The Rise of the NormieNet

Echo chamber politics

Nov 27, 2024

“One can understand that at this time preoccupations are elsewhere,” begins a pamphlet circulated around Paris in late February, 1848.

This brochure, in the works for months, was published just as France was in the throes of yet another revolution. The radical artists behind the document noted, that in the capital of late, “talk was replaced by rifle shots.”1

This was no call to arms to overthrow the monarchy, but an appeal to oust the academic juries which governed the Salon de Peinture et de Sculpture.

This salon, and others like it — what we’d call today museums or art galleries — had become the jewel of French society. While they were distinct from the informal salons which had helped spur the French revolution a half-century prior, they were a similar product of the enlightenment and critical in the birth of the third place.

But these juries, which selected successful candidates to line its walls, had become enemies of change. In some years, nearly half of all submissions to the juries were rejected — amongst them impressionist genius Édouard Manet and avant-garde renegade Paul Cézanne. The revolutionary openness of these salons had been corrupted by these formalist juries.

“For them art is reduced to exterior line, and what line, great Gods!” the pamphlet-writers howled. “This fatal preoccupation with line dries up their hearts, freezes their imagination and makes them commit systematically the grossest mistakes.”

These artists were pushing against these doors, and the public was behind them. “People want to know everything, or rather talk about everything, and to appear ignorant of nothing," writer Antoine Joseph Pernety had opined some decades earlier.2

Both the royalists and the revolutionary republicans seem obsessed with keeping the juries in place, to cool public emotions. And the juries had become obsessed with the rules and expectations which came to define their salons. “By dint of seeing, even without noticing it, one shapes oneself according to what one has seen,” philosopher Charles Batteux remarked of the increasingly-blasé salon walls.

It would take until Napoleon III’s reign as emperor that the public outcry over these staid institutions would finally be rectified. But, a classicist himself, Napoleon would not dismantle the jury system. Instead, he opened a competitor: A Salon des Refusés, a spectacular showcase of all those refused admission from the stuffy academic salons.

The response was polarizing. One viewer reported watching patrons “giggling, choking, bursting out laughing, holding their sides.” But, they wrote, “If there was the best, there was also the worst.”3

The rebel salon catapulted the impressionist movement into the mainstream, even if it continued to be locked out of elite recognition. Soon, more rival salons opened up and demolished the juries’ monopoly on taste.

These opposite clubs, of romanticists and modernists, are what we might call today: Echo chambers, self-sorting groups of like-minded individuals aligned in their views on how to change the world.

This week, on Bug-eyed and Shameless, a treatise on what it means to be in an echo chamber, the threat of closed information environments, and the promise of the coming internet.

This is a dispatch about Bluesky, sort of.

Over the course of this newsletter, I’ve touted all kinds of alternative social fora — Mastodon, Post News, Sprout, Threads, Substack itself — never quite sure which would take off, which would become niche concerns, and which would collapse entirely.

Mastodon continues to be a good place for certain nerds, and the technology underlying it continues to be interesting. Substack is the go-to place to publish independent long-form content these days, although its social network features are a bit insular. Threads is a Potemkin village of a social space, a series of branded façades designed to deliver the illusion of human interaction whilst primarily serving as one big interactive billboard.

Of all the options of social media platforms, the past week has clearly crowned a winner: Bluesky.

If you’re already there, follow me! If you’re not, but are looking for a new social space to follow, discuss, debate, and find interesting new people, ideas, music, art, and everything else: I strongly recommend you try it out. (I even made you a starter pack!) If the entire concept of digital social networking repulses you, either on principle or from experience, I get it. Read on anyways.

Bluesky has exploded in the past two weeks, quickly eclipsing Threads’ growth and seriously challenging Twitter’s incumbent supremacy as the place where it all happens. Its userbase now sits at over 20 21 22 million, and continues growing rapidly.

The Bluesky netizens have bristled, however, at an oft-made allegation — by non-users, by the media, even by some fellow skeeters — that Bluesky is an “echo chamber.”

“The flight to Bluesky is a concern, not because X is the superior platform,” Globe & Mail columnist Phoebe Maltz Bovy wrote last week, “but because at a moment when persuasion and communication are vital, Mr. Trump’s loudest and most influential critics are building an echo chamber.”

I want to spend time with this idea, because I agree with part of it: Bluesky is an echo chamber. But I reject the idea that it is a bad thing. I will argue that Bluesky is really just a very modern Salon des Refusés.

Fly the Blue Skies

Last year, working on a project trying to diagnose the state of political polarization, the team at the Public Policy Forum put together a series of roundtables of youth from across Canada. Listening to them, I was gobsmacked at the degree to which they were hyper-connected into a 24/7 online discourse that demanded their engagement. As one participant told us:

It’s like anybody, at any time of the day, can be called out. And you have to take a stand, even if you've never heard of this thing until this day — even if you work a 9-to-5! If you don't spend your life on the internet, you have to take a stand. Even if you don't know every single detail that makes someone problematic, if you don't know every single detail like that happened two years ago, five years ago. It's like a pressure cooker of anxiety. I feel like you are always on the spot. I think it has evolved to where us, as consumers, are almost held as accountable as people in positions of power.

Much of this pressure cooker of anxiety has been blamed on the youth themselves, as though they designed the systems that enable this kind of systemic bullying. But of course they’re not responsible. It was Facebook’s owners, not a 15-year-old high school student, who decided to secretly game the algorithm to promote anger-inducing content. It was Tiktok, not a teen, which designed a for-you page to optimize for time spent scrolling. It was Twitter, not a GenZ, which gave a class of blue checkmark trolls the power to drive the collective conversation.

People, by nature and inclination, are realizing how much they hate being shunted into massive networks determined more by potential advertising revenue than social value. That puts them at odds with Big Tech behemoths, whose value directly correlates with user volume and levels of engagement. This tension would be managable, if users owned real market power. If users could simply switch to a competitor, it would force platforms to balance user happiness with advertising revenue. Unfortunately, by making themselves integral to the infrastructure of the internet and to our IRL social lives, whilst crowding out alternatives, platforms grew disinterested in users’ wants. The decisions they made to maximize revenue, engagement, political impact, and growth all worsened the user experience and drove social discord. Users began to believe that they were, in fact, trapped in these misery-inducing, adversarial, distrust-fomenting adversarial social networks. (Dispatch #64)

Social entertainment platforms have always been upfront about juicing the algorithm to maximize engagement, but Twitter — as the “digital town square” — was supposed to be a neutral and shared platform for news, politics, and culture. This is a refrain that is repeated constantly, including by many supposedly serious columns. But this is unbelievably naive. (Dispatches #26, #32, #63, #69, #80, #116)

Musk’s Twitter is an echo chamber masquerading as an open platform.

This month, researchers at the Queensland University of Technology published new data from their study of Twitter’s recommendation algorithm. After analyzing changes in engagement over recent months, they found “a potential algorithmic adjustment that preferentially enhanced visibility and interaction for Musk’s posts” and, just this summer, “a group-specific boost, with Republican-leaning accounts exhibiting a significant post-change increase relative to Democrat-leaning accounts.”

We can now see clearly how Twitter is being calibrated to serve a particular political outcome: Support for Donald Trump.

Just as the consequences of this change were becoming clear, through the clouds peeked out Bluesky.

It succeeded because it looked and felt like Twitter, from an era before Musk, and because it already had a sizable, active, and fun userbase. It became immediately popular because it was familiar and inviting.

But those criteria could be applied to Threads as well. So why did Bluesky pop off?

I think that has to do with the other half of Blueksky’s appeal, an under-appreciated part of what our netizens want. People want to be in control of their own conversations.

Meta, as a company, runs on the idea that its value isn’t just in its ability to provide technology — a platform, a video-sharing feature, the ability to poke your former coworkers who you haven’t spoken to in a decade — but in its ability to make sense of messy conversations. Facebook wants you to think of it as a filter, an unseen force ensuring that your conversation is a pleasant one.

To that end, Bluesky’s appeal is simple: Its main feed is reverse-chronological. It is not trying to sell you anything. It is not asking you to trust it to manage your online life. What’s more, most of its curation efforts are artisanal — its beloved ‘starter packs’ are prepared by and for other users. Its lists and feeds are largely designed by power users, allowing you to customize what, if any, algorithm is used to filter the conversation.

While monetizing Bluesky may ultimately prove a problem, the very core of the platform operates on the idea that users should be allowed to customize their own conversations.

So, is it an echo chamber?

The short answer is, yes, of course it is. Because Bluesky’s core userbase is made up of those who are fed up with the weaponized anger of Twitter and the liminal space of Threads, they automatically have a unifying identity. They are more likely to be liberal and progressive, to work in news or be obsessed with current affairs, be distrustful of Big Tech, and be intolerant of hate speech. If this cadre were to stay on Twitter, they would be unhappy and ill-served; and any platform they flock to would, by virtue of their mass, be changed by their joining.

But none of those constituent characteristics are actually bad. And by defining themselves against the externalities of the platforms which they’ve fled, they’ve managed to create an actively user-focused culture.

Insofar as this is a problem (as I’m about to argue, it’s not really one), it’s one created by the philosopher-douchebags who pushed them off a platform they once liked.

And, most critically, Bluesky comes equipped with the infrastructure to allow users to break out — or into — any echo chamber they want. Users, for example, could build a news feed which prioritizes Bluesky’s most-banned users, or privileges content with few likes and reposts. You could generate a newsfeed that over-compensates by promoting conservative voices.

So there is a profound naïveté and hypocrisy embedded in the idea that we must be worried only about the echo chamber of Bluesky, the only platform letting users control their own feed.

Echolocation, echolocation, echolocation

In their 2008 book, Echo Chamber, researchers Kathleen Hall Jamieson and Joseph N. Cappella put a microscope to Rush Limbaugh, the Wall Street Journal, and Fox News. And while they were clearly worried about the impact of hyper-partisan media on the national discourse, they didn’t actually blame the echo chambers themselves.

Jamieson and Cappella: On the plus side, the echo chamber encourages engagement and increases the audience's ideological coherence; on the downside, its balkanization, polarization, and use of ridicule and ad hominem rhetoric have the potential to undercut individual and national deliberation.
Is there a way to salvage what is valuable, without undercutting the formula that attracts audiences to these media? If so, it would be a way to fairly present the opposing point of view and then fairly rebut it without use of ridicule or ad hominem. We see no reason one cannot meet these standards while also entertaining and framing from one ideological perspective rather than another.4

Critically, these authors found that participation in this right-wing echo chamber did not correlate to a disengagement with mainstream conversation. Quite the opposite: Limbaugh’s listeners were voracious readers of the New York Times. A ton of research backs up this phenomenon in the digital age. Echo chambers are just one space in which people live, not the only one.

So Limbaugh didn’t need to convince his listeners to distrust or discount the mainstream media, or to demonize the Democrats with wild lies. But he did. Why? A possible explanation lies in what Jamieson and Cappella highlight as the “ways the conservative media perform actual party functions.”

Limbaugh didn’t just set out to make his show a place where he served his listeners. He wanted to achieve Republican success. He was incentivized to weaponize his show.

Incentives are a useful thing to think about. Elon Musk had enormous incentive to use Twitter to secure a personally-profitable victory for Donald Trump. Mark Zuckerberg’s empire earns 98% of its income through advertising, and is therefore incentivized to run a platform which hoovers up as much data from its users as possible.

Sometimes people like to disconnect the motives of a corporation from the activities of its users, but I would suggest that is a very stupid thing to do.

A good, if troubling, example can be found in the world of gambling.

Over the past three years, there has been a massive boom in online betting — thanks, largely, to liberalization that has happened rapidly across the United States and Canada, a rise in shady offshore gambling companies, and the proliferation of cryptocurrencies. You may have noticed the sports betting boom, but you may have missed the explosion in online youth-targeted betting networks.

Consider Aiden Ross. A prolific streamer, he amassed millions of subscribers for livestreaming videogames on Twitch. During the pandemic, he went big into online gambling. With scores of his young fans watching, Ross would log onto gambling platform Stake and pony up hundreds of thousands of dollars. His young fans watched on one night in particular, as Ross poured money into online blackjack. “I’ve lost so much money, guys, it’s insane,” he said. “It’s like, now I can’t stop.” One of the other influencers on his stream pops in: “I told him to quit.” Ross responds: “I know you did, bro.” Rubbing his temple, Ross says: “Now I’m down 1.8, bro. I was at 700k, all I had to do was stop.” And then, surprise. Ross wins a million bucks.

The moral of the story, I guess, is that gambling always pays if you stick with it.

To their credit, Discord has tried to crack down on these gambling streams. But Kick, one of their competitors which signed an exclusive 2023 deal with Ross, is much more permissive. Perhaps that’s because Kick is owned, at least in part, by Stake. And Stake has gone on a binge of hiring streamers, like Ross and Drake, to market their gambling products on Kick. (If you recognize Ross’ name, it’s because of the mega-viral stream he did with Donald Trump.)

Oh, and if you haven’t already guessed: It’s all a sham. The gambling sites are letting influencers gamble for free, and are perhaps even rigging the tables and machines so they can win big.

This is a dangerous echo-chamber. Companies, platforms, and influencers are conspiring to dupe unwitting users — including literal children — that gambling is fun, cool, and profitable. Stake made $2.6 billion in gross revenue in 2022. And the users are actively participating in it.

Corporations are profit addicts. If they hit on a revenue source which begins generating money, they will find ways of continuing to extract that profit, even if it means getting teens addicted to gambling. The vertical integration of allowing gambling sites to also operate the platforms on which youth congregate is obscene and dangerous. (Twitter is, of course, rife with online gambling content and Elon Musk himself regularly promotes PolyMarket, a betting market rival to Stake.)

It’s worth comparing these incentives to Bluesky and Substack. These platforms have relatively few staff, and either earn profit through taking a cut of peer-to-peer transactions, or hope to charge power-users for premium features. In other words: These platforms profit from their core functions, not through seedy externalities and hidden motives.

So whether Bluesky becomes more or less of an echo chamber is entirely up to its users. But its odds of becoming a toxic echo chamber remain very low, so long as the incentive structure remains the same. If Bluesky opts to become an open platform, meant to incorporate international, national, regional, and local public discourses from across the political spectrum, so long as the discourse follows certain rules and norms, I think it would be a welcome and positive addition. In so doing, it will undoubtedly spawn a bunch of micro-echo chambers within it, just as Twitter did. Black Twitter, Weird Twitter, and other self-selecting groups which thrived on pre-Musk Twitter are fascinating because they conducted both insular conversations whilst interacting with the broader discourse. They are echo chambers where particularly virulent ideas, memes, and culture break out into the mainstream. Their only objective is to engage in discourse.

Reddit operates on exactly this premise. Each subreddit is, to varying degrees, an echo chamber. But because they cohabitate, it creates something resembling a cohesive identity.

So the issue is not that echo chambers exist. It’s that echo chambers can be weaponized. The problem is not that Bluesky may be an echo chamber, it’s that we have lost our useful shared platforms to mediate those echo chambers.

To borrow an axiom from systems theory: The purpose of a system is what it does, not what it claims to do.

"The Organization is a Future Adversary"

“You build the technology knowing you won’t be staying at the company forever, and you signpost the things that protect users / your-future-self,” Paul Frazee wrote on HackerNews in 2023. “This includes open sourcing everything, moving specs to standards bodies when they stabilize, and building the network around low switching costs. I can’t predict the future. We may screw it up. I’m trying to protect the community from us if we do.”

Frazee, one of the founding developers at Bluesky, began his comments with an axiom that I love: “The organization is a future adversary.”

His comments are both a sign that Bluesky is doing things the right way, an invitation for us to never fully trust Bluesky.

A year ago, I tried my hand in a bit of futurism in imagining The Bird Internet (Dispatch #47) — a vision for a World Wide Web where personal choice, small groups, and reasonable inter-personal connections form the basis of our online relationships. In it, I tried to challenge the idea of the internet as a very static thing. I wanted to get you, dear reader, (and myself) disabused of the notion that our digital institutions and organizations are so big, so rich, so powerful that they are simply immovable.

Because it’s exactly that kind of thinking that got us into this mess in the first place. It is that mindset that led users to give up their market power and accept technical solutions that they not only didn’t want, but which actively drove them mad.

If we are to create an online economy that values users, we need to claw back our market power. Decentralization is one way to do that, by allowing users to disengage and exit whenever they stop having a good time.

Bluesky, and its AT Protocol, claims decentralization as a core value and says it wants to let users exit — and bring their posts, followers, and conversations with them — whenever they want.

“Even though the majority of Bluesky services are currently operated by a single company,” the founding paper on Bluesky’s standards reads, “we nevertheless consider the system to be decentralized because it provides credible exit.”5

The problem is, Bluesky isn’t really decentralized. As any fire marshal would tell you: A “credible exit” is not the same as an actual exit.

Christine Lemmer-Webber, software engineer and co-editor of the ActivityPub standard (on which Mastodon runs) recently teased out Bluesky’s promise of decentralization and gave it a thoughtful mixed review. I encourage you to read her fascinating blog if this interests you, but I’ll skip over the technical bits to get to her core point: Bluesky, because it is still run by a single company, could still let profit pervert its motive, or face a hostile takeover from a bad actor, and trap its users in its walled garden. It could, in short, still weaponize its echo chamber.

“Bluesky is built by good people who care, and it is providing something that people desperately want and need,” Lemmer-Webber writes. But, she continues: “Bluesky is not meaningfully decentralized.”

But what’s fantastic is that Lemmer-Webber accepts what Bluesky gets right, and is actively calling on her fellow developers to reimagine ActivityPub to respond to it. In other words: The market power of users who wish to have a healthy, functional, non-toxic online space is incentivizing a movement to continue building out space for them.

And as she considers the challenges for the decentralized system on which she works, she really sums up what I’ve been describing in this (increasingly lengthy) post. I will close out on her explanation of it:

Lemmer-Webber: Conway's law applies in both directions: a technical system reflects the communication and social structure of those who build it, but the communication and social structures that we have available to ourselves are informed by what technology is available to ourselves.

If normies, rather than billionaires, can be allowed to build and influence the technology and culture which manages our shared conversations, our echo chambers will cease to be toxic, adversarial, destructive, paranoid, and exploitative and will start being productive.

Our Salon des Refusés will take the art world by storm.

Be sure to share this post on a rage-inducing platform weaponized by bad actors to profit off your misery. Or, alternatively, somewhere else!

That’s it for this week (which should have come out last week.)

I hope to be back in your inboxes later this week with a punchy new piece on the incoming Trump cabinet.

For my Canadian readers, I’ve got a piece in Saturday’s Star on Pierre Poilievre’s obsession with bug-eating.

I hope my American readers enjoy their inferior Thanksgiving.

Until next time!