Can ActivityPub save the internet?
The hottest new thing in social isn’t vertical video, and it’s not AI-driven algorithmic feeds. Instead, it’s a little-known, years-old protocol called ActivityPub that could help rewire the entire social fabric of the internet.
In recent months, a number of tech companies have thrown their resources into ActivityPub and what’s now known as “the Fediverse.” Tumblr is working with ActivityPub, as are Flipboard, Medium, Mozilla, and even Meta. There’s now an official WordPress plug-in for ActivityPub, which will enable the protocol for something like half the internet all at once. Developers are using ActivityPub to build new and different takes on YouTube, Instagram, and much more. ActivityPub is everywhere! ActivityPub!
And, of course, there’s Mastodon, the ActivityPub–powered platform that has become a haven to Twitter Quitters all over the internet. But ask around the tech industry, and there’s a growing set of people who will tell you the future isn’t Mastodon but what it represents: a scaled ActivityPub-based social platform.
So what is ActivityPub? It’s a technology through which social networks can be made interoperable, connecting everything to a single social graph and content-sharing system. It’s an old standard based on even older ideas about a fundamentally different structure for social networking, one that’s much more like email or old-school web chat than any of the platforms we use now. It’s governed by open protocols, not closed platforms. It aims to give control back to users and to make sure that the social web is bigger than any single company.
ActivityPub is not a perfect protocol, and there’s a lot of work left to do to improve it. There’s also a lot that could go wrong and a lot of ways for its potential to be snuffed out by corporate interests or bad technology. And there will be plenty of competition in the race to reinvent social media: social upstarts like Artifact and Substack Notes are building their own closed platforms, and Bluesky, Farcaster, Nostr, and others are building their own open protocols that also aim to decentralize social networking entirely.
But the people who have been working on the web for decades, who have seen the energy around decentralization come and go so many times, claim it’s going to be different this time. “I think this year could be the breakout year for the Fediverse,” says Steve Teixeira, Mozilla’s chief product officer. “It certainly stormed in: I’ve had my Mastodon account since like 2017, and I hardly used it until last year.” Mike McCue, the CEO of Flipboard, echoes the sentiment: “I was there in the early days of the web, and this whole thing with ActivityPub is as big a deal as HTML was back then. This is the single biggest opportunity I’ve seen for the web since the dawn of the web.”
For most of the last 15 years, the social web has felt like a settled market. Facebook and Instagram won, Reddit and Snapchat were around, and everything was shifting toward algorithmic entertainment anyway. TikTok’s explosion changed the landscape, but then everything turned into TikTok anyway. If you want to use the internet to keep up with your friends and interests, you’ve been stuck inside the walled gardens of closed social platforms for a long time.
For most of the last 15 years, the social web has felt like a settled market — and then Elon Musk bought Twitter
And then Elon Musk bought Twitter. For years, it had been kind of a mess but chugging along — in many ways, the default answer to lots of questions about where to quickly reach an audience. Musk thought he could save Twitter, but it turns out he may have saved the idea of an open social internet instead. When Musk spent $44 billion to acquire Twitter and then systematically destroyed everything people loved about the platform, users went looking for something better. Seeing demand in the market, developers set out to build products to fill it.
Before we go too far, it’s helpful to understand what this vision for a better future of social actually is. “Decentralized social networking” is a heady concept, and it’s quite different from the way the internet works now. But here’s the simplest way I can think to explain it: to decentralize social networking is to completely separate the user interface from the underlying data. Any time you sign up for a new social app, you won’t have to rebuild your audience or re-find all your friends; your whole following and followers list come with you. Those things should be part of the internet, not part of an app.
Email is the best example of how this system works now: it’s based on open protocols that lots of services tap into, so while there are many email apps with different features and quality levels, your contacts carry over and will always work. (Can you imagine if you needed an Outlook address for your Outlook-using colleagues and a Gmail address for your Gmail-using friends, and then a Hotmail account just to talk to your aunt Gertrude? Well, that’s currently how social works now.)
Facebook is an even more helpful counterexample. Your friends on Facebook are your Facebook friends. You can’t export the list to use it in another app or easily follow all those same people on a separate platform. If you want to read Facebook posts or create your own, you have to do it on Facebook. This is an excellent situation if you happen to be in charge of Facebook, and it’s how Facebook became a cash machine for nearly two decades. Platform lock-in has always been the most profitable strategy.
But if our current social system was decentralized, you’d be able to post a picture on Instagram and I could see it and comment on it in the Twitter app. Your friends could read your tweets in their TikTok app. I could exclusively use Tumblr, and you could read all my posts in Telegram. Different apps would have different strengths and weaknesses, different moderation policies and creator tools, but you’d have the same set of followers and follow the same accounts no matter which platform you use. There would be no such thing as “Facebook friends” and “Twitter followers.” The social graph and the product market would split completely.
“We had this vision of a more peer-to-peer internet,” says Christine Lemmer-Webber, one of the co-editors of the ActivityPub standard. “But at the very least, if a server goes down, it should not be catastrophic to you.” Your social world shouldn’t live inside an app, she says, or depend on a company staying solvent. It should, and could, be much bigger than that.
You really cannot overstate how old an idea all this is. ActivityPub has been a finalized standard since 2018, but its roots go back almost as far as the web itself. “I have spent more than 15 years working on distributed social network protocols!” says Evan Prodromou, another of the co-editors of the ActivityPub standard. For years, he ran an open-source project called StatusNet and its flagship product Identi.ca, which aimed to be… basically a decentralized Twitter. He says he’s thrilled to see these protocols finally taking off: “I think that there’s been a lot of opportunity in this space, and I think a lot of people are looking and seeing that opportunity.”
Over the years, Prodromou says, there has been a parade of protocols aiming to open up parts of social networking. OStatus; pump.io; Open Social; Pubsubhubbub; WebFinger; ActivityStreams; XMPP; RSS; OpenID. There are plenty more, and you’ve almost certainly never heard of most of them. Prodromou worked on a lot of them. At various times, even some of the biggest companies in tech have been behind the protocols. “Remember Google Buzz?” Prodromou asks. “That was compatible with OStatus, which was wonderful.” (Buzz launched and closed in under two years.)
Standards like this are typically overseen by the World Wide Web Consortium, better known as the W3C, which is essentially the web’s governing body. There have been “social on the web” groups working on this stuff for decades. In 2008, Marc Andreessen gushed about the launch of something called OpenSocial, writing that openness “will always swamp anyone’s attempt to wall off a proprietary world with tight controls and sharp limitations.” It… did not work out.
In July of 2014, a new group convened, known as the “Social Web Working Group,” that was explicitly tasked with figuring out federated social networking.
The group fought and debated for the next three and a half years. “At one point,” Lemmer-Webber says, “it looked like we were going to run out of time and ActivityPub wouldn’t happen.” But when the group finished up in February of 2018, it delivered a handful of new ideas back to the W3C, ActivityPub among them.
Right now, if you’ve encountered ActivityPub on the web, it’s almost certainly because you’ve used Mastodon. The app — essentially a federated Twitter clone — has been around since 2016 and has used ActivityPub as its primary protocol since 2017. “I’m a realist so I don’t think that it will be able to compete with Twitter,” founder Eugen Rochko said on the Hacker News forum when he first launched the service. He’d been inspired by TweetDeck, and he hoped he might be able to give people a choice of tools for socializing online.
But over the last few months, Mastodon has emerged as the clearest alternative to Twitter. “Social media websites come and go,” Rochko told us last month. “They die. Even if they stay around, like MySpace did, nobody’s talking about MySpace like it’s a relevant thing anymore. It is a graveyard, and I never believed that Twitter would be completely immune to that.”
“For now, at least, Mastodon is the flagship of the Fediverse,” says Anil Dash, the head of Glitch at Fastly and a long-time web executive. There are a few other growing players in the space, like Pixelfed (a decentralized Instagram) and PeerTube (a decentralized YouTube), but most of the action in the Fediverse is happening on Mastodon.
Exactly what that means is hard to tell: the platform recently passed 10 million registered accounts, but that’s a tricky number to match to real-world usage, and by some metrics, the first Twitter-exodus bump in Mastodon activity has simmered a bit. Even in the rosiest of measures, it’s still several orders of magnitude smaller than Twitter (under Musk, still in the hundreds of millions), and Facebook’s not even in the same universe (close to 3 billion).
Whatever the numbers, though, the vibes are good. “It just has it,” Dash says. “It has this legacy of 20 years of us blog-nerd-dads wishing for that internet. And it has a real problem to solve, which is: I’m Starbucks or I’m Nike, and I’m not fucking interested in Twitter anymore for my brand.” Mastodon may not be the future of social, but we might not get to the future without it. The fact that Mastodon was ready and able to provide a new home for users ready to leave Twitter was a big deal.
If you’re a company looking for a new place to hang your social shingle, you could do one of two things: set up a Mastodon account or build your own platform that integrates with ActivityPub and thus can interoperate with Mastodon. Many companies are choosing to do both.
Flipboard set up a Mastodon server at flipboard.social and began inviting some of its users and curators to post there, in addition to on the main platform. Medium did the same. “I wanted to launch early because it’s really important for Mastodon to grow,” says Medium CEO Tony Stubblebine. “Because it can’t all be vaporware — people have to show up and deliver and bring the users and make the whole ecosystem actually better.”
Flipboard is also beginning to support Mastodon inside of its app, thanks to ActivityPub. But that doesn’t just mean embedding posts on Flipboard pages, like the app used to do with Twitter before its API was cut off. If you like a Flipboard post that comes from ActivityPub, that like appears in the creator’s Mastodon app. If you comment, it shows up as a Mastodon reply. This is not embedded content; it’s actual interoperation. “If we create a future where a whole bunch of social networks are interoperable based on ActivityPub, it’s a great future,” Rochko says.
PeerTube is the Fediverse’s answer to YouTube. And as you can see, the Fediverse is kind of obsessed with itself.
Ultimately, the bet all these developers and companies are making is not just that Twitter will die and hundreds of millions of users will need a place to go. (Though they’re definitely betting on that.) It’s also that the possibility of a TikTok ban will open people’s eyes to how fragile these platforms are and how foolish it can be to give your audience and content to a platform that could simply disappear. And that after years of hearing about Instagram’s effect on young people, users might want a tool that gives them more control over the content they see and avoid. ActivityPub, they think, can enable all that.
“A year ago,” McCue says, “if you talked to people, you’d say, ‘Do you need an alternative to Twitter?’ And most people were like, ‘No.’” But after Musk took over and began to change the company’s moderation policies, fired most of the staff, and presided over what felt like a suddenly dying service, the answers changed. Flipboard, like many other companies, began to think about embracing a new era of social.
“A year ago, if you talked to people, you’d say, ‘Do you need an alternative to Twitter?’ And most people were like, ‘No.’”
Mozilla’s Teixeira sees it the same way. “For all its faults, Twitter had worked really hard to build a trustworthy ecosystem where you could tell, despite the challenges, they were trying to act in a trustworthy way.” Now? “With that whole ecosystem kind of under threat and eroding, there was a window of opportunity that we saw.”
But ActivityPub is a protocol, and you can’t download a protocol. For this new future to take hold, it needs a killer app. I asked almost everyone I spoke to for this story, Do you think Mastodon is going to win? Is it going to be the next big thing in social? Most said probably not. In part because, if ActivityPub really takes off, there might not need to be a next big thing at all. A couple of dominant platforms would be replaced by millions of interoperable smaller ones — and they argue that’s a better outcome anyway.
If you want a reason to bet against ActivityPub taking over the web and bringing us a better, more open future, I’ll give it to you in one word: money. The path to success for many social platforms has often gone the same way. A startup takes VC money, spends its way to massive growth, raises more money, spends to more growth, tries to corner the market and crowd out competitors, and eventually, maybe, makes some money back. (Even some successful platforms never hit that last step.)
The best and worst thing about the Fediverse is that its values don’t allow for most of that. “That’s kind of a no-go zone for me,” Rochko said on Decoder. “It’s a trap. It’s clearly against our project’s ideals.” Most people currently building in this space are doing it because they believe in it, not because it smells like billions. But over time, that scent usually wins.
There’s also the problem of the existing social networks, the giant thriving businesses with little incentive to play nice. You could certainly argue that the major social networks are all past their peak and that a new generation is all but guaranteed to come along. Many of the people working on ActivityPub argue just that. But right now, billions of users are still happily logging into Facebook, Twitter, Instagram, and the rest. Getting them to switch is hard, and getting the platforms to support a more open model might be impossible.
Even in the early days of ActivityPub, the protocol’s creators tried to get the big social networks on board with a more open social web. “We were actually told, ‘You’ve got to try to convince Facebook and Twitter, etc., to get interested in this,’” Lemmer-Webber says. “And we reached out to them, and… silence.”
The feedback they got, if they got any, was essentially, we’ve heard this story before. Federated social was an old idea, and it was never going to work or make any money. Those companies had little reason to play nicely with the open web, either, when all the money was in building all-encompassing ecosystems and walled gardens.
Right now, the best reason to bet against Mastodon specifically is because Mastodon is hard. Using it exposes the biggest problem with ActivityPub: the protocol doesn’t allow for any kind of universal system of identity. When you join Mastodon, you have to pick a server — which is a lot of work and requires a lot of conceptual understanding — and then find people to follow without any kind of cross-server directory. There’s no good tool for verification, either, so if you want to find NPR on the Fediverse, all you can really do is guess which account is the real one.
These are things ActivityPub will have to learn from protocols like Bluesky, which everyone agrees handles questions of identity and discoverability much better. “We’ve designed a protocol that has three big things we think are missing from the Mastodon ecosystem: account portability, global discoverability, [and] composable, customizable curation and moderation,” Bluesky CEO Jay Graber recently told The Verge.
Any Mastodon user will tell you that private messaging is clunky, too. And given how many Mastodon servers are just run by regular people running hardware in their homes, the bigger the network gets, the more brittle it will become. “If [Meta] were to go live and all of a sudden dump all of those users into the Fediverse as it exists,” McCue says, “with 30,000 privately run servers, usually by like one person, it could completely crush the Fediverse.”
Of all the things left for the Fediverse to figure out, content moderation will be the thorniest
But of all the things left for the Fediverse to figure out, content moderation will be the thorniest: it’s an expensive, complicated thing to get right, and without good content moderation, social platforms simply don’t work. In many ways, content moderation is the primary product of any social business, and decentralizing that work means only a few large players will be able to afford to do it well. Some people I talked to think it will become its own separate industry, available to lots of platforms at once; others think user-moderated small communities are the answer. Nobody knows how it’ll work at scale.
One possible future for ActivityPub and the Fediverse is that a single platform becomes the default, the way Gmail is not the only email provider but it’s the one you go to if you need a new address. Mozilla’s mozilla.social certainly aims to be something like that: “We want it to be true that if you have a Firefox account, you can use that to sign into our Mastodon instance,” Teixeira says. Others like it will surely crop up, too — even Meta is working on an ActivityPub-based tool, and the company knows more about the power of usernames than just about anyone.
But the advice you’ll hear from most people in this space is this: own your own domain. Don’t be john@mastodon.social or anna@facebook.com. Have a space that is yours, that belongs to you, a username and identity that can’t disappear just because a company goes out of business or sells to a megalomaniac. If it works, if ActivityPub becomes the underlying infrastructure of the social web, your identity becomes your identity for everything. It’s your YouTube channel name and your TikTok username and your Instagram handle and your phone number and your Twitter @, all in one name.
“If you solve identity with domain names, it makes things easier because it fits the way the web has been for 20 years,” says Manton Reece, who runs an ActivityPub-supporting microblogging platform called Micro.blog. “On the other hand, no one understands DNS, no one understands how to configure your domain name.” That’s why some people are so excited about WordPress’ support for ActivityPub — you might soon be able to turn your personal website into your entire social identity online. Some domain registrars are setting up similar tools, and some entirely new companies like Superlink are springing up to offer the same.
Most people in tech seem to think these are solvable problems and that the usability of the decentralized social web will get much better quickly. (Teixeira likes to talk about the “loadout” of the social web and what it will take to make things like verification and moderation easier for anyone to set up.)
The bigger question looming is more existential: can ActivityPub grow without getting lost? There’s a familiar pattern with protocols like this, known as “Embrace, Extend, Extinguish.” Step one: start an app on an open protocol, grow quickly because it’s easy to adopt. Step two: add new, platform-specific features, usually while complaining that the open protocol isn’t powerful enough to keep up. Step three: bail on the open protocol altogether, saying it simply didn’t serve your users’ needs anymore. Microsoft did it with the early internet; Google Talk did it to the open XMPP messaging standard.
In theory, a company like Meta could run that playbook here. It could launch Fedigram or Fedbook or whatever, gain a lot of users, and then slowly shut out the broader ecosystem. But most of the people I talked to are optimistic that won’t happen. In part, that’s because Mastodon is already big enough to have some sway in the ecosystem. But it’s also because of the nature of a protocol-based system, Dash says. “Podcasting is the analogy, right?” he says. “The end of every podcast ever is, you know, ‘subscribe to us on Spotify, iTunes, or your favorite app.’ And the reason people say it is because podcasting is a protocol, not a company. And because it’s an open protocol, you can take your podcasts and leave.”
There will be big players in the ActivityPub-led future, Dash says. Maybe not Facebook big, but big nonetheless. (And maybe Facebook big wasn’t such a good idea anyway.) The difference is that no one will be tied to a platform, and any platform that tries to tie users down will lose them instead.
After nearly two decades of fighting for this vision of the internet, the people who believed in federation feel like they’re finally going to win. The change they imagine still requires a lot of user education — and a lot of work to make this stuff work for users. But the fundamental shift, from platforms to protocols, appears to have momentum in a way it never has before. For decades, the open web has always run into endpoints: things were handed off to SMS or email or a third party like Facebook or Twitter. Thanks to standards like ActivityPub, Dash says, “now it’s the web all the way down.”
And no one, not even Elon Musk, can get in the way.
Comments
Post a Comment