Bluesky founder and CEO Jay Graber illustrates what it looks like to free our social identities from large walled gardens and leverage open source software for a better internet.
There’s a reason journalist and Bluesky board member Mike Masnick calls the platform “the most interesting experiment going in social media.” Originally launched as a project within Twitter in 2019, Bluesky has since become an independent company intent on making social more like the web.
What does that mean, exactly, and why does it matter? Bluesky founder and CEO Jay Graber says social media is stagnating because “we're in this trap where users are locked in and developers are locked out.” It’s time to open things up again, she states, like in the innovative early days of the internet.
Highlights of this conversation:
• Bluesky’s origin story
• The case for decentralization — and Bluesky
• Developer activity and other “wacky experimentation”
• Workings of identity online and DIDs (decentralized identifiers)
• Bridging AT Protocol and ActivityPub
• Bluesky’s exciting cultural moments
Mentioned in this episode:
🔎 You can find Jay at @jay.bsky.team
✚ You can connect with Mike McCue at @mmccue.bsky.social and at @mike@flipboard.social
This transcript was generated by AI, which may affect its accuracy. As such, we apologize for any errors in the transcript or confusion in the dialogue.
Bluesky started as a project within Twitter in 2019. Its goal was to develop a decentralized protocol that would allow multiple social networks to interact with each other.
Over time, however, circumstances changed. Bluesky spun out and built its social platform on top of the protocol originally developed for Twitter, called the AT (“at”) protocol.
Today, journalist and board member Mike Masnick calls Bluesky quote “the most interesting experiment going in social media.” Why is that? How is Bluesky integrating with the rest of the social web? And what’s really cool about how they’re thinking about identity online?
Welcome to Dot Social, the first podcast to explore the world of decentralized social media. Each episode, host Mike McCue talks to a leader in this movement; someone who sees the fediverse’s tremendous potential and understands that this could be the Internet’s next wave.
In this episode, Mike’s talking to Jay Graber, the Founder and CEO of Bluesky. Jay says one of Bluesky’s biggest principles is to make social like the web, and she believes open source software is the key for a sustainable and healthy digital commons.
We hope you enjoy this conversation.
Mike McCue:
Jay Graber, welcome to Dot Social. It's so good to have you here.
Jay Graber:
Good to be here. Thanks for having me.
Mike McCue:
So when did you start the Bluesky project?
Jay Graber:
Bluesky kicked off within Twitter as a Twitter project in 2019 I got involved as an outside expert on decentralized social that they pulled in. There was a matrix chat room they created with about a dozen people. And so it was me, people like Jeremy Miller, who's now on our board. We're also in that room, and we just advised Twitter on how they could do Bluesky. But in 2020 the pandemic hit, and so that froze up everything that could happen for a while. And then when they chose me as lead, it was about mid 2021 I spun out a separate company at the time, because Twitter just moved so slowly. It taken them like, you know, year and a half to even choose a lead. And so I felt like it would take a long time to get anything done within Twitter, also incredibly neutral open protocol requires some distance from like, a large company that has very entrenched existing incentives. Like, it would just be impossible to build it within that structure. So we spun out Bluesky is a separate company, and then Elon bought Twitter, and that sort of changed everything. After that, we were off on our own. He cut all he cut the whole company, almost all everyone who worked there, all the contracts. So we were off on our own. And then we built the Bluesky app on top of the app protocol, which we built for Twitter, yeah.
Mike McCue:
And what was that like when Elon came in and took over Twitter?
Jay Graber:
It was pretty chaotic. I mean, we'd been working through the, you know, working our way up the chain at Twitter, talking to people up and down the organization, trying to socialize ideas around Bluesky, getting feedback on design documents, on architecture documents, all this stuff. It kind of took months to get a full round of review on anything from within Twitter, even like a press release or something. So it was taking a long time, but once Elon bought Twitter, things slowed down even more, because nobody within the org was sure what they had the mandate to do or not do. And so it was just kind of this weird period of limbo where they tried to, you know, keep calm and carry on. But everything was strange. And so we were finding out about everything, like from reading the news, like we didn't know anything internally, because we're not a part of Twitter. It's like anything that we even heard from meetings with Twitter was like, what they might tell like an external contractor, which is what we were, right? So the day that Elon finally bought Twitter, we found out in the news with everyone else, and we're like, well, all right, that changes things for us, you know. And basically the we'd been set up for some a scenario like, sort of like this, because of my sense that centralization has this inherent risk to it, right? Like at the time when early on in Bluesky, they proposed having Bluesky be like a subsidiary of Twitter, or like within Twitter, we were earmarked much the same as birdwatch, the team that built community notes for Twitter. And so it was that kind of an experimental project. We might have been within Twitter, but then I looked at this, and I thought, Look, Jack is our biggest champion there. If Jack leaves, then the project support could fall apart. And, you know, the captain can sink the ship, is something that I always say. And so it's like, you can have a captain that steers you true, or you can have a captain that steers you into an iceberg, and that's just the high variance nature of centralization. Things can go very well or very poorly. And so I wanted some resilience against that outcome. And so I didn't know Elon was going to buy Twitter, but I knew that maybe Jack could leave, maybe some chaos could happen over the five years that Twitter had committed to the project. And like, five years in Tech is a long time, so things just played out a lot faster than I even expected.
Mike McCue:
And how big was your team at that time?
Jay Graber:
Three people, me and two other folks.
Mike McCue:
So that required quite a bit of leadership to get through that period of uncertainty. When, you know, I think it is pretty clear that that whole change with Elon taking over Twitter gave the concept of decentralized social media a whole sort of reason for being right. You know, I think that it's, it's kind of the height of irony.
Jay Graber:
It did, because you can tell people centralization can lead to all sorts of bad outcomes. It's high variance. Like, sure, you might, like, who's running Twitter now, but that could quickly change if you have the structure that's set up for that. But it's all hypothetical until it actually happens. People are like, Yeah, but like, when would that happen? You know, like, things are fine, right? And, like, that was the sentiment before, and I've been working on decentralized social for years before and explaining the benefits of decentralization ultimately, you know, there's just a few very concrete things that it gives you. One is experimentation and open innovation, and like, parallel experimentation. And then another is resilience and anti fragility, right? But like that resilience, you never know you need it until you do. Even with, like, supply chains during covid, we didn't know that. You know, auto parts supply chains were so centralized until suddenly there's a bottleneck in one of the critical points, and now it's like, oh, well, I wish we'd gotten those from somewhere else as well. You know, so social and decentralized social is the same. It really needs to be built out for those moments when, over the long course of history, there will be moments where political power and technological power like become, you know, there's like sharp disruptions, and you will want a communications infrastructure that is more distributed than one person, one billionaire, one autocratic entity being able to control all of it.
Mike McCue:
I listened to your interview on hard fork with Casey Newton. I thought that was a fantastic episode. Definitely recommend people who haven't heard that go back and listen to your interview there. I One of the things I love from that was you talked about how centralization, decentralization, decentralization, people don't really know that that's important, or even why that's you know why that matters, until they need it. When you start to see these kinds of things happen, and you realize, you know, having a decentralized model helps protect against a rogue, you know, CEO or a, you know, other kinds of problems that can happen with a single business controlling everything that's happening on a social platform.
Jay Graber:
Yeah, I mean, I think, like, everything that's happening with X kind of proves the case for why you need something with Bluesky. And there's also, there's like, the downside prevention of decentralization, the risk mitigation of things going wrong with a centralized company. But then there's also the upside that people don't understand because they don't see its benefits in a very centralized architecture. And so, for example, the web is an open protocol, and so much experimentation can happen there, so much new stuff is built. And when you have something where something where people can just experiment, and developers can jump in and solve problems, you get this multifaceted sort of diverse ecosystem of different sorts of things happening. When you have a situation where only one company can make changes, and if you want a change to your feed, your algorithm, your user experience, you have to rely on that company to decide to make that change, then you're in a situation where innovation gets really bottlenecked. And so opening things up to innovation is, I think, the upside of that. And so that's why we've tried to show users look. You can pick your feed, you can pick your algorithm, you can customize all sorts of things about your experience. And like, you know, we are still a company building a client that has, you know, limitations in terms of, like, what it can do, but anyone else can come along and build another client, anyone else can, like, hack on the open ecosystem and build new experiences. And then that's also a way to, you know, not just give users and fun new experiences, but also, I think, a way to address some of the problems that we see in social at scale. And so if it's concerns around mental health or concerns around like politics and censorship and like other things, like being able to experiment with different ways to address these problems, is I think, the best way forward, rather than saying there's like, you know, one man with the plan and one company is going to fix it, like, that's probably not a tenable solution, or it's not a tenable path towards a solution.
Mike McCue:
Yeah, I think that's so well said. You're decentralizing the innovation. It's not just coming from one or two or three people. It's coming from a whole host of different people with different backgrounds, different perspectives. I had this real aha moment when, as a developer for Bluesky, Flipboard has been developing on Bluesky for a long time now, and it's a, it's a really elegant API. There's, we've also created custom feeds for Bluesky. There's, it's a really fantastic platform. I had this aha moment when I saw, I forget the name of the app, but it basically was just a website that you could go to and you could see the see the entire Bluesky fire hose in real time happening. It was fire sky.tv, yes. Amazing, absolutely amazing. And then, as a developer, you think, okay, what can I do with all of this, everything that's happening here, all this data that's traversing the network? You know, it used to be like Twitter would never give the fire hose to anyone ever right? I think maybe Google had a deal for a while where you could get access to the Twitter firehose of data. But to, you know, to fire up this app and to see the Bluesky fire hose in real time was really something. And you know, then I've really been blown away by some of the other you know, projects. And there's sky grid, which lets you make these really compelling custom feeds. And super, super cool. What are some of the things for you, as you're looking at developer activity on your platform that you're particularly excited about?
Jay Graber:
Yeah, like seeing, you know, Sky feeds and some of these others take off with custom feeds was really exciting to see. You know, people at first were asking, who's going to make their own algorithm? Who's going to like, that seems so difficult. And really it's just, you know, developers started beating us to the features that we thought we could build out eventually and just building out the, you know, simple web interface for building your own feed, and then people have created a bunch of feeds. Like, there's 50,000 custom feeds that have been created, and most of those are built by people who don't know how to code. Like, there was just an interface that emerged that let people tinker around with it. And there's even more complex interfaces that people could build over time. Like, people are, like, what if you could just describe what you want, and then you could, you know, machine learning could, like, you know, extract some set of features for your feed. It's like, yeah, like, you could build that. And, like, that's something any third party dev could build. There's other examples of, you know, third party developers just beating us to features and, like, building things out that people ask for. I think that's really cool. Like, that's, you know, the whole ecosystem can evolve based on what users want and are asking for, not just through us. But, you know, if we don't have the capacity or the time to do it, or we didn't prioritize it, but someone else wants it, someone else can build it. And so this means that innovation doesn't have to happen in one company. You can just anyone. Can just build, you know, whatever they want. There's been some really wacky experimentation too. Like, one of the things I really love to see was somebody threw up a app that put your Bluesky posts in virtual reality, sort of a open space, and then you had a little anime avatar that summarized your timeline, and they could, like, walk around and explore your feed in 3d and this is something where I don't think it's like most consumers will, like, want to use it this way, but it's pretty cool, and you can just experiment with doing wild things like that, even within the app right now, there's really deep, deep integrations with third party feeds and moderation labelers. They're like a third party can build one of these, run it on their own servers, but it shows up in the app, like it's just part of the app experience. And so people have built things like a geeky versus Buba labeler, which just like, looks at your profile and tells you if you're Hiki or bubo, which is, like a weird little thing that's like, you know, just people's perceptions of shapes and sounds. And another thing is, is like people playing made a TTRPG game that, like, is based off this labeling system. And you can, like, roll for a character class, then people are, like, dueling each other on the network. And so these little emergent games can happen in an ecosystem where anyone can build,
Mike McCue:
Yeah, it's really cool. Yeah. You know, for a time, Facebook had their open social graph API, and there was a lot of innovation that happened in the Facebook newsfeed. You know, Farmville was a notable example of that. And you know, then, of course, the social graph API got closed down, and, you know, all sorts of changes happened, but there was this sort of brief moment of experimentation, and some some big new companies emerged from that. You know, I think Spotify, Pinterest, Zynga kind of came from those days of like having at least some sense of an open access to a social graph that a lot of people were participating in, and we're able to build social experiences on top of that,
Jay Graber:
Yeah, part of the reason I think something like Bluesky needs to exist now is 10 years ago, you know, social was a much more open place. APIs were more open. There's more experimentation happening. And even not too long ago, there was stuff like block party being built to augment your Twitter experience help you moderate, you know, your replies and like, you're just experience better. And this is an example of, you know, innovation coming from somebody who experiences the problems with the online harassment directly and just wanted to build a solution for it. But, you know, they've pivoted their business, and they're not doing that anymore, because now they're priced out of the APIs and things like TweetDeck, you know, were acquired and and there's really like, not that open space for these kinds of third party clients or third party experiences or augmentations to your social experience to exist anymore. So even, you know, creators who are upset about, you know, changes to the algorithm affecting their whole livelihood on, you know, TikTok or Instagram, there's no alternative for them. And things that came along, like there was an app, og app whose founder worked with us for a while, and he built an alternative to the Instagram timeline that just gave you more control over what kinds of feeds you could see, because people wanted, you know, more control for their mental health, for like, their livelihoods. And they got, you know, 40,000 downloads in a day, and then Facebook shut them down, and like drove them out of business. Yes, you know. And so this is a very monopolistic position that we're in right now, essentially, but I don't think regulation is going to get to it in time, or necessarily tackle it the correct way. And so we need people just sort of taking a grassroots approach from the bottom up, using and building on open ecosystems.
Mike McCue:
I've been struck by the terminology that you've used in on your website. And as you talk to people about your vision, you talk about the social internet. And, you know, really, it seems as though you think of social and the internet as really being kind of completely integrated that, you know, there's sort of like part of the same thing when you look at, you know, how the world of the social internet is going to evolve, and then you look at other protocols, like activity, pub, and what's been happening on that front, how are you seeing this when you step back, you know, how are you seeing this starting to shape up?
Jay Graber:
Yeah. I mean, I think one of the observations at a very high level is that, you know, the web linked data, but it didn't link people very well, and so all of our social identities gathered into these large walled gardens. Like, the reason, you know, Facebook and these other things really came to be is that, like, when you wanted to have, like a follow relationship, or like the likes and like all these like social primitives that emerged. These were taking place on centralized sites, and you could just put up a website, and, you know, RSS let you, like, pull other people's blogs and stuff, but that wasn't really enough for this new richly interactive model of socializing online, and so we really wanted to build a protocol that would like, capture some of those primitives in a way that is composable and extensible, and so that you can mix and match these into new ways of doing social one of the things is, just like our philosophy of user data is it's we were inspired by like Git and GitHub, so you can, like, have all of your follows, your likes, all your data you create in something like a Git repo that, like, moves around with you, and then you're able to put those on different social sites, just like GitHub lets you display your Git repo, but that's not like the canonical source of your data. Like the data that I create as a person interacting with other people on the web is like within App protocol, all within my repo. And then, you know, this could extend to all sorts of new types of data in the future, like the lexicon system for specifying like how you extend the schemas in that protocol lets you mix and match different ones, like new modalities. And then our hope, our hope is that all these pieces can evolve separately as people find them useful. So things like the identity system of app protocol, we built off of D IDs, decentralized identifiers, which is a standard that was being worked through by people for several years before us, and we wanted to build off something other people were using. But, you know, take an approach that served our use case, but could be extensible to others, and then ideally, people start to converge around something like a D, I D, that will let all sorts of different apps interoperate. And our idea for that protocol is that someday you just have a home on the web, which is like your social identity. It's an open protocol. It can be under a domain that you own. So building off the open systems of the web as it was already built out. So like, I could buy my domain name and then under there, you know, there's all my links to different social sites. If they're all, you know, compatible through the same protocol, they can share pieces in terms of, like, you know, curation or moderation or anything else, and then it can interoperate. And so, like, we don't have to necessarily be the stewards of every piece of this ecosystem, even it's just that the pieces can plug together to create a rich social experience.
Mike McCue:
And that's what you mean by composable when you when you use that phrase. So, so, yeah, the concept of these different components being able to be plugged, plug and play to create different kinds of social experiences and allowing different people to innovate and advance those whether it's for moderation or identity, you know, etc, that that's a really incredibly smart way to approach things.
Jay Graber:
Yeah, maybe a concrete example would be so there's a blog that someone on our team runs, and several other blogs have adopted this approach, but something that she did to show what was possible is that she took the blog comments for her blog post and made those replies to a Bluesky post. And so when she posted the blog post, all the replies of the Bluesky post are being fed into her blog commenting system as the comments on that post. And then what that does is let her blog comment system now borrow from the moderation system of what she's built for herself on Bluesky. So if she's blocked people, you know, where she's like, you know, as part of some groups or not part of others, like she can all that can get pulled over into the commenting system. And so that's something where you can have something that builds off something else, like the blog post itself doesn't necessarily have to be a part of that proto it could be, but it could also be separate, and then you're just pulling a piece of it. Another example could be like your username for your Bluesky account can be a domain name that you own, and then people have built a browser extension that checks out domain names and sees if there's a Bluesky account associated with them. And so. You know, bob.me could link to my Bluesky profile, and then you could start having social interactions, like, based off my blog post, because you immediately see that I have Bluesky profile.
Mike McCue:
When you talk about this social internet that's that really is like, you know, you type in, you know, a domain into a browser and it lands on a Bluesky profile page, and that is the website, right? That is, that is really, really powerful. And, of course, with all of the developer capabilities that you have, you could make that website look and feel less like a page out of Twitter, and more, you know, a profile on Twitter, and make it more like a, you know, fully rich website, but that also has this inherent social layer underlying it all, yeah, and
Jay Graber:
like the idea of your app protocol identity is, you know, bob.me let's say that's my domain name. That could be sort of a link tree type of landing page. That's my personal home page and blog, and it links all my social profiles. And then you can go to my Bluesky profile, like my micro blogging profile, and I'm also bob.me there. And then you can go to my, you know, TikTok style app profile, and I'm also bob.me there, and as the creator, can really own their identity. Then across all the platforms, like now, if you are a new music artist or something, to come up with a cool name, you really, it's a real struggle to make sure that you get something resembling your name, because, you know, it's the namespace for each app is unique and separate. Here we're building off the namespace of the web itself. So if you own like, you know, T swift.com like that can be you everywhere you go.
Mike McCue:
Which is pretty amazing, yeah. Right. Now I have in Flipboard the ability to log in to my Mastodon account, which is Mike at Flipboard Dot Social, and then my Bluesky account and also my Pixel fed account. So I have three different accounts right now, and of course, I have my Flipboard account as well. So now I have four accounts, and we are recently started to federate Flipboard. So if you search for me, right, you're going to see like, at least four different versions of me on four different, you know, kinds of app experiences. And you know two of them, or three of them, I guess, are on ActivityPub, and one of them is on at proto for normal mere mortals, like my mom, if she wants to find me and see my stuff and follow me, right? How do you see this kind of playing out, where it becomes simpler to have my identity be connected to all these other apps I might be using to post. What's your sense there of how that's gonna shake out? Yeah,
Jay Graber:
I mean, long term, the way we'd like this to work is when you meet someone and you're exchanging info right now, it's a bit like, are you on Instagram? Oh, no, okay, well, are you on Twitter? Okay. Well, right text, you know, but the idea of the at protocol is your at can be your identity on the social internet. And so it's like, what's your at? You would say, i@bob.me'm that's a domain you own. You go there, and then that links to everything else, and that is across all of your at protocol apps. And if a lot of social apps are compatible with that protocol, then suddenly you just have one name as something that you know, it's it's not like a namespace that we control. It's built off the web itself. And then that's essentially back to everyone having their own blog again, except in this, you know, new, fully featured social way, where you don't have to even learn how to run a web server or anything. So, you know, the goal would be like your mom would have her own domain. Doesn't necessarily think about it that way. That's her at handle now everywhere, and then you can find her across all the social sites from that she wants linked to that handle.
Mike McCue:
So it's sort of like a it's sort of like linktree, but built on an open protocol. Yeah,
Jay Graber:
exactly. And it's kind of wild to me that linktree is, you know, a fully funded company just to give people a personal landing page that aggregates all their social links again, because things got so fragmented, you need to pull them together again, but then that's another point of centralization. And so I think pulling them together around an open standard where things can, you know, gracefully come together, and, you know, fragment apart, but under a cohesive way of being able to reference things that's going to make this whole ecosystem play out better. Well, I
Mike McCue:
remember when back in when email was still evolving on the internet, and you had pop three and you had IMAP, two different protocols, two different ways of getting your email and sending email. But you didn't have to think about that as a user, right? You would just, you know, plug in your email account. It, maybe it was powered by IMAP. Maybe it was powered by pop three. You didn't really, you didn't really, you know, notice it the client just sort of like, smooth that out. And so for normal people, they were able to just, you know, send emails and receive emails. Do you see that happening with things like activity, pub and and at proto?
Jay Graber:
Yeah, that's a good comparison, because I think something like that is already happening. And open protocols definitely have the ability to merge together like this. And so right now, there's a bridge that's being run that bridges Bluesky and master. On and essentially at proto and ActivityPub, they're not like fully compatible, but this means that, as a user, that's hidden from you, it's the fact that people are on two different protocols talking doesn't really show up. And so now, when I post about Mastodon or ActivityPub on Bluesky, I get people in my replies commenting from Mastodon because this bridge is mirroring posts across them. And I think the same is true for noster and a few other open protocols where people have built bridges. But, yeah, this bridge, when it was created, was it received some pushback from Mastodon community, because there's just, I think, cultural resistance to Global Discovery, actually, which is, you know, one of the one of the things there that like meant that we wanted to build a protocol that really embraced Global Discovery as a core piece of it, but it's still possible, and like now, this bridging is happening, and so I think we'll see how consumer behavior evolves. But in the future, it's possible that your clients just speak several different protocols under the hood, and they all sort of come together and do a unified experience, as long as they're open protocols.
Mike McCue:
We talked with Ryan who built bridgey fed a few Dot Social episodes ago, and it was really interesting to hear how he was thinking about it, and the experience he had in building that and releasing it, and the expectations that people have for, you know, bridging to other protocols or other social networks. And you know, it is. It's kind of a new concept for people, but it is a remarkably simple thing. At the end of the day, I follow you from my Mastodon account today, and it works great. I also follow Mike Masnick. Mike sometimes posts on Mastodon, sometimes he posts on threads, and he's federated on threads, and then sometimes he put any lot of times he posts on Bluesky, and so I can see wherever he posts. I get it, I still see it in the client that I'm using, which is really, I think the way it should be, right. Mike should be free to post from whatever app he wants to for whatever reason, and still allow his audience to see his post, right?
Jay Graber:
Yeah. And like, this is what creators and users want. You know, the fact that we get siloed off into apps is not something that is a good for UX or the user experience. It's really just the fact that these companies have the network effects, and it's hard to get away from that. And they've, they've been able to close down the API's because of that. And I think in a world where, you know, consumer preference drove things more, we would have more interoperability and more ability for people to, you know, have different preference. People do have different preferences on the UX if they want on, you know, the sort of moderation preferences they want, and they could interact with each other through different clients that offer different services, and the service isn't getting in the way of the interaction. Like right now, the service you use to talk is in the way of talking, which should not be the case. You should just be free to post and talk to your friends wherever they are, and they might be using something else, you know, part of a different community, part of a different service, but you can still reach them, and that's what an open protocol is going to give you. Like, I can email you at your Yahoo account or your personal email server, even, and it's the email is still going to go through, because email is an open protocol, and I can be on Gmail, and you can be on your university email, you know, yeah,
Mike McCue:
well, you know, today, what's interesting to me, if you zoom out and you look at this whole space, The concept of open, decentralized social networking. There are at least three major players taking fairly different approaches, but with the same vision, you've got what you've been doing with at proto you've got what has been happening on the ActivityPub front, with Macedon and other people that are federating into the ActivityPub world. And then you have threads which is also adopting ActivityPub, but also has a lot of people who haven't federated. It's not it's an opt in experience. So you have to know that you want to federate, which creators, I think, increasingly, will believe that's important thing to do. But there's an awful lot of people who are on threads who are not federated. So what you have now is some, you know, fragmentation that's starting to develop, right? And some siloing, right? You know, do I join Bluesky? Do I join threads? Do I join mastodon? Do I join, you know, Flipboard, where? Where do I Where do I set up my sort of social networking presence? So I It does seem as though there's beginnings of some interoperability coming together, like bridge efed, what? And we talked a little bit about this kind of idea of a common identity that could span all of these systems. What? What are some of your other thoughts about how, how ideally would you like to see these worlds start to come together, as opposed to, you know, further fragmenting?
Jay Graber:
Yeah, I think you need an open identity standard that is fairly flexible. So we picked the D ID standard because it is fairly flexible. I would like to see more people converge around it. It's something where you know you can support several A client can support several D ID methods, and just like not having to know what email protocol you're running. You don't have to know which you know Id method your people you're talking to are necessarily using. And then you can start to make identity something that also has that freedom to evolve protocol wise. So I'd like to see that happen. And then I'd also want to make sure that we are using a system where users have the right to leave. And so one of the reasons we didn't build on ActivityPub is your identity. And all your data gets very tied to your server, and you can move, migrate your account, but that requires your old server cooperating to redirect you over, and then you have to, like, take all your data, move over. It's not a very seamless process. I think giving users the ability to leave with a clean user experience, so that everything is bridge and interoperable will allow this evolution of social protocols to take place more fluidly, the threads version of Federation right now, since you know inActivityPub identities are all tied to your account or something, if they just like, shut that down, then you collapse back to being a centralized service, because these pieces are not separate, right? It's just a way for centralized it's essentially a way for centralized services to each other, which does give you Federation and interoperability, but because the layers of identity and data and other things aren't separate, you aren't able to really have the flexibility to move those pieces around. So that's one of my concerns. Is that things could Facebook could decide tomorrow, like, ah, you know what? This Metaverse thing isn't working out. Let's shut it down. Boom, everyone's back on like, a centralized server. We would like to have an identity system and like the core primitives of like user data, like your data and identity being in a system where, you know, one company doing that shuts down a service, but doesn't shut down the open ecosystem.
Mike McCue:
You know, this particular point you're making, I think, is one of the most important things for technologists and product people to understand, you know, this idea that your identity currently is attached to a social platform of some kind, and to move that, to move around, you kind of have to walk away from that identity. The thing that I think is really interesting is that you guys have been thinking about identity built on this decentralized identity protocol from the W, 3c, this is something that has been people have been talking about working on for quite some time now, right?
Jay Graber:
Yep, it's, you know, started years before us people. There's, like, hundreds of did methods at the time that we got started. We looked through all of them, did an analysis. We ended up creating one called PLC is the public ledger of credentials, and it's just a very simple one that served our use cases. But, you know, the idea is that anyone can create a did method, try to get adoption around it, and then if it gets enough traction, you know, other clients can implement it, and like the W 3c, maintains this universal resolver that tries to, you know, resolve all of the hundreds of did methods, but you don't really need them all. Many of them have, like, zero to no usage. So the goal is, just like, you know, early on in the web, there are a lot of competing protocols, lots of computing standards, and then we kind of ended up standardizing around a few. The same thing is starting to happen in the digital identity space, and we'd like to just see that standardization happen around a few sensible identity standards that give people a choice in terms of how they want to manage their identity. For example, in Bluesky, you know, we have PLC that, like, you know, we manage, but anyone else can replicate, or you can self host your identity. So we use did web, which means that you can set up your identity on your own domain, and then just say, I'm going to maintain it. Don't worry about it. I have the technical skills keep my domain up, and if it goes down, then we, you know, Bluesky's not gonna be able to help you with that, but that gives you more control. And so I want users to be able to choose where they want to fall on this spectrum of convenience to control. And we have a few did methods to support that, and then we can let the protocol ecosystem evolve towards what people want to use the most. If
Mike McCue:
we all were supporting this did standard. Then, in theory, I could set up shop on any one of these apps and any one of these platforms. They might use different protocols underneath, but I could still nevertheless move from one platform to the next platform to the next platform. And let's say I set up shop on threads and you were following me from Bluesky. What that follow is, is some unique identifier to me, and if I decide to move my identity to another app, or maybe have multiple apps that I'm using with the same identity, you're still that connection is maintained between you and I, right? So, right, that's a big because right now it's mike@threads.net or mmccue@threads.net but if I leave threads, that connection's broken forever, right? Unless there's some software written to, like, somehow migrate that name over to some forwarding server or something, I guess, kind of like email forwarding, but that's kind of clunky if, instead it's a if I have a unique identifier, and I can basically use that across all these different apps, independent of the namespace, independent of the server that I'm on, and I guess even independent of the protocol, then you have a social graph that can span all. All these different services, regardless of the protocol that they're using underneath, whether they're using App proto or activity, pub, in theory, right? Yeah,
Jay Graber:
that would be the goal, like an OpenSocial graph, that any new social app can come in and build on, that users and creators can maintain their relationships as they switch services and switch apps. And then, you know, you can see a life cycle of different apps come and go, different trends in social come and go, but you could keep your username, which is your domain, and you could move across these different things without restarting your digital life over each time. Actually, there's a way that this has played out with threads. Actually, is because meta, the parent company owns both Instagram and threads. They were able to let you log into threads with your Instagram username, and they ported over the Instagram social graph and so you can follow all your Instagram friends like you're constantly being pushed into your Instagram social graph and threads in order to grow that new network. And they could do that just because they own both of those properties. This is actually quite monopolistic behavior, where it's just like, well, I own it all, so now I can have interoperability between the two things I own. But what we're trying to do is build an OpenSocial graph where, you know, a new app could come along and build an Instagram, like, photo sharing app on the App protocol, and it's not run by our company, but they can pull from the same social graph, because our relationships shouldn't be owned by one company. And like, if the company wants to grant us, you know, the option of using a new app that they developed, then they let us forward our social graph. Otherwise they disallow it, right? Like, that's the current status of things, but this is our relationships with other people in the world. Why is it all being mediated through one large company? It should be something where, if I want to try a new app and I want to continue talking to my friend, I should have the right to do that. That like right of free connection and communication is something that I think we've really lost in the world where there's just a few large companies that allow or deny interoperability, depending on how it serves their business interests
Mike McCue:
Exactly. It sort of reminds me of phone number portability. Remember how you if you signed up for, you know, singular or AT and T wireless or Verizon, you get a phone number for your for your cell phone, but if you wanted to move to a different network, you had to get rid of that phone number and get a completely new phone number, right? Which, yeah, that's pretty hard, right? Could you imagine everybody's phone number is always changing when they you know, and that's, that's the way it used to be. And I think it took a regulatory, you know, effort to actually set it up so that, no, no, you will have number portability. And so if I move from AT and T to Verizon, my phone numbers doesn't change. And so now there's not this unnatural moat that gets created. It allows for better competition and allows for users to decide which service they want to use based on where they're living, how different plans they want to sign up for. And that's sort of like that, I guess, for social media, right? If we all had these D IDs, right? Independent of any of any one social platform, I could move around. I could use multiple platforms but still have that same identifier. Yeah, totally.
Jay Graber:
Emails and phone numbers are one of the examples we use when we show people. Look, you have this portability and interoperability in other parts of your life, like people are really, you know, take for granted and like the fact that I can keep up my phone number even if I've moved between T Mobile of Ryzen, I'm just choosing a different service provider, and those companies are still able to make money without, you know, monopolizing our social graph. There's just this balance that we found between a person's right to, like, keep their identity and keep their relationships with their friends and keep their address book and contact list intact, and the company's ability to make money. And I think that balance is off in social now, like the companies that come between us and our ability to keep our identities and talk to our friends, and we need to have those rights, and then the companies can still provide services. And you know, that's, you know, something where it's like, we're providing a service, not losing our whole identity every time we go in. Another example I use is like, if you went to a store and you were shopping there, but as soon as you leave, like, they take your passport, and you know, you have to, like, go, start over and, like, reregister, go shop at a different store. Might never leave the store just like, Fine, I'll just Walmart forever because, like, they have a passport, you know, like, that's kind of the situation, right? And you want to be able to have the freedom, but just move between, like, Walmart and Target, and like, not have to worry about, like, restarting your life and like, building it out from scratch again.
Mike McCue:
When you think about Bluesky and or at proto and activity, pub, I think a lot of people obsess about the two different protocols here, and why aren't they? Why are they different? And you know, the reality is, is that protocols can be bridged. The thing that's the hard problem is really the identity, right? The fact that identity is tied to a particular type of social network is the real problem to solve, and that is a problem that no one's really tackled across protocols. You guys have really innovated in this space and using, using the what I think is a very solid foundation, which is the W 3c standard for decentralized IDs. Yeah, and that's something I think, you know, I'm excited about evangelizing that more, because I think if threads Macedon, you know, other act, if we could get this to be an activity, pub, W, 3c, initiative to kind of bring identity to the forefront and make a standard that really worked across all these different apps independent of the protocol, that would be a big deal. Do you see is that? Is that wishful thinking? Do you, I mean, you know at protocol, you obviously know at proto, you know, activity, pub, do you see a pathway, technically, to be able to make something like happen, assuming that there were willing participants?
Jay Graber:
Oh yeah, absolutely. That's part of the reason we picked the DI D standard, because it was flexible enough to allow different opinions around identity to coexist and like different implementations to coexist, and we wanted to, we've tried to build that protocol in such a way that pieces can evolve separately, so we don't have to build one big system that everything was right on the first path, you know. And so the D ID standard, I think, allows different people to push for different, you know, di D methods that they think are like the right way to do identity, and then different clients to choose which ones they're going to support. So they, you know, you might get some inconsistencies in the period where things are shaking out, where a client doesn't support all of the D ID methods that, like this other one does, but I think those are resolvable problems. And, you know, there's just kind of a like, we picked what was right for ourselves. We didn't want to go down the rabbit hole of solving identity, you know, divorced from any use case, because a lot of people try to do that and then just end up in the abstract, you know, you know, the realm of abstraction, essentially. So like, we built a real social app, and we picked an identity method that worked for us, but within a larger standard that allows other people to come in and work off it and build on it. And then, you know, there's, there's already pre existing work in this space. There are people, you know, pushing for the IDs for, like, standardization, for different things like this. And so we'd like to continue to build off that, and continue to see the momentum in this space, we'd be happy to participate. But we're kind of just a sort of go ahead and show we're really big on the show, not tell philosophy. So just showing people that, for example, D IDs can work at scale, that you can build a system that works like this at scale. There wasn't that much adoption of D I Ds a few years ago. I think there's starting to be more. We might be one of the biggest deployments of them now. So I think it's important to really just get working code out there, get working apps that people like using, and then go from there, and then try to build the technical standards on a solid foundation that can create enough consensus to work as like one ecosystem across different protocols.
Mike McCue:
Yeah. And this is all with the goal that I can be, you know, Mike McHugh, broadly, anyone can search for me, follow me, and I can post from Bluesky, Mastodon threads, Flipboard, pixel fed wherever I want to future apps and people who are following me can still see those regardless of the app that they're using, and really independent of the underlying protocol, whether it's at protocol or at proto or ActivityPub, because, you know, the thing that there is, you do have the bridging. The problem is the is the identity, the identities and the social graphs are separate and distinct for each different app. You know, even though pixel Fed is built on ActivityPub, I still have to separate. I have a separate pixel fed account with a separate pixel fed social graph that's different than my Macedon social graph and and profile. And I can see why there that could be useful in some ways, because people that you know are want to see my photos are different than people that want to see other things that I post on mastodon. But I do think, though I'm still the same person, you know, there are different dimensions of me, right? And so how we bridge all this, I think, is one of the most important problems to solve going forward for social media to work the way we envision it.
Jay Graber:
Yeah, I would love to see that. And you know, that's one of our biggest principles, is to make social like the web. And you know, the social web should be working much like the web itself, where you have your domain, you know, used to be your blog, you could, like, follow someone's blog, and now we just have more functionality than that. And so you should still go to someone's domain, and instead of just their, you know, blog sitting up there, maybe it's like an at proto landing page, and it connects you to all sorts of different things. Or, you know, maybe that proto connects you to also you could, like, if you have a D, I d integration, it connects you to, like, your master account or something, you know, and you have all the things that live under your identity, the spaces online where you've chosen to build your presence, but you can be found through something that is really in your hands. You know, built off the Domain system and links you to all the other ways that you want to engage with the digital world. So
Mike McCue:
now you are gearing up to launch, you know, I looked at the roadmap that you guys published, you know, a few months ago. It's been great to see you continually adding more. More. Tell us a little bit about where you see Bluesky going. What's up next for you in that roadmap? What are you excited about right now?
Jay Graber:
Yeah, so we've been building both a social app and that protocol, and a lot of the protocol work is, I think, starting to reach a stable 1.0 so getting to something that we're confident to call at proto 1.0 is, you know, definitely on the horizon. We're super excited about that. There's, like, a few more things that we're, you know, getting in there. You know, an OAuth method is coming out soon where you can, like, easily log in with stuff that's really exciting. We're working on polishing up our APIs and SDKs, just making it easier to build, and then making it easier to build new apps with different modalities, like people have built blogging apps on that proto other things. But, you know, the lexicon system is how you specify this it it all links together, but in a ways where we haven't, like super polishing up yet. So getting to that level of polish and stability is something that's near on the horizon, and I'm super excited about because we're starting to reach out to devs more ask people to build. And there's been an effect where people have been just pulling things out of our hands, like, they build features before we get to them. They, like, you know, build on. They build new forms of, you know, app, proto apps, before we're even ready for it. But that's been great to see. And so, like, just, you know, putting more fuel into the fire and then getting video in. So we're putting video in the Bluesky app that's coming soon. That's awesome. And, yeah, like, I would love to see different versions of not just the, you know, centralized social apps we're used to exist like Instagram and TikTok or whatever, but also new versions that we haven't thought about that might take some of these pieces and recombine them, or, you know, throw them up in virtual reality, or, like, just playing your social games. And I think this is something that you know, just thinking through, how do we enable that kind of innovation? Is a really it's a challenge that the team is really excited to tackle, because we all came at this from, you know, the source most of us are, like open source developers by background, and we just love to see people building and innovating in an open ecosystem.
Mike McCue:
One of the things that you have on your roadmap is custom feeds, you know, enabling, I guess, is it the concept of end users being able to make custom feeds, not just developers, so
Jay Graber:
They can make those through third party apps right now, right, like Sky fees and stuff, but we've watched the patterns of adoption emerge, and we'd want to make it easier to make not every type of feed, but like, You know, there's like, kind of one sort of golden path of a community driven feed that people would probably like to make in app. And like, that's the kind of thing that we'd like to make easier and just pull into the app. But then if you want to build a more complex feed, like, there's these really cool other apps out there that let you do that. And so, you know, our goal is to, like, let the whole ecosystem flourish and let people build things that are, you know, there's like, power tools that are under the hood, and you can really get into custom feeds and, like, build your own with Sky feeds. But if you just want to, like, make a little community that's, like, these are my gardening friends or something, we want you to be able to do that. And, like, if you click some Yeah. So that's like, the kind of functionality that we're seeing emerge, and then following along our product development
Mike McCue:
path got it so for things like lists and things like that, that are a little bit more of a straightforward kind of custom feed where I'm just bringing multiple accounts together into a feed,
Jay Graber:
yeah, yeah, based off, like, more simple heuristics like that. Like, if you want to build a fancy feed that, like, pulls in certain keywords and emojis or like, trends and whatever, and creates a feed out of that, then you're either going to do it on your own as a developer, or, like, build off something like Sky feeds to do it. But if you just want to create a simple feed, that's like, essentially like a subreddit, it's like, these are essentially functioning like subreddits or like discords. They're like communities. They're not private communities, but they're like public subreddits, essentially where, like, people create them and curate them, and it's little community space. And so, you know, it's right now, we're in the position where it's almost like we're the custom feed marketplace is, in some ways, like a Reddit, where you can't create a subreddit from within Reddit. You go to another app. One of
Mike McCue:
my favorite custom feeds is, I think it's called Quiet posters. It's like, you know, people who haven't posted in a long time, when they post, that's what you see in this feed. People who don't, who aren't they out there screaming all the time, just every now and then they have something super interesting to say that they think is post worthy. Yeah.
Jay Graber:
Another thing we'd like to do is actually, like, maybe group some custom feeds, so like, there's like, a cozy mode where you can just see, like, your mutuals and the people who don't post that often, and you're not tuned into the larger discourse. And if you combine a few feeds, like, now you have a sort of like a cozy little corner of the social network that's just, you know your friends and like your people you actually know in real life, or, like a few quieter followers who come out with the gym, like once a month, but don't post all the time, and like having that sort of be surfaced, not through one master algorithm, but just sort of like a cluster of different feeds, like, this is some cool stuff that could be composed. Like, I just like the fact that people can experiment with how they want to engage with social and it's under their control again. So if you just want to look at pictures, even just pictures of cats, there's feeds for that, you know. If you just want to see what your friends post who don't post that often, there's a feed for that. If you want to see what was most popular in the last 24 hours, there's the catch up feed. You know, and so you can go from like, very global, like keeping up on the global discourse, to very local and just paying attention to your little corner
Mike McCue:
of the world. I love it. I love it. And, you know, from a developer point of view, are there recommendations you have for folks when they want to get started to experiment with all of this? I know everything is an open source. First of all, that you guys have done, I assume that's something that's a kind of a first principle for what you guys are doing. Everything you build is just going to be open sourced. Yeah,
Jay Graber:
we want people to be able to build off what we're building. We want the ecosystem to evolve freely. And, you know, we want, like, part of this is also like, building a foundation developers can trust. Like, Twitter used to have open APIs, and like, you know, yank the foundation out, right? And so, like, a lot of this is about resilience, like making sure people are able to self host, making sure that, you know, their code is open source, so that, like, this is an ecosystem that's guaranteed open. The APIs are locked open, and, like, you can build on it. And if you, you know, don't like what the client is doing, you can build your own client. Like these are just like principles that we're staying true to. Like social should really kind of make it more of a Digital Commons. Again, you know, open source code is a way to do that. And then, yeah, we're running a service on top of that and building a social app that a lot of people are using, and it's still, still a lot of people end up using the thing that, you know, like the early companies build, so definitely still possible to, like, make a living within it. But I think having an open ecosystem is like the foundation of that trust for the dev ecosystem.
Mike McCue:
Well, you know, before we go, you know, one of the things that I think I'm really interested to hear from you as a CEO and a founder and somebody who's been passionate about this vision for many years beyond just what the work you've done at Bluesky. Have you had kind of a moment where you were like, oh my god, this is actually starting to come together. I assume it's happened probably multiple times, but yeah, what's been like most notable for you there?
Jay Graber:
Well, early on, like we talked about some of these milestones, we're like, wow, I would just, like, love it when, like, drill starts posting on Bluesky, right? Like, we were very online people. We had, like, a book of drill tweets and, like, laughing about it at two weeks, and then drill started posting on Bluesky, and it was, like, a great, hilarious moment, you know, just kind of like an icon of deep Twitter lore, like, outside Twitter, I guess I don't know him, but it's just so funny, so and he's been there, like, with his, like, very like, snarky posts and, like, you know, big moments, like our launch moment, he's like, I can't believe that they, they let the commenters in or something, you know. And then the other is, we talked we like, I just love to see these ideas play out in reality when we've talked about them in theory for years. Like, you know, in theory, you could have a social post that, you know, people are like, Oh, how do we verify politicians and stuff? We're like, Yeah, but like, you already have, like, official government websites under.gov What if you just let them use their domain name as their handle? Then you know, it's that politician. And now we have Senator Ron Wyden posting regularly on Bluesky under wyden.senate.gov and you just know that this is clearly Ron White. And we didn't have to, like, run a verification apparatus to do this. They self verified by setting it up. It gives, like, only the Senate controls that domain name. And like, that's an open way to do it doesn't verify everyone in the world, but it verifies those with well known domain names, which are often sources that you really want to be able to trust. And for example, the New York Times is posting under NY times.com, you know, and like, that's what you want to see. You want to see the established reputation that web 2.0 is already built up. All this credibility around, like, domains that have become really powerful, you know, presences online, like, let's just build off that in the social web.
Mike McCue:
I think that's absolutely huge. You know, the whole all the verification challenges that exist. And you know now you can just pay to be verified on on x, and people actually don't want to be verified because it, you know, looks like they're paying to be verified. It's just a weird kind of confluence of expectations there that just don't work. The fact that Ron Wyden is posting from his Senate address is absolutely the way things should work. I think that is that's a huge milestone. And I love that particular example, because it gets back to, how do we where are we going with all of this? You know, what are some of the big things we've learned? And you know, what you guys have done around identity? I just think is so incredibly important, and I hope we can, we can build around that to create this really, truly decentralized, truly open social media network that is just a part of the internet and that developers can build on in these ways that you've envisioned. I it's awesome. I
Jay Graber:
think it'd be better for users, developers, online creators, people making living online, and enable a lot more companies to emerge. Like social is really starting to stagnate, because we're in this trap where users are locked in and developers are locked out, and we need to open that up again, and then we'll see a lot more. We'll see new companies. You know, more innovation and, like, a new era of, I think, fun experimentation and also solving some of the social problems that have emerged through innovation being bottlenecked on a few big companies.
Mike McCue:
That's so well said, Yeah, users are locked in and developers are locked out. That is so well said, Jay, yeah, I think the the work that you've done here to create that excitement around the possibilities is is starting to pay off. It's really exciting to see it come together. And I think there's going to be a whole new chapter here as we start to like when we look at what's happening with threads, lot of people are now on a platform where they can choose to federate that. That's the that's an introduction to the concept of, you know, being in a decentralized world with the benefits of that. So that's a great, you know, kind of overall thing that's been happening on the side. And I think that the work that you've done, the work that Eugene's done, there is an opportunity to bring this together and take us to a whole new level that, I think, is just, you know, it's already, you're already getting glimpses of it happening on Bluesky right now. And I'm sure that's why Mike is so excited about being on your board, because he's seeing it actually really happen here.
Jay Graber:
Our goal is to try to show you the world that's possible. And we can't, you know, realize it all at once, but we can show you a window into it. Like, you know, the domain name verification through of a senator. And like, the ability to pick your own custom feed. Like, have a feed that's just cats, or just the last 24 hours, or just your friends. Like, this is the ability to control what shows up in your timeline. And like, it's crazy that we don't have that anywhere else. Like, your attention is the most valuable thing you have. Actually, it's like your time and your attention, and we just give it to companies to, you know, do whatever they want with for the purposes of ad revenue, as opposed to being able to have some agency over it. And I think in general, people's dissatisfaction with social has resulted right now in a trend of decreasing use across public social overall. And like people are treating into WhatsApp groups and discords and things where they have more control, but if we give people the control over how they spend their attention on public social platforms, I think that people love having the ability to connect with anyone, even if sometimes they just want to talk to their friends. And so like creating that agency and fluidity of for engaging with public social is the way that we will put some life back into this ecosystem again. Otherwise, I think it's just gotten pretty toxic, and a lot of people are tired
Mike McCue:
of it. Well, you know, that is, that is such an awesome vision. You are fighting the good fight. And I want to just say thank you. On behalf of a lot of people, lot of developers, a lot of a lot of people who've helped to, you know, build, you know, around this space. You're an inspiration. What the work you and your team have been doing, you guys are, you know, really staying true to a clear vision and trying to make the best decisions possible for enabling this future to happen, and you've done amazing work, and I can't wait to see where this all goes in the near and further future. So thank you for everything, Jay, and thanks for being on Dot Social. Yeah,
Jay Graber:
Thank you so happy to talk with you and hear your vision of this as well, because it takes more than one person or one company to build, so I'm glad you're pushing on things too.
Well, thanks so much for listening! You can find Jay at @jay.bsky.team
You can follow Mike at @mmccue.bsky.social and @mike@flipboard.social and @mike@flipboard.com
Big thank you to our editors, Rosana Caban and Anh Le.
To learn more about what Flipboard's doing in the fediverse, sign up via the link in this show’s notes.
Until next time, see you in the fediverse!