I believe the Bluesky experiment is already over, and it was not successful.
1. Bluesky was advertised as a very open, resilient platform because of AT, impossible to fully censor, with moderation at the user level, not platform level. In reality, "Bluesky Social" is what everyone uses (not AT), which performs bans at the core levels (PDS / Relay / user account level). So it's just like any other social media platform. Still better than the Mastodon Reddit-style mod abuse problem, but not fundamentally more open than other social media platforms.
2. The second promise is that if anyone hosts their own PDS, anyone else can build their own relay and their own social media app that gets around the bans. In reality, I don't see much appetite for using someone else's distributed key-value JSON schema design. The value add isn't clear enough. Other social media platforms build their own systems, which is an intuitively natural choice.
If Bluesky is not a highly open platform, and AT has low adoption (no killer app, hasn't realized the promise/benefit of being able to aggregate data from multiple publishers) then I don't think it has a future.
I think a response to this comment have to be built on two axes:
1. Is the anything goes Happy Hour of Bluesky over?
2. Is the remnant of happy hour useful?
And, IMO, the answer to both are yes. I bet it'll coast for a decade or two and I bet it'll do tolerably; the social media wing of The Internet was given an gratuitous 1UP.
Seems rather premature to call the experiment over when Bluesky/atproto has only been generally available for a year since they opened up registrations to all in February 2024. That’s a very short amount of time to explore the affordances of a brand new platform and ecosystem.
The development of that other protocol that was also blessed with a money injection from Jack, is NOSTR.
This protocol and all the clients and relays went the full decentralized route.
This had up and downsides. Not all clients support all features. Not all relays support all features, and the quality between these apps is greatly differing.
But the upside is: the development is truly decentralized. Nobody is in control and every developer chooses which of the many NIP (Nostr improvement proposals) they want to incorporate or ignore.
Off course, most users will just use the most popular clients (like https://primal.net/home and https://damus.io) but simply thaking your crypographic keys and moving clients, or using a whole bunch of clients next to each other, is just such a vastly better experience then logging on to one platform, either BlueSky or X.
Why should the experiment be considered over? It seems like there's still work in progress? Is there something preventing a "killer app" from becoming popular someday soon?
Bluesky is currently a terrible product with a paltry number of users. The small dev team can’t compete with social media platform features. React Native means bad mobile UX.
Higher bit-rate video, faster load times, app stability, and edit functionality (which many people repeatedly request, despite it being arguably low-priority because you can just delete and repost), just to name a few. I'm practically a Bluesky fan, and I again disagree with Steve's characterization, but even I see many things that have room for improvement. Granted, Twitter/X should crush them on this alone and they've arguably leveled the field by making their platform worse over time, so heh.
if that's the case, I'd think Bluesky has an advantage. Because there's open API and third-party apps actually works, compaired to very closed nature of X.com
at least I can read posts from public officials without an account on Bsky. representatives and public organizations using sites that require an account & agreement with a third party private Terms and Services just to read their posts should be illegal
I'm of two minds about this. I don't really hold a super high opinion of Jack Dorsey, but regarding why he left Bluesky I think he actually had a point. It's not really clear what Bluesky actually practically does other than directly compete with Twitter.
The thing that steams my rice personally is that I am friends and acquaintances with a variety of different folks who use social media. I don't think Bluesky would have seen nearly as much adoption if it didn't position itself as having solutions to the inherent problems of centralized social media: the same promises are what propelled growth on Mastodon, too. But see, even though Mastodon is flawed in many ways, it really does solve some of the problems of centralized social media. It lacks some of the theoretical capabilities that Bluesky and AT proto offer, and I absolutely think what Bluesky and AT proto offer on paper is amazing. Censorship resistance! Host your own PDS! Relays keep working when PDSes go down! Sounds good, whereas on Mastodon your identity is tied to where your data is hosted and there's nothing good to do when an instance goes down. However... In practice, none of it matters because indeed, AT proto adoption seems rather unimpressive. The Bluesky app view offering a single URL regardless of where the data is hosted is great, but... In practice, it centralizes all of the moderation for all users. The way Bluesky operates today makes it more like Twitter with extra steps. And knowing some of the moderation woes, I'm not sure I'd categorize Bluesky as unilaterally better than Mastodon; in my opinion, the main advantage of Bluesky over Mastodon here is the lack of instance-level blocks, which I think are grossly overused in the fediverse. OTOH, though, the ability to choose an instance means you do have a choice, however inconvenient it is. Bluesky having one appview, there is one choice. You get to see what they let you.
They built something impressive, and sure, it has its fans, but I worry it attracted people under somewhat dubious pretenses. It wasn't really decentralized in any meaningful way when users originally flooded in, and to this day the actual degree to which it is decentralized is quite questionable. I understand they have valid reason to fear giving up control in any truly meaningful way, but if they can't actually do that, then is there really a point to this? In practice, if Bluesky was completely centralized, it would basically make no difference. It wouldn't change much about moderation or how censorship-resistant Bluesky is, or any of that. It'd probably just make the damn thing a lot less operationally complex, and it would effectively be Twitter but for people who are left-leaning.
Thankfully for me, I don't really want any particular Twitter-like, but it's still hard to deny the importance of Twitter. A lot of the internet's personalities and creative talents are hanging out on these crappy social media platforms. I have my misgivings with the Fediverse, I've written at length about it in the past... But if I had to pick one, right now I'd pick the Fediverse over AT proto, scale be damned, subreddit moderator problems be damned, because it offers practically useful decentralization today and not "maybe some day."
This is the clear and present problem in my eyes. I disagree with your first premise. I feel like it's unfair to call it over and unsuccessful. Many people are investing directly in the problem we're discussing. It's my hope this is the core focus of everyone working on AT at the moment: incentivizing some kind of competition in the ecosystem and creating and actively demonstrating that value add. Many want Bluesky to be viable because the alternative publishing platform is a dumpster fire. However, it's only fair to admit this doesn't magically make it so. I'd really like to see some more concrete details on how they plan to generate revenue.
I don't know about you but I think trying to build a business on "Maybe (hopefully?)" sounds pretty scary. I'd much rather be literally giving away money (if that's what it takes) to get someone to be the Pepsi to my Coca-Cola on the platform.
For me, the most meaningful test of decentralization is whether one player can decide what most of the users can see. And that is entirely the case with Bluesky. You can own your information alright, but what is it good for when the single centralized app can decide what to do with them? If the single centralized app gets sour, you will be left with the impossible task of coordinating all your contacts to move elsewhere. If the single centralized app is down, nobody sees anything. How is this progress?
Meanwhile, the Fediverse has dozens of thousands of independent nodes and dozens of independent implementations.
The issue is that “BlueSky” is, for most of its users, the centralized official app. If that gets bought and starts blocking or demoting content from “non-official” clients, switching to a third-party client will not help you since you would lose your contacts.
And the longer this centralized situation lasts, the harder it will be to get back to a healthy ecosystem where nodes are truly peers and none of them can hijack the network on its own.
(Also, I am ignoring the complexity and cost of running an AppView. Running a Mastodon server is a $5/month affair, running a BlueSky AppView is several orders of magnitude more expensive.)
Agree. I am so happy that Dorsey's other open protocol (nostr) went another route. There is no 'official' client, there is just a whole bunch of competing clients, some of which have even attracted some small funding strategies to maintain a development team for the forceable future.
But none of these clients are big enough to push through locked-in features. They all have to play nice (for now, like SMTP and IMAP once did) and be interoperable.
I still immediately think of the Hayes (AT) command set every time I see "ATProto". I suspect the former will outlive the latter, so I'm not making much effort to break the habit.
I know next to nothing about Bluesky or ATProto outside of what this article told me, but: as I was reading the part about the need to trust who's controlling the PLC directory, I kept thinking that this could be a proper problem for trust-less cryptographic data storage mechanisms (like cryptocurrency ledgers) to solve. Isn't that right? You would shift the risk from "the entity hosting the PLC directory goes rogue and you can't update your data anymore" to "a sybil attack can boot you off the network"
I think maybe a globally consistent ledger ( blockchain ) isn't exactly what we need though, and what we actually want might be something closer to pkarr.org on top of the BitTorrent / Mainline DHT.
Because I think what we need is discovery and persistence of many self-verifying records that are independent from eachother.
In other words, each user's own identity record log is it's _own_ "blockchain" already, and we have no need for large, expensive, coordinated consensus, because each user only needs a single signature from one of their authorized keys to make changes.
So if we can just reliably distribute your identity records, then I think we've mostly solved the problem. Having the plc.directory as a standard, primary way to get those records doesn't seem like a bad idea, but having an official fallback to something like the Mainline DHT might be able to make it 100% decentralized in the event of a compromised plc.direcotry.
It's still a fuzzy idea in my head but eventually I'll do some serious thinking about this and try to get some actual proposal or community feedback if it seems like there's promise.
I believe the Bluesky experiment is already over, and it was not successful.
1. Bluesky was advertised as a very open, resilient platform because of AT, impossible to fully censor, with moderation at the user level, not platform level. In reality, "Bluesky Social" is what everyone uses (not AT), which performs bans at the core levels (PDS / Relay / user account level). So it's just like any other social media platform. Still better than the Mastodon Reddit-style mod abuse problem, but not fundamentally more open than other social media platforms.
2. The second promise is that if anyone hosts their own PDS, anyone else can build their own relay and their own social media app that gets around the bans. In reality, I don't see much appetite for using someone else's distributed key-value JSON schema design. The value add isn't clear enough. Other social media platforms build their own systems, which is an intuitively natural choice.
If Bluesky is not a highly open platform, and AT has low adoption (no killer app, hasn't realized the promise/benefit of being able to aggregate data from multiple publishers) then I don't think it has a future.
I think a response to this comment have to be built on two axes:
1. Is the anything goes Happy Hour of Bluesky over?
2. Is the remnant of happy hour useful?
And, IMO, the answer to both are yes. I bet it'll coast for a decade or two and I bet it'll do tolerably; the social media wing of The Internet was given an gratuitous 1UP.
Seems rather premature to call the experiment over when Bluesky/atproto has only been generally available for a year since they opened up registrations to all in February 2024. That’s a very short amount of time to explore the affordances of a brand new platform and ecosystem.
A couple days ago one of the most prominent alt-networks in the ATmosphere, Blacksky, received a $30k dev grant: https://bsky.app/profile/rudyfraser.com/post/3lk6gfblhdk2b
In a week from now the first ever atproto-specific conference of its scale will be taking place: https://atprotocol.dev/atmosphereconf/
The development of that other protocol that was also blessed with a money injection from Jack, is NOSTR.
This protocol and all the clients and relays went the full decentralized route. This had up and downsides. Not all clients support all features. Not all relays support all features, and the quality between these apps is greatly differing.
But the upside is: the development is truly decentralized. Nobody is in control and every developer chooses which of the many NIP (Nostr improvement proposals) they want to incorporate or ignore.
Off course, most users will just use the most popular clients (like https://primal.net/home and https://damus.io) but simply thaking your crypographic keys and moving clients, or using a whole bunch of clients next to each other, is just such a vastly better experience then logging on to one platform, either BlueSky or X.
NIPs feels a lot like XMPP XEPs )
“In reality, I don't see much appetite for using someone else's distributed key-value JSON schema design. The value add isn't clear enough.”
the value is that you’d start with an already existing full network of users
Why should the experiment be considered over? It seems like there's still work in progress? Is there something preventing a "killer app" from becoming popular someday soon?
bluesky has a very bright future because x.com sucks more.
Sometimes you don't have to excel at everything, just be better than your competitors.
Bluesky is currently a terrible product with a paltry number of users. The small dev team can’t compete with social media platform features. React Native means bad mobile UX.
What features are there? (Serious) Like, isn't it mostly just make a post (+media?) and catch likes/comments? That's some.
Is it discovery? Algo? Some AI shit?
Higher bit-rate video, faster load times, app stability, and edit functionality (which many people repeatedly request, despite it being arguably low-priority because you can just delete and repost), just to name a few. I'm practically a Bluesky fan, and I again disagree with Steve's characterization, but even I see many things that have room for improvement. Granted, Twitter/X should crush them on this alone and they've arguably leveled the field by making their platform worse over time, so heh.
> React Native means bad mobile UX.
if that's the case, I'd think Bluesky has an advantage. Because there's open API and third-party apps actually works, compaired to very closed nature of X.com
at least I can read posts from public officials without an account on Bsky. representatives and public organizations using sites that require an account & agreement with a third party private Terms and Services just to read their posts should be illegal
I'm of two minds about this. I don't really hold a super high opinion of Jack Dorsey, but regarding why he left Bluesky I think he actually had a point. It's not really clear what Bluesky actually practically does other than directly compete with Twitter.
The thing that steams my rice personally is that I am friends and acquaintances with a variety of different folks who use social media. I don't think Bluesky would have seen nearly as much adoption if it didn't position itself as having solutions to the inherent problems of centralized social media: the same promises are what propelled growth on Mastodon, too. But see, even though Mastodon is flawed in many ways, it really does solve some of the problems of centralized social media. It lacks some of the theoretical capabilities that Bluesky and AT proto offer, and I absolutely think what Bluesky and AT proto offer on paper is amazing. Censorship resistance! Host your own PDS! Relays keep working when PDSes go down! Sounds good, whereas on Mastodon your identity is tied to where your data is hosted and there's nothing good to do when an instance goes down. However... In practice, none of it matters because indeed, AT proto adoption seems rather unimpressive. The Bluesky app view offering a single URL regardless of where the data is hosted is great, but... In practice, it centralizes all of the moderation for all users. The way Bluesky operates today makes it more like Twitter with extra steps. And knowing some of the moderation woes, I'm not sure I'd categorize Bluesky as unilaterally better than Mastodon; in my opinion, the main advantage of Bluesky over Mastodon here is the lack of instance-level blocks, which I think are grossly overused in the fediverse. OTOH, though, the ability to choose an instance means you do have a choice, however inconvenient it is. Bluesky having one appview, there is one choice. You get to see what they let you.
They built something impressive, and sure, it has its fans, but I worry it attracted people under somewhat dubious pretenses. It wasn't really decentralized in any meaningful way when users originally flooded in, and to this day the actual degree to which it is decentralized is quite questionable. I understand they have valid reason to fear giving up control in any truly meaningful way, but if they can't actually do that, then is there really a point to this? In practice, if Bluesky was completely centralized, it would basically make no difference. It wouldn't change much about moderation or how censorship-resistant Bluesky is, or any of that. It'd probably just make the damn thing a lot less operationally complex, and it would effectively be Twitter but for people who are left-leaning.
Thankfully for me, I don't really want any particular Twitter-like, but it's still hard to deny the importance of Twitter. A lot of the internet's personalities and creative talents are hanging out on these crappy social media platforms. I have my misgivings with the Fediverse, I've written at length about it in the past... But if I had to pick one, right now I'd pick the Fediverse over AT proto, scale be damned, subreddit moderator problems be damned, because it offers practically useful decentralization today and not "maybe some day."
> AT has low adoption
This is the clear and present problem in my eyes. I disagree with your first premise. I feel like it's unfair to call it over and unsuccessful. Many people are investing directly in the problem we're discussing. It's my hope this is the core focus of everyone working on AT at the moment: incentivizing some kind of competition in the ecosystem and creating and actively demonstrating that value add. Many want Bluesky to be viable because the alternative publishing platform is a dumpster fire. However, it's only fair to admit this doesn't magically make it so. I'd really like to see some more concrete details on how they plan to generate revenue.
How many systems did SMTP start with? Maybe (hopefully?) AT just needs more time for more implementations to emerge?
I don't know about you but I think trying to build a business on "Maybe (hopefully?)" sounds pretty scary. I'd much rather be literally giving away money (if that's what it takes) to get someone to be the Pepsi to my Coca-Cola on the platform.
> I don't know about you but I think trying to build a business on "Maybe (hopefully?)" sounds pretty scary.
Isn't this, quite literally, "business as usual" for the vast majority of entrepreneurs?
For me, the most meaningful test of decentralization is whether one player can decide what most of the users can see. And that is entirely the case with Bluesky. You can own your information alright, but what is it good for when the single centralized app can decide what to do with them? If the single centralized app gets sour, you will be left with the impossible task of coordinating all your contacts to move elsewhere. If the single centralized app is down, nobody sees anything. How is this progress?
Meanwhile, the Fediverse has dozens of thousands of independent nodes and dozens of independent implementations.
I haven't experimented much, but I think BlueSky has customizable feeds?
https://docs.bsky.app/docs/starter-templates/custom-feeds
and with an open API, I think 3rd part apps can do their own customization?
The issue is that “BlueSky” is, for most of its users, the centralized official app. If that gets bought and starts blocking or demoting content from “non-official” clients, switching to a third-party client will not help you since you would lose your contacts.
And the longer this centralized situation lasts, the harder it will be to get back to a healthy ecosystem where nodes are truly peers and none of them can hijack the network on its own.
(Also, I am ignoring the complexity and cost of running an AppView. Running a Mastodon server is a $5/month affair, running a BlueSky AppView is several orders of magnitude more expensive.)
Agree. I am so happy that Dorsey's other open protocol (nostr) went another route. There is no 'official' client, there is just a whole bunch of competing clients, some of which have even attracted some small funding strategies to maintain a development team for the forceable future.
But none of these clients are big enough to push through locked-in features. They all have to play nice (for now, like SMTP and IMAP once did) and be interoperable.
For info, check out Lopp's blog: https://blog.lopp.net/why-nostr-matters/
despite being crypto-adjacent, nostr seems to have mostly solved this problem at least.
I still immediately think of the Hayes (AT) command set every time I see "ATProto". I suspect the former will outlive the latter, so I'm not making much effort to break the habit.
I’d encourage everybody to read the article, which gives a fairly thorough and nuanced view of atproto in its current state
I know next to nothing about Bluesky or ATProto outside of what this article told me, but: as I was reading the part about the need to trust who's controlling the PLC directory, I kept thinking that this could be a proper problem for trust-less cryptographic data storage mechanisms (like cryptocurrency ledgers) to solve. Isn't that right? You would shift the risk from "the entity hosting the PLC directory goes rogue and you can't update your data anymore" to "a sybil attack can boot you off the network"
I think maybe a globally consistent ledger ( blockchain ) isn't exactly what we need though, and what we actually want might be something closer to pkarr.org on top of the BitTorrent / Mainline DHT.
Because I think what we need is discovery and persistence of many self-verifying records that are independent from eachother.
In other words, each user's own identity record log is it's _own_ "blockchain" already, and we have no need for large, expensive, coordinated consensus, because each user only needs a single signature from one of their authorized keys to make changes.
So if we can just reliably distribute your identity records, then I think we've mostly solved the problem. Having the plc.directory as a standard, primary way to get those records doesn't seem like a bad idea, but having an official fallback to something like the Mainline DHT might be able to make it 100% decentralized in the event of a compromised plc.direcotry.
It's still a fuzzy idea in my head but eventually I'll do some serious thinking about this and try to get some actual proposal or community feedback if it seems like there's promise.
I'm now the ATmosphere's biggest fan of Charlie the Owl. He's a weird one.