There’s a crucial difference between federatable and federated.
“Are you on Bluesky?”
Friends, colleagues and strangers have emailed me to ask whether I’ve set up on the new, federatable social media incubated at Twitter and spun out, which many view as a viable Twitter successor.
“Are you on Threads?”
I get this one too, from people who want to know if I’m going to sign up for the Instagram-derived insta-success that leveraged Facebook’s network effects to achieve an explosive launch (reports of Threads’s subsequent waning are overblown — that kind of “scalloped growth” is normal for new services). Threads, too, is federatable, with a promise to connect it, eventually, to the Fediverse — the collection of services built atop the ActivityPub standard, of which Mastodon is the most famous.
I’m not on either service. I’m honored that people would like me to join their conversational spaces, and I’m grateful to the people who offered me signup codes for Bluesky, but I have no plans to join either, and for a very simple reason.
I don’t trust ’em.
It’s not (just) that Bluesky’s board includes the guy who sold Twitter to Elon Musk. It’s not (just) that Threads is owned by Facebook, a cancer of a company founded to nonconsenually rate the fuckability of Harvard undergrads (incredibly, that was Facebook’s least terrible moment. It actually got worse after that).
It’s that neither of these services are actually federated. Both claim that this is coming. Both claim that they have the technical hooks to federate with other services when the time is ripe. But that time isn’t now.
Look, I’m done. I poured years and endless hours into establishing myself on walled garden services administered with varying degrees of competence and benevolence, only to have those services use my own sunk costs to trap me within their silos even as they siphoned value from my side of the ledger to their own.
The only thing worse than having wasted all that time and energy would be to have wasted it — and learned nothing.
As George W Bush told us, “Fool me once, shame you. Fool me —you can’t get fooled again.”
Wise words.
Look, I’m 52 years old. I used to be way more interested in how things worked. Now I’m interested in how they fail. I don’t care how good the administration of Bluesky or Threads is today— I care about what happens if it sours tomorrow.
Facebook has broken so many promises. Remember when Facebook opened up to the general public in 2006 with the promise of being the pro-privacy alternative to MySpace? Remember when they told us they’d never collect and mine our data? They are liars, and we shouldn’t trust them.
But a company doesn’t have to be run by venal scumbags to put its own interests ahead of its users.
Many is the tech CEO who reasoned that selling out their users was the moral thing to do, because the alternative was firing dozens or hundreds of people who trusted them, quit their jobs, and jeopardized their mortgages and kids’ college funds to come work for the company.
Seen in that light, selling out your users is actually an act of noble self-sacrifice, in which your loyalty to your friends trumps even your pride in delivering a high-quality product.
You might be tempted to do the same thing. I know that I’ve been guilty of motivated reasoning in the past. Being a moral actor lies not merely in making the right choice in the moment, but in anticipating the times when you may choose poorly in future, and taking steps to head that off.
That’s where Ulysses Pacts come in. In Ulysses’ world, sailors who ventured through the sea of sirens would stop up their ears with wax so they couldn’t hear the siren-song that would lure them to jump overboard and be pulled down to their deaths.
But Ulysses was a hacker and he wanted to hear the sirens’ incredible song. So he had his sailors lash him to the mast and told them that no matter how he begged, they were not to untie him until they were clear of the sirens.
Ulysses didn’t have the strength to resist the sirens’ song, and he had the humility to recognize his own weakness. So he took measure to prevent himself from being tempted — he used his strength to shore up his weaknesses.
We make little Ulysses Pacts all the time. If you go on a diet and throw away your Oreos, that’s a Ulysses Pact. You’re not betting that you’ll be strong enough to resist their siren song when your body is craving easily available calories; rather, you are being humble enough to recognize your own weakness, and strong enough to take a step to protect yourself from it.
Free and open source software licenses, as well as Creative Commons licenses are Ulysses Pacts. These are irrevocable licenses, which means that you can’t decide to take your code or art back from the commons once you put it in there.
That means that no matter how desperate your business becomes, no matter what pressures your investors or creditors put on you, that measure is off the table. They can pressure you all they want, but you literally can’t give them a proprietary future for your open materials.
A federatable service isn’t a federated one. The fact that Bluesky and Threads claim that they will someday be federated is an ocean away from actually federating.
I have learned my lesson. I have no plans to ever again put effort or energy into establishing myself on an unfederated service. From now on, I will put more weight on how easy it is to leave a service than on what I get from staying. A bad service that you can easily switch away from is incentivized to improve, and if the incentive fails, you can leave.
“Fool me once, shame on you, fool me — you can’t get fooled again.”
I hear that the Bluesky developers are having conversations about how they design their service to protect themselves from their future selves. That shows an admirable degree of humility and foresight.
But you know what would show an even greater degree of humility and foresight?
Being federated.
Mastodon is far from perfect. But I only have so many hours in the day, and only so many days left in my life. I would much rather spend those precious hours making an open service better than using a temporarily superior closed one. I have seen that movie. I know how it ends.
The more effort we put into making Bluesky and Threads good, the more we tempt their managers to break their promises and never open up federation.
For the owners of a service, the fact that it is thriving as a standalone is a powerful siren song: why open this service up and share its bounty when it’s doing just fine as a walled garden?
Even if every decisionmaker at Bluesky and Threads was personally known to me as the world’s most thoughtful, moral actor, I wouldn’t join their services.
Enshittification isn’t merely the result of greed or foolishness — it is the inevitable consequence of a captive userbase.
A service that isn’t federated is unfit for purpose, and the managers who decided not to federate it demonstrate by their acts that they are not worthy of our trust.