Against Bluesky (and Blueskyism)

Politics, echo chambers, and the difference between persuasion, propaganda, and performance on social media

Shortly after Charlie Kirk’s horrific assassination, Ezra Klein published the following thoughtful and magnanimous remarks:

“You can dislike much of what Kirk believed and the following statement is still true: Kirk was practicing politics in exactly the right way. He was showing up to campuses and talking with anyone who would talk to him. He was one of the era’s most effective practitioners of persuasion. When the left thought its hold on the hearts and minds of college students was nearly absolute, Kirk showed up again and again to break it. Slowly, then all at once, he did. College-age voters shifted sharply right in the 2024 election.”

Although Klein acknowledges that he is not the right person to eulogize Kirk, he notes that he still

“envied what [Kirk] built. A taste for disagreement is a virtue in a democracy. Liberalism could use more of his moxie and fearlessness. In the inaugural episode of his podcast, Gov. Gavin Newsom of California hosted Kirk, admitting that his son was a huge fan. What a testament to Kirk’s project.”

In some ways, Klein’s remarks are too magnanimous. One can and should condemn Kirk’s murder, and political violence more generally, whilst acknowledging the truth: that in some ways Kirk was simply a MAGA apparatchik and propagandist. To take only one example, supporting Trump’s “Stop the Steal” campaign to overturn the results of the 2020 presidential election was not exactly “practicing politics in exactly the right way.”

At the same time, Klein is undoubtedly correct that Kirk’s actions exhibited both courage and an admirable focus on political impact. Unlike many who treat politics as a largely inward-looking, masturbatory indulgence of tribal instincts, Kirk turned up at hostile college campuses and sought to change people’s minds through persuasion. And to at least some extent, he succeeded.

Klein is right that liberalism—more precisely, liberals—could use a lot more of this moxie and fearlessness.

There was no better vindication of this than the characteristically hysterical reaction to Klein’s remarks on Bluesky.

Contrary to prominent falsehoods spread by conservatives, it’s simply not true that every post on Bluesky celebrated Kirk’s murder. Although there was some celebration, and any celebration is too much, most prominent accounts categorically condemned the assassination, as did almost every prominent liberal politician and pundit more generally. In so doing, they exhibited a commitment to liberal-democratic ideals not mirrored by most prominent right-wing politicians and pundits, from the dangerous opportunism of Trump’s own speech to the fascist outbursts of figures like Elon Musk, Matt Walsh, and Laura Loomer, all chomping at the bit for a Reichstag fire event.

Nevertheless, the sheer rage among so many liberals and progressives at Klein’s article is telling. Bluesky, and more broadly the progressive culture of “Blueskyism” examined by Nate Silver, is the implicit target of Klein’s article. Whereas Kirk turned up at hostile campuses to persuade those with different views, Blueskyism involves the self-indulgent performance of political persuasion in communities policed to ensure everyone shares the same views.

For this reason, it seems a good time to release this article on Bluesky and Blueskyism that I originally published for paid subscribers back in February. Although it focuses on Bluesky, I like to think it also has something broader to say about politics, the importance of persuasion, and the dangers of two toxic doppelgängers of persuasion: (1) politics as propaganda, the dominant mode of communication today among the far right, and (2) politics as performance, the dominant mode of political communication among many progressives.

Against Bluesky

Over the past decade or so, many academics and pundits have looked to social media to explain the alarming surge in popularity of illiberal, right-wing, populist politics throughout the Western world.

In such analyses, two concepts have been highly influential: “misinformation” (see also “fake news” and “disinformation”) and “echo chambers”.

The first draws attention to the prevalence and dangers of false, misleading, and incendiary content in driving declining trust in institutions, attacks on democracy, conspiracy theorising, hateful ideas, and other troubling developments. According to this analysis, social media is distinct from legacy media in that it provides an ideal breeding ground for such misinformation, which allegedly spreads faster than accurate information online. By taking advantage of this fact, sinister foreign influence campaigns and domestic demagogues have managed to infect the masses with bad ideas, driving them to support dangerous politicians and policies.

The second draws attention to one of the alleged features of social media platforms that greatly exacerbates this problem and creates new ones: their tendency to enclose users within information silos in which they are shielded from contrary perspectives, driving polarisation and radicalisation. Here, the core concern is that the engagement-maximising character of social media algorithms interacts with psychological factors like tribalism and confirmation bias to pull communities into comforting epistemic bubbles, exacerbating their “susceptibility” to congenial misinformation and driving them to develop beliefs increasingly unmoored from reality.

As I have pointed out in previous writings, neither of these explanatory frameworks has been very helpful in understanding troubling political developments in recent years.

At least when it comes to very clear-cut misinformation, most people encounter very little of that content online, and the fringe of active social media users who encounter a lot mostly do so because it aligns with what they already believe.

When it comes to echo chambers, there is simply not much evidence that the problem is worse online than offline, and the genuine—and genuinely worrying—echo chambers that do exist online seem to be driven more by self-selection into such communities than by “the algorithm".

For these reasons, much of the panic surrounding social media and misinformation has been excessive. Contrary to conventional opinion in some quarters, online misinformation is not the most serious near-term global risk, and it does not seem to be the driving force behind complex, large-scale sociopolitical trends.

Nevertheless, the fact that popular discourse surrounding these topics has been alarmist does not mean there is nothing to worry about. Online misinformation and echo chambers are genuine problems, even if the scale of these problems has frequently been exaggerated.

Moreover, it is reasonable to worry that both problems have worsened considerably since Elon Musk took over Twitter (now “X”), not least because Musk himself—one of the most powerful people on the planet and the biggest account on the platform—incessantly posts and amplifies false and misleading content. (Even his own “non-woke” large language model, Grok, is forced to acknowledge this).

Given this, it is puzzling that many of the same people who have spent nearly a decade worrying about online misinformation and echo chambers have responded to this development by abandoning X en masse to create their own progressive echo chamber on Bluesky.

The puzzle of Bluesky

First, if you think that X increasingly functions as a dangerous vector for right-wing or outright fascist misinformation, presumably you would think it is important that informed progressives spend time on the platform refuting (“fact-checking”) that misinformation and improving the quality of information there.

Second, if you think echo chambers are dangerous, presumably a process in which progressives abandon one of the world’s most popular social media platforms en masse greatly increases its tendency to function as a right-wing echo chamber favourable to the spread of misinformation and radicalisation. Likewise, you would presumably be worried about inhabiting a progressive echo chamber of your own.

However, the puzzles do not stop there.

If you spend time on Bluesky, as I sometimes do, you will find an endless stream of posts documenting the sheer awfulness of many of the actions and trends associated with right-wing populist, authoritarian movements around the world. These posts are generally high-quality and insightful. Reading them, I often find myself agreeing. Apparently, however, so does everyone else on Bluesky.

And this makes me wonder: Who are these posts for?

Persuasion, propaganda, performance

In thinking about political communication, it is helpful to contrast a kind of ideal speech situation with two prominent forms of defective speech.

In the ideal situation, people participate in the public sphere by sincerely expressing their viewpoints to try to persuade those they disagree with, keeping an open mind that they might be mistaken—and being more persuasive precisely by projecting this openness. This is politics through rational persuasion, a precondition of which is that speech is directed at those who hold different views. It involves respecting one’s audience as fellow democratic citizens, from whom one might have something to learn.

Politics as propaganda

One alternative to this model maintains the focus on persuasion but abandons any concern with sincerity, truth-seeking, rational argument, or respect. It involves politics as propaganda.

There is no better example of politics as propaganda than Elon Musk’s approach to political communication. It is evident to any moderately well-informed, fair-minded person of any political viewpoint that Musk’s only goal in posting or amplifying content is to spread whatever narrative he thinks would advance his interests. He exhibits a psychopathic disregard for ideals of honesty, rationality, evidence, or intellectual virtue. There is nothing particularly new about this style of politics (hence why the popular label “post”-truth is misleading), but it is real nonetheless.

Politics as performance

Another alternative maintains the focus on sincerity found in rational persuasion but abandons any concern with persuasion. Here, the goal of political communication is not to change an audience’s mind but to advertise the state of the communicator’s mind. It is politics as performance, treating political communication as a tool for signalling the communicator’s virtues, social identities, or loyalty to specific political factions and causes.

There is no better example of politics as performance than Bluesky.

Given that the audience almost entirely includes people who endorse the same narrow range of political views, it makes little sense to understand people’s activity there in terms of rational persuasion or propaganda, at least when it comes to most explicitly political content on the platform.

Instead, it often seems to be about expressing viewpoints to people who already hold them.

On one level, this seems bad because it seems pointless. With almost every political post I encounter on Bluesky, I think, “Wouldn’t it be great if someone who didn’t already agree with this post saw it?”

However, it also seems bad for the same reasons that echo chambers, in general, are bad. When people disproportionately encounter evidence and arguments in line with what they already believe, they tend to “go to extremes”, increasing their confidence in those viewpoints or adopting increasingly radical versions of them. The same is true when they encounter viewpoints that merely echo (and, hence, seemingly corroborate) their own.

Moreover, once echo chambers arise, and politics becomes more about ingroup signalling than rational persuasion, the deep-seated human drive to manage our reputation and win esteem drives a cluster of dysfunctional group dynamics: people not only self-censor to avoid eliciting social disapproval but also participate in the enforcement of shared beliefs to signal their sincere commitment to such beliefs.

Such cancel culture dynamics were ubiquitous on progressive-dominated Twitter before Elon Musk’s takeover. They made progressive politics seem puritanical and insufferable, hurt the public reputation of progressivism, and drove many extremely online progressive journalists and politicians to vastly overestimate how left-wing the general, mostly offline population was.

To the extent that Bluesky is recreating many of the same dynamics in an even more extreme form, that seems bad.

Responses

Based on encountering responses on Bluesky to others who have raised the worry that it functions as a progressive echo chamber, I know many will hate this post. (Search “Jemima Kelly” on the platform and spend a few minutes scrolling to get a feel for the explosive and frequently unhinged reaction to her article on the subject).

I will end by considering five objections.

1. “It’s not an echo chamber to avoid interacting with Nazis!”

First, some people deny that Bluesky is an echo chamber. The typical argument of this form is something like, “It’s not an echo chamber to avoid interacting with Nazis!”.

Being in an echo chamber is a bit like having bad breath: it is often only obvious in others. In all the time I have been on Bluesky, I can honestly say I do not remember encountering a single conservative viewpoint on any topic. Even liberal centrists (i.e., people who are broadly progressive but dissent from progressive orthodoxy on a couple of topics) are widely attacked and blocked on the platform. (I am currently blocked by hundreds for the sin of following Jesse Singal).

Recently, I logged on to discover commentary on a post from Peter Singer, someone to the left of 95% of the population on almost every issue. I initially thought this might be a sign of increasing viewpoint diversity given that Singer is famously controversial in ultra-progressive spaces, only to discover that the main post was someone saying he should “get fucked” (over 700 likes) and “swallow cyanide”.

2. “Musk won’t give us a fair fight.”

Second, some argue that there is no point in trying to address misinformation on X because Elon Musk is influencing the platform in ways that mean it will not be a “fair fight”. For example, Charlie Warzel, a staff writer at The Atlantic who has made a career pushing maximally alarmist commentary on the topics of social media and misinformation (recent headlines include “The Internet Is Worse Than A Brainwashing Machine” and “I’m Running Out Of Ways To Explain How Bad This Is”), has advanced the following argument in response to the fact that Elon Musk blocks links to platforms like Signal:

“[I] genuinely don’t get why anyone who is not subscribed to Musk's political project is throwing content up there. This and all other evidence dismantles the argument that one needs to 'stay and fight' the war of ideas (lol) on a platform that is making it clear as day it will not give you a fair fight.”

I am puzzled what the argument is even supposed to be here. It is frustrating that Musk blocks links to competitor platforms (including Substack), and it seems extremely likely that he manipulates the platform to promote the spread of his preferred viewpoints. Nevertheless, it is obviously still possible to publish accurate information there and to call out lies and misinformation in ways that reach a large audience.

For example, in response to Donald Trump’s recent stream of falsehoods about Russia and Ukraine, many people from across the political spectrum published posts condemning them and providing more accurate information. These posts undoubtedly reached many people who did not already agree with them or who were simply uninformed. It is difficult to say the same about Warzel’s articles in The Atlantic or his posts on Bluesky, which largely preach to the choir.

3. “You can’t change their minds anyway.”

Third, there is a related argument I have seen some people make that it is pointless to try to address misinformation on a platform like X because it is impossible to change people’s minds. This often goes hand in hand with the attitude that “fascists” or hateful people more generally exist in a space beyond reason and persuasion. The best you can do when confronted with such people is to try to defeat them.

This attitude is mistaken. Rational persuasion is hard but possible. Facts matter. People will often try to ignore or deny factual information in tension with their preferred views or interpret it in ways consistent with such views. Nevertheless, as Hannah Arendt points out in her article on truth and politics, factual truth places important limits on such propaganda and self-deception. There are hundreds of millions of users on X. The idea that all or even most such people are beyond rational persuasion is absurd.

Moreover, I find it odd to hear this argument from many people who have spent the past decade worrying about online misinformation and pushing for interventions like censorship, content moderation, and institutionalised fact-checking to combat such misinformation. Is the idea that some people can only be persuaded by misinformation, not true information? Do some progressives only support efforts to address misinformation if they involve the technocratic, top-down regulation of social media platforms? Is the idea of addressing misinformation using the same tools as ordinary users beneath them?

4. “Leaving X signals disapproval of Musk.”

Fourth, some people argue that we should leave X as a way of signalling disapproval of Musk or that staying on the platform supports the platform and perhaps Musk himself (for example, by keeping the platform alive).

I sympathise with the impulse behind these arguments. We should signal disapproval of Musk, and I support efforts to boycott his products in other domains. I think it should be shameful to buy a Tesla, for example. We are not dealing here with ordinary political disagreements or unconventional or even heretical views. We are dealing with somebody who lies systematically and repeatedly in egregious ways.

Nevertheless, the best way of expressing disapproval of Musk is clearly on X, given that this is the platform used by him, his supporters, and—most importantly—a large, more diverse user base. To a first approximation, everybody on Bluesky already sharply disapproves of Musk.

Moreover, Musk has likely benefitted from the mass exodus of informed progressives from the platform by greatly diminishing the number of people motivated to call out his lies. For Musk, the value of the platform has never been financial. It rests on its potential for propaganda. Decamping to a progressive echo chamber on Bluesky does more to help that propaganda than hinder it.

5. “X is too unpleasant to stay.”

Finally, some argue that the best reason to leave X is simply that it’s an incredibly unpleasant place saturated with hate speech, Nazi content, snuff videos, and much more.

This is undeniable. Moreover, some people have been subject to systematic, coordinated harassment and abuse on the platform. I am not suggesting those people have a responsibility to stay on it. And if you are only on social media for enjoyment or personal utility, then the fact that X is a cesspit of unpleasant content is a legitimate reason to leave.

However, at least many who engage in political debate on social media are concerned with loftier ideals than pleasure and self-interest. If that includes you, and you are not subject to harassment and abuse, then considerations of pleasure and self-interest do not constitute a persuasive argument.

For such people, decamping to a comforting progressive echo chamber looks like an abdication of political and intellectual responsibility.