Ghost founder and CEO John O’Nolan wants to build a tech company that plays by different rules. Learn how ActivityPub is enabling a publishing platform with sustainability at its core.
John O’Nolan, the founder and CEO of Ghost, calls himself “the inverse Peter Thiel.” That’s because he wants to build a tech company that bucks the usual narratives, with as few monopolies as possible. His open-source publishing platform is structured as a nonprofit and is integrating with the ActivityPub protocol, giving creators digital sovereignty. No longer do writers have to perform for an algorithm to succeed or get stuck inside closed systems that monetize off their backs.
Does this scenario seem too good to be true? As you’ll hear in this conversation with Flipboard CEO Mike McCue, John doesn’t think so. There’s still a lot to be figured out, but both entrepreneurs are here for whatever this next phase of the internet brings.
Highlights of this conversation:
🔎 You can find John at https://john.onolan.org/
✚ You can connect with Mike McCue on Mastodon at @mike@flipboard.social or via his Flipboard federated account, where you can see what he’s curating on Flipboard in the fediverse, at @mike@flipboard.com
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.
There's a common narrative in tech. Someone has a great product idea, and they raise a bunch of money, they say they're going to change the world, and then at some point, there's an exit. They sell the company, it shuts down, it goes public, you know the drill.
Well, what if you were to build a company that did the opposite of this norm? What if that company does what it says it's going to do at the expense of everything else. How does something like ActivityPub enable this scenario?
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.
Today, Mike's talking to John O’Nolan, the founder and CEO of Ghost, an open source publishing platform for professionals. John believes that to have journalism with integrity, you have to have technology with integrity. Learn how he plans on living up to that promise. We hope you enjoy this conversation.
Mike McCue:
John O’Nolan, welcome to Dot Social.
John O’Nolan:
Thanks for having me.
Mike McCue:
It's fantastic to have you here. There's a lot to talk about you've you're the founder and CEO of Ghost, which is a incredibly well respected platform for publishing, and we're going to talk all about that, but, but I thought a good place to start is you recently announced that you're going to be federating over ActivityPub. And this is a huge, huge announcement. It was generated a lot of excitement, certainly from from me and many others, we are incredibly excited about what you're doing here. So maybe a good place to start is, why? Why are you federating over ActivityPub?
John O'Nolan:
Yeah, well, first of all, thank you for those kind words. We've been an independent, open source publishing platform for 10 years, and over that kind of decade, we've had various competitors all along the way, and they've been different platforms that have been trendy and popular at different times. But one thing that they've all had in common has been a social interaction layer that has allowed their publishers to interact with one another, to follow, to like, to subscribe, to respond, and that has always been the hardest sell about having an independent website and being on your own infrastructure, your own domain, has been it's been lonely. You know, people set up on the open web, maybe with WordPress in the past, maybe with Ghost today, and they're publishing, but who's reading, who is, who is engaging? And for a long time, you know, you didn't know early in the blogosphere days Google Analytics, traffic was exciting, and then it became less exciting, right? As social networks got big. And so we always kind of dreamed of trying to have some sort of ability for our publishers to interact and be connected to one another, but we couldn't figure out how. And at some point in 2017 This was way before activity pop we had thought about we dreamed up our own open API idea that was sort of an abstraction on top of RSS that would allow for some of these things to generate a decentralized newsfeed. But we were too small to ever turn that into reality. And then couple years later, ActivityPub came out looked very interesting, but we couldn't quite figure out how to take the spec and turn it into a product, and we've been watching the space closely, and obviously, Mastodon an amazing job. You yourself and Flipboard have done a lot of work recently, threads, suddenly, lots of platforms getting into this space makes it feel like ActivityPubs having a moment. So this year, we're looking at the space again. I was talking to Chris Messina, who's very vocal advocate for ActivityPub, and it just felt like the time of the fediverse now, you know, it felt like this is the time to do it, and so we thought we'd test the waters by saying that we're going to do it and that we're going to experiment at the very least, and see how strong the interest is. And answer, very strong, lot of interest,
Mike McCue:
Yeah, from from users and creators alike, I assume, right? You know, famously, Casey Newton switched over to Ghost from Substack. What about six months ago? Or something like that? And platformer dot news is a fantastic site running on Ghost. It's, it's, you know, every bit as good as anything he had on Substack. It's a very clean user experience, but you're right that he doesn't know well who is reading his stuff. Now he does have a subscription model, so that obviously helps with the email connection and so on. But you know, did you talk to Casey or. The Creators about ActivityPub, and what did you hear from them as you were thinking about this?
John O'Nolan:
Yeah, talk to some before and some after. We're kind of talking about getting into the space. There's a lot of interest, which I would describe as tentative, and sometimes apprehensive. There's definitely a sentiment within, outside of, you know, our space, the kind of product builders make a space, the sense of, I'm not completely sure I understand what it is. Can you tell me again what it is? You know, that's the most common conversation I have. And so there's that, but there's also this aside of everyone being exhausted by the antics of algorithms and social networks, and everyone hearing that this fediverse bus thing, like threads is doing something with that, and Mastodon like that's an alternative. So everyone's got this, this little sense that there might be a possibility of something special, that's different, that's driven by different incentives. And these two things are a little bit in tension with one another at the moment. So it's kind of like, tell me what it is, but what is it? I don't understand it now. I don't it's too much effort, but what is it again? And so that feels like a really special moment, an opportunity, if we can harness it, if we can create good products, products that people don't necessarily have to understand the technical foundations of. We could be, we could be on something, I think, but it's, it's a real, it's a real inflection point.
Mike McCue:
I think it is. And when you look at, for example, with some of these writers that were on sub stack prior and have now since moved to Ghost. Presumably, one of the things that they're telling you that they miss is, like the discovery capabilities that they saw on sub stack, which kind of in the sub stack wall garden, they've been innovating a lot on, you know, how to recommend other blogs to people who are reading one blog and lots of their, you know, kind of the Substack notes capability and so on. So when you look at ActivityPub, and you look at that kind of discovery realm that that Substack has been working on, what are your thoughts there and how that could happen on Ghost.
John O'Nolan:
Yeah, I think there's a lot we can do there. So we have a similar recommendation feature. We also have a directory called Ghost Explorer, which is similar as the way Substack allows you to discover other writers. I think the ActivityPub layer is going to these are very structured forms of discovery, right? You have to go somewhere, you have to find a list, and then you, sort of you manually discover stuff. What ActivityPub does that's interesting is it creates an organic web of discovery, which, in my mind, I think Twitter and then Tumblr, like, really nailed. This is when someone else has retweeted or commented on that you've never heard of, and you're like, that's hilarious, or that's interesting, and you click on a profile that you weren't looking for, that just came to you because someone who you already know shared it. I think it potentially unlocks some of that. And we can still augment that with a directory. We can still augment that with recommendations manually, either from us or from publishers to each other, but that organic piece of just I like this, you might like it too. There's something magic there. When you can, when you can get it right. And I think that's, that's exciting.
Mike McCue:
That is exciting, you know, one of the things that I'm always, constantly amazed by is I'll find some podcast, or some writer, some blog that I had never heard of before, and I and it's amazing. And bunch of other people have already been like, yeah, of course. You should be, you should be listening to this podcast. You should be, you know, reading this writer, and it's amazing to me that I still sometimes don't discover these things. And I do think that like that is a big need, especially now with, you know, all the concern about SEO going away, and you know, how the search engines are going to be, you know, gradually bringing all of this content up into AI summaries. How do you discover some really awesome writer, you know, in this new world?
John O'Nolan:
Did you see that Google search algorithm leak yesterday that came out as well?
Mike McCue:
I did. I did.
John O'Nolan:
I mean, that just underscored, like, for anyone who doesn't know, there was a big leak of Google's search algorithm that was published by Rand Fishkin, who is the founder of Moz and it sort of underscored that for years, Google has been not penalizing but not promoting small sites. Specifically, it has been ranking larger sites higher. So if it has felt to you over the last number of years like your ability to discover weird, little crazy sites on the Internet has gone down in you know, when you search for, search for stuff, and you click on something and you kind of just end up on wire cutter and Amazon, over and over and over again, it's not just in your head. Now, we know that that's true, and that's, I think that's really sad. I think that it's not what the web should be. You know.
Mike McCue:
I know, for a lot of people who've worked at Google in the past, and I know that there was a significant amount of emphasis placed on discovery of smaller sites, but I don't know you know how that's played out recently, and you know that, I think that that concern of you know, if you're a blogger, you're a podcaster, you're a YouTuber, creator. How do you get discovered in this new world? And, yeah, it's already been on a decline to your point, right? And I do think that ActivityPub really, you know what it does is, is it establishes a human connection that's incredibly needed now as a source of discovery, right? I trust you. You really think, you know, a blog post is really great. You post that I'm following you. I now see that blog post. I discover that blog or, let's say, you comment on a blog post. You didn't, you didn't even, you know, boost it, you just commented on it. But I'm following you, and I see that comment, and I now discover that blog post, right? So that that is the beauty, I think of, of ActivityPub. And you know, the fact that you're baking that in, at a core level, into Ghost, is incredibly exciting for discovery.
John O'Nolan:
Absolutely. And as we one thing that we're doing slightly differently in how we're thinking about it to what seems to be the default. So there's a number of publishing platforms with long form content who've implemented ActivityPub, and so far, they've done so in a kind of broadcast manner. So you're sending you can send your posts out into the ActivityPub network, and people are messaging could follow, but you don't really know what happens when you do that. You know, you're just sending stuff out. And I think that's great. It's a step in the right direction, but it misses a really important part of the story, which is ActivityPub is about a two way connection, not a one way broadcast. You know, RSS is a one way broadcast, and you just put stuff out into the world. People can subscribe. But what we're focusing on is a reader as well as a publisher, so that when you put something out into the world, you can see what happens, and people can respond back to you, and you can interact with them, and you have a more direct feeling of being part of a network, not just pushing stuff into a network.
Mike McCue:
And I think that is where the real potential is, right? So it's not just a two way protocol, it's a two way experience, yeah, yeah, that's a big deal. So in your mind's eye, you know, as you're, as you're thinking through the product, you know, presumably, you're in product meetings every day, working on this, right? Sure, what? What is your ideal experience, you know, let's say you come to, you know, a blog on Ghost. How is someone who doesn't know ActivityPub even exists? You know? How are they going to experience this?
John O'Nolan:
Yeah, this is the product problem, I think that we think about the most, and I think the complete answer is not one I can give now, because I think a large proportion of this is going to come out through discovery, and we are going some of our hypotheses will be proved wrong, some hopefully will be proved right, but definitely. But the way we're thinking about it so far is to try as much as possible to have you not need to know what ActivityPub is to get started. But if you already do know what ActivityPub is, then it should feel familiar. And so one of the things ActivityPub has in common with email is an address, right? You're a user at a domain, we know that it's very recognizable. It's very simple. It's very similar to email. And what Ghost already has when you visit a website, say, platformer dot news, is this subscription form where you can sign up, but you can also sign in with an email address. And what the direction we're going to try and go is to see if we can have that one box that input box support both email addresses as well as activity per handles, so you could sign up or sign in with either one of those, and that's just a case of where do we send the magic link, or where do we send the content. But either way, we can send it to where you're expecting to receive it. So that's the idea whether we can pull it off,
Mike McCue:
I mean, and let me just, let me just try this out for for a second. So I'm going to be able to come to a blog that's running on Ghost, and I'm going to be able to say, I want to subscribe to this, to this newsletter, I can type in Mike at Flipboard dot social, which is my fediverse address, and the right thing will happen now. Now, effectively, it's as if I just hit follow on that blog, Mastodon, and now I'm seeing it in my home timeline. When, when new posts come,
John O'Nolan:
yeah, and the challenge for us will be, how seamless can we make that experience so that it feels like you just put in one thing? And push the button and what you expect it to happen happens. I think there's some technical challenges in making that feel really simple. That we've kind of seen some of the follow buttons and other activity products, right? Sometimes you have redirects and pop ups that immediately start feeling confusing. And so to the extent that we can make those go away, I think will determine how user friendly it feels,
Mike McCue:
And then I guess when I now, let's say I've put in my ActivityPub address, Mike at Flipboard, dot, social, you now determine, Okay, this guy's actually on ActivityPub, he's got an account on the Fed. Averse, now I can, presumably, I would be able to reply to one of these posts on this website and type in my reply, and then anyone following me would see, oh, Mike just commented on this article.
John O'Nolan:
Yeah, yeah, exactly. So the Ghost has memberships, which is why an email sign up is also a login. So if you put in your email address, we'll send you an email and say, click this button to sign in. Once you're signed in, you're now an authenticated user. You can leave a comment, you can like post, you can like comments, you can interact. We can do the same thing, we think, with ActivityPub handles. It depends somewhat on, you know, the private mention or or DM encrypted DM features that we can utilize or work with that are in existence in other ActivityPub products, as long as we can find our way around those to be able to send a magic link to authenticate, then, yeah, you'll be able to sign in, comment, reply, interact without needing a password, without needing to remember details, just through JSON Web Tokens,
Mike McCue:
So that that's fantastic. So basically, if I have a fediverse account, wherever I created that, maybe I created it on Macedon dot social. Maybe I created it through Flipboard. Maybe I created it on, you know, some somewhere else. You know, I could basically put in my my address, into this one box and then go through an authentication process where now full blown ActivityPub client experience. I could boost, presumably, posts. I could favorite, I could reply.
John O'Nolan:
I can follow people as long as long as your activity platform of choice has a way of receiving private messages of some sort. I think we can do it, and then if it doesn't, we might have to fall back to email or something like that. This is where we we still have a lot to figure out. There's a lot of theoretical ideas in here, but that's the dream. That's what we want to get to. Get to, and we think it could be really special if we can have, you know, a platformer, dot news community that is its own community, but it is also part of and connected to a broader network of individuals that might be made up of all kinds of different communities across the web that is going to be an absolute game changer for discovery, right?
Mike McCue:
Because I bring my whole follow graph to Platformer dot news, and then I reply to a story you know on that, on that website, and now people will discover not just my comment, but also that that article even exists, right? What you were commenting on? Yeah, yeah. And that that's just such a big deal, right? Hitting boost, you know, it's such an easy you know thing to do, like, oh yeah. I just want to share this with my followers. And of course, that's for for any you know blogger, for any you know newsletter owner. This is amazing, right? Because now they just have a bunch more people discovering their content coming into the top of the funnel, they can then decide whether they want, you know, how much of that content they want to have be free versus, you know, behind some sort of paywall.
John O'Nolan:
How's that going to work in Flipboard? Where do you think that'll that'll sit?
Mike McCue:
Well, I think that, you know, just as you're saying that you want to kind of effectively embrace existing fediverse accounts, right? And that that kind of account model behind the scenes, that's also something that we're thinking a lot about as well. So as we look forward to kind of the next generation of Flipboard, we would prefer not to have any Flipboard accounts at all and just have it all be accounts that are in the fediverse, right? One of the things that I think, you know, Eugene built, that a lot of people don't fully realize, Eugene rochco, for the creator of Mastodon, is that it wasn't just the sort of micro blogging, sort of Twitter esque experience. It's also all of the infrastructure underneath that it, you know, enables accounts, account management, authentication, all of these, all of these capabilities integrated deeply into ActivityPub and and that's a very powerful thing that, I think is something that you know, we're looking at building on. And, you know, effectively, just not. No longer having our own walled garden account system, right? Obviously, we'll have, you know, metadata attached to an account, right? So just as you said earlier, email addresses can unlock it can basically, you know, be one way you sign up for a newsletter, but So could a fediverse address, right? And that that concept, I think, is absolutely brilliant, and that is definitely something that we'd want to be able to do on Flipboard as well.
John O'Nolan:
Yeah, yeah. I think that could be really special. And Flipboard could be this amazing gateway for people to read, ActivityPub, you know, to discover, to explore the network and see what's there and get an understanding of what ActivityPub is beyond the confines of microblogging or an Instagram type thing, because lots of people have a preconceived notion of ActivityPub being one format, and the reality is it can be so many formats, and having an agnostic Flipboard reader would allow you to mix and match and and have more of a chaotic experience in a way that, I think is what makes the web magical.
Mike McCue:
Yeah, exactly, yeah. Good, good form of chaos, right? Discovery through, through just lots of different, you know, pieces of content that are relevant for some thematic reason or because you know they're they're shared by somebody that you trust, you know, whatever the however you discovered it, being able to see all that in one place is something that you know is really great. And to the point you made earlier, I think is a really important one that is bears repeating, which is ActivityPub is not just about short, micro blogging, kinds of user experiences. It is also about long form writing and federating, that it's also ultimately about podcasts a two way relationship. Podcasts are amazing and that they're built on RSS. Love the whole, you know, ethos. I think it was a Neil dash who talked about the power of just being able to say, you know, get this podcast, wherever you are, listen to us, wherever you get your podcasts. That power of that, but then, but then complementing that with a two way relationship, so you know who's listening to your podcast, right? And then enabling more discovery around podcast videos. Another thing right that you know, there's constantly YouTubers. I'm discovering it. You know, realize, my God, I wish if only I'd been, you know, listening to this, you know, a year earlier. You know, product leaders, fantastic lectures. You know, fantastic. You know, YouTubers that are, you know, really putting out amazing content, but maybe they put it out a year ago or two years ago, right? The algorithm doesn't surface it. Exactly. The algorithm doesn't surface it. YouTube has one algorithm to fit everyone and and I think that's a big that's a huge change with ActivityPub, you can apply any algorithm. There can be, you know, all sorts of different algorithms, even in a single product, there could be multiple algorithms. They could be built by anyone and implemented and then utilized against this whole array of content.
John O'Nolan:
It would be amazing if we end up with ActivityPub, search engines like a Google for ActivityPub, with one algorithm, but it's not just returning results. It's returning content as well. That'll be for just different explorers and different algorithms that maybe biased towards different formats or types of content or types of creator.
Mike McCue:
So many interesting opportunities there you're right? I mean, it's, it's the it's a whole new social web, right? And the web browsers we have today, the search engines we have today, aren't really the right type of product to effectively surf that social web, right? And that's the thing that is extremely exciting. Also, it's not just going to be one search engine, right? There's going to be lots of different ways to discover this content. So when you look at, when you think about all the writers on Ghost, you have discovery algorithms today, right on. Have you thought about how that's going to apply to, you know, the ActivityPub realm, as you start to federate a little bit, but not a huge amount,
John O'Nolan:
I think we're, we're thinking most of the ways in which we think about these types of things are very organic. Can you know, how can we allow people to surface what they like to each other, rather than how can we push discovery in a particular direction? But honestly, we're not quite deep enough in to have fully formed thoughts on what the best way to do that is yet, but it's interesting what you just gave me the thought of you. I wonder if we'll end up with a different type of browser. I wonder if there will be a ActivityPub browser, rather than a web browser, that is a different experience based on the way in which this type of content evolves and grows.
Mike McCue:
That's a fascinating rabbit hole to go down, well, and in particular, because, you know, from what I understand in reading the your blog posts, you've been thinking about actually embedding the whole article in the ActivityPub post, right? Which is that's kind of a mind blowing thing, right? It's not just a link to an article that lives somewhere else, like, the whole article is in the ActivityPub post.
John O'Nolan:
That's right, yeah. And this is, this is, it makes perfect sense, and yet is somehow completely new, which is weird. So, you know, we've had blog posts for ages, the whole blog post on the web, and we had RSS feeds, which were often truncated because people wanted the web traffic so you get a preview, but then they wanted people to click through because they want the web traffic because maybe you've got an ad. As we got into email newsletters, things like Ghost and Substack that are about sending articles over email became popular to kind of receive the entire article in an encapsulated form and read it where you want to read it in your inbox. We're doing the exact same thing with activity partners. Just send the whole article. But the funny thing, from a technical point of view, is that with email, there's this middleman, right? So you write an article, you send an email, it has to go through a whole load of web servers that someone else operates. They decide whether or not it's spam, maybe they stop it halfway through, and then maybe some proportion of people ultimately receive the email. And so you have this big middleman cottage industry that's very expensive for sending emails with ActivityPub, you have one Creator that sends the whole article to someone else's inbox, and there isn't a middleman. I mean, there is, in the sense that there's a server somewhere, right? But it's, it's a much more streamlined approach, and that's an interesting new dynamic. And it comes with complexity. There's more onus on spam prevention, on the recipient, but it comes with a lot of dismantling old gatekeepers in terms of the distribution potential. And that's exciting. I like that a lot.
Mike McCue:
That is really exciting, man. I mean, when you look at what can be in that ActivityPub, post, so obviously you can have formatted text. Can you have images, video, how? How are you looking at those things?
John O'Nolan:
So activity was fairly agnostic. You can, in theory, shove almost anything into it. The question becomes, what is the client on the other side prepared to receive? Can they understand the formatting? Do they have some way of displaying it? And there's going to be a little bit of a push and pull there around. I can see this evolving a bit like how Twitter at mentions got invented. You know, people just started saying, hey at Mike, what do you think of this? There was no notification, there was no link to your username. It's just how people started using it. So I think as if we get platforms in the ActivityPub network that start innovating with content types and pushing it out, and it doesn't work in all clients, if they're popular enough, it might cause those things to be adopted, and those things might end up driving the standard and what is possible to display. So I think there's going to be some experimentation, some innovation, some things that do work, some things that don't work, and a little bit of boundary pushing around. How much can we include? How dynamic could this be? Because I want to receive an article with one of those cool New York Times dynamic charts where I can click and drag right, but that requires an object with JavaScript and other things. I don't think we can do that yet. But can we do that in future?
Mike McCue:
Maybe, yeah, you know that concept of pushing the boundaries, and, you know, doing some collaborations with other folks that are building readers and building clients. And, you know, I think that is a really important moment that we're in, you know, in the fediverse, it's, you know, I sometimes tell folks I feel like I'm on Team fed averse, and not just like the CEO and, you know, product person at Flipboard. I feel much more like, you know, I'm collaborating with other product builders to build the right overall experience and ecosystem for the world, right? And I'm one of those, one of those players, and you know the fact that you're building this, you know, much richer ActivityPub post, you know, I can tell you, like that, we will make sure that we can support that directly in Flipboard, right and and help further push those boundaries. Because I think that is exactly what's needed. Now that, you know, we have an opportunity, and it doesn't come along very often, to say, You know what, let's just kind of like, take a giant step back. Back and think about, how should the web work going forward? And that really is what we're talking about right now. You know, making RSS more of a two way relationship instead of a one way relationship, fundamentally improving discovery, improving the reading and consumption experience, these are all possible right now,
John O'Nolan:
I'm curious if there are any parallels, you know, from your early work in browsers and being a part of a set of web standards that came before this one. Does it feel very different? Does it feel the same? Does it feel it feels
Mike McCue:
it feels remarkably similar. John, it's it, is it? Oh, yeah, it is like, for example, when we were building the early versions of Netscape Navigator, one of the things that we wanted to do was make HTML more app be able to support applications, more sophisticated applications, right? So, for example, having a toolbar where you could, you know, click on buttons, or maybe drag the toolbar around, or drag, just drag and drop inside of HTML, right? So we needed to create an event model and actually make that part of HTML, right? So there was some work that we did, and some work that Microsoft did to basically build an event model, and those ideas were not part of any sort of W 3c you know, kind of, there was some, I there was some work that was going on W 3c but it wasn't, it wasn't really kind of at the point where it's, like, practically implementable. So we've just pushed the boundaries, right? And we did some stuff, Microsoft did some stuff. People did complain that, like, Wait, this isn't really part of the standard, you know, what's going on, right? But then we gradually coalesced around a standard, you know, together and move the world ahead. And that was, that's the kind of thing that I think you have to be able to do in order to advance things, because otherwise, we'll just stay in this ethereal, theoretical land all the time, right? And we won't, we won't move things forward.
John O'Nolan:
That's right. It's amazing to me that we have these two really clear, obvious examples of this working really well and really badly. You know, web browsers evolved and got better and got better and got better, and we slowly dropped the things that were holding everything back. Like IE six is obviously the classic example from that era, but email HTML the opposite path right today. It's still tables, it still doesn't support Web Fonts. It still has all these crazy limitations where you're still injecting CSS inline because the rendering engines of Outlook are based on Microsoft Word If I write correctly. And Google decided we're going to innovate with Chrome, but we're not going to innovate with Gmail. And so the entire what you can do with an email newsletter stagnated completely and never moved. And that's an interesting little historical artifact of just having these two touch points of ActivityPub could go either one of these ways, depending on whether we as products manage to push the spec forward based on what people want it to do, not necessarily just on what the technology is doing at its starting point.
Mike McCue:
Now, exactly, you know, I think, I think you need to have this kind of spirit of collaboration across multiple companies and people in the standards bodies and, you know, and also, critically, importantly, you know, creators and publishers and people who are actually, you know, building businesses on all this stuff and, and I think that, like, you know, I sometimes use the analogy of, like, you know, when you're writing an email and You've got it, you know, maybe some long emails, three, four paragraphs, and somehow your computer crashes. Now you got to go write it again. You write when you write it the second time, it's like, almost always like, way better, right? Because you kind of already thought it through, right? You know that? Yeah, and so. So I think that's kind of where we are with the web. We actually have this very rare chance for some of the people who actually built the web, who are were there in the early days, but then also a whole new generation of people as well, to kind of say, look, here's what we love about the web, and then here's the things that just aren't working. They haven't scaled. They're creating problems around business models that are leading to bad behaviors. And right? So let's take a step back and rethink this and and that that is, that's the thing that is so exciting to me, because it is very reminiscent of the days when I was at Netscape, because it feels like we can go anywhere. We have a blank sheet of paper here, right? And as long as we stay collaborative with other key people who are actively building in this space. I'm incredibly optimistic that we're going to actually be able to solve some of these big problems.
John O'Nolan:
The other thing that's cool to your point is no one has won this space, right? There isn't a Google, there isn't a Microsoft, there isn't a Yahoo. There. There's no giant 10,000 100,000, person company who has the definitive thing. It's all very we're at a very grassroots stage of mix of hackers and enthusiasts collaborating to make the thing they want, not the thing that has the most funding. And I love that that that speaks to my heart.
Mike McCue:
I know, I know it really is. It really you're so right about that. And you know, with by creating this ecosystem, we have an opportunity to prevent some of the bad things that can happen when you do have big players emerge onto the scene and create kind of the best practices, the best behaviors, so that, you know, even when really large companies start to play in this world, right, it's still actually a good thing overall for everyone.
John O'Nolan:
Yeah, yeah. I think a big piece of that is going to be keeping as much decentralized and having multiples as we can. You know, as soon as the search engines all narrowed to Google, the diversity of finding things on the web dramatically reduced, and by all accounts, because Google was very good at what it did for a long time, but we lost a difference of thought, and obviously now we have DuckDuckGo. And there's some interesting remember that new paid search engine? It's really cool, but I always forget the name of it. There are some alternatives again.
Mike McCue:
Oh, right, the new paid search engine, right. Yellow background. We'll put it in the show notes. I know exactly what you're talking about, and people have been, yeah, David Pierce was saying that he's switched over to this.
John O'Nolan:
That's right. Casey's been talking about, I tried it recently. It was amazing. I think that has legs. But, you know, there's this, they're small. All the alternatives are small. So I think in activity, probably in the same way the behaviors we I want to encourage is going to be to try and have as few monopolies as possible. That I think is meaningful. I'm an inverse Peter Thiel on this. I do want competition. I don't want monopolies.
Mike McCue:
Yeah, the decentralization is key, as you say, that is, that is how you make that happen, right? We have lots of instance owners. We have anyone can build an ActivityPub server, front end service, and that's huge. And that is the thing, I mean, that is the thing that helped make the web happen, right? You, you effectively decentralized all that stuff. The problem was for some of the core functionality that you still needed discovery and a social layer, for example, those because those things weren't baked into the web to begin with, then you ended up having those things effectively get centralized. And the thing that I think is such a big deal here is what you're the most important thing that we're decentralizing is the human connection, the human relationship, that fact that nobody actually owns that, in terms of companies don't own that, but also, inversely, the creators own the relationship with their own audience, Right? The people who own the relationship are the ones who should own that relationship. That is, like, epically huge for the future of the web to actually formalize that architecturally, right? And I think that when you do that business models, browsing, search, all these sort of ways we think about navigating and using the web today, those will be very different, you know, because it will be based on now all of these human connections and a whole myriad of different apps running on the same protocol. It just, it's just a completely different experience. And yeah, so I do feel like that by decentralizing that human connection, you're going to end up having it be much harder for these monolithic, centralized forces to really take hold.
John O'Nolan:
I mean, a really underappreciated thing so far, and I think this is going to change slowly at first, and then rapidly, is the scale of that decentralized human connection, because at the moment, I think people have this perspective of Mastodon has 10s of millions of users. That's a bunch. There's a few other platforms that have single digit millions, and then there's a long tail and 1000s. And people kind of maybe look at those at the moment and think it's just not a very big network. But the underappreciated thing is that's going to just continue to grow, because it's not one company that has the network. It's every single company on top of one common network. So whereas in the past, you might have, you know, built up a million subscribers on YouTube, but then, oh, the algorithm changed, I don't get any views anymore. Now I'm going to switch to doing video on TikTok. You start from 00, you build up from scratch, ActivityPub, it's all there. So if you move between them, you're all your subscribers that they're still the same. There's still the same people with access to the same content.
Mike McCue:
So you're just posting from different apps, right? And growing the same subscriber and audience base from different apps.
John O'Nolan:
And so now, today, it's commonly known in marketing, but the most valuable thing you can have is an email address, because an email address is forever, right? No matter what is platform agnostic, I think ActivityPubs gonna head the same way. And, you know, one of the stats we use in our landing pages, there's four and a half billion people using email. That's more than any social network in the world. It's more than the top three combined. I think ActivityPubs going that way. You know, it's going in the in the direction of email. It's going to outgrow, I think every single social network out there eventually. And the reason I, I feel so certain about that is because it doesn't depend on one company. It doesn't depend on you doing great stuff with Flipboard. It doesn't depend on us doing great stuff with Ghost it depends on everyone, and that de risks the chances of success dramatically,
Mike McCue:
Exactly. Yeah, this concept of, you know, network effects, where, when people join a network, that network becomes more valuable, right? And, but what's happening is it's not just one of many networks, it's a overall network that people are joining, which then what's also interesting is that as companies federate like Flipboard is federating our whole user base, all of our curators, so they'll be joining this network. You have entire networks joining network, this big network, right? So, so you're right, it doesn't take very many of those before you end up having, you know, something that's absolutely massive.
John O'Nolan:
That's right. This is another example of maybe something people don't completely realize yet, is Ghost has hundreds of 1000s of sites, but all of those sites, many of them, have millions of existing subscribers, so the scale of bringing users to the ActivityPub network is far larger than the number of Ghost sites. It's the number of communities within each Ghost site. And the more we can connect the different humans within each of these sites, whether they're the authors and the publishers or the readers and the commenters, the the faster that that whole situation grows, which you know, I think perhaps we're a little bit guilty. I know I am of I feel like I can see where this is going. I'm excited about it, but we're not there yet. And I don't know about you guys, but we're not that close yet. We're kind of just getting started, so at risk of coming off a little bit too optimistic, I know there's a lot of work between where we are now and getting to the, you know, this utopian vision of what we're talking about. But if we can get there, the potential is very special.
Mike McCue:
Yeah, yeah, I agree, yeah. It's interesting too, because you know, when you imagine a world where all these different apps are connected by common social layer, right now, what becomes really important, and I think will be very healthy, is that people have to specialize. Companies have to specialize. When you're building on ActivityPub, right? Because what you don't have to do anymore is build the entire stack, like what we see today happening, where everyone's just becoming another version of Tiktok. They have their own social graph, and then they're doing vertical video. Everyone becomes versions of themselves, you know? And you just have these like impenetrable vertical stacks that aren't compatible with each other. And they're it's all, it's all a mutually exclusive, you know, either these guys win or those guys win, kind of model versus a scenario where you have some a company, a product team, focus on what they are particularly good at, right? The stuff that they are uniquely good at. They don't have to build a follow graph, they don't have to build a whole social layer, and they don't have to build a Tiktok thing just to be competitive with other social platforms, right? They they can focus on what they are particularly good at, and that's a whole other way of thinking. I'm curious, when you think about Ghost, what are the things that are unique to what you're building? You know, what? What is it that's special specifically about your approach with Ghost independent of ActivityPub.
John O'Nolan:
Yeah, I mean, there's a lot and both the model of the product as well as the model of the business the products we created at a time when the dominant open source publishing platform WordPress was starting to move more. Towards this website builder space competing with Squarespace, and we felt like it was important for the web, important for journalism, to have a platform that was open source, decentralized, freely available to everyone, to level the playing field of technology for what was available to publish online with that was purely focused on publishing, not on website building, not on real estate websites, but just the best possible publishing platform. And that's what we have been on a mission for 10-11, years to create. And, you know, some people say we've done well, other people not as much, but that's what we care the most about, and that holds a very special place in my heart. I think to have journalism with integrity, you have to have technology with integrity. And in my mind, open source is the way to have technology with integrity. And I want the best journalism to win because it's the best journalism, not because they have the best platform. And I think, like BuzzFeed, especially early BuzzFeed, was a great example of the best technology winning and Upworthy. You know, that was the best technology winning. They figured out how to crack the algorithms the content, aside from Buzzfeed News, which was incredible at one point, was not good. It captured hearts and minds, not necessarily in a good way, and it was because their technology was great. So I would like for local a local newsroom, to have the same quality of technology as the New York Times, and for them to be on the same playing field. If we can do that, we've gotten somewhere, and that's different to the majority of our competitors, and how they're thinking about it. The other thing that's a little bit unique and weird about us is everything I just said has been wrapped in what I've various points referred to as a thought experiment, which is, what if you were to build a company that does the exact opposite of every other company? And you know, my frustration with technology over the years has always been it's the same story. Every time someone has a great product idea and they raise a bunch of money, they say they're going to change the world, and then at some point, you point they sell it, it shuts down, it goes public, it screws over its users, it sells data. It's just the some ulterior motive that is predicated on creating extreme wealth. And so my thought experiment in creating Ghost was, what if you had a technology company that does what it says it's going to do at the expense of everything else. So the Ghost is set up as a not for profit organization, primarily to ensure that it will always and forever be independent.
Mike McCue:
Ghost is all about providing an open publishing platform that publishers, big and small, can count on forever
John O'Nolan:
True independence. Yeah, true. That cannot be taken away from you, that cannot be subverted. And you know, one of the critical things we've seen the degradation of platforms over time, right, especially as the internet has matured, and at various points, we've seen some that were good suddenly turn very bad. And there's recent examples, and there's less recent examples. The critical thing with open source platforms is, if that happens to an open source platform, you can say, Nope, not following, not listening to you guys anymore, you've lost the plot. We're taking all your code, we're going over here, and we're going to continue the way we want to continue. And as a community of users, that can be decided for zero cost, for zero legal liability, you just rename it and take it in your own direction. And that is interesting. One, because it's possible, and that's a good thing, but two, because it incentivizes the original platform not to be bad in the first place. Because, you know, if you do something bad, your users can just leave. They can leave, they can go out the door. You don't have a moat. You don't have a monopoly.
Mike McCue:
And I think technology, moatless technology, moatless technology, I
John O'Nolan:
think, is something we should aspire to fill the canals with grass.
Mike McCue:
Exactly. Let people out, let them in or out. And you know, do you, when you look at publishers, another thing that you seem to have done a really nice job is enabling them to have kind of a payment model, a business model around their their content creation. So can you explain a little bit about you know, what you do now and how you see that evolving a bit in the future?
John O'Nolan:
Yeah, absolutely. Um, I mean, the simplest way I think most people have heard of Substack at this point is we enable the same business model that Substack has made very popular within the publishing space. We started on it, actually before Substack existed, and the idea was to bring the software as a service business model where people are pretty comfortable paying a monthly subscription for something that they really care about and value, bring that to creators, to publishers. We have quite a strong focus on news and journalism, but we also have a lot of new media creators, podcasters. As YouTubers, also a lot of business journalism, lot of companies use Ghost power their blogs. And I think aligning the incentives is so is so critical. We've we're in this the end, it feels like of this era of clickbait journalism, where publishers used to make money from newspapers. That money went away, then they had to make money from web advertising. Google and Facebook were taking all of it so they needed more traffic, so they started writing headlines that would get more traffic, so that they could get more views on the ad, so that they could just stay alive and not do yet another round of layoffs. And it did not work, and it did not work, and it did not work, and it led to the polarization of a lot of things, it led to the lowest level of trust in news that we've had since records began, as far as I know, and the and the economic incentives of who are you publishing for? Is it to get traffic for advertisers, or is it to serve your audience? Had a lot to do with that. I think it had a huge amount to do with that. And so what this business model of simple membership subscriptions unlocks is for publishers, particularly news publishers, but really everyone, to treat their audience in the same way that Netflix treats its audience, in the same way that Spotify treats its audience. It has to build a product they want. It has to create content that they want. And if they don't, if they're not impressed and they don't think the quality is high enough, they will press cancel, and that's when revenue goes down. So it incentivizes publishers to create content that's aligned with their audience, that their audience needs, that their audience wants, that their audience cares about, and typically that content is much higher quality. Go figure, you can produce less of it. You can spend more time on it. You usually people who will pay for engagement with your content usually have much more respectful engagement with it, for want of a better description. And we've got so many publishers now that have proved that this is a really viable model for, you know, creators, as we call them now, publishers, as we've called them in the past. Ben Thompson was one of the early movers with Stratechery, an analyst in the Tech Tech News space. And that model is repeating and through Ghost and through platforms like it. I think it, it offers a really promising future for the economics of creating content online and having, you know the Kevin Kelly, 1000 true fans, if you can find 1000 people who are willing to pay for your work in some capacity, who really, really love it. It can be sustainable, and it doesn't take us. It doesn't it doesn't take as much as you would think to make that possible. You need a passionate audience that really cares. You know, you can't just be writing random news or random posts. You have to be on a topic that people are like, I want this, and I can't easily get it somewhere else. If you fulfill that, it solves the economics in a really meaningful way, in a really sustainable way. That gives me a lot of hope for the future.
Mike McCue:
That is really fantastic. The the concept of having a a subscription button that someone can press, or there's, there's follow, but then there's also subscribe, right? So have you thought about federating the subscription button, right? So like, if I, if I'm in an ActivityPub, experience some, some, maybe it's Flipboard or some other product. And I'm looking at, you know, a post on, would I be able to say, Okay, let me buy, let me subscribe right here. I don't have to go to some other place to do it. I can just subscribe right here, and then it unlocks and I just see the rest of it.
John O'Nolan:
Yeah. I think, I think we can do that where I might, you know, someone might take this recording and years from now, you know, that sort of remind me, and whatever we're like, look what you said, completely wrong. But so far, we have multiple ideas for how we can make this possible, and we think it should be doable. And in short, you know, it'll be an encryption handshake between the client server. Is this person authorized to receive the full post or part of the post, and then an API response to either deliver the all of the content or a truncated piece of it. It's very exciting. I think as far as we can tell, these technologies are compatible, and that offers a lot of interesting additional possibilities. One of the reasons people use Ghost over, perhaps some of the centralized platforms, is when they're trying to create a media brand. And a media brand is about brand, right? You you want to be the New York Times. You want to be the Economist. You don't want to be columnist on platform name, someone on Tumblr, someone on Reddit, someone on sub stack. And a problem with things like sub stack is you have to down your readers. Have to download the sub stack app, and in that sub stack app. App, or is every other publisher, and what we find is independent media brands, they want either their own app or they want something that isn't going to dominate and take away from their brand. So I could see Ghost publishers telling their readers go download Flipboard. You want to read all your paid content in a beautiful UI, in a nice reader experience on a iPhone or a tablet. Use this reader app. It's fantastic. Equally, the publishers themselves might have be able to log into Ghost and do that, but it opens up. You could consume paid content in multiple different places. Not only, you don't have to have it all in your email inbox. You know, if you don't want it all there, you could get it out.
Mike McCue:
This is a problem I think social media has in general as well, which is this over rotation around a for you feed or a following feed, and then that's the only way you consume and the there's many problems with that model, but one of the one of those problems is that it takes away the sort of deep dive intentional use of social media for learning about a thing deeply, right? You because you're just for the nature of following, and for you is they're totally random, right? Just like my email inbox, right? And so like, what I want to be able to do is say, All right, I really want to go big on learning about, you know, I don't know how to sail. And now I want the best content, the best writers, the best videos, the best podcasts, all in one place, and then deep dive on that, versus this sort of snacking use case that most social media exists in today is completely random. It's kind of mindless. There's no intention behind it. You just absorb whatever comes to you, right?
John O'Nolan:
I think you know the old saying, first we make our tools, and first we shape our tools, and then our tools shape us. I think there's a there's a parallel here, where you know first that we shape our algorithms around our content, but eventually we shape content around algorithms. And so what happens so often with decentralized networks is there's a lot of creativity at the beginning, and then this, this concentration to a baseline average, when Mr beast decides what you do is thumbnails like this, and then suddenly everyone's got a thumbnail like that, and then the titles have to be caps. And now suddenly all titles are caps. There's this deviation towards the mean, where everything ends up looking the same, because people are trying to be successful, and in order to do so, they are optimizing to the algorithm, whether it's Google or TikTok or YouTube.
Mike McCue:
And then meanwhile, the algorithm changes, and it moves somewhere else, right? And then they got to all change well, and I guess now we have to do YouTube shorts. All right? What's, how are we going to make that happen? Right?
John O'Nolan:
And so what I hope we get to with ActivityPub is a is a diversity of algorithm where there's not one dominant one so that, so that people can be more creative. They don't have to perform for an algorithm to succeed. They can just do whatever it is they want to do and succeed, because that's it's possible to win that way.
Mike McCue:
I think we're going to get there. You know, I think what Bluesky has already done, really has done some, I think, very promising and innovative work on the concept of custom feeds, those concepts, I think you're going to see come this year, into the into the fediverse. And I think that's going to be really exciting, so that we aren't just all hinged to a following or for you feed, a one size fits all feed, yeah, and I think you know, coming back to, you know, this idea of federating, the subscription to a creator on Ghost, that is another really big enabler, right? Because, again, you're what you're really doing is you're trying to put the best content in front of its natural audience, right? And when people can appreciate that great content, the first thing they want to do is get more of it, and there's far more likely, because they're in that intentional space right, that they're want to pay right there. And one of the problems that I see, you know, it's been, been a big problem on the internet for a long time, is like, okay, great. Go make a subscript, go get a subscription, sign up, and think about how hard that is, right, and how many steps you have to go through just to make a subscription work, right? And you got to create a whole other account with a different password, and, like, figure out what Patreon is exactly. It's just, yeah, I, you know, I support a number of creators on Patreon, but I'm not going to go to patreon.com to go see what those creators are posting. And, like, a completely different place I got to go to now, right? That should just be federated. And so I really am excited about your thinking here around kind of federating that subscription model. And I'll tell you that is definitely something that I'll work with you on in Flipboard to enable inside the reading experience seamlessly, right? You see a great piece of content. Yeah, that's something you really care about and make it. Incredibly simple to just buy it right there and see the rest of the content right there. You don't lose your place. It's just a simple, easy experience, Dude, we got to make that happen.
John O'Nolan:
We might have to get stripe involved in that conversation as well, because I have a feeling that's going to be key. But yes, I think that would be amazing.
Mike McCue:
Have you looked at other types of micro payments or other kinds of models for supporting creators, any experimentation or thoughts on that front?
John O'Nolan:
We're doing quite a lot with different payment methods. So in the US, obviously, credit cards the primary payment method, and buy now, pay later services are also getting quite popular. But outside of the US, payment methods are vastly different across Europe, South America, and fortunately, Stripe allows a lot of different relatively easy experimentation to have payment methods that are more suitable for international audiences. So a lot there less in the specific models. I think micropayments are something that has been tried more times it is failed 100% of the times it's been tried. And it has been tried more times than anyone can count. And the idea that it's going to work is always the same, but the reality is also always the same, which is that it's not meaningful enough to work out, and so I'm not super optimistic on micropayments. I am willing to be proven wrong at some point as someone comes up with a great solution. I'm interested in other types of models, though, where you can get communities collaborating, or maybe have having non monetary based contributions that allow special access. I think that's interesting. Obviously, there's there's crypto is unsolved and interesting, which I think suffers is in a similar place that they haven't got out of. You know, there's this early, chaotic era of open technology, which is it's exciting, but no one really understands what to do with it, and some technologies transcend that and move out of it. And crypto has not yet, and I'm still somewhat optimistic that it will find mainstream, easy adoption, where people really don't need to know what a blockchain is. But we haven't. We're not quite there yet, but I'm still optimistic that that might lead us somewhere interesting.
Mike McCue:
Have you thought about bundling models, or other kinds of ways for creators to, kind of like, pool together and maybe have some sort of collective or yeah,
John O'Nolan:
This is the the, one of the most common questions we get was we call people? Call people. Call it subscription fatigue, right? It's this theory that's once everything's a subscription. God, you don't want any more subscriptions. We have thought about it, but what we have seen play out so far is it doesn't really seem to happen ever, or at least not in the space that we operate in. And the reason for that is that it's not really 50 people subscribing all to the same three publications and or same 10 publications. It's 10 publications and 50 people, and each one of them subscribes to something different. And you might have you know, these handful of really popular people that everyone subscribes to, but broadly, the way we see people interacting with subscriptions is, it's far more niche. They're all subscribed to different stuff. There's not necessarily a giant overlap in, you know, it's not quite like having an Apple TV and a Netflix and a prime and A whatever else subscription. It's kind of just these three individuals that are my weird hobbies that, no, I'm not forgetting that content anywhere else. I can only get it from this person. It's a long tail of that will that eventually, could that eventually get to a size where there is subscription fatigue, even in the within creators? I think it's possible. And at that point, bundling gets interesting. I don't think we're there yet.
Mike McCue:
Well, I really appreciate how deeply you're thinking about this stuff as a product builder and as a CEO with with some phenomenal publishers that are doing incredibly important work on your platform. It's, it's easy to me to see why ActivityPub is such a no brainer for you, right? I mean, the whole mission and ethos of Ghost being an open platform to advance the world of publishing now embracing an open protocol to enable those publishers to have direct and, you know, audience relationships that they can own forever as well. I mean, that is, I mean, that's just poetically, you know, perfect. I mean it really is. I mean, really is the right exact combo for for, you know, where you're headed,
John O'Nolan:
Yeah. I mean, thank you, thank you for the kind words I'm. We're going to try, we're going to do our best. I feel, I feel quite a lot of responsibility to do a good job for it, and quite a lot of uncertainty about, you know, the details of how exactly we're going to execute everything. You know, there's, there's a lot of people have been in the activity space for a long time, and we are going to depend a lot on how well we're able to lean on the spot and learn from a lot of the people who've already been in the space for a while, who've already built incredible tooling and infrastructure that we have the benefit of now being able to take advantage of, because we're coming in a couple of years into a lot of the early movements. But the fit is is perfect for what we're what we want to do, and yeah, I really hope we can do a good job of it. We're going to do our best, that's for sure.
Well, thanks so much for listening.
Well, thanks so much for listening! You can find John at https://john.onolan.org/
You can follow Mike on Mastodon at @mike@flipboard.social and @mike@flipboard.com
Big thank you to our editor, Rosana Caban and Anh (Anne) Le (Lay).
To learn more about what Flipboard's doing in the Fediverse, sign up via the link in this episode’s description.
Until next time, see you in the fediverse!