Is it possible for a social media platform to plan for its own death? I’ve been thinking about that while enjoying Bluesky’s “fizzy, fuck-around energy,” as my WIRED colleague Kate Knibbs described the (still) invite-only Twitter alternative a few months back. The shitposting, the earnestness, the bad takes from novices trying to figure out the vibe. (Friendly advice: Don’t come here to say bad things about the public library.) I wasn’t there for the birth of Twitter, but I’m told Bluesky feels a little like the rowdy early days of the platform it’s meant to supplant.

Bluesky users have become ardent defenders of this gorgeous chaos—especially this week, as millions signed up for Meta’s new Twitter clone, Threads—but they are also purveyors of tough love. My feed is cluttered with grumbling about bugs, debate over what constitutes a ban-worthy death threat, and demands for answers from the platform’s CEO about why the terms of service read like you’re signing away your life rights. The grousing, I think, shows people care. And for the most part, so far, Bluesky’s leaders have listened. The decline of Twitter marooned us all on a desert island. Now we’re collectively building a ship to get off. Some of the volunteers might seem a little high, but it’s a good time on the beach, and she’s looking more and more seaworthy by the day.

Funny then, that this surging social media app isn’t really supposed to exist. In 2019, when then Twitter CEO Jack Dorsey first tweeted the name Bluesky, the idea was not to build a platform that would compete with Twitter but a decentralized protocol that Twitter, among others, would adopt. That would create a system of “federated” servers or platforms, operated by different organizations with different rules and policies that could nonetheless interoperate. Users on Twitter could talk to people on other services compatible with Bluesky—and, if necessary, move their digital identities elsewhere.

There were plenty of reasons to be skeptical. The idea that Twitter would voluntarily cede its power to control (and monetize) its users seemed laughable. To me, then a cynical crypto reporter, “decentralized” suggested a weird Bitcoin thing. (Dorsey was especially jazzed about digital currency at the time.) During the 2020 election season and the attendant calls to regulate social media, it sounded more like a handy way to shrug off tough moderation decisions.

Then came Musk and the new Twitter. The stakes changed. For one thing, the destabilizing acquisition brought home that whenever we sow our seeds in new digital territory, it’s important to keep the garden gate open. It also cut the Bluesky project loose from Twitter, so its leaders began work on their own social app to get people using its federated network, known as the AT Protocol. Suddenly, they were building what was essentially Twitter 2.0, while at the same time still building the protocol that would, in theory, ensure Twitter 2.0 didn’t meet the same fate as the original.

One word for that fate—used, I’ve noticed, by some Bluesky developers—is “enshittification.” Writer Cory Doctorow coined the term last year to describe the way for-profit digital platforms such as social networks or online marketplaces end up choking themselves. First, often flush with money from investors, a platform treats its users nicely. It helps you build a following and yoke yourself into a network. Then, once its user base has become comfortably ensconced, the platform changes the rules, aiming to maximize its gains. The difference between tech monopolists and, say, railroad barons, Doctorow explains, is how fast they can twiddle the dials. Post-Musk Twitter offers examples: Suddenly you lose your ability to reach your audience without paying a fee, like for Twitter Blue, or you can no longer see what you want because of a clutter of ads. There is no recourse. Users are abused in more damaging ways as the platform seeks greater profits, until a breaking point is reached and it becomes unusable. The platform kills itself off.

The point of federated social media protocols is to plan for a future when a platform is no longer responsive to its users. As Bluesky’s own culture doc puts it, “The company is a future adversary.” Once the rescue ship has launched and the castaways are at the mercy of the captain and the crew, it needs lifeboats, in case it becomes necessary to escape the escape vessel. In theory, platforms may be more cautious with twiddling the dials of enshittification if users are able to easily take their data, networks, and attention elsewhere. Emphasis on “in theory,” because Bluesky (the beta app) launched before the ability to federate on AT Protocol did.

I called Doctorow over the July 4 holiday weekend, and I was surprised to hear that he wasn’t on Bluesky yet. A perennial first-adopter (including, more than a quarter-century ago, Twitter’s beta), he had finally reached his limit. “I’m too old for this,” he says. He was, to say the least, disappointed that the project had ended up launching the platform before the protocol. Bluesky had jumped the gun on a good idea, he thought, because of the temptation to capitalize on Musk’s failures and build an audience. With roughly a half million followers on Twitter, Doctorow knows the perils of getting marooned. “I’m not going to stay in a hotel that doesn’t have fire exits,” he says. Call him when it’s ready for prime time.

In theory, things are moving in that direction. The tools for federating with other apps went into “sandbox mode” for developers last month, and they have been designed to allow you to take all of your data with you from one server to another. (That’s a key difference from Activity Pub, the protocol underpinning Mastodon, where you can take your identity to a new server, but your data gets left behind.) That gives teeth to the threat that disgruntled users will go elsewhere, and it also makes the process of doing it more seamless for users. (A lesson learned from Mastodon’s fleeting Musk bump is that offering a variety of federated servers can be confusing and cumbersome; Bluesky’s goal is to make it invisible.)

The key, though, isn’t just the technical ability to move your data between platforms—it’s having viable places to go when you want to. “There’s a lot of daylight between federated and federate-able,” Doctorow says. Bluesky CEO Jay Graber has said that she envisions not just various competitor social platforms, but also users choosing from a “marketplace of algorithms” that curate their feeds in different ways.

That part I worry about more than the code. Bluesky’s app, of course, will have a head start over its future competitors—just how big it is will depend on how soon federation materializes. Earlier this week, the company announced an $8 million seed fund to support scaling up its app and building the AT Protocol, as well as its first paid product: a service to help register your internet domain name as a Bluesky username. The company says it doesn’t want to use ads, because that creates bad incentives to abuse users, but that could change, of course. Others might start there. Many will fail to grow. A few may grow into giants, bolstered by features and incentives only they can offer—or potentially overnight if they are launched by a major company.

Threads illustrates that danger. Meta’s version of the public square comes with its own promises about escape hatches—in particular, a claim that it will eventually, on some unspecified date, support integration with Activity Pub, the protocol that underlies Mastodon. Why would a company that has so far built walled gardens now want to federate? I suspect it’s because Meta, with its billions of users, ability to rapidly roll out new features, and ecosystem of other apps, doesn’t see opening up as a real threat. (Note that for now, at least, Threads accounts can’t actually be deleted without deleting your Instagram account too.)

The launch of Threads might be a sign that Bluesky was right to start bootstrapping when it did. But the challenger’s challenge is to keep the app weird and good as it grows while also making it as easy as possible to ditch. Until then, enjoy these fun and fizzy beginnings, but also know that it can all go away with the turn of a digital knob. Doctorow compares Bluesky’s situation to a game of chicken: an app running full steam ahead toward the forces of enshittification. Depending on your type of thrill, it’s a fun ride until it isn’t, when the headlights coming at you grow too bright for comfort. What’s the best way to win? Hopefully you ripped off the steering wheel as early as possible.

As Doctorow points out in his essay on enshittification, the interoperability wars have been raging for decades and across all layers of the internet. In 1996, Steve Steinberg wrote about the battle between two groups known as the Netheads and the Bellheads. The latter argued that information networks moving your data around should be centralized and under the control of corporations, not flexible and open to additions from just about anyone. That would’ve sucked.

The Internet … doesn't require connections to be set up ahead of time. I can just send a packet to my friend without waiting for permission. How the packet gets there depends on the second-to-second status of the network's topography. Internet packets act like tiny autonomous agents that are able to find their own way. Both the advantages and disadvantages of packet switching are due to the difficulty of keeping track of traffic that can come from any direction. It makes packet-switched networks harder to censor (or destroy), but it also makes them harder to manage.

"Traffic management" is a phrase with almost totemic weight among telecom engineers. So it's not surprising that when they looked at data traffic, their first thought was to make the packet switching of the Netheads act more like the circuit switching traditionally used for voice. Or as David Sincoskie, one of Bellcore's key people who helped make ATM [asynchronous transfer mode] a reality, tells it: "Our goal was to prove we could achieve the same quality of service as circuit switching, but with packets." Which is exactly what ATM does. Voice and data are both split into small, equal-sized packets. But a virtual connection, called the VC, must be established before these packets can be sent. That's because ATM packets are lobotomized: instead of knowing where they want to go, as in a true packet-switching network, they just know the ID number of the virtual circuit they belong to, and they must stick to that predetermined path.

On Twitter, @binreminded asks if police will use satellites to solve crimes such as robbery or murder.

Thanks for this question, which is sure to become more relevant over time. My first thought is that high-resolution satellite imagery is far more limited than film and TV can suggest. Satellites peer down at a particular region only briefly, and they are limited in geographic scope. That’s why it’s probably not practical to try to hunt down imagery to find a getaway vehicle from a robbery in Peoria or catch a murder by Convict Lake as it happened last Tuesday. I can’t rule out that it’s happened before, but a cursory search suggests that requests for satellite images of crime scenes haven’t been that fruitful.

That said, there are other offenses where satellites are already a prime form of evidence. Think of war crimes—for example, in Bucha, Ukraine, last year, where the bodies of local residents killed by Russian soldiers were left out in the street for days. Or territorial incursions like last year’s Chinese spy balloon mystery. Or environmental crimes like illegal logging. These are crimes where the evidence is visible and long-lasting.

I can imagine that changing as satellite imagery becomes more comprehensive. Private satellites will become more common, high-res cameras cheaper, and artificial intelligence will help speed up the search for evidence. Baltimore residents had a taste of this future a few years ago when police circled a plane with a dozen cameras in an effort to establish a persistent surveillance program—where all things and people outside could be tracked at all times. Crime-solving potential notwithstanding, the question is whether that’s a future we want everywhere.

You can submit questions tomail@wired.com. Write ASK LEVY in the subject line.

Malaria was eliminated from the US in 1951—now it’s back.

Don't miss future subscriber-only editions of this column.Subscribe to WIRED (50% off for Plaintext readers)today.