Lukas Rosenstock's Blog

Lukas Rosenstock's Blog

Meta, the company behind Facebook, Instagram, and WhatsApp, is about to launch a new social network. Its working title is “Project 92” (it used to be “Project Barcelona”, and may be named “Threads”), and they position it as a direct Twitter competitor. In other words, it’s a heavily text-focused network compared to the visually focused Instagram and the general-purpose Facebook. The other exciting aspect of this project is that they’ve now confirmed - on a screenshot shared by Verge - that it will support ActivityPub, the standard used by, among others, the federated Mastodon network.

I’ve observed people on Mastodon and the fediverse having mixed feelings about it, but the negative prevails. Too many fear that Meta will do an EEE, an embrace, extend, extinguish move with that project. An EEE is when a product first embraces an open standard and extends it with further functionality not fully compatible with the standard. Then, they eventually extinguish the standard by achieving market dominance. Some Mastodon instance admins have already announced that they plan to “defederate” Meta, meaning they won’t allow the free flow of content between their instance and Meta.

I want to share why I believe Facebook is starting this project and supporting ActivityPub and why I think it’s unlikely they can kill the fediverse with this move. On the contrary, they could boost the federated network. Either I’m right, or in some time, you can pull out this post and show me how overly naive I’ve been.

Through its ownership of WhatsApp, Instagram, and Facebook, Meta dominates the messaging and social networking market. Some activists and politicians have criticized WhatsApp and Instagram acquisitions and asked for a rollback. Meta must play nice in this climate and show the market they’re not too dominant or anti-competitive. What could be better than supporting an open standard and decentralization?

Twitter never had the mainstream appeal of other networks, but it has a critical cultural impact because it attracts influential people. It’s also escaping Meta’s sphere. They couldn’t purchase it earlier, and now it’s sure that an acquisition would not meet the regulators’ approval. However, Twitter’s ownership did change when Elon Musk bought the network, and his erratic leadership in the first weeks and months made people lose confidence. It’s been a good time for Twitter’s competitors, which are primarily decentralized networks, the ActivityPub-enabled fediverse (with Mastodon and others), and the newer ATProtocol-enabled BlueSky network.

In the current social networking ecosystem, the following is true: It’s doubtful that a centralized Twitter competitor will emerge. People will stick to the original unless they have good reasons to switch. And decentralization could be one of those reasons. The fediverse has already moved beyond being a nerds’ toy, with larger organizations setting up their instances in the network to gain some independence. Hence, if Twitter fails somehow, the replacement wouldn’t be a single service but a network of services. The network comprises Mastodon (and Friendica, Pixelfed, Misskey, etc.) instances, individual WordPress sites, and some centralized services like Tumblr. If Meta can’t wholly own this substitute, they can at least be part of it, which is better than nothing.

Meta has a competitive advantage with “Project 92” due to the integration with Instagram, which no third party can replace, as the company might restrict the necessary API access. They are counting on users with an active Instagram presence to move to their new text-based social network instead of joining or creating, e.g., a Mastodon instance. These users get full access to the fediverse and their existing Instagram followers, a win-win situation for them and a lock-in effect for Meta.

It’s tough to convince users who are not active on Instagram or are critical of the whole company to join the new network. Thus, it’s still better for Meta to get some access to these users through federation instead of none at all. And, honestly, they probably still have their grasp on them through WhatsApp.

If “Project 92” doesn’t work out because users are not interested or it’s not making the company any money, they might shut it down. However, what could they do if it’s mildly successful, but they want more? They could shut down ActivityPub, but I don’t see why they would do it. They’d degrade the experience for their users who followed federated users and give them a reason to leave their app and join a Mastodon instance. They could subvert the protocol subtly to make remote users click through to content on Meta servers, where they can show ads or convince them to switch. Again, they rely on the goodwill of fediverse admins, who have already demonstrated they might defederate at any moment if they see foul play.

Many services like Tumblr have announced that they’re working on ActivityPub integrations but have yet to launch them. When Meta launches, those services get another push to ship support, which will take little time because they’re already working on it. ActivityPub is the future. Meta will stand against an entire unified ecosystem that may be stronger than them and has the Meta-critical public on their side.

Fediverse admins should be careful, of course, and they should not let Meta get away with sleazy moves. However, they should not defederate prematurely. Of course, they can do that if they want an alternative, corporate-free ecosystem for a small niche. However, this is an opportunity if they believe that the fediverse can become a mainstream social network and that decentralization isn’t just for nerds but everyone. They could be the ones to onboard the people who aren’t very active on social media, don’t necessarily want to use a Meta-owned app, but want to follow some of the influencers on that major network.