After many years of progressive discourse hegemony, we are witnessing the bursting of the woke bubble. Throughout the West, the wave of support for the new right—which the left calls the ‘far right’—is growing relentlessly. The collapse of the house of cards, both in the streets and at the ballot box, is happening so rapidly that there is an entire generation of leftists who have never seen anything like it before, who grew up believing that it’s normal for everyone to be ‘good—that is, left-wing—and no one to be ‘bad’—that is, right-wing. They are incapable of self-criticism and seek almost esoteric explanations for the consequences of their own mistakes. The latest idea has been particularly amusing: theorizing that Elon Musk is to blame for deceiving users of X, and dressing up this hypothesis as scientific research endorsed by none other than the journal Nature.
The controversial research, “The political effects of X’s feed algorithm,” is allowing the Western mainstream press to publish reports like the one just released by the Spanish newspaper El Mundo, with this astonishing headline: “Neither ‘fake news’ nor muting accounts: this is how Elon Musk’s social network X makes you right-wing without you realizing it.” The authors of the article claim that their thesis is based on “a trial with thousands of Americans” that confirms that X is capable of “moving users toward more conservative opinions” through the “for you” tab.
I have taken the trouble to read the 23 pages of the study published by Nature, something I doubt those who wrote reports like the one mentioned above have done. I dare say it is the best example of the century of how to try to manipulate people by convincing them that they are being manipulated. Let’s consider some details.
The report finds a “conservative effect”: activating the algorithm (the “For You” tab) increases exposure to conservative content by 2.9% and drastically reduces exposure to traditional media by 15.5%. Any discerning reader will spot the first trap: what leads the researchers to assume that “exposure to traditional media” equates to neutrality in political opinions? It is well-documented that traditional media in the United States has a clear liberal/progressive bias, something that also occurs in Europe. Presenting traditional media as the epitome of ideological neutrality should instantly invalidate the rest of the study.
The next finding seems like a joke: they claim that the algorithm suggests conservative activist accounts that users begin to follow, but that even if you turn off the algorithm and switch to the chronological view, those accounts continue to appear in your feed, topped by the most recent tweets. Apparently, none of the four researchers in the study stopped to consider that this is precisely the most basic function of all social networks: that you freely follow a user and then see their posts. If those accounts disappeared when switching from the ‘For You’ tab to the ‘Timeline’ tab, X would be flooded with messages from users complaining about a stupid programming error and the disappearance of accounts they had chosen to follow.
One of the most striking aspects of the research is that it denies the foundational fact of the human species: the freedom to choose. They assume that Twitter users change their political opinions based on what their feed shows; moreover, they go even further by suggesting that all this happens without them realizing it. In short, the study treats the user as a passive entity, almost a zombie pressing buttons because the algorithm tells them to. They don’t even consider the possibility that the user is actively and freely seeking alternative opinions to the unified narrative of traditional media, which is, moreover, the most likely scenario.
Furthermore, what the study denounces as X’s conservative persuasion is probably nothing more than a correction of the prevailing leftist bias in mainstream media—which, incidentally, has never been the subject of a Nature study along the lines of “this is how traditional media make you left-wing without you even realizing it.” When one says “the media,” one could just as easily say “Netflix,” or even, in this case, the journal Nature.
If you’ve read this far, you’ll have had a chance to understand the study’s weaknesses. It seems designed solely to be amplified in traditional media and used once again against Elon Musk and his support for free discussion on Twitter, and of course, as a weapon against the right wing in the ongoing culture war. However, you’re still missing the most serious point: the researchers don’t seem very interested in letting it be known that the study was conducted three weeks after the iconic photo of Trump in jail, following his surrender in Fulton, after being indicted for election conspiracy and the events of January 6th at the Capitol. Amid this climate of extreme political tension, and with pro-Trump activists denouncing a witch hunt against the Republican leader left and right, the researchers believed it was the perfect moment to analyze the neutrality of X—that is, the neutrality of its users.
Somehow, the researchers treat user opinion as something malleable by X’s algorithm, ignoring that in August 2023, the country was engaged in a national debate about the politicization of the judiciary, and that most Americans already had their opinions before logging onto any social network. Thus, the researchers convey the feeling that if you don’t think like the New York Times, it’s because the algorithm has “manipulated” you.
In short, conducting a study on “algorithmic persuasion” during the height of the Trump indictments is like studying the calm of the ocean in the midst of a tsunami. The results are tainted by the event, not by the code of Elon Musk’s social network.
Nor should we overlook another aspect of the research that is sociologically surprising. Not only did the researchers choose their random sample at a time of maximum political tension, with the entire system and the media attacking the world leader of conservatives, but they had also never done so before. Although they mention a 2016 study that already detected conservative bias on Twitter, there is no rigorous direct comparison with the ‘pre-Musk’ era using the same methodology.
It is clear that part of the authors’ intention was to refute a large-scale study conducted in 2020 by academics in collaboration with Meta, which demonstrated that experimentally replacing the personalized feed with the control feed produced no significant effect on users’ polarization or political attitudes.
Finally, the fact that the study focused exclusively on opinions about Trump already places the focus far from neutrality, given that they never conducted a similar study, for example, focused on opinions about Joe Biden during the height of Hunter Biden’s legal controversy, which could have balanced the results and offered data with a minimum of scientific and sociological relevance.
Although the study provides ammunition for left-wing media—who are not overly concerned with rigor when it comes to throwing red meat to their uncritical audiences—the truth is that left-wing elites will have to keep searching, and do so more convincingly, if they genuinely want to discover why the new wave of the right is gradually conquering the whole of the West. Here is a hint: they will find no algorithm, no magic potion, and no grand enchantment to blame.
Did the Algorithm Shift Users to the Right? The Flimsy Case Against X
euconedit
You may also like
UPDATE: Meta Reinstates Our Facebook Page After Suspending It
A June 2025 post promoting our article on age verification for pornographic websites ended up labeled as “nudity and sexual activity” eight months later.
The Limassol Carnival: A Living Tradition in a Changing Europe
As cities increasingly resemble one another and cultural influences circulate rapidly, local festivities like the Limassol Carnival serve as anchors of continuity and self-understanding.
An Obituary for Stefan Niehoff—and Free Speech in Germany?
The German justice system used a law designed to prosecute Nazis in order to prosecute and convict someone who was plainly and obviously anti-Nazism.
After many years of progressive discourse hegemony, we are witnessing the bursting of the woke bubble. Throughout the West, the wave of support for the new right—which the left calls the ‘far right’—is growing relentlessly. The collapse of the house of cards, both in the streets and at the ballot box, is happening so rapidly that there is an entire generation of leftists who have never seen anything like it before, who grew up believing that it’s normal for everyone to be ‘good—that is, left-wing—and no one to be ‘bad’—that is, right-wing. They are incapable of self-criticism and seek almost esoteric explanations for the consequences of their own mistakes. The latest idea has been particularly amusing: theorizing that Elon Musk is to blame for deceiving users of X, and dressing up this hypothesis as scientific research endorsed by none other than the journal Nature.
The controversial research, “The political effects of X’s feed algorithm,” is allowing the Western mainstream press to publish reports like the one just released by the Spanish newspaper El Mundo, with this astonishing headline: “Neither ‘fake news’ nor muting accounts: this is how Elon Musk’s social network X makes you right-wing without you realizing it.” The authors of the article claim that their thesis is based on “a trial with thousands of Americans” that confirms that X is capable of “moving users toward more conservative opinions” through the “for you” tab.
I have taken the trouble to read the 23 pages of the study published by Nature, something I doubt those who wrote reports like the one mentioned above have done. I dare say it is the best example of the century of how to try to manipulate people by convincing them that they are being manipulated. Let’s consider some details.
The report finds a “conservative effect”: activating the algorithm (the “For You” tab) increases exposure to conservative content by 2.9% and drastically reduces exposure to traditional media by 15.5%. Any discerning reader will spot the first trap: what leads the researchers to assume that “exposure to traditional media” equates to neutrality in political opinions? It is well-documented that traditional media in the United States has a clear liberal/progressive bias, something that also occurs in Europe. Presenting traditional media as the epitome of ideological neutrality should instantly invalidate the rest of the study.
The next finding seems like a joke: they claim that the algorithm suggests conservative activist accounts that users begin to follow, but that even if you turn off the algorithm and switch to the chronological view, those accounts continue to appear in your feed, topped by the most recent tweets. Apparently, none of the four researchers in the study stopped to consider that this is precisely the most basic function of all social networks: that you freely follow a user and then see their posts. If those accounts disappeared when switching from the ‘For You’ tab to the ‘Timeline’ tab, X would be flooded with messages from users complaining about a stupid programming error and the disappearance of accounts they had chosen to follow.
One of the most striking aspects of the research is that it denies the foundational fact of the human species: the freedom to choose. They assume that Twitter users change their political opinions based on what their feed shows; moreover, they go even further by suggesting that all this happens without them realizing it. In short, the study treats the user as a passive entity, almost a zombie pressing buttons because the algorithm tells them to. They don’t even consider the possibility that the user is actively and freely seeking alternative opinions to the unified narrative of traditional media, which is, moreover, the most likely scenario.
Furthermore, what the study denounces as X’s conservative persuasion is probably nothing more than a correction of the prevailing leftist bias in mainstream media—which, incidentally, has never been the subject of a Nature study along the lines of “this is how traditional media make you left-wing without you even realizing it.” When one says “the media,” one could just as easily say “Netflix,” or even, in this case, the journal Nature.
If you’ve read this far, you’ll have had a chance to understand the study’s weaknesses. It seems designed solely to be amplified in traditional media and used once again against Elon Musk and his support for free discussion on Twitter, and of course, as a weapon against the right wing in the ongoing culture war. However, you’re still missing the most serious point: the researchers don’t seem very interested in letting it be known that the study was conducted three weeks after the iconic photo of Trump in jail, following his surrender in Fulton, after being indicted for election conspiracy and the events of January 6th at the Capitol. Amid this climate of extreme political tension, and with pro-Trump activists denouncing a witch hunt against the Republican leader left and right, the researchers believed it was the perfect moment to analyze the neutrality of X—that is, the neutrality of its users.
Somehow, the researchers treat user opinion as something malleable by X’s algorithm, ignoring that in August 2023, the country was engaged in a national debate about the politicization of the judiciary, and that most Americans already had their opinions before logging onto any social network. Thus, the researchers convey the feeling that if you don’t think like the New York Times, it’s because the algorithm has “manipulated” you.
In short, conducting a study on “algorithmic persuasion” during the height of the Trump indictments is like studying the calm of the ocean in the midst of a tsunami. The results are tainted by the event, not by the code of Elon Musk’s social network.
Nor should we overlook another aspect of the research that is sociologically surprising. Not only did the researchers choose their random sample at a time of maximum political tension, with the entire system and the media attacking the world leader of conservatives, but they had also never done so before. Although they mention a 2016 study that already detected conservative bias on Twitter, there is no rigorous direct comparison with the ‘pre-Musk’ era using the same methodology.
It is clear that part of the authors’ intention was to refute a large-scale study conducted in 2020 by academics in collaboration with Meta, which demonstrated that experimentally replacing the personalized feed with the control feed produced no significant effect on users’ polarization or political attitudes.
Finally, the fact that the study focused exclusively on opinions about Trump already places the focus far from neutrality, given that they never conducted a similar study, for example, focused on opinions about Joe Biden during the height of Hunter Biden’s legal controversy, which could have balanced the results and offered data with a minimum of scientific and sociological relevance.
Although the study provides ammunition for left-wing media—who are not overly concerned with rigor when it comes to throwing red meat to their uncritical audiences—the truth is that left-wing elites will have to keep searching, and do so more convincingly, if they genuinely want to discover why the new wave of the right is gradually conquering the whole of the West. Here is a hint: they will find no algorithm, no magic potion, and no grand enchantment to blame.
Our community starts with you
READ NEXT
Judicial War in Poland: Nawrocki Pushes Back as Tusk Government Cries “Polexit”
Fear of the Domestic Mujahideen Is Suffocating Britain’s Foreign Policy
How Europe Made Itself Irrelevant on AI