Mic check. Mic check. Yeah? Okay, cool. All right. All right. Hello everyone. I'm Rudy Fraser. And brings me a lot of joy to uh be presenting on this topic here at the second Atmosphere conference. It's also been it's been really cool to be able to ask people if this is their first Atmosphere conference. And if this is your first Atmosphere conference, welcome. And yeah, I hope that there's something here that you can take away and ship something and build something for the community and ecosystem. Alright, so uh as I said, I'm Rudy Fraser, I'm the founder and CEO of BlackSky Algorithms.
I am previously a fellow at the applied social media lab at the Berkman Klein Center. I'm the board chair of PAC Collective, which is a nonprofit that fiscally hosts uh about 25 uh mutual aid groups in New York City. Uh and I'm an organizer with We the People NYC. Um BlackSky launched about three years ago uh now to create and we we started off as just a custom feed on the Bluesky app that and now has been uh used by uh over two million people. The feeds are used by a quarter million uh people every month, uh, and we are a member supported organization.
Um and yeah, and that that member support is really important because the last time that I was here at Atmosphere Conference, uh it was just me on this slide, but now uh it's uh our entire team and our entire team is here today.
So if you see folks with a BlackSky shirt on, uh tell them what's up. Uh yeah, and um, yeah, I think the the thesis for BlackSky is that we really do believe that um app protocol has these unique properties that allow folks to uh it it helped us bootstrap uh a pretty large community and we think that um we're we're really interested in in helping other folks basically do the same thing uh leveraging the the stuff in the protocol. Okay, uh so the title of this talk is Groundings with My Siblings. Um and I want to start by explaining where that comes from.
Walter Rodney was a Guyanese activist scholar, uh pan-Africanist, uh, and I think uh most important for us today, he was a practitioner. Uh he's best known for how Europe underdeveloped Africa, but his earlier work, The Groundings with My Brothers, is one that changed how I think about what I'm doing with BlackSky. Rodney was 26 years old, uh fresh off his PhD teaching at the University of the West Indies in Jamaica in 1968, and he did something that the university and the Jamaican government could not understand. He left the campus. He went into the uh into the communities uh and sat down with uh Rastafarians.
Um and I think a lot of people have an understanding of like Rastafarians in like Bob Marley's sense, but Rastafarians in Jamaica are considered like they're like houseless people. They're like the the they're like gutter punks, like uh they're outcasts. And um, and yeah, and so he he sat down with these folks and they reasoned together. And that's what grounding is it's a Rastafari concept, uh sitting down together to reason. Uh and the part that I keep coming back to is that Rodney insisted that he was the student uh in those exchanges. He said, I got knowledge from them, real knowledge.
He went in thinking he had something to teach, and he came out understanding that the people closest to the problem already had the analysis. They didn't they uh they just didn't have the platform uh that he had. Uh his widow uh Patricia described grounding not as an event but as a practice, a way of living, uh one where academics and activism were integrated and inseparable. I'm borrowing uh this framework because over the last year I feel like I left the code base. Um and uh went to kind of did the inverse, went to college campuses, went to conferences, went to community spaces, uh sat down with people, spoke, listened, um, probably talked to the most people I've spoken to in my life in the last year.
Um and like Rodney, I learned more from those conversations than I could have given back in those moments. And so that's what this talk is about. It's like what I heard, what surprised me, what we got wrong, and kind of yeah, what we took away from from these moments. So here's what that grounding looked like in practice. 15 different cities, several college campuses and schools, a lot of times sitting with people who have never heard of app protocol, never heard of Bluesky, but who care deeply about the same problems that I believe everyone in this room is building towards.
Really I should be probably sitting amongst y'all, right? Like it's uh instead of just like speaking from the from the podium, but uh I I want to give you what Rodney gave the people he sat with, which is just honesty and his experience. There's a tendency in tech circles and communities to only share the wins, only share the big fundraises. And I think if you follow me on Bluesky, BlackSky, I tell you when I think we're doing something wrong. And and yeah, and so I don't think you learn from just wins or highlight reels.
I think you you learn from losses and from pain. Uh the open source community is at least adjacent to the ideal that information wants to be free. And so this is my giving this information to you for free so that you can build something uh that folks find valuable and you don't have to go through the same pain, pain points that we had. Okay, first up, PDSs. Uh I think port portability is a core premise of the protocol. Uh, but there's a pattern in how it actually becomes real for people. So in this chart, you can see so over time, right?
This is this is a chart pulled from uh Makuba, I think it's their name. Uh they have this dashboard of independent uh weekly users posting from non-blue sky PDSs. And that first gray circle there was I I kind of unintentionally posted about giving away invite codes. I was talking about giving away invite codes to our mods, and then BlackSky was like, there are invite codes. And then what are the invite codes for? Um so and so, yeah, so I did the like Oprah meme, and then we got like 200 replies asking for invite codes. And uh at the time, I remember being pretty frustrated with where the Atmosphere community was at on data portability.
Uh I feel like there are times when stuff is like treated as a solved problem. Uh and so I made a post about it. And uh a day later, I made a video showing how to migrate using the GOAT command line tool. Uh that video now has one point like a thousand video a thousand views on YouTube of just like a 45 minute video where I'm just like, this is hard, you probably don't want to do this. Um then the very next day, Bailey Townsend um shipped version one of PDS Mover.
Um and yeah, and so and like you can kind of clearly see that that makes that makes a difference, right? Uh the purple circle is some bann controversies uh that we won't get too too deep into. Um but you know it it was a it was again a real pain point. I think sometimes people see things on the internet and they think it's like uh it maybe doesn't affect real life, but again, I'm I'm in all these different places. I was at a conference in uh the Bay Area, this like observatory, and multiple times in the night, people kept coming up to me is like, so what's going on with Link?
I'm just like, do you know him personally? Um but they just heard about the story, and for us it was very painful. It brought our entire mission, like sometimes people were like, why do you even need a PDS? It brought our whole mission into question of like why do you even exist then if this could happen? Um, and it was and it had a real dollar value to it. Like we ended up shelling out 14,000 to like set up an app view server just to like bare metal, um, and then went, and that's not counting the labor to like get all that stuff um built and shipped.
And then the blue circle, which is the big spike, is Euros Sky launching. Shout out to them.
So you can see all this happened at the end of last year. And there were just like very specific things people did to change the narrative. And with uh, as I mentioned to Bluesky folks, like with all due respect to Bluesky, like this was a fully community-driven effort. There was no like outside initiative, it was just folks saw a problem, they worked on it. I'm sure I know Bailey was like going back and forth with folks, figuring out how to make it better. We were going back and forth. I was every invite code was done via my DMs at one point.
Um so I was literally talking to hundreds of people to like coordinate this. Um and so the pattern there is is the story is really just like there was a tool, but it wasn't accessible to folks, right? The command line tools didn't move the needle. Um and then it went to this client-side Chromium-based tool that was mobile friendly, and that was what folks needed. Um yeah, and so it was just like to me that felt very clear. The users that we knew, like they they weren't ever gonna like download the source code and like build um build something to use it.
You know, it needed to be really clear and concise. Um we found out from just doing that, like there are people using all types of browsers I've never heard of personally. Um in different scenarios. So imagine you got some one of our first users who migrated, it was like a half a gigabyte blob storage. Imagine if you try and do that client side uh over like a cell phone network with some random browser. You know what I mean? Um so and yeah, and so and then what I think I tilt I tip my hat to the Eurosky folks, what they did was they did it server side.
Uh and so it was like it just happened in the background for the user. Um and it became really turnkey. So we took that inspiration from that and built what became move.blacksky.community, which followed that same pattern. Um also because we have this community space, we're able to like analyze the this 20 million record feed database, and was able to find like the real people had real fear. They had they thought that their accounts were gonna get deleted if they did the migration process. Um and so like another thing that we did was like add social proof.
So like when you go to move.blacksky.community now, you see how many people are on uh the PDS. You see, and it's sorted by people with big followers because people are like, oh, if I have a bigger account, how will this work for me? Da-da-da. So having that kind of social proof. But the but we we found that that was the effective way to help people get through the migration. Um yeah, now our PDS has over 30,000 accounts on it. Um, all right, the next one, uh favorite. The app.bski uh compatible app view. Um I wanna this is like an important one for me because I think sometimes people point out, I guess, where we're at and not necessarily where we want to be going.
Um BlackSky is not trying to be a microblogging app. We're not trying to like compete with like Bluesky on that. Um the the app modality or medium is not important for us. Uh to fulfill the mission of reaching millions of people but being independent, we needed to do this. We needed to be able to say that we could do a backfill of the entire network. It needed to be possible. People have said it was done. We needed to like actually do it. Um more than anything for me, I needed to prove to myself this could be done in a way that people would actively use it, that I would be able to log in, open it up every day, and like it would be an effective experience for me.
And it was hard. And the things that I knew were gonna be frustrating for people were the things that ended up being frustrating. Some people like to act on the internet that like follower counts don't matter. The biggest complaint is the follower accounts not matching Bluesky. Um to the point where we built a feature that says backfill in progress that analyzes what our count says compared to what Bluesky says compared to what microcosm says, and if it's a big enough difference, then we show that badge to people to try to get ahead of that um and like communicate that they're this we're working on it.
Right. Uh profile pictures not loading. People are like, I can't recognize, I don't know this person by their handle. I know them by their profile picture. So like if you don't have the profile loading, now it's like a weird experience when you're scrolling your timeline. Stuff like that matters a lot. Another thing that we found out that I don't think Bluesky knew themselves is that uh if you're banned from Bluesky, you aren't banned from the image CDN. So your profile picture can load, but you are banned from the video CDN. So you can't upload videos.
And so now we're like, oh we got an app view, boom, we unbanned you, like we reverse the decision, and then they tried to upload a video, but we were still using their video CDN, and now once again we look like we haven't fulfilled our mission. Um it's important to to know these things. I have on the the the kind of um just like I think what people think it's like setting up an app view, I just want to like highlight some of that stuff is like uh they think the top is just like record hydration.
Then it's like there's 42 million people, but we don't care about all of them. So like how do what's the order that you backfilled the the people's accounts? Then there was like stuff was taking too long, and then people wanted to use Skeets app with our app view, and then Skeets app had like a timeout on it so that that was like shorter than our timeout. So then it was a bunch of 502 errors if you try to use Skeets. Um notifications is its own thing, there was a bug at one point where it was giving you duplicate notifications.
That's annoying. Um post aggregation, profile aggregation, those are the follow counts. Uh handle resolution, um, being able to like include the germ DM, uh get timeline, it the the further down is like the harder the things got. Um and so get timeline like your following feed is like really hard. Mod appeals, so if Bluesky bans them and and they're like we got people were reaching out to us like which one of you banned us. Um it's very tricky and complicated things to then answer, and then get back to people about like you know, and again, every single time something like this happens, it's like if you can't if like why do you exist if we're still you know experiencing these things.
Umath. You found out that OAuth doesn't allow you that if you do a social app fork, you're you're used to being able to like delete your account, deactivate your account. We included OAuth to make it like really easy for anyone, like even if they're not on our PDS to be able to log into BlackSky, but then you couldn't, the OAuth permissions don't allow you to delete uh your account. And so then we're like, then we had to like hack on PDS gatekeeper to like do some things where like now we ask you if you're on our PDS, we ask you for the password, and that allows you to delete the account.
Uh and if you're not, we send you to the account management page, but we know that they're the account management page also like doesn't have like all the things that you need, right? So um pain, all this is pain. Um again, you like you learn from the pain, but we experience the pain, so you now if you want to do this, you know what's kind of coming, right? That's the whole message of this. Um post nerd app proto. We're um so something that I yeah, I keep coming back to is like uh and I got actually got this from Marissa from our team is like this concept of like the job to be done.
Um when I'm speaking to people, it's not to do app proto advocacy. Like someone took their time out of their day to hear if I can solve a pain point for them. I'm not gonna pitch you on, you know, like I'm not gonna put you on the cause. I'm supposed to pitch you on like I can actually fix this thing that you're experiencing. Um if you you know a lot of the a lot of the app protocol devs are the community, we're talking about we're talking about record hydration, we're talking about Postgres optimizations. You're you've gotten those conversations are very far away from what the user is experiencing, unless your user is a developer, which they could be, but for us, we're trying to onboard communities of storytellers who are not sold on decentralized social media, they're sold on community building as a tool, and then we said that this protocol allows us to get the things that you want out of it.
Um yeah, so I spoke to many groups, um, many groups like this, along with um AJ from from Northwestern. We spoke with Movement for Black Lives, we spoke with Slow Factory. These are people who like they're they're activist organizations, community builders with millions of followers on other platforms. They've had people on their teams who've like built they were like VPs of engineering at like social media companies. So they understand social media, they understand decentralized social media to a degree. Couldn't close those deals.
There were young folks at our South by Southwest talk who came to learn how decentralized social media can be a part of their content creator journey. And if you and this, which was like a little weird for me because it's like you're in college and I think maybe do something else. But and then I gotta imagine how disappointing it was for you to like want to be get you want to get lit on the internet, and then you we we're talking to you about podcasting and RSS. But you know, we can't write uh we can't write people off.
Uh if we want the stuff to grow, right? Like another uh I posted about this one example, like there's a media organization, they wanted to reach a particular audience. Um they want to reach black Chicagoans. Uh we did not lose this deal because uh the complexity of app protocol that wasn't like hard to explain to people. We didn't lose it because they they were like, oh, master don't exist, that's not why we lost it. Um we lost it because they had a specific need, they wanted to reach black Chicagoans. It's a decentralized network, we don't have all that information.
You know, it's disparate. You may be on our PDS, but like maybe some of you are on a PDS, other people on other PDSs, you're on the feed, maybe you're doing using a different client app, all these different things, it makes it hard. Like, even if even if we kind of knew these people existed, there wouldn't be a way to like do the same, because they were able to then go to Facebook and Facebook's like, give us 20,000 dollars, we'll get you in front of a million people in the next year. Can't promise that to people.
Um yeah, and we've only uh we've only ever done inbound. We think that um some things that maybe we have some like theories that if we build out this project called BlackSky Cash, uh, it will make the pitch more compelling to people, it'll make a clear monetization strategy. So maybe even if some things aren't necessarily all the way there yet, it's like, oh, here's like a way that you can build community, do democratic governance, and have a way for you to fund things. Um but the job to be done is never use app proto, right?
The job to be done is solve the problem that I have. Um now and this one's very important because we're pitching community builders, we are pitching the Atmosphere in these spaces. Like, I'm not pitching BlackSky necessarily. I'm pitching everyone's projects that are here. I know like I sometimes talk kind of crazy on the internet. Uh, but I am a lot of people's cheerleaders, um, cheerleader. Uh I can't do all the things that y'all have done. Um, but I do tend to get people to open up about their problems, and when they do, I pitch your products.
Um an example is like at South by Southwest, Karen Attia who was fired from the Washington Post because of a post on Bluesky about Charlie Kirk. We were talking, and she was like, you know, I wish there was a way to go live and stream for my Bluesky account. And I was like, there is. Have you heard of Streamplace? I was like, it's another app. And she's like, well, why is it another app? Um, and so we've had this pitch of like you have, you know, there's the whole login with Bluesky, login with app protocol.
That's even that's even like a second step. It's just like, why do I have to go to another place if you're saying that I can use this same thing, right? Um so to me, this is like a tension in the Atmosphere right now. Uh you can log into a dozen different apps with a single identity. That's great to us, but to everyday people, being told use this app for microblogging, use this one for live streaming, use this one for long form content, it doesn't create delight. It feels like what they already have, a bunch of apps.
And there's no even way to discover all the apps. There's not an app store for uh app protocol. Um so the odd, there's no aha moment there for people. Like we have the aha moment, we look at the fire hose, we look at your repository, we know that all this data is on one account. The user is just like, okay, this is another, this is another skin of a thing, right? Um so for me, what would be compelling uh uh and I think uh to the people that I've been talking to, is something like an app protocol browser and app protocol app store, something that turns login with your identity across many apps from developer premise or a promise to a user experience.
So that's the the question mark there. It's like I think we we talk about app protocol as like the protocol and we use the comparison to the web. Had debates about that even just in this trip. And I'm like, so what is the well the web has Firefox, and I can go to any website with Firefox, and it feels really simple, intuitive, right? There's no universal app protocol client. If you had to download, the equivalent is like downloading an app on your phone, but then again, now we're just back to downloading a bunch of apps on your phone.
And if you have login with Google, then like as a user, I'm like, what's the difference? All right, finally, um, there is a book by Douglas Hofstadter called I'm a Strange Loop. Uh recognition. Nice.
Of yourself perceiving the world, and that model changes what you perceive, which changes the model. Um so when we started using moderation, we used Blue Skies Ozone. Uh it's open source, it works. Uh, but it wasn't built for us, right? It was built for Blue Sky's trust and safety team. Uh our moderation team, it started out as like very community moderation. Uh and we still want that to be that way. We're even building more ways that the community can participate in moderation. It's more like Ozone was built for a particular world view, not necessarily a scale.
Um, because we can do peer moderation at scale. Reddit does peer moderation at scale, it's just a different opinion of how things work. Um and so for us, our our like Dr. K, JD, they're community members. They weren't like reviewing content for like a platform, they were protecting a space that they're a part of. Um and so Rishi on the team built Acorn and literally uh and from as I understand it because Rishi is very interested in co-working with people because we're all remote right now. Uh he and JD got on calls every day and just work side by side.
And when JD ran into a problem, Rishi built something for it. Um, and so yeah, and so what you're looking at on the screen is like the first thing a moderator sees when they open a session. Not a queue, a check-in. How are you feeling right now? Do you have water nearby? You can step away. Your well-being comes first. There are session goals. It could be time-based or item-based. Um, moderators choose how much weight they can take on. Because you could just loop, right, and just keep going. Uh, and that can be really really heavy. Um we we're building collaboration tools in it.
We're building ways for us to govern our community from it. We're we're gonna build ways to issue badges to our community through it. Where we oza we use uh our feed kind of works like a subreddit. So Acorn is gonna be built to like manage a subreddit where you can remove members, you can remove posts. You could do this from your phone. Our our mods, our mods want to moderate from their mobile devices. And uh, yeah, Ozone doesn't, it just doesn't work like that, makes it really difficult to. Um and again, that's not a that's not just because it's some big platform, some billion person platform, it's just us we built it the way our users, our community, our mods do things.
Um that's the strange loop. Like you are the user, you're also the builder. Uh and the thing you build changes changes your user experience, which changes what you build next. Um Tony Morrison said if the book you want to read doesn't exist, you have to write it. Um but it's deeper than that. So once you write it, now it's out in the world, now it's coming back to you, right? So the gift and the recipient keep changing each other. So every lesson today comes from the same place. We are the user, right? I was the first one to migrate to the BlackSky PDS.
I also had the pain when it was just GOAT. I knew the the app view had to work because uh like the way that the community would expect it. I knew follower counts had to be right. Like the first app view that I used that was independent from Bluesky, it was like half my followers. No, this is weird. So you just you can predict these things, right? Um I know we have to we have to get beyond just the dev community, right? Like if we want it to be a big thing, like it has to be fun, it has to be cool, it has to have different, it has to have that delight, right?
It has to have these aha moments. Um that's the that's the kind of thing we we became we become who we build for. Um and that's not like a bug. That's like the whole point, and that's the our practice of how we we work on things. Thank you.