Damn it's almost as if this was a conversation we should have had around 2018 at the latest. Like evident bias in social networks was a real concern all along.
What was the response every single time? Ah yes. "Get your own Twitter".
I despise most social networks, this one included, but I can't overstate the degree to which the audience deserves this.
Absolute no lessons learned though. Maybe Musk has to buy Reddit too for people to get a clue.
As far as I can tell, the only way Twitter has substantively changed over the past few years is that a small (but non-trivial) number of politically obsessed left-wingers have voluntarily self-exited. Everything else is narrative.
Every social media site is like this now. They fear the scrapers. Most of the people complaining about having to log in on X, they intentionally have no account in the first place because they don't like the site's owner or something.
I say this as part of the exception. The only reason I don't use X is because they banned me for no reason. Probably not the common case.
That's what some motivated cable news researchers decided to abuse statistics to say. To be generous, I would say the article makes a significantly stronger causal claim than the experiment supports. The best description is a good exploratory algorithm audit with weak statistical and causal inference.
The analysis rests on just nine synthetic accounts, treats thousands of highly correlated feed impressions as if they provide robust independent evidence, lacks a proper baseline for the political content available on X, and reports no meaningful uncertainty, significance testing or adequately powered replication.
If you find yourself either ignoring or logging in to view Instagram posts, then yes. But history is path dependent, and this problem mainly occurs on X.
> right-wing content was shown most prominently, regardless of users' political leaning.
> Left-wing users saw almost the same amount of left-wing and right-wing content, even though they only followed left-wing accounts. But only 14% of the political content sent to our right-leaning users was left-wing.
> The neutral users saw twice as much right-wing content as left-wing content. And barely any of the political content shown to our X users was non-partisan.
> Over half of the political content shown to our new X users came from accounts categorised as extreme. Of the content from extreme authors, 72% of posts came from right-leaning accounts.
Before X flagged me as a bot for no reason, I had a mostly unused account and can confirm it was showing me a ton of right-wing content. Also lots of stuff related to Elon Musk in particular.
On the flip side, X has less moderation, and it's pretty even-handed. They're allowing a lot of pretty mild right-wing content that Twitter was banning people for.
They didn't release the ranking algorithm as open source, but they released enough of a pretense that people got some hints, and we know that whether a post is by Elon Musk is one of the most important ranking factors.
So journalists ran a study with extremely poor sampling, base rate, and classification methodology designed by left-wing researchers and activists groups (including groups explicitly opposed to X), found the conclusions they wanted to find, and spent the second half of their article decrying the lack of heavy-handed moderation that pre-Musk Twitter used to control speech.
All we can actually derive from that article is that X’s new user recommendation system exposed a small number of near-default, low-feedback, synthetic accounts to more content that left-wing “misinformation” researchers classified as “right-wing”.
Researchers use whatever tools are available, you cannot complaint about imperfect methodology when Elon refuses to provide researchers access to their APIs in order to do a better job at assessing the state of the system.
But for whatever it's worth I had made a new account around the same time and my experience was very similar to what was reported.
Again, Comcast is not an "activist group" and the study literally got awarded by the Royal Statistical Society.
As a cherry on top, they do owe researchers in the EU (though this was done in the UK, I know Sky has an EU presence as well) API access or similar under Article 40 of the DSA.
You’re just being brazenly disingenuous at this point.
The publication is left-leaning, regardless of the distant parent company.
The experts they used to design and run the study were literally left-wing political advocacy groups, and a self-declared left-wing disinformation researcher.
The award is an empty appeal to authority.
But none of those facts matter as much as their obvious selection, base rate, and classification methodology failures.
It’s little odd that you replied to the least substantive fact. Disingenuous even.
> "brazenly disingenuous" in such context is a ridiculous and emotive over statement.
“… behavior that's not totally honest or sincere. It's disingenuous when people pretend to know less about something than they really do.”
Rather like ignoring the methodology errors, or the fact that the classification model was developed by an “unapologetically progressive” political consultancy, or that their own experts rated its accuracy poorly.
You keep claiming all these statistical failures, yet you fail to produce evidence for any of them.
I'm sorry, but I trust actual statistics experts more than some guy on HN that makes bold assertions with no proof. To say "nuh uh" and then not back it up with anything is below the standard of debate I expect.
What are these "obvious" failures? Why have you failed to point them out concretely over several comments now? What statistical knowledge do you possess that makes them obvious to you but not to one of the most prestigious statistics organizations in the world? Please, do share with the class.
Your ideological attempts to poison the well by ignoring objective data because you don't like the messenger are not an argument.
To call me disengenous when you spent the last several comments failing to engage with the data and trying to hand-wave it away is ridiculous.
It is getting very hard to assume good intent at this point. This series of comments paints a picture of somebody starting from ideological priors and then trying to bend reality to arrive at their premade conclusion. I would be happy to be proven wrong here, but so far you have failed to do so.
> This series of comments paints a picture of somebody starting from ideological priors and then trying to bend reality to arrive at their premade conclusion.
The irony is palpable.
> I trust actual statistics experts
I listed the issues, but if you need an expert to explain them to you in detail, then perhaps you have neither the statistical nor media literacy necessary here.
1) Selection. Only 9 accounts, no interaction, short period of time. This is not a sufficiently large or realistic sample. That is also not how users interact with the product — for example, things like dwell time directly and materially impact the algorithmic weighting. They specifically avoided any measurable interactions.
2) Base rate. They did not know if the results reflected the base rate across the platform, or algorithmic bias. In other words, did X itself skew right, or did the algorithm.
3) Unreliable classification. Their own experts rated its accuracy poorly, and those are the people most ideologically aligned with the progressive political consultancy that trained the classification model in the first place.
Never mind the “propaganda” claim, which would require showing that not only was the algorithmic bias real (not demonstrated), but that the intent was propaganda instead of — for example — simply the result of maximizing for user engagement. None of which was shown.
> I already am eating from the trashcan all the time. The name of this trashcan is ideology. The material force of ideology - makes me not see what I'm effectively eating. It's not only our reality which enslaves us. The tragedy of our predicament - when we are within ideology, is that - when we think that we escape it into our dreams - at that point we are within ideology.
Some food for thought from ideology expert Slavoj Zizek.
> I listed the issues, but if you need an expert to explain them to you in detail, then perhaps you have neither the statistical nor media literacy necessary here.
You didn't list any of those before, no need to get personal. You could just argue on the facts instead.
Regarding 1 - You still haven't established why 9 accounts are insufficient. No interaction seems to be exactly the right move if the point is to establish the baseline behavior.
Your argument 2 is kind of self-defeating, don't you think? Either it's a right-wing propaganda machine because its algorithm emphasizes right wing propaganda or it's a right-wing propaganda machine because it's full of right-wing propaganda. I don't see how this argument helps your case that it isn't in either case.
Re 3 - now you're getting back to attacking the messenger. Attack the actual model. Which classification do you disagree with?
> Never mind the “propaganda” claim, which would require showing that not only was the algorithmic bias real (not demonstrated), but that the intent was propaganda instead of — for example — simply the result of maximizing for user engagement.
You haven't established why it would require any of that. As a counterexample, MoMA funded abstract art in the 50s as what they thought was supporting artistic freedom. Turns out it was CIA-funded propaganda. [0] It didn't require any intent from MoMA. It still was a propaganda campaign.
Andrew Anglin, the subject of the first article, was re-banned in 2023 for the same reason that crimethink, inc was banned.
The subjects of the second article are particularly odious right-wing conspiracy theorists and racists that lost their reach, which they believe was connected to arguing with Elon.
That seems like an action you’d be in favor of, and also it counters the propaganda claim — but regardless, their claim that they were shadow-banned, that the cause was direct intercession by Elon, and that the motivation was their merely disagreeing with him, is unsupported by any evidence.
I'm not sure where you got any of that from. Where are you reading anything about the researchers working for Sky News, a Comcast subsidiary (!) being "left wing"? What are you basing your criticism of sampling and classification on?
I provided a study that literally won the Royal Statistical Societies' "Statistical Excellence in Journalism" award [0]. You provided unfounded dismissal and conspiracy. The burden of proof is on your side.
Congratulations, Elon Musk has just forced you to either feed his propaganda machine or successfully locked you out of the conversation.
Damn it's almost as if this was a conversation we should have had around 2018 at the latest. Like evident bias in social networks was a real concern all along.
What was the response every single time? Ah yes. "Get your own Twitter".
I despise most social networks, this one included, but I can't overstate the degree to which the audience deserves this.
Absolute no lessons learned though. Maybe Musk has to buy Reddit too for people to get a clue.
Yeah, most of the people are upset because they're Democrats politically on Twitter, and the old one catered to them.
As far as I can tell, the only way Twitter has substantively changed over the past few years is that a small (but non-trivial) number of politically obsessed left-wingers have voluntarily self-exited. Everything else is narrative.
It's also become a whole lot more hostile to anonymous visitors, hence the topic under discussion. That's not just narrative but a fact.
Every social media site is like this now. They fear the scrapers. Most of the people complaining about having to log in on X, they intentionally have no account in the first place because they don't like the site's owner or something.
I say this as part of the exception. The only reason I don't use X is because they banned me for no reason. Probably not the common case.
That's not what the data says: https://news.sky.com/story/the-x-effect-how-elon-musk-is-boo...
That's what some motivated cable news researchers decided to abuse statistics to say. To be generous, I would say the article makes a significantly stronger causal claim than the experiment supports. The best description is a good exploratory algorithm audit with weak statistical and causal inference.
The analysis rests on just nine synthetic accounts, treats thousands of highly correlated feed impressions as if they provide robust independent evidence, lacks a proper baseline for the political content available on X, and reports no meaningful uncertainty, significance testing or adequately powered replication.
You mean like Zuck prevents me from seeing any Instagram content without login?
If you find yourself either ignoring or logging in to view Instagram posts, then yes. But history is path dependent, and this problem mainly occurs on X.
Yes, and? I can't bend reality to make it otherwise.
XCancel can.
"XCancel Taken Down Again" is the title of this thread.
Indeed. But they've been doing good work and fighting is better than just giving in to apathy and losing by default.
Given its extremely liberal content moderation policy, how is X a propaganda machine?
algo is what can make or break a propaganda machine.
Here's a pretty in-depth investigation into the workings of X as far-right propaganda machine: https://news.sky.com/story/the-x-effect-how-elon-musk-is-boo...
> right-wing content was shown most prominently, regardless of users' political leaning.
> Left-wing users saw almost the same amount of left-wing and right-wing content, even though they only followed left-wing accounts. But only 14% of the political content sent to our right-leaning users was left-wing.
> The neutral users saw twice as much right-wing content as left-wing content. And barely any of the political content shown to our X users was non-partisan.
> Over half of the political content shown to our new X users came from accounts categorised as extreme. Of the content from extreme authors, 72% of posts came from right-leaning accounts.
Before X flagged me as a bot for no reason, I had a mostly unused account and can confirm it was showing me a ton of right-wing content. Also lots of stuff related to Elon Musk in particular.
On the flip side, X has less moderation, and it's pretty even-handed. They're allowing a lot of pretty mild right-wing content that Twitter was banning people for.
They didn't release the ranking algorithm as open source, but they released enough of a pretense that people got some hints, and we know that whether a post is by Elon Musk is one of the most important ranking factors.
So journalists ran a study with extremely poor sampling, base rate, and classification methodology designed by left-wing researchers and activists groups (including groups explicitly opposed to X), found the conclusions they wanted to find, and spent the second half of their article decrying the lack of heavy-handed moderation that pre-Musk Twitter used to control speech.
All we can actually derive from that article is that X’s new user recommendation system exposed a small number of near-default, low-feedback, synthetic accounts to more content that left-wing “misinformation” researchers classified as “right-wing”.
Researchers use whatever tools are available, you cannot complaint about imperfect methodology when Elon refuses to provide researchers access to their APIs in order to do a better job at assessing the state of the system.
But for whatever it's worth I had made a new account around the same time and my experience was very similar to what was reported.
[flagged]
Again, Comcast is not an "activist group" and the study literally got awarded by the Royal Statistical Society.
As a cherry on top, they do owe researchers in the EU (though this was done in the UK, I know Sky has an EU presence as well) API access or similar under Article 40 of the DSA.
https://news.ycombinator.com/item?id=49700433
https://www.allsides.com/news-source/sky-news-media-bias
You’re just being brazenly disingenuous at this point.
The publication is left-leaning, regardless of the distant parent company.
The experts they used to design and run the study were literally left-wing political advocacy groups, and a self-declared left-wing disinformation researcher.
The award is an empty appeal to authority.
But none of those facts matter as much as their obvious selection, base rate, and classification methodology failures.
Ever so slightly left leaning by your source.
Centralist and unbiased by another: https://mediabiasfactcheck.com/sky-news/
That's quite the achievement for a group that was once firmly a propaganda wing of the global Murdoch PoV.
"brazenly disingenuous" in such context is a ridiculous and emotive over statement.
> Centralist and unbiased by another
It’s little odd that you replied to the least substantive fact. Disingenuous even.
> "brazenly disingenuous" in such context is a ridiculous and emotive over statement.
“… behavior that's not totally honest or sincere. It's disingenuous when people pretend to know less about something than they really do.”
Rather like ignoring the methodology errors, or the fact that the classification model was developed by an “unapologetically progressive” political consultancy, or that their own experts rated its accuracy poorly.
You keep claiming all these statistical failures, yet you fail to produce evidence for any of them.
I'm sorry, but I trust actual statistics experts more than some guy on HN that makes bold assertions with no proof. To say "nuh uh" and then not back it up with anything is below the standard of debate I expect.
What are these "obvious" failures? Why have you failed to point them out concretely over several comments now? What statistical knowledge do you possess that makes them obvious to you but not to one of the most prestigious statistics organizations in the world? Please, do share with the class.
Your ideological attempts to poison the well by ignoring objective data because you don't like the messenger are not an argument.
To call me disengenous when you spent the last several comments failing to engage with the data and trying to hand-wave it away is ridiculous.
It is getting very hard to assume good intent at this point. This series of comments paints a picture of somebody starting from ideological priors and then trying to bend reality to arrive at their premade conclusion. I would be happy to be proven wrong here, but so far you have failed to do so.
> This series of comments paints a picture of somebody starting from ideological priors and then trying to bend reality to arrive at their premade conclusion.
The irony is palpable.
> I trust actual statistics experts
I listed the issues, but if you need an expert to explain them to you in detail, then perhaps you have neither the statistical nor media literacy necessary here.
1) Selection. Only 9 accounts, no interaction, short period of time. This is not a sufficiently large or realistic sample. That is also not how users interact with the product — for example, things like dwell time directly and materially impact the algorithmic weighting. They specifically avoided any measurable interactions.
2) Base rate. They did not know if the results reflected the base rate across the platform, or algorithmic bias. In other words, did X itself skew right, or did the algorithm.
3) Unreliable classification. Their own experts rated its accuracy poorly, and those are the people most ideologically aligned with the progressive political consultancy that trained the classification model in the first place.
Never mind the “propaganda” claim, which would require showing that not only was the algorithmic bias real (not demonstrated), but that the intent was propaganda instead of — for example — simply the result of maximizing for user engagement. None of which was shown.
> The irony is palpable.
> I already am eating from the trashcan all the time. The name of this trashcan is ideology. The material force of ideology - makes me not see what I'm effectively eating. It's not only our reality which enslaves us. The tragedy of our predicament - when we are within ideology, is that - when we think that we escape it into our dreams - at that point we are within ideology.
Some food for thought from ideology expert Slavoj Zizek.
> I listed the issues, but if you need an expert to explain them to you in detail, then perhaps you have neither the statistical nor media literacy necessary here.
You didn't list any of those before, no need to get personal. You could just argue on the facts instead.
Regarding 1 - You still haven't established why 9 accounts are insufficient. No interaction seems to be exactly the right move if the point is to establish the baseline behavior.
Your argument 2 is kind of self-defeating, don't you think? Either it's a right-wing propaganda machine because its algorithm emphasizes right wing propaganda or it's a right-wing propaganda machine because it's full of right-wing propaganda. I don't see how this argument helps your case that it isn't in either case.
Re 3 - now you're getting back to attacking the messenger. Attack the actual model. Which classification do you disagree with?
> Never mind the “propaganda” claim, which would require showing that not only was the algorithmic bias real (not demonstrated), but that the intent was propaganda instead of — for example — simply the result of maximizing for user engagement.
You haven't established why it would require any of that. As a counterexample, MoMA funded abstract art in the 50s as what they thought was supporting artistic freedom. Turns out it was CIA-funded propaganda. [0] It didn't require any intent from MoMA. It still was a propaganda campaign.
[0] https://news.artnet.com/art-world/artcurious-cia-art-excerpt...
Musk also uses plenty of heavy handed moderation; it's just almost all directed at the left. For instance: https://crimethinc.com/2022/11/25/elon-musk-bans-crimethinc-...
They were banned for literally advocating for political violence. The fact that they hadn’t been banned earlier is the only concerning fact here.
The current top article on their website argues for punching their political opponents in the streets.
And yet right wingers openly calling for the west to ethnically cleanse the non white races get reinstated.
https://www.rollingstone.com/politics/politics-news/elon-mus...
And simply criticizing Musk is enough to get a shadow ban
https://futurism.com/twitter-shadow-ban-elon-musk
Andrew Anglin, the subject of the first article, was re-banned in 2023 for the same reason that crimethink, inc was banned.
The subjects of the second article are particularly odious right-wing conspiracy theorists and racists that lost their reach, which they believe was connected to arguing with Elon.
That seems like an action you’d be in favor of, and also it counters the propaganda claim — but regardless, their claim that they were shadow-banned, that the cause was direct intercession by Elon, and that the motivation was their merely disagreeing with him, is unsupported by any evidence.
He had already been banned for the same reason, and Musk unbanned him in the same wave that had him ban crimethinc.
I'm not sure where you got any of that from. Where are you reading anything about the researchers working for Sky News, a Comcast subsidiary (!) being "left wing"? What are you basing your criticism of sampling and classification on?
I provided a study that literally won the Royal Statistical Societies' "Statistical Excellence in Journalism" award [0]. You provided unfounded dismissal and conspiracy. The burden of proof is on your side.
[0] https://rss.org.uk/news-publication/news-publications/2026/g...
[dead]
X is extremely aggressive in banning anti-Zionists, both left and right wing. It has some of the most anti-free speech policies on the internet.