leftwing, and prior to the transfer virtually every video was discourse on ICE. Following the transfer, I get content that is all over the place. At one point, I got 8-9 tiktoks in a row of obviously bot-created rightwing text.
At least on the surface level, I could believe this is just a full algorithm reset and they are having problems with it. But even after other algorithm resets that I believe I've experienced, Tiktok figured it out extremely quickly. If this continues, I will believe in the heavyhanded censorship theory.
As a former employee in TikTok US Data Services, the division that was stated working to separate the US TikTok service and infrastructure, I wouldn’t be so quick to dismiss the algorithm reset theory, because the reset this time is qualitatively unlike the others.
Since 2020, when the first Trump administration attempted to ban TikTok, work has been ongoing to separate out the TikTok US business in all aspects from the parent company Bytedance. Setting up dedicated infrastructure for US users was already accomplished by 2023 for the most part, and this included the restriction that US user data couldn’t be used to train or otherwise influence non-US user operations.
However, there were a few major caveats. First, all the actual videos, at least if they’re public, were considered to be Bytedance data, even those created by US users (although the actual user intent signals - likes, watch behavior, and the like, were considered to be US user data). This allowed them to be used to train the main Bytedance-owned algorithm. Second, the Bytedance-trained algorithm continued to form the basis of the algorithm to serve US users. At least when I was there, US user data was used to tune the algorithm for US users, but the algorithm was not necessarily trained from scratch only on US user data in practice.
One of Bytedance’s main conditions for the TikTok US sale has always been that they own the algorithm (both the code and the models) and would not transfer it to the US, so this was definitely a foreseeable issue. With the chaos of the TikTok Us divestment between the Biden and Trump administrations, though, I suspect that it was hard to hire and retain ML engineers that could build a proper replacement for the algorithm in time for the divestment, let alone build one that matches the behavior of the previous algorithm.
If the algorithm is separated between US and non-US as strictly as the TikTok US Data Services mission always aimed for, then TikTok for US users is in many respects a new service entirely that shares the same UI and features. I also don’t know how US users get trained on non-US content, or if they’re even exposed, nor if any other countries use the US algorithm. So this change in content may last at least into the medium term, if not permanently. The question will be if you start seeing more left-wing anti-ICE content in the coming weeks or months.
Allow me to offer some words of wisdom. If you help building weapons to be used against $currently_designated_bad_people, you can rest assured that given enough time, those weapons will be used against you. I am watching all this with a mild sense of bemusement.
It looks like some are moving over to upscroll, anyone know anything about upscroll? what other apps are you using?
I remember when everyone migrated from MySpace to Facebook and I assumed everyone was going to just keep moving over to the next big thing every few years but that actually didn't happen. Facebook became an institution.
When I was 11, on 17th Nov 1989, in Czechoslovakia, my father was watching the evening news on our (black and white) TV, as usual.
There was a protest and the state media was reporting on it. When the reporter said, "our camera broke down and we can only show black and white pictures", my father IMMEDIATELY jumped up and angrily said, "that's bs, you don't want to show how they [the protesting students] got beaten up [by the police]!"
This was an interesting life lesson. So yeah, sure, technical difficulties..
It feels like federated networks with open-sourced feed algorithms are the best path forward.
If AI removes any technical limitations, and automates content management, what's stopping a content creator from owning what they create and distributing it themselves?
141 comments
[ 3.4 ms ] story [ 134 ms ] threadNo, at least during the Biden administration when the law was passed, it wasn't.
This shit is a lot more complicated that a hot take based on today's news.
The TikTok algorithm in South America. Content about Tiananmen Square and Tibet gets filtered out. Content about the United States government rolling through protesters in armored vehicles, killing people in Venezuela with bombs, and threatening Greenland, straight to top of feed.
The most brutally honest propaganda is always the most effective propaganda.
Without antitrust regulation, TikTok would have been sold to Meta, and that would be it. We'd have an even worse monopoly (which is not a good thing), but at least we wouldn't have this. With such regulations present, the US government both forced a sale and disallowed a sale to anybody who they didn't like, basically forcing TikTok to choose a government-approved partner. What did that partner do to become government approved? We'll never know.
Antitrust in the US (and GDPR in Europe) give regulators wide latitude over who to prosecute and for what. This makes it much easier to do under-the-table deals to achieve objectives that you can't or don't want to achieve by regulation, like restricting free speech.
Subjecting companies to such regulation was ok when it was about transporting cattle or selling bricks, but giving governments the ability to regulate companies that have a wide impact on speech, even if the regulations don't seem to have anything to do with speech, is just asking for trouble.
I had expected an Orbanisation (aka, what happened to the media sphere in Hungary after Orban took over and his cronies bought up almost all media) of Tiktok, but not that fast, it's like less than a week after the deal [2].
Scary shit if you ask me, and it's made scarier by the fact that Tiktok has already been changing the way our youth speaks due to evading censorship (e.g. "graped" instead of "raped", "unalived" instead of kill/murder/execute/suicide).
[1] https://x.com/krassenstein/status/2015911471507530219
[2] https://techcrunch.com/2026/01/23/heres-whats-you-should-kno...
It's sad that certain topics (anti-ICE, Epstein) neutered on a social media platform, but this went on for years when the politics were reversed.
Let everyone have their say, I say.
Something definitely broke.
Meta and Google (including Youtube) kowtow to the administration in what speech they promote and suppress in the exact way the administration (both parties) says China might theoretically do in the future.
[1]: https://theintercept.com/2024/10/21/instagram-israel-palesti...
[2]: https://www.middleeasteye.net/news/metas-israel-policy-chief...
[3]: https://www.972mag.com/social-media-ukraine-palestinians-met...
[4]: https://www.hrw.org/report/2023/12/21/metas-broken-promises/...
Edit: here’s a link to an example https://bsky.app/profile/seaniebyrne.bsky.social/post/3mby7j...
Reminds me of how someone lately was going crazy about weird video-artifacts on Youtube. It was fixed (for his videos) after contacting somebody on the technical side of Youtube, but there was never an explanation AFAIK of what actually went wrong, so I was left pondering if that could be a result of some more ambitious ML-experiments in attempt to improve compression rates or something, but never found out conclusively.
The same with X and, before that, Facebook.
TikTok has never worked for me though so maybe there's no real equivalent alternative. Maybe time to make one if not?
To me it says something about the public, but I'm not sure what. I'm tempted to attribute it to indifference or complacency but I'm aware of network effects and the reality of alternatives.
Sometimes I feel like education and theory about security practices needs to extend beyond micro-level phenomena like passwords, to things like administrative conflicts of interest and strength in decentralization and competition. Private monopolies and quasi-monopolies aren't just economically bad, they're bad for privacy and security, and make the public vulnerable through lack of choice. In important ways it doesn't matter if it's the government or a private company; whenever power concentrates it is easier to align and abuse.
Going by the comments, people on TikTok seem very fast in seeing conspiracies, when many problems can be simply explained with normal problems or human failings. And it's good to be critical and aware of dangers, but I fear if they are so easy to call out problems, it will wear of fast, and people will start to ignore real problems again, like they used to be.
They ALL do incredibly corrupt things
it's just retaliation
and obviously, trump will play into this
Note: They also are having "technical difficulties" transmitting DMs with the string "epstein" in them.
https://gothamist.com/news/ai-videos-of-fake-nypdice-clashes...
I suspect these are some of those that have been banned from TikTok, and there's probably heightened moderation around this content at the moment since people are sharing AI-generated propaganda and riling others into violent confrontation with ICE.
At least on the surface level, I could believe this is just a full algorithm reset and they are having problems with it. But even after other algorithm resets that I believe I've experienced, Tiktok figured it out extremely quickly. If this continues, I will believe in the heavyhanded censorship theory.
Since 2020, when the first Trump administration attempted to ban TikTok, work has been ongoing to separate out the TikTok US business in all aspects from the parent company Bytedance. Setting up dedicated infrastructure for US users was already accomplished by 2023 for the most part, and this included the restriction that US user data couldn’t be used to train or otherwise influence non-US user operations.
However, there were a few major caveats. First, all the actual videos, at least if they’re public, were considered to be Bytedance data, even those created by US users (although the actual user intent signals - likes, watch behavior, and the like, were considered to be US user data). This allowed them to be used to train the main Bytedance-owned algorithm. Second, the Bytedance-trained algorithm continued to form the basis of the algorithm to serve US users. At least when I was there, US user data was used to tune the algorithm for US users, but the algorithm was not necessarily trained from scratch only on US user data in practice.
One of Bytedance’s main conditions for the TikTok US sale has always been that they own the algorithm (both the code and the models) and would not transfer it to the US, so this was definitely a foreseeable issue. With the chaos of the TikTok Us divestment between the Biden and Trump administrations, though, I suspect that it was hard to hire and retain ML engineers that could build a proper replacement for the algorithm in time for the divestment, let alone build one that matches the behavior of the previous algorithm.
If the algorithm is separated between US and non-US as strictly as the TikTok US Data Services mission always aimed for, then TikTok for US users is in many respects a new service entirely that shares the same UI and features. I also don’t know how US users get trained on non-US content, or if they’re even exposed, nor if any other countries use the US algorithm. So this change in content may last at least into the medium term, if not permanently. The question will be if you start seeing more left-wing anti-ICE content in the coming weeks or months.
I remember when everyone migrated from MySpace to Facebook and I assumed everyone was going to just keep moving over to the next big thing every few years but that actually didn't happen. Facebook became an institution.
There was a protest and the state media was reporting on it. When the reporter said, "our camera broke down and we can only show black and white pictures", my father IMMEDIATELY jumped up and angrily said, "that's bs, you don't want to show how they [the protesting students] got beaten up [by the police]!"
This was an interesting life lesson. So yeah, sure, technical difficulties..
They thought they were free.
https://en.wikipedia.org/wiki/They_Thought_They_Were_Free
If AI removes any technical limitations, and automates content management, what's stopping a content creator from owning what they create and distributing it themselves?
How can centralization continue to survive?