Post

Every explanation, every justification, every "here's why we did/didn't ban this person" becomes an admission that you're playing the old game. You're the arbiter. You're Twitter.

Once you're in that position, you're subject to all the same pressure campaigns, advertiser boycotts, and political football that comes with it.

Some of you, though, are real fuckin' stupid.

People want to play the pressure campaign game and are mad the levers they're used to pulling simply don't exist here. Like seriously, go back to twitter and go cry to Elon. Oh, right, he was the reason EVERYONE LEFT. Just start your own platform. With blackjack! And hookers!

Fleeing a HOA-controlled suburb for rural living and then immediately organizing HOA-style rules. You left because of the thing! The thing was bad! Why are you trying to rebuild the thing! You idiots!

You can't have it both ways. Either you want distributed, user-controlled moderation with all its tradeoffs, or you want a benevolent dictator and you're just shopping for one you like better. Pretending you can have the latter while demanding the former is just...

There's a mistake in thinking the exodus was actually "we need a different model because the centralized one is fundamentally broken", that's creating this very culture clash. A lot of people actually WANT a centralized model but with better management. That's not this. Maybe another PDS could be.

Don't look at me, though. I'm not handing out invites to juche.social.

Frankly the only real alternative to building your own now is to give into the the greater devils of our nature and figure out ways to programmatically bully people off platform, as opposed to simply yelling at them until they leave. Mob justice with extra steps.

"Bullying works" is something we've seen before. If there's no central authority to pressure into banning people, the crowd develops its own enforcement mechanisms. Coordinated harassment campaigns. Mass reporting. Doxxing. Making someone's experience so miserable they leave "voluntarily".

The irony, of course, is that this isn't "Blueskyism", it's just .. classic online dynamics with laissez faire moderation. People acting like this is some novel phenomenon when it's literally just how every platform has worked since forever whenever there isn't "pro-active" moderation.

It isn't a bug of decentralization, it's what happens everywhere when there isn't strong centralized enforcement and people are left to their own devices. It's NextDoorism, if it's anything. Bluesky didn't invent it. ATProto didn't enable it. It's simply what humans do in online spaces.

The main difference is that ATProto is being honest about it in terms of context and trade-off; traditional platforms pretended they were moderating effectively while this stuff STILL happened anyway, just with higher boiling points.

Individual users mobbing up and bullying each other off the platform is annoying and ugly, but it's arguably small potatoes. The real question is: what happens when someone with actual resources decides to manipulate the system?

Quis custodiet ipsos custodes, but make it distributed systems architecture.

Frankly, the idea of architectural pluralism it could facilitate makes it more compelling to me, even I may have my own issues with the design assumptions. Fragmentation is here whether you like it or not. The polynopticon is ugly, but at least ATProto is honest-- more than we can say for Twitter.

Underneath, Twitter was always a polynopticon in drag: splintered realities held together by the illusion of unity. The algo fed different slices to different cohorts. Moderation rulings were selectively enforced, uneven, sometimes political, always subject to leadership changes.

Whole communities operated as if they were invisible to each other, until suddenly they weren’t.

Anyway. My worry about ATProto and this place now is not the pissant shitposter crybully who has to continually update their blocks and keyword mutes because they had to see something offensive to their sensibilities, it's the brittleness against state and corporate-level power.

Completely different threat models with different defenses. And yes, I'd include our own wonderful home country among those threats. It's concerning that the operational posture is "we’ll figure it out later". That’s not going to hold against anyone who isn’t simply a crank with a bot farm.

I mean, forget observation. Observation's a foregone conclusion with the firehose design. But the admin could pressure Bluesky PBC to filter content at the relay level before it hits the firehose.

Or, if you want an end-run around 1A, which could be more likely, demand the DID:PLC directory resolve identities differently for "national security". You're not taking down speech. You're making identity resolution work differently. For national security.

The posts still exist! The data is still in the user's PDS! We didn't censor anyone! Except nobody can find them. Their identity doesn't resolve. Their handle points nowhere. Or it points to someone else. Or it resolves differently depending on who's querying.