The Great Bluesky Migration: I Answer (Some) Of Your Questions
For the last year and a half, us Bluesky users have frequently reminded one another that we are merely posters on a niche microblogging website.
It's intended as a warning about hubris. A protective against the ever-present danger of poster's madness. A nerd-ass version of the Ancient Roman custom of assigning a slave to whisper "Remember you are a man" (Hominem te memento!) into the ear of victorious generals as they rode gilded chariots through the streets of the Eternal City during triumphant parades.
It's a valuable reminder, but there's a problem. The "niche microblogging website" part may no longer be true.
As I write this in November 2024, Bluesky has just reached a roster of 20 million users, after a massive wave of new-sign ups. In the last few weeks, politicians, big organizations, musicians, journalists (including some really awful ones) and even Polish Prime Minister Donald Tusk have either started Bluesky accounts or revived long-dormant placeholders. Amidst the recent flurry of mass-media articles about the rise of Bluesky, Slate was even forced to issue a correction concerning Alf's genitalia.
We've reached some kind of tipping point, some kind of collective herd-momentum away from the shambling corpse of Twitter. Or what was Twitter before Elon Musk, in the fashion of a parasitoid wasp, laid his eggs all up in it so that his hideous MAGA offspring could gnaw their way out of the hollowed-out husk.
This influx of new users also means I'm getting a lot of questions about how Bluesky works. This post is my attempt to answer some of them. And refer to the official, actual Bluesky FAQ for more information.
Is Bluesky really less hostile and take-driven than Twitter, in such a way that it's making America's overpaid pundit jerks nervous?
I'd argue that it absolutely is. For now, and I think this will continue.
This is because Bluesky, unlike Twitter, is not designed around a central ethos of "putting spiders in a jar and shaking it up to make them eat other."
This is just one of the many ways that Bluesky is different from Twitter, but it's a really important difference.
Bluesky has been designed to give users a lot of granular control around what they see and who they interact with. When you block someone on Bluesky, unlike on Twitter, you can't easily choose to see what that person is up to. All connection between you and them is severed.
All interactions you've had with that person are made illegible in Bluesky (although if you unblock the person, they'll become visible again). While it's true that that person can use an alt account to view your posts anyway, or access Bluesky via other means - after all, Bluesky posts are public - the so-called Nuclear Block has proved to be a surprisingly effective tool for shutting down pile-ons, fights, and other conflict-driven social media behavior.
Bluesky also offers a remedy for quote-dunking. If someone quotes your post to make a nasty comment on it, you can detach the quoted post entirely. (And then you should block the jerk).
Perhaps most importantly, Bluesky has developed a culture that is very pro-blocking, free of the social shaming that was once associated with using the block button on Twitter. It's an era I remember vividly, because I'm one of those degenerates that's been on Twitter since I was a teenager in 2008.
Many prominent voices on Twitter, who not-at-all coincidentally happened to be mostly conservative-leaning straight white men, believed that blocking was a coward's move, an admission that you lacked the mental fortitude to be exposed to Valuable Opinions Outside Your Bubble. Even if those valuable opinions happened to include lots of slurs.
In the wake of the once-and-future Trump era, most of Bluesky's users have figured out that this approach to blocking on social media is stupid bullshit that benefits centrist columnist mediocrities and right-wing assholes and is miserable for everyone else.
Generally speaking, Bluesky adheres to the same ethos that most people adopt in the real world. You are under zero obligation to listen to an asshole verbally abuse you just because you both happen to be at the same grocery store. You are allowed to tell him to shut up and to leave the premises. Same idea.
This is why assholes, including mediocre centrist assholes with prestige columnist jobs, are getting nervous. And, in some cases, having spectacular public meltdowns about how unfair it is that Bluesky users can easily choose not to listen to them.
People are figuring out that they aren't morally obligated to listen to assholes on social media. And Bluesky is actively helping them do that.
How can I block miserable assholes in large quantity on Bluesky? For is it not true that the world contains a tremendous multitude of miserable assholes?
There are indeed more assholes in heaven and earth than are dreamt of in your philosophy, and for this, Bluesky offers interesting remedies.
They call it composable moderation.
In practice, Bluesky currently offers three main asshole control mechanisms:
- You can manually block assholes. This is self explanatory.
- You can quickly block or mute a group of assholes by means of a list that you create or by using one created by someone else. When you subscribe to such a list, newly-added assholes that meet certain asshole-determination criteria will be automatically blocked or muted for you. You can find lists to subscribe to at this website.
- You can use labeling services created by you or someone else to report and categorize assholes and the asshole content they create. Think of these as user-created moderation services, which you can choose to subscribe to if you want. They're intended to complement Bluesky's own moderation services, giving the user community a lot more collective control about what they see on their feeds.
These services can also be paired with lists. It works like this: when you see an asshole post, or an account run by an asshole, you can report the post or the account to the person running the labeling service. They can then decide to add an Asshole label to that account, and they can also decide to add the account to the linked block list.
Labeling services don't all concern assholes. Many are simply useful, like the labeling service that allows you to set your own pronouns and see those other people use, or the labeling service that alerts you to the presence of sensual pictures of 1980s TV aliens. Others are fun, like the labeling service that allows you to play an RPG (no, seriously).
If you want to browse currently available labeling services, check this website out. More are added all the time.
A Warning About Block Lists and Labels
Don't blindly subscribe to block lists and labelers. Look into who's running it, and the accounts they're blocking.
Some list and labeler makers are actively malicious. (In recent days, we've seen a few networks of assholes make fake and misleading block lists to trick Bluesky users into blocking innocuous accounts run by trans people). Others may be working out personal beefs. And finally, there's list and labeler makers who are decent people but simply don't share your instincts around what accounts should be blocked.
Here's a great thread on how to vet Bluesky blocklists and filters from Rahaeli, a long-time trust and safety expert who describes this stuff way more coherently than I can. Please give it a read.
I know. This is a lot. I will be the first to admit that Bluesky really needs to put out more documentation describing how this stuff works for normal, non-professional tech nerd audiences.
In the meantime, feel free to ask the Bluesky team if you have any questions. They're communicative and answer questions. The broader Bluesky community in general has a good track record of helping one another figure out how these tools work, and there are a lot of friendly people who will help you.
You can even tag me on Bluesky, and I'll try to get your question seen.
How can I find people who aren't assholes to follow on Bluesky?
Use Starter Packs.
These are a neat Bluesky feature that allows users to create lists of other users, often aligned around a certain theme. Other users can then choose to automatically follow everybody on the list (or follow the accounts one by one, either way works).
You can find and search a directory of Starter Packs here, covering everything from Arctic and Antarctic historians, journalists who are active on Bluesky, UX and UI design, stormchasers and weather photographers, and a whole lot more.
More of these are also being added all the time. Consider making your own.
How does Bluesky's algorithm work? How can I win its favor?
Bluesky doesn't have an algorithm.
Well, technically it has a bunch of them. And you can decide which ones you want to use. They're called Feeds.
Users have made a lot of custom feeds already, and you can subscribe to and use whichever ones you want. On the Bluesky app and website, the Feeds tab has a search box. You'll also get recommendations for existing feeds. Let us say you want to see lots of posts about cats. Run a search for cats, and you'll see a lot of options, as in the picture above.
Bluesky does maintain the Discover feed, which is provided as a default option when you start a Bluesky option. The Discover feed is algorithmic, in that it's customized on the basis of what it thinks you'd like to see. However, you don't have to use Discover: you can delete it and forget it ever existed.
Unlike on most other big social media websites, Bluesky doesn't constantly try to force you to use its single, centralized algorithm. You will not be punished or cursed with total obscurity by the capricious Algorithm Gods if you decide not to use it. I don't use it! I just use my Following feed.
You can even make a feed yourself. I've made a few that capture terms surrounding natural disasters.
Hopefully, Bluesky will soon introduce new tools to make it even easier for people who don't have code experience to create their own feeds. There are also third party tools that allow you to make your own feeds.
Will Bluesky punish us for posting links?
Of course you can post links on Bluesky without your content or your account being banished into the outer realms.
It is in fact objectively insane that Elon Musk now actively punishes people for posting links to other websites on his dumb shambling husk of a social media network.
Post all the links you want.
Here's what Bluesky CEO Jay Graber has to say on the subject:
How private is Bluesky?
Bluesky isn't private at all.
Blocks absolutely do reduce friction and help shield your posts from most less-committed assholes. But a dedicated asshole who's determined to mess with you will be able to find a way to view your posts on a public social media website, even if you've blocked them.
Other users also can see who you've blocked, which is a tradeoff that comes along with Bluesky's intentionally open source and decentralized design. Bluesky's own app and website don't allow users to view other user's blocks, but third party websites and apps do provide this functionality. (I am not going to tell you where to find them).
So far, I haven't seen the public nature of blocks widely used as a vector for harassment. But that absolutely doesn't mean it's impossible. Keep this in mind going forward.
Currently, Bluesky also doesn't offer locked accounts - this is technically difficult due to the way it's designed. Scroll down for more on that.
If you need to communicate sensitive information online, please, I beg of you: use Signal for that. Don't post sensitive information on a public social media website.
Personally, I'm OK with the public nature of Bluesky. I think that social media is inherently a public thing. I do worry that people may be lulled into a false sense of security on social media websites in general by means of features like blocks and the ability to lock accounts. Bluesky has done a decent job of communicating this to users, but I think it's time for another public-awareness blitz on the subject.
Bluesky does have direct messages, but they aren't very secure - IE, they aren't end-to-end encrypted. Bluesky is working on adding encrypted messaging, but it hasn't rolled this feature out yet.
Again, use Signal for anything sensitive.
Isn't Jack Dorsey involved with Bluesky?
No. Jack Dorsey flounced off of Bluesky - and quit the board, too - in a fit of pique in May 2024.
Probably because too many people were calling him Jack Dorky, and also because people wanted to use Bluesky to talk about stuff that wasn't cryptocurrency and nasty-maxxing the appearance of one's beard. (I bet he's got chia seeds stuck in there).
He is not missed.
You can also refer to this post from Bluesky CEO Jay Graber, with regards to his involvement:
Bluesky is fun, but won't it inevitably be ruined by horrible billionaires like the other social media websites I've used over the last 15 years?
I share your trauma.
I've watched multiple formerly beloved social media online-homes crash and burn over the course of the last 20 years. I've also spent a lot of my working life since 2010 being professionally mad at sinister tech companies. I don't think I can be reasonably accused of being a wide-eyed and naive optimist about the relative perfidy of people who run social media companies.
I absolutely can't guarantee that Bluesky and its team won't turn cartoonishly evil.
However, I think there are some genuinely compelling reasons to give them a chance. The first, and most important reason?
Bluesky's team are trying to build a federated (or distributed) social media network. Yes, I know: bear with me.
Technically, they're building out something called the AT Protocol, which you can think of as a set of rules that allow different systems to communicate and exchange data with each other.
This means that Bluesky itself - the website and app you and I post on that looks kinda like Twitter - is built on top of the AT Protocol, and is intended to be a real-world demonstration of what the protocol can support.
For the 99% of people who aren't hopeless tech dweebs, these words are effectively meaningless. For now, you just need to know this part:
Bluesky is intentionally designing a social media network that will be a lot harder for a single hyper-rich asshole to buy and then completely fuck up.
Think of your Bluesky account as a little website, to paraphrase this really great, if somewhat technical thread from Bluesky team member Paul Frazee.
The idea that the Bluesky team is building towards is that even if Bluesky's central server (which almost everyone is using right now to store their little websites) goes down or becomes outlandishly evil, you'll still be able to move that little website - which Bluesky calls a personal data server, or PDS - to another server that's run by someone less evil.
You'll also be able to move all your existing posts and data to that other server. And because of the way the protocol that Bluesky is run on works, other people will still be able to find you and your posts. To paraphrase Paul again, it's a lot like how you can switch from using Google's search engine to Bing's search engine and still access the same Internet.
Here's another metaphor.
Imagine you own Trainbutts.com. You get fed up with your current web hosting service. So you move your website's data to a new web host, and you make sure that you transfer the Trainbutts.com domain name to your new hosting service. Voila: your established Trainbutts.com users can type in the same URL they've always used to access your website. The data just lives on another server now.
The idea behind the AT Protocol and Bluesky is that this should be how your social media profile works. People should be able to find you and your data, even if you decide to move somewhere else.
It's true that Bluesky is not yet fully decentralized.
The team is still building this out, and moving towards making it easier for this to happen in practice (such as, for example, making it easier for users to host their own servers, or use a service that will do it for them). But I do believe that this is what the Bluesky team genuinely wants, and I believe they're trying hard to make this happen. Which is why I'm sticking around to see if it does.
This process of building towards greater decentralization is also hastened by the fact that Bluesky runs on open source code, which means that other software developers are free to tweak, modify, and improve what's already been built.
There's a few more reasons why I don't think Bluesky is going to turn outlandishly evil in the near future.
For one: Bluesky is structured as an independent public benefit corporation (PBC) which means the corporation is permitted to spend some of its profits or resources to support a certain public benefit or mission, not just to advance shareholder's interests.
Another description of how this works comes from this source, concerning PBCs subject to Delaware General Corporation Law, which Bluesky is: the "directors of a PBC are tasked with managing the corporation’s business and affairs in a manner that balances the stockholders’ pecuniary interests, the best interests of those materially affected by the corporation’s conduct, and the public benefit identified in its certificate of incorporation."
This is a relatively new corporate structure, only introduced in 2010, and companies like Patagonia, Warby Parker, and Allbirds are now organized as PBCs.
It's true that PBCs actual adherence to their social mission is not exactly stringently enforced at this time. However, I think that Bluesky's decision to begin with this structure is a good sign, and an indicator about how they view themselves and their mission.
Finally, I genuinely like the Bluesky team and find their thinking to be well-aligned with my own.
I don't know them personally outside of the context of shitposting at each other on Bluesky since April 2023, but I've enjoyed my interactions with them. I think they have interesting ideas and a compelling vision for the future of social media.
I've been impressed with how many people on the team have deeply engaged with literature on the ethics of technology, and I've noticed that they seem invested in learning from the many, many horrendous mistakes that other social media companies have made over the last 15 years.
I've been consistently impressed with the choices they've made and how they've handled genuinely hard problems like content moderation and scaling the website up so it doesn't crash and burn whenever a bunch of new users join after yet another Elon Musk Event. They do a surprisingly good job of listening to and engaging with users. They don't fill me with a sense of ineffable and uncanny-valley unease like a lot of other people who run tech startup companies do.
Again, this could all change. I'm not ruling out Bluesky turning evil. Far from it.
To paraphrase "Heat," when it comes to my online life, I firmly believe in not getting attached to anything I'm not willing to walk out on in 30 seconds flat if I feel Mark Zuckerberg is coming around the corner.
But it's a risk I'm willing to accept on Bluesky, at least for now.
Here's some interesting reading on how Bluesky works, and how leadership are thinking about it:
Protocols, Not Platforms: A Technological Approach to Free Speech - Mike Masnick. A good explanation of why protocols - which is what Bluesky is really building - matter.
Bluesky, decentralisation, and the distribution of power - Laurens Hof, The Fediverse Report
Bluesky, and what Bluesky is not. - Alexia, WhiteWind
Some thoughts on what a potential Elon Musk-level Evil Event might look like on Bluesky from engineer Dan Abramov (and an interesting talk he gave on the AT Protocol that Bluesky runs on).
A conceptual model of ATProto and ActivityPub - Laurens Hof, The Fediverse Report.
Notable Bluesky facts from Bluesky CEO Jay Graber.
Won't Bluesky eventually need to make money? Which will then turn it evil?
Yes, that's a real concern.
The Bluesky team got into the Profit Question in this October 2024 blog post, describing their latest Series A financing round. I suggest you read it.
The main headline: Bluesky will soon be rolling out subscriptions. But don't panic.
These subscriptions won't impact how people talk to each other, or how visible certain accounts are, unlike Elon Musk's foul paid blue-checkmark system - and the team emphasizes that Bluesky will always be free to use.
Subscribing users will simply be entitled to some aesthetic perks, like higher quality video uploads and the ability to further customize the appearance of their profiles. (And if Bluesky really wants to make a billion dollars, they should let its nostalgic millennial user base make their profiles look like the awful ones we all had on Myspace circa 2005).
Other monetization ideas in the immediate pipeline include charging users for custom domain name registrations, and developing ways for users to pay content creators from the app. I'm OK with all of this.
Bluesky CEO Jay Graber outlined her thinking on how Bluesky might make money further in a Wired magazine interview, which is worth reading (note that this took place before Jack Dorsey left the board).
In the long term, I think there's a lot of fascinating non-egregiously-evil businesses and applications that could be built on the technology that Bluesky is developing. I believe that this could point the way towards a sustainable profit model for Bluesky, and for the further collective development of the AT Protocol.
I also think that maybe, just maybe, this could all be the start of returning to a less hideously corporate-dominated model for the social Internet. Which is cool.
Ah, but didn't you see that post alleging Bluesky's dark and illicit ties to crypto ghouls/Russia/China?
Far as I can tell, these rumors are bullshit. And are largely being spread by people who have an interest in seeing Bluesky fail. Here's a solid debunking of some of these rumors from Greg Wolf.
Bluesky does not rely upon or use blockchains or cryptocurrency, and the team have consistently and constantly stated that they're not interesting in "hyperfinancializing"the platform.
Some members of the team, including CEO Jay Graber, have worked for cryptocurrency companies in the past. I do not find this inherently unusual or notably suspicious for millennial and Gen Z tech workers in 2024. (Graber has also worked as a digital rights activist at non-profits, focused on protecting net neutrality and privacy).
Obviously, this could change. But prioritizing shady crypto scams is not what the Bluesky team are doing now, and I haven't seen any indications they're moving in that direction any time soon.
Bluesky does use cryptography, in the sense of using math to establish and verify the identity of a user, a device, or a system. This is not the same thing as cryptocurrency.
This is a pretty good explainer of the difference.
When Bluesky does turn evil, you'll come crawling back to Mastodon. Which I bet you don't even know about, you absolute rube. You'll see! You'll all see!
Look, I'm aware of Mastodon.
I've got nothing against Mastodon.
I think Mastodon, and its ActivityPub protocol, and the fediverse network that's been built around that protocol, is conceptually great. I'm happy it exists.
I'm glad that it's part of a global movement of people working very hard on building a decentralized social online world that's less vulnerable to dead-eyed billionaire freaks. I very much want Mastodon and the fediverse to continue to exist.
I also tried using Mastodon for a while and just couldn't get into it.
It moved too slow for me, the culture around content warnings was intimidating, and the overall tone was a bit too serious for my taste. I'm an inveterate shitposter, a horrible Internet gremlin: Bluesky just feels more accommodating of my personal style.
If I'm going to get more specific:
I don't like how ActivityPub doesn't allow you readily move your old posts from one server to another. The AT Protocol is at least intended to allow for "portable identities." In other words, I want to be able to take all my shit - posts, followers, blocks- and move to another server, if the server I'm on isn't working for me anymore.
As a consequence of this, I don't like how ActivityPub puts you (and your posts) at the relative mercy of whoever is administering the server you're on, for matters that range from moderation choices to de-federation with other servers. Again, I just prefer the AT Protocol vision for how federated social media should work. And I really, really like the AT Protocol vision for composable moderation, among many other things.
What's with the Alf stuff? Why am I seeing pictures of Tom Bombadil wearing a bralette on my Bluesky feed?
Yeah, sorry. That was me.
Faine Greenwood
I am not actually sorry.
