All right. Just as we were standing back there, my watch pinged me and I looked down and it said, “Your body feels a little more stressful than usual,” so we will get into this, too. Um, I thought to start out, I wanted to actually talk about the very first time I met Jay and how this whole conversation is something of coming full circle as we get into the conversation about Bluesky, which is that, as some people may know, in 2018 I wrote a paper which was released in 2019 called “Protocols Not Platforms,” which was a discussion about, “Hey, we’re living in this world right now where we have a bunch of big tech companies that seem to have an awful lot of control, and there are a lot of concerns about that control and what the people who control it and run it might do. And maybe there is a better way.

Maybe there was a way to use protocols to actually do something valuable. And when I released that paper, I heard from a mutual friend of Jay’s who said, “There’s somebody out there who’s thinking about these things just like you are, and you should talk to them.” And so we ended up connecting over email, and we went and had lunch, and Jay had read my paper and started talking about it. And I sat there and I said, “This person understands my paper better than I do.” This person has the vision to understand what it means to build a system based on protocols, instead of focusing on a centralized platform. Since then, a whole bunch of things have happened.

And now Blue Sky exists, and it has, you know, 32 million users at this point, and all sorts of other stuff has been happening. And so, let’s get into the conversation.

I did want to ask, for as much as we can see it, how many people in the audience have an account on Blue Sky? If we can see it. I see a lot of hands. I hear some screaming. Excellent! That’s what we always want to hear.

But to start out, to explain to people in the audience who might not be as familiar with it, what is Blue Sky? How is it different? How is it not just another social media property? Yeah. Thanks, Mike. Blue Sky is an open social network, so our goal has been to make social media more like the web. And that’s the goal of, you know, building a protocol instead of a platform.

And, you know, the way that this plays out is, for example, we don’t down rank links because we see ourselves as a portal to the open web, sort of a gateway where you can access the social web. And our goal is focusing on user choice and developer freedom.

So users have a choice in what they see and interact with, and developers can freely build without us getting in the way. And the key to this openness is really the protocol.

Which is the foundation that Blue Sky is built upon. This is an open protocol for building social networks, and this is what guarantees openness over the long run. And so right now, as I mentioned, there are over 30 million users.

What? What kind of users are there now? What kind of communities are using Blue Sky? A lot of people who want a better social experience. So we started off early on with a lot of developers, artists, journalists, scientists, people who are interested in trying out new social modalities. And recently we’ve had a lot of new communities come on.

So the sports community is growing. For example, since we added short-form videos, there’s a lot of overlapping communities on Blue Sky because you can really customize it to what you want it to be. If you want to just see, you know, science news and cat pictures, you can make that your corner of the network. And so there’s a lot of communities that overlap in interesting ways.

So, you know, the Blue Sky team is a fairly small team. How big is the team right now? About 21 people. 21 people, 30 million plus users. Obviously, it’s been growing a lot. Yes. Cheers. Yes, yes. So as a small team, as a rapidly growing social Network, but also a protocol.

And you’re doing both of those things at once. What have been sort of the biggest challenges that you’re facing as a company? There have been a lot of challenges. I would say actually, that tension of building a usable social app that preserves what people like about social and is as usable and convenient as people expect. While scaling up on an open protocol is one of the core tensions and core challenges.

So we’ve grown from 0 to 32 million users in two years, building any sort of social network is hard to do, because every time you grow by an order of magnitude, you have to rethink how you’re doing things, both organizationally and technically. And so the technical challenges have been huge. But at every stage, we’ve rethought how we’re doing things, and we’ve made sure that we have grown in such a way that we preserve openness while scaling up.

So that’s involved some trade-offs, like early on in 2023, when we first announced that we had an app in beta, there was a lot of excitement. A lot of people wanted to come on, but at that time, our choices were to scale up in a more traditional way and, you know, go scale things out on AWS, which would be very expensive and would sort of tie us into a more centralized application. Or we could go ahead and execute on the original plan, which was to build out the Federation behind the scenes so that our internal services were federated and then open it up to others.

So we went ahead and did that. That took a few extra months. Then, when we opened up in February 2024, we had a federated internal network and we opened it up to federation with external parties. Now you can self-host your data and interact with the rest of the network. But that was a trade-off. Doing that while building an app where you’re using, constantly trying to iterate on user feedback and grow in response to people’s questions, comments, preferences means that it’s a bit like building the wings on an airplane once it’s already flying. A little later in the conversation, I want to dig in a little bit more on the decentralization, the nature of federation, why that’s important, and how that all works.

I did want to address one question that comes up a lot around Blue Sky because it originated as an idea at Twitter. With Jack Dorsey announcing it and endorsing it, and um, interviewing you to sort of take the lead, there are a lot of questions about what the relationship is between Blue Sky and Twitter and Jack Dorsey as well. Do you want to address that? Yeah, Blue Sky is fully independent. It always has been, but it started off as a project internal to Twitter. In 2019, I was a decentralized social network researcher. I was building my own social network, and I saw on Twitter, Jack Dorsey posting that Twitter is funding a protocol that Twitter will someday run on.

And I’d been building decentralized social protocols, doing a lot of research in this area. So I got really excited and I DM’d the Twitter account, and then they put me in a chat room with some other folks, I think, including yourself. And for about a year and a half, that was the state of the Blue Sky project. Just a chat room. And in mid-2021, Twitter finally did some interviewing, chose a lead, chose me to lead that project. And then, because Twitter had moved pretty slowly getting up to that point, I decided it’d be best if we were an independent organization, and then we were a contractor with Twitter.

So we were contracting to build the app protocol, which Twitter would someday run on. But things changed very quickly, and it’s good that we had that independence. That’s an understatement. Yes. Yeah. So. So that independence allowed us to go ahead and build what we originally framed as a reference client. So Blue Sky was an example of us dogfooding the protocol we built to test it out, make sure that we could actually build a scalable social app. But then it became more and more of a real product because we had all this demand. As soon as we put up a waitlist saying we’re building an app.

We had a million user signups within a few weeks. And so we knew then that, all right, we should launch this thing and build it for scale. But it’s entirely independent of Twitter. It has no connection. Yeah. Entirely independent. Early on, Twitter had a board seat, but then that ended when the contracts were cut. And since then, we’ve raised money on our own and continued to build an open social network, carrying through that original vision of being a protocol that builds a microblogging app.

And so, as you said, there was that chat room that I was in and I was sort of watching it. And, you know, I actually started to lose hope a little bit because it went on for a really long time. Jack announced it and I think it was December of 2019. And then obviously there was a pandemic. Some of you might remember that happened in between, which I understood slowed stuff down, but they didn’t seem to really get serious about interviewing until early 2021.

What made you stick with it over that period of time? Why? Why was this project so important to you? Yeah, I’ve been interested in this problem for years. I started off my career as a digital policy activist, and we were organizing our campaigns on Facebook, and I just thought that it would be great to have an alternative to Facebook.

At that time, there was diaspora. It was a small, early predecessor to networks like Mastodon, but it was an open network and I was really inspired by that. So when I became a programmer around 2018, I circled back and started building decentralized social protocols. I think that society starts to reflect the structure of its dominant form of communication. And so you need communications infrastructure that is democratic and gives people a choice, lets people intervene and change it to suit their own needs and preferences.

Building open networks, I think, is very critical to the problems we face in the world. If you have an open network, then you have a form of communication that people can modify and make their own. So what are the questions that always come up when thinking about any social media network? Are questions around content moderation, what’s sometimes referred to as trust and safety?

I know that The Verge, at one point, referred to social media saying content moderation is the product, and there have certainly been questions and challenges that Blue Sky has faced around trust and safety and content moderation. But how does Blue Sky look at those challenges? How does it think about questions around trust and safety and content moderation?

Yeah, moderation is governance. It’s choosing how you want your digital space to be governed and the rules you want people to play by. And so we have a multi-layered approach to moderation.

So, within the Blue Sky app, we have the Blue Sky moderation service. And on top of that, you can stack other moderation services. So if you’re a developer or a community builder, you can create an independent moderation service. We call it a labeler, and then you can install those within the app. And so these can do things as specific as labeling AI generated images.

So, you know, if you’re looking at something by a human artist or an AI-generated piece of art, or it can be something like a politics labeler. So, I use one of these when I want to just scroll through and see what’s going on in the network. But I don’t want to get distracted by every political post. So it’s hiding and labeling everything.

And then when I want to see what’s going on, I turn this off so I can just see everything else in the network. So this is an example of, within the app, this stackable approach to moderation is what we call it. And then, outside of the Blue Sky app, it’s a composable approach.

So you can pick and choose what pieces you want to pull into your own app if you’re building your own application on the network. So that means if you’re building a new app, you can say, “I want to import Blue Sky’s moderation policies and build on top of it.” Or maybe I just want the spam policies, and then I’ll roll my own for the rest of it. The service that we use to moderate is called Ozone, the protective layer for the atmosphere, is the app protocol.

This is something that’s open source. Other people can run it and use it to build their own services, and people are doing that. So, a lot of these third-party moderation services are running through ozone, and then they get their own content moderators, their volunteers, or their people who want to moderate in this specific field, so they can run it themselves.

One of the things that often comes up when I’m talking about Blue Sky or talking to people about it, and the sort of open, composable moderation system, is people worry that it enables bad people. The sort of summary is to do stuff. And so, the example that comes up a lot is Nazis. Nazis are a problem. What happens if Nazis come along and create Nazi Sky, and are using the same protocol and the same system? How does Blue Sky deal with that?

Well, a protocol is a language for computers to connect, and it allows you to speak, but it doesn’t force you to speak. You’ve seen this play out in other open source ecosystems, for example, in the Mastodon ecosystem, which is built on this open protocol called ActivityPub. Gab came along and built an instance that operated by very different norms and rules than the rest of the Mastodon ecosystem did. As a result, many Mastodon services decided to not speak to that instance.

And so essentially, they’re still operating on the same protocol, but they’re not necessarily connected. This can play out in the app protocol ecosystem as well. It’s open source code. Anyone can pick it up and run it. But it’s really an open marketplace approach where services can choose who they’re going to connect to and who’s going to be sort of within the parameters of the norms that they’re setting for their community. You know, running an app in the App Store, you have to comply with Apple’s policies.

A lot of our policies flow downstream from that. If you’re building your own application, you’ll contend with those as well. That kind of sets some parameters around how moderation can work in the broader social ecosystem. One of the other questions that comes up in this space is whenever you have a large enough community of folks, there are always concerns about toxicity and the nature of the conversation and where that goes. You know, there’s always some element of that when anything gets large enough. How do you think about dealing with toxicity on a social media platform? Yeah, our goal is to build a choose your own adventure experience at Bluesky. So if you want to get in there and just talk to your close friends and have a more, you know, safe, closed experience, you can customize that. You can do things like only use feeds that show You decide what you want to see.

Install moderation services that filter what you interact with, and then even do things like close off your replies to just people who you’re following or who you nominate to speak to you in the post, and do things like get out of quote posts. So if you’re dragged into a conversation you don’t want to be in, you can get out of that. These are all things that we’ve built to give people more control over how they’re interacting online, and create a less toxic space because you can choose to define whether you’re going to go into the social app with a more PvP (player versus player) mentality, where you’re going to argue and dunk, or you can go in with a, you know, I want a close circle, close friends experience and just set my default interactions going forward to everyone who I follow.

And those are the only people I’m going to talk to. We think it really should be up to the user how they want to interact, and they can choose how they want their space to be. Do you think that, I mean, just how does that deal with the toxicity issue? How do you—how does the sort of user choice aspect of it get at the concerns about toxicity? It means that if you want to have an anti-toxic experience, that you sort of put yourself in a space where you know that’s your experience as much as possible. And then, of course, the Blue Sky moderation service also sets the parameters for discourse on the app. So we have, you know, anti-harassment policies, things like this.

But then, you know, there are things that go beyond harassment, which people see as toxic discourse, like getting into quote posts like dunk sessions between users. This is a really common dynamic on microblogging apps. Some people actually really like it, so it’s really hard to say. It’s really up to the user whether they want to engage in that or not. Some people find that it’s not for them, and they want to be in a space where they’re just reading more informative content, maybe having a more calm experience, and that’s something that’s up to them to create by customizing their settings.

One of the things, and I’ve been talking about this a little bit, and I think it’s sort of relevant to this discussion on the toxicity, is that I think for the past, I’ll say decade or decade and a half, however long you want to look at it, users have been, you know, put in this mindset of what I’ve been referring to as learned helplessness when it comes to social media. The platforms are, as Cory Doctorow famously said in defying this sort of natural state of things where, you know, as a network gets big enough, it bends. The network itself has so much control that they start to make things worse and worse for the users. And users have no real way to deal with it other than to complain to billionaires or to the government. Right now, both of those are the same thing, apparently. I’m not in a great situation.

So one of the things that you’ve said, and I think you Guys now have some ads making this claim: is that Blue Sky billionaire proof?

I know that a lot of people have been concerned that, oh, well, Blue Sky is going to follow this same path. It’s going to defy its founding principles and billionaires are going to come in and they’re going to take it over and do all the same bad stuff that they’ve maybe done with some other platforms that will remain nameless. Um, so how can you say that Blue Sky is billionaire proof? What does that even mean? Yeah. So the core thing here is where we’ve tried to rethink social from first principles and build an open protocol.

This is what guarantees that users will always have the right to leave. That means that all the code is open source. So right now, if you want to clone Blue Sky and start over with a new version of it, you can do that. You can run your own version of it or even within it. You can run your own feed. You can self-host just like a piece of it. That means that depending on what aspect you’re dissatisfied with, you can find alternatives. Or if they’re not there, you can build them. So if you don’t like the main timeline, let’s say that we started shoving in ads in between every other post and it got really annoying. Don’t do that. But that’s an example of that certification path that we’ve seen play out.

Right. But if we did that, you could choose a different timeline. Just even within the app, you could uninstall the feed that we’re putting all these ads in and install a different one. And so the fact that users have that choice means that we’re incentivized to keep serving users, because if we change things to a degree that people didn’t like and didn’t tolerate, they could switch off to another alternative. And that’s true at different layers. So if you didn’t like the app anymore, because let’s say we made it a very obnoxious color and put flashing ads across the top of the app, then you could use a different app and access the same network.

You would keep your identity, you would keep your data and just move to a different application that doesn’t do that. And so that actually aligns our incentives, because we want to keep things in a place where users want to continue using the app, they want to continue using the timelines and the features that we’ve built. And that’s one of the core principles of alignment. Now, another part of that is we’ve structured ourselves as a public benefit corporation.

So that means in addition to being a company that’s making money, we have a mission. And so that’s to build an open protocol for public conversation. And that means that whenever we’re staying focused on that mission, the board and the team are all aligned behind that.

But ultimately, in the long run, we’ve seen companies, social companies change; services come and go. But the open protocol is what ensures that there’s a layer that sticks around, that lets services change. Users change their preferences and still have these core components of social be available to them, which are their data and their identity.

How does that address the billionaire issue? That means if a billionaire came in and, you know, bought Blue Sky or took it over, or if I decided tomorrow to change things in a way that people really didn’t like, then they could fork off and go on to another application. There are already applications in the network that give you another way to view the network.

Or you could build a new one as well. And so, that openness guarantees that there’s always the ability to move to a new alternative, still using the same, you know, network; all the data and all the stuff that you’ve already posted onto Blue Sky will stay around even when you move to a new thing.

Do you think that it also, um, almost acts as a poison pill to avoid the billionaires coming after you in the first place? I think it definitely decreases the incentive to do so, because if you know that if Blue Sky, the company, were to radically change things in a very heavy-handed way, then there would be this smooth migration path.

And so that means that there’s much less tolerance in the ecosystem for a very heavy-handed change. And like subtle changes can happen, and we can respond to user feedback like we’ve done.

The way replies sort or like show up in the following feed. And users have a lot of opinions about that, but as long as it’s within a tolerable range, then we iterate on that and improve. But if it’s something where overnight you change the application in a way that people didn’t like, and to a very large degree, then there would probably be a mass migration to another application. Yeah.

I mean, I sort of think of it; you can compare it to the other experience that many of us felt, which was Elon obviously buying Twitter. And then everybody had to leave and, you know, reform somewhere. But the idea is that with Blue Sky and the protocol, the reforming somewhere else is much easier. Yeah.

So protocols allow this more seamless fragmentation and coming back together, so you can move off to a different application. And then maybe the Blue Sky app says, okay, we’re backing off that change. And then people could come back, but you don’t lose your identity in that migration. And so right now the barrier to leaving is very high because you essentially have to start over from scratch, rebuild your entire follower graph, rebuild your entire post history, and start over with a new identity.

So that’s essentially migrating to a whole new place and having to start over again. But if you can start over with the core pieces of what you’ve already built on a social network, then the social part The network stays intact.

What you’re building is essentially just a service that gives you access to that network. You’ve talked a little bit about how Blue Sky and the protocol effectively enable user choice. Can you go into a little bit more detail? I think this is part of it. You know, a lot of people come into Blue Sky and the first thing they say is, “It looks like, feels like what everyone is used to, as sort of a microblogging service.” They don’t even necessarily see or recognize some of the other elements that you’ve enabled and the affordances that it allows.

So do you want to describe some of what user choice means on Blue Sky? Sure. Yeah. So concretely, some of the ways that this shows up are custom feeds, composable moderation, self-hosting and alternative apps. Custom feeds mean that when you sign up for Blue Sky, you get two default feeds: following and discover, one’s algorithmic, and one’s chronological. Then you have a marketplace of over 50,000 feeds that you can choose from. That means you can install a feed that’s just science posts or just cat pictures if you want to see that. There are many you can choose from. If you don’t see what you like, you can create one. An example of composable moderation is, as I mentioned, Earlier, you have a US politics labeler or an AI art labeler.

And that means, as a user, once you install this custom moderation service, you can go in and just warn or label on all images of AI art, or you can hide them entirely if you don’t want to see them. These are built by third parties, and it’s something where if you don’t see what you want, you can build a new one. And then self-hosting means that if you want to make sure that no matter what happens, you keep control over your data and you keep a username that you want. You can run your own personal data server and self-host your data.

So people are already doing this in the network, and they interact with the rest of the network seamlessly. You can also bring your own domain name. So, for example, you can come on with a j.com and have that be your username. And that’s also your website and links over to your website. That means if you use a different service, you still have the same user handle and it’s one you control. And then the last one I mentioned was alternative clients. And so this means if you want to try a different way to access the same network, you can install an app like Flashes, for example, and that’s like a photo sharing app on the Blue Sky network that lets you see things laid out, all your photos laid out in a grid format. And if you want to move over to that and use the Flashes app as your main art portfolio, you can do that and still talk to your Blue Sky friends.

So those are all examples of user control and customization that you can do. But I’ve also seen you’re a huge Blue Sky user. Mike, what are your favorite customizations that you use? Oh gosh, I’m constantly experimenting with stuff and sort of playing around. And you know, one of the custom feeds that I use, that I also think is like a really good example of how it’s different. A lot of the custom feeds are just sort of, you know, like collections of interesting users and sort of, you know, so, you know, on a particular topic.

So like tech journalists, right, is one that I’ll use to sort of keep up on what’s going on with tech journalists, but you can do more interesting things with it. So one of the ones that I really like is a custom feed called Quiet Posters, which is one of these things, which is, you know, if nobody has time to read everything. I don’t know, maybe some people have time to read everything of all the people that they follow. I guess if you follow a small enough audience, but you know, there’s always important content that I feel like I miss. If I’m just doing that. And Quiet Posters somehow sort of recognizes, you know, the users that I follow who don’t post very often and I can just go in and then see a listing of things that were recently posted by them. And I felt like I discovered so much interesting stuff that way that I never would have seen otherwise, because it would have just, you know, been lost in the stream or atmosphere.

I don’t know how you describe it, like I would have missed it. And so I think there are some really interesting things, and I’m sort of excited to see people beginning to experiment. And because anyone can build these things, you know, it’s exciting to see more stuff come out of it. Yeah, definitely.

There are a few that I like, like ketchup, which is the most popular post from the last 24 hours. So if I just want to catch up on what happened, I can check that. And I really like the moss feed, which is just pictures of moss in mossy forests, so it’s very relaxing. I like to go there after too much browsing of other timelines. I should have done that right before coming on stage. Just calm. Calm.

Um, so you mentioned flashes being this new app. This is kind of an interesting development that they’re coming along. They sort of wanted to build a kind of Instagram TikTok thing. The fact that they can do that and it’s not you guys, right? It’s entirely independent. Yeah, it’s an independent developer. We heard about it when they said, “Hey, I’m building this app called flashes.” And the cool thing is you can sign in to flashes with your Blue Sky account, and then all the pictures that you’ve posted to your Blue Sky account show up in this nice grid format.

Or you can post new pictures through flashes, and they have some custom filters and stuff that we don’t have on photos. And then you can upload things with custom filters. But also, if you want to just create a new account, keep those identities separate. You can do that too. So you can either have a new account and say, “All right, this is my art account, and I’m going to keep all my art over here and talk to people over on Blue Sky.” Or you can unify them and say, “You know what, I want it to be a consolidated username across all these sites.” You know, I’m J.B. SkyTeam on Blue Sky, and I want to be that on flashes and just have a different format to post my pictures, so you can do both. But an interesting thing about this is it’s basically treating the open data network as a shared network that anyone else can build on as well.

And when Facebook launched Threads, or I guess Meta launched Threads, they let you sign in with Instagram; or they actually, I think, made you—they forced you to sign in. They had you create an Instagram account to log into Threads. And those two apps were connected. But the only reason that they did that is they own both of those apps. And so they’re creating one unified ecosystem among the things that they own, so that you’re all on a Meta property, no matter which one of those apps that you use. An open network lets you do that.

Except it’s not one party. It’s not just one company. So anyone can come along completely permissionlessly build a new app and have it connect with your Blue Sky username. And we don’t even know about it necessarily. It’s something where you can just create an app, launch it, and make it a specialized app, maybe just for photographers or something like that. And you don’t have to rebuild the social graph from scratch.

So when I was building a social network back in 2019, I was building an alternative to Facebook events, essentially, and it was really hard to get the initial network density to get enough people on this app. And that’s the challenge most social network builders face. But if you have this shared network you can tap into, you can freely build, and all of a sudden you’re tapping into a user base of 32 million users, all the things they’ve ever posted, all the connections they’ve already made, and you can build apps then that don’t require getting from zero to a few thousand users to even get over the barrier of relevance.

I think that that point is actually really important, and it’s worth thinking about because historically or for the last, however many years, decades, you know, if somebody was building any kind of social-related app because you had that challenge of bringing in all these people, you sort of had to build for a mass audience. You had to build to the largest extent possible and try and come up with ways to actually bring in that audience with something like App Protocol, where you have an audience that you can Just sort of bootstrap on.

It sort of feels like that is like the community layer becomes infrastructure on which other people can build and do interesting things. And it doesn’t have to be for a mass audience, right? Some of them obviously will be. So just as one example of this, I mentioned this to you right before, and Jay had not heard of this. So I’m assuming a lot of people have not heard of this. I just this weekend saw somebody had created another uh, app on that protocol. Not even bootstrap to Blue Sky, I think, but directly on that protocol with their own app view. Um, not called flashes, but flushes or flushing. Im, I think is what it was. And the way it was described was, um, the designer of it put an NFC sticker on their bathroom door, and as they walked into their bathroom, they could tap it with their phone and it would post to this brand new social network all about, “Am I in the bathroom or not?” You can really build anything on an open network. And people are using it.

If you go look, it’s I its flushing Im. If you look, there are people there now posting whether or not they’re going to the bathroom. Now this seems very silly, but it’s like the kind of thing that you can do right without this. You know, nobody would build that just on their own and have to attract users, but because it’s built on that protocol, because you already have this user.

Based on the community connections, people can build those kinds of things. Yeah, I mean, we’ve seen lots of experiments pop up. One that I was trying out recently was an app that assigned you an animal based on your posting style. It would take in, you know, all the kinds of posts that you’d made and the interactions you’d had and say, “Oh, it looks like you’re a performative peacock based on your posting history.” So, so fun.

Apps like this can kind of come and go, and people can pick them up and play with them. And if you’re a developer, it’s an open API. That’s what an open protocol is essentially. So you can just come in and build something new and try it out and see if people like it.

One of the questions, and I know this has certainly come up a lot with Blue Sky, you’re talking about user choice and all the different apps and different things that people can build. There’s a lot of excitement about AI tools and LLMs, but also concern, and in some cases hatred. Um, it’s been an issue that a lot of people are discussing, particularly because we’ve seen some other nameless apps kind of feeding all of their content directly into their own LLM to do something with. How is Blue Sky thinking about that question along the lines of user choice and, you know, users having control over their own experience?

Yeah, we really believe in user choice.

And so we’ve been working with some partners on developing a framework for user consent for how they want their data to be used for generative AI. So for example, this is something where you could specify how you want your data to be used. And it could be something similar to how websites specify whether they want to be scraped by search engines or not.

And so, you know, search engines can still scrape websites whether or not you have this because websites are open on the public internet. But in general, this robots.txt file gets respected by a lot of search engines. And so you need something to be widely adopted and to have users and companies and regulators go with this framework. But I think it’s something that can work here. If you have a framework where you can get user consent specified at the account level, or maybe the post level, and then have companies respect that. So it’s a proposal that we have on GitHub. We’ve been working on it with other people in the space concerned about how AI is affecting how we view our data, but I think this is a promising direction to take.

Yeah, it’s interesting because, like the robots.txt standard, which is for searching, which a lot of people are familiar with, like there’s no legal basis for that. And yet it is widely respected. And anyone who tends not to respect the robots.txt, they’re, you know, ostracized to some extent. There’s some sort of, you know, social mechanism for dealing with bad actors in this space.

Do you think that this sort of approach creates a similar kind of thing? But for, you know, and I know that people have started to use things like robots.txt or similar tools to prevent scraping for AI. Do you think that this could work at the level of an individual social media user account? Yeah, you need social consensus for this to work, but it’s something that we’ve seen play out before. I think it’s an emerging conversation right now that needs more clarity and transparency around how data is used, and our existing frameworks don’t fully cover it. The app protocol has all data being public, but we really want to respect how users want their data to be used. So I think having a consensus framework for how user data specifies its intent can be very important in the long run.

So we talked a little bit about the billionaire issue and the acidification concerns. Another question that certainly comes up a lot is that Blue Sky is a relatively new social media network that doesn’t really have a detailed business model and things along those lines. There is always concern that as soon as money becomes a part of it, and obviously you’re a company, you need to be sustainable in some form, that inevitably leads to certification. So how do you view the issue of monetization and the concerns that some people have that it’s going to lead to, you know, either closing down some of the openness or, doing things that you think justify it for the company, but not necessarily for the users.

Yeah. Our principles are keeping our business model aligned with users in the long run. And so the first thing we’re exploring is subscriptions. We mentioned this last year, but there’s been a lot of growth recently. Stuff has happened. Yeah, we’ve been keeping up with scaling. But subscriptions are, I think, the first step in this direction where we are providing value and then having users be directly the customers.

Another thing that we’re exploring is developer services. So this keeps our incentives aligned with developers in the long run as the dev ecosystem grows. You know, it’s an open network, but we’re providing a lot of the services that people will use and build on. And then another thing is the marketplace approach. So as people are building this ecosystem, there’s third-party feeds, there’s moderation services. You know, we want to look at how we can offer value and then maybe take a cut of that if people are making transactions.

But this is something that we think, over the long run, is just about keeping our values aligned with the ecosystems. We’ve said that we believe money follows value. First of all, you have to prove that this model is valuable, that people want to post here, that they want to build here. And then from there, you can go into what other business models can you pursue that start to monetize off all.

The transactions and sorts of forms of value that are flowing through the network? Do you think it’s possible to keep those things aligned, like there are always pressures and challenges along those lines? How do you make sure that you keep the values aligned to the business model and the users so that they are aligned and not in conflict with each other? Yeah. I mean, as I mentioned earlier, the open protocol is one of the deepest guarantees here, because if we stray too far from our principles, then users could decide that it’s not worth the trade-off for them anymore.

They might use a different service that comes along and offers something different. And so we’ll be watching the ecosystem, seeing how it evolves. If we do things in a way that certifies too much, then users have the option to leave. But I think in the history of open protocols, there’s been a lot of big companies built. The web itself consists of open protocols. You have email, you know, search engines, large companies built and new things can come along and disrupt that. So, perplexity and new ways of doing search have come along in recent years with new technology. And in open source ecosystems, you have companies like WordPress and Red Hat that have been built that are essentially providing services to open source ecosystems.

Users can spin up their own website, and if you’re really technical, you can host it yourself. Do everything yourself. But if you want an easy way to run a blog, you tend to use a blog service. You use a website builder or something else. And so, you know, even on Blue Sky right now, we sell you the ability to set your custom domain. You can bring your own domain if you know how to set that up and purchase one through any other domain service. But if you buy it through Blue Sky, then you’re getting something set up really easily and we’re providing that service.

One of the other things that always comes up is this idea that, you know, it is still impossible to really get to critical mass for a new social media network. I mean, if you go back, you know, there was a period of time not that long ago where people were saying there would never be another, you know, that the social media wars were over. You know, Meta had won. You also had like, you know, sort of Twitter hanging out there. Um, and so there are always like there have been other sort of upstart social media networks that have come along over the years. But what do you think is necessary to actually take this to the same kind of scale and reach of some of these other networks? I mean, in the history of technology, there’s been A lot of dominant companies that have come and gone, and I think there’s a real opening in the market right now where people are looking for better alternatives.

They’re looking for forms of social that don’t feel as toxic, that give them more control, that let them build ideas that they’ve had for a while that they haven’t been able to build. And so I think that we’re providing a new paradigm that gives people the ability to do that, and we’ll see how it plays out.

But I think there’s been a lot of promising uptake of all the things that we’ve put out there. And, you know, the history of technology is long and filled with many changes and surprises. Yeah, I always remember there was an article in, I think it’s 2007 declaring the social media wars over, that there would never be another successful social media company because Myspace had one and no one could ever compete with them.

It does mention Facebook in that article. I pull up this article way too often just to remind myself. It does mention Facebook but says they have no chance against Myspace. Things change. Uh, it’s. Yeah, I think things are changing. And yes.

I think that people trying out Bluesky have seen that. You know, here’s a social space where if they don’t like it, they can customize it. And that customization is getting more and more elaborate over time as we hone in on what we can offer.

And other developers come in and build it, because it’s not just bottlenecked on us to provide the innovation that drives this forward. The open ecosystem means that if somebody else has an idea, they can come in and build it, launch it tomorrow, and we might hear about it like after it’s launched. But that means that people can launch things for their communities, build things that work for them, and I think people are best poised to know what’s going to serve their needs more than a company that is trying to serve, sort of trying to be everything for everyone.

Like, we can try our best to be everything for everyone, but then also creating the open interfaces where people can build things that work for them is actually the best way to provide innovation. And that’s the kind of competition that drives things forward. Because, you know, if we’re not doing it, somebody else will come along and do it. And we’re already doing things that the other social companies haven’t done. And now there’s more pickup of the kinds of things that we’ve done, like custom feeds or starter packs, these new ways that, you know, allow people to connect and find each other on social that previously weren’t tried because there’s no incentive to innovate when you feel like all of your users are locked in and they can’t leave. Yeah, I think starter packs is we haven’t spoken about them, but I think that was an amazing example of an innovation.

Companies are copying. Um, how did that even come about? How were starter packs created? Well, we were hearing from users that, you know, they love Bluesky, but they’re like, oh, well, just my friends are over here, and this, you know, somewhere else, and I can’t bring them over. And it’s hard to get them to join one by one and to find them when they join and, you know, plug them in.

So we created essentially a way to build a list of all the accounts that, you know, are on Bluesky and have somebody share it with someone through a link or a QR code, and then you can follow all in one go. And so people have built up their communities very quickly on Bluesky using this.

So they share a link to, let’s say, the Marine Biologists Starter Pack. And then if you’re a marine biologist, you can just join through that link, click Follow All. And then suddenly your entire network of everyone that you used to follow is replicated over on Bluesky. That’s also within Bluesky, a great way to find new people to follow.

So if you’ve picked up a new interest recently, like I’ve seen a lot of really niche starter packs like the Commodity Traders Starter Pack, and maybe you just got into commodities trading and you want to figure this out. You just click Follow All, and then suddenly you’re following. Your feed is full of this new group of accounts that You decided to follow.

So let’s build out a follow graph really quickly. One community of users that I think has definitely really embraced Bluesky and had previously embraced Twitter is media folks. Um, and, you know, I think that the reaction that I’ve seen at least, and I am, you know, media, and we’ve certainly seen that links to Techdirt articles get way more traffic and way more interest than any other social media network. Um, how are you thinking about dealing with media users?

Yeah, we’ve seen a lot of uptake, and we keep getting the word out that we are essentially a portal to the open web, which means other sites on the internet, so your links don’t get downranked. If you put a link in a post, it can get picked up not just in the main feed, but even in specialized news feeds. So there’s news-focused timelines out there that just let you catch up on all the articles from news organizations that are put on the list for this feed.

And so that means that people can use it essentially like a newsreader and just catch up on links. There’s also link-only feeds where you can see all the links that your friends are sharing. And so this makes it really useful for being a way to catch up on information or share information, because it’s not focused on keeping you on the app; it’s focused on delivering value to users. Yeah, speaking of custom feeds that I use, one of the ones that I find really useful is the I forget exactly what it’s Called, but it’s like the gift link custom feed, because obviously, a lot of media properties now have paywalls.

And so, a lot of them will allow subscribers to post gift links that enable you to bypass the paywall. Someone created a custom feed that collects anytime someone posts a media link that is a gift link, allowing you to get past the paywall. It all goes into a single feed. It’s very useful for keeping up on the news from sites that you don’t have subscriptions for. It’s pretty stunning.

One thing that I think has been central to the idea of Blue Sky and the ad protocol, going all the way back to the beginning, that we haven’t really discussed here is just the very concept of decentralization, um, which I think right now is more important than ever again. We’re living in a world where we’re noticing the consequences of very centralized power and control, and the ability to manipulate things from a centralized position. We are seeing what the end results of that are.

So how does Blue Sky, as a still centralized company, think about the importance of decentralization itself? Yeah, I mean, we use decentralized protocols every day. The web is a decentralized protocol. Email is a decentralized protocol. For example, with email, you can use Gmail and still communicate with somebody using ProtonMail.

And if you decide that you don’t like the way Gmail handles your privacy, you can move over to ProtonMail and still message all of your friends from before. One way that we made sure to build the app protocol is to have a portable identity, so you can keep your username when you move between services. With email, you don’t really have that, but the underlying principle is similar, which is you can have the same identity, communicate across service providers, and you know the email service is there helping you communicate, but it’s not controlling your whole experience. It’s not the only option you have. And so Blue Sky is like that. It’s a service that lets you communicate with the network, but it’s not the only option you have. And new stuff is being built every day. I think this is what keeps the door open to innovation and letting people build better social experiences down the road.

So what does the future hold in your ideal world? What do you see? Blue Sky. And at protocol. Looking like five years, ten years, 15 years down the road. Well, I hope that people don’t see social as inherently toxic, because there’s a lot to love about social networks, and there’s a lot that can be done to improve it. And if innovation is just coming from one company, there’s a lot that can be done. Like if you focus on user control and user choice…

You can do so, but also, if you want to have a different experience, you can just come in and build it. If there’s a possibility to do so. And so, I’d like to see a world where people don’t feel like their only alternatives are, as you mentioned earlier, yelling at the company or yelling at the government to do something about the company, but actually feeling empowered and actually having the ability to go out there and build what they want, whether that’s a different way to approach moderation or a different way to build a custom feed that uses a different algorithm.

This is something where I’m already seeing pickup from, for example, academics who are researching ways to do pro-social feeds. That’s something that they can do directly on Blue Sky, and they can build it in and then experiment with whether this produces better outcomes. And you don’t have to wait until the company makes a change in order to integrate that feature. You can just build it directly in and see if users will adopt it. So my hope is down the road that we have a much more diverse and varied social ecosystem. There are all sorts of different apps, all sorts of different experiences within the apps, different feeds, different moderation services, and people can find what they want.

If they have a complaint, they know that they’re empowered to go build it. They can just shop around and find something out there that meets their needs better than the main The thing on the market.

Yeah, I think it’s interesting. Um, it’s easy to sort of get keyed into, like where we are right now, and there’s a lot of negative stuff happening in the world, and a lot of it is sort of tied back in some ways directly, and in some ways indirectly to social media. But I do think it is really important for people to think back. I mean, Hugh mentioned it in the intro, how much excitement and possibility and opportunity a lot of us saw in social media in the early days and how much it really did empower people. And it did create connection and community and change. You know, I mean, there are the stories of the Arab Spring and Black Lives Matter and the MeToo movement. I don’t think any of those really happened without social media and the ability to connect that way.

And one of my fears is that, you know, what’s been happening as people sort of get upset about where social media is today is that we begin to throw out that power as well, that we get concerned. And so one of the reasons why I’m excited about Blue Sky is the fact that it is sort of looking at this in a, you know, through that prism of how do we bring back that kind of social media. Yeah, that was good and was valuable. And part of that is not by letting you control it all, but you sort of creating these affordances that allow anyone to do it. So we’re starting to see, you know, people getting past that learned helplessness, I guess, as I described it before. Yeah.

Well, it is what we make it. And Bluesky is more of a community project than any other social media, because it’s open source, because it has an open protocol, and it has all these open interfaces. Our goal has been to make it more and more easy to build things on it and lower the threshold to participation. So if you’re a developer, of course you can get in there and build whatever you want. But even if you’re not a developer, people have come along and built no-code custom feed creators. So you don’t have to know how to code to build one of these feeds.

Actually, most of them are built by people who don’t know how to code. And there are new things like that emerging every day where you can participate. If you’re just a community builder or you’re just somebody who wants to tweak things in a different way, you can. If you have an idea and you want to make it a reality, you can use one of these third-party feed builders to build a new version of a timeline for yourself. And so I think, you know, just the philosophy of keeping things open, letting people build, and then moving incrementally towards a world where more and more people are able to make the changes that they want is a really great way to approach things. You know, one way that we could use LLMs to actually make better feeds could be to maybe just describe the kind of ideal feed, and then you have a service that transforms what you just described into a feed for you.

And then you can say, “Show me more or show me less” until you get the feed that you want. That’s something that hasn’t been possible, but now can be built. And you know, you can build this in the ecosystem today if you’re technical enough to build and launch that. Yeah. I mean, it’s interesting because there’s also been this sort of pushback on algorithms in general that we’ve seen. And you know, the thing that I keep thinking is that it’s not the algorithms that people dislike. It’s the particular algorithms and the fact that who controls those algorithms and how they’re used for manipulation.

But algorithms can be good, right? And can do exactly what you said and sort of be used for pro-social or helpful reasons. Right? Well, yeah. An algorithm is just a way that a social app shows you posts that can be really simple or really complicated. But the main problem I think people have experienced with algorithms is that it’s just one algorithm that’s a black box. You don’t know what it does and you don’t have any control over it in any meaningful way. And if you can choose what the algorithm does or if you can choose between algorithms, you have an open marketplace where you can decide, does this one suit me? If it doesn’t, then I’ll use a different one. And so I think the algorithms that have produced a lot of Bad outcomes are you only have one of them, and they’re mostly driven towards an ad-driven feed model.

And so that means that you’re getting stuff that’s more engagement-based than anything else. And even if it doesn’t make you feel good, you don’t have an alternative other than just not use the app entirely.

And so having open choice means that you can design things that work for you. If it’s not there, you know, then think of a way to put the idea out there. Maybe somebody else will make it. Having that openness is really what makes better outcomes.

Yeah, I mean you talked about sort of like the no-code examples of creating custom feeds. There’s one service that I’ve been playing around with called Graze Social, which I think also came out of that, like somebody sort of said, “This is what I really want,” and someone else is like, “I kind of want to build that.” And they started talking and making it. It’s this sort of amazing, amazing service.

Yeah. One analogy I use sometimes is like social media algorithms have kind of turned into like high fructose corn syrup, where it’s all like high dopamine engagement bait. But you know, you don’t have to just say no algorithms at all. I only want chronological, which, by the way, is also an algorithm, just a very simple one. You can have essentially a farmers market of algorithms where people are bringing their own. And so our custom feed marketplace is kind of like a farmers market.

There’s everything spanning from new complex algorithms that developers have built to something really simple, which is a user that just used a no-code tool to say, “You know what?” This algorithm is only going to show pictures of cats. That’s an algorithm. You can make the feed created that way, and then you can shop around and just design your custom social experience with, you know, the five, six, or ten algorithms that you like, built by the people who are showing you things you want to see. And that’s a way to experience algorithms in a much healthier way because you have control over whether you want to be in the high fructose corn syrup mode today, or if you want to go and choose a more healthy option.

All right, so I know we’re running out of time. We have basically less than a minute left. So very, very quickly for folks who heard this or have been thinking about this, but watching this and are excited about both that protocol and Bluesky, how do they get involved? What can they do? What should they do very quickly? Yeah, well, there are tons of ways to participate. I mean, first of all, just use the app, download it, check it out, maybe follow people through a starter pack, search for a starter pack that shows things in your interests. If you’re a developer, it’s open to build on, so go ahead and try it out. Build something. You can build really simple apps or really complex ones.

If you’re a community builder, get involved. Help shape the culture. You know you can do this by posting or by running a moderation service or a feed. And if you’re a creator or somebody who just wants to establish a presence on Bluesky, start posting or contact the partnerships team at partnerships@blueskyweb.xyz and we’ll help you get started. Really, this is a community project. As I said before, it’s open source. It’s designed to build communication infrastructure that stays around and gives people control over their experience. So it is what we make it. Get involved. Great, everyone. Let’s thank Jay.