Articles fromjason.xyz

The real story isn’t about Bluesky’s supposed failures—it’s about how these critiques fundamentally misunderstand what people want from social media and who gets to decide what constitutes healthy discourse.

So the “bubble” critique is empirically wrong. But even if it were right, it misses the more important point: this isn’t really about ideological diversity. It’s about who controls the microphone. When critics argue that people should have stayed on ExTwitter to “engage across difference,” they’re ignoring a fundamental reality: Elon Musk controls the algorithm and actively throttles content he dislikes.

So when McArdle suggests that “liberals” made some mistake by leaving ExTwitter, she’s essentially arguing that people should willingly subject themselves to algorithmic suppression by someone who has explicitly welcomed extremist content back onto the platform.

The “bubble” framing also fundamentally misunderstands what most people want from social media. When you go to a knitting circle, are you disappointed that most people there want to talk about knitting? When you join a book club, do you complain that everyone seems interested in books? Pundits and politicians may want to broadcast to the largest possible audience, but most people are looking for community, not maximum reach.

Most people aren’t looking for a debating arena. They want to talk with people they like about topics they care about—whether that’s knitting, local politics, or professional interests.

The difference is that Bluesky users have actual tools to address these issues themselves, rather than begging platform owners to fix things for them.

This is the fundamental point that critics miss: Bluesky isn’t just another Twitter clone. It’s a demonstration of what happens when you give users actual control over their social media experience instead of forcing them to rely on the whims of billionaires.

Most users don’t actually need to know about this, and they don’t need to buy into the ideology of decentralization and user empowerment, but it’s really all about giving the users more control over their social media experience whether directly on a single platform like Bluesky (with things like custom feeds, custom labelers, self-hosted data servers) or through the rapidly growing set of third-party services and apps, some of which have nothing to do with Bluesky.

This represents a fundamental shift from the past decade of social media, where users had to conform to whatever made billionaires happy (posting to the algorithm, accepting whatever content moderation decisions were made) to a system where users can customize their experience to work for them.

The “Twitter competitor” framing is the Trojan Horse. Bluesky demonstrates just one type of service that can be built on an open social protocol—but the real revolution is in returning agency to users.

The data totally undermines the “dying platform no one uses” narrative: multiple media properties have noted that they get way more traffic from Bluesky than sites like Threads and ExTwitter (both of which throttle posts that include links). And a recent Pew study found that so-called “news influencers” are increasingly on Bluesky.

So we have a platform that publishers say drives more engaged traffic than the “mainstream” alternatives, where news influencers are increasingly active, and which generates enough interest that major media outlets regularly write trend pieces about it. This is not what “failure” looks like.

And Bluesky lets people have way more control over those norms and experiences than any other platform and doesn’t support fascist billionaires at the same time.

The promise of constant connection turned out to be a cruel trap. A generation with access to billions of people say they feel lonelier than pensioners.

for example, cofounder of DeepMind Mustafa Suleyman recently called for sales of chips to be restricted to firms that can demonstrate compliance with safe and ethical uses of the technology,

Amid a steadily growing push to build AI at larger and larger scale, access to compute—along with data and skilled labor—is a key component2 in building artificial intelligence systems. It is profoundly monopolized at key points in the supply chain by one or a small handful of firms