Can Twitter Ever Be Decentralized?

Jack Dorsey’s plan is long on ambition and short on detail.

Photo illustration of decentralized Twitter networks
Photo illustration by Slate. Image by Twitter.

On Dec. 11, Twitter co-founder and CEO Jack Dorsey announced Bluesky, a project to develop an open and decentralized standard for Twitter and the social media landscape. Presently, social media users can only interact with other members inside the same platform. Dorsey suggests his new standard would allow communication to flow across social networks, similar to how email works. Beyond that, though, Dorsey didn’t say exactly what any of this means.

If Twitter really wants to decentralize, it can tap into an existent ecosystem. A range of services for decentralized social media already link together in the Fediverse—a collection of interoperable social networks that use open standards and protocols to talk to one another. But instead of integrating with current solutions, Dorsey wants to spend years developing a new standard. There is good reason to be skeptical.

Twitter presently offers a centralized platform that generates profit by monetizing surveillance of its users for targeted advertising. Decentralized social networks, by contrast, are designed to minimize data collection, cut out advertising, and democratize ownership and control of social networks. The Mastodon social network, which has more than 2 million users, provides a glimpse of how that works. It closely resembles Twitter, except that users can host their own mini social networks on a server, called an “instance,” that can talk to users on other Mastodon instances.

Whoever creates the instance sets their own code of conduct. For example, a person may create a Mastodon instance dedicated to puppies. Another instance may be more general-purpose, but prohibit hateful speech and graphic violence. Users can talk to people in other instances, but if they violate the rules, instance moderators can block them or kick them out.

Mastodon not only gives communities authority over their own spaces; it also provides valuable privacy options. Instances can be made private and restricted to, say, a network of schools, and users can sign up without disclosing their real names or linking their accounts to personally identifiable information. Server administrators can see user correspondence on the network, so trust in admins is critical to users. There is also no advertising—targeted or otherwise.

Because Mastodon is decentralized, there is no centrally controlled instance, and it interoperates with other social networks and websites across the Fediverse: Mastodon and Pleroma for social networking, PeerTube for video sharing, and PixelFed for social image sharing (akin to Instagram). Users can thus post content like videos and interact across websites and services without having to separately log into each service.

By contrast, big social networks like Facebook and Twitter are built to keep users locked into their own corporate universe. Twitter does not talk to Facebook, which does not talk to YouTube, and so on, because those companies choose not to allow their users to talk to members of other networks. Imagine, for example, using Gmail, but only being allowed to email other Gmail users.

Decentralized services, by comparison, are designed to shatter the social media world into a fragmented landscape of community-owned and -controlled networks. They rely on open protocols—a set of rules and procedures for transmitting data—to facilitate interactions across services. ActivityPub is a popular protocol for decentralized social networks, but others, such as Matrix, are used for things like chatrooms and chat apps.

This is where Twitter’s announcement becomes interesting. Decentralized social media offers thousands of connected networks designed to provide an ad-free, privacy-respecting experience. If users discover a network is abusing their data, they can move to another service. Big social media services can get away with spying on their users because everyone’s friends are locked into their centralized networks.

It’s difficult to imagine what Dorsey envisions for “decentralized Twitter.” The very nature of a truly decentralized Twitter would undercut the company’s business model. He briefly mentions the possibility of integrating with ActivityPub, but does not say whether he would stop serving ads and monetizing data, like the others across the network.

Decentralized social networks date back to the Diaspora social network. Founded by four NYU students in 2010, Diaspora was inspired by a speech delivered by Columbia law professor Eben Moglen, who proposed the creation of a personal cloud server called a “freedom box” to help re-decentralize the internet. The FreedomBox project was launched shortly thereafter. Crucially, it provides infrastructure for decentralized services like social networking, email, chat apps, and calendars. Users can send their data through trusted avenues to friends in a decentralized or peer-to-peer fashion. This cuts out surveillance intermediaries like Facebook or Twitter, which sit in the middle and spy on people while they use their platforms.

The Diaspora project made an initial splash but eventually lost steam. Other technologists took up the mission in the ensuing years, with Mastodon being the largest success story to date. Standards and protocols like OStatus, Matrix, pump.io, and ActivityPub were developed to power federated services. Presently, these projects are mostly hosted on community-controlled servers rather than the personal cloud devices Moglen envisioned. However, FreedomBox is working on Fediverse integration, which would help further drive decentralized server hosting to the edge users of the internet, and similar software called FreedomBone offers apps for services like PeerTube and Friendica.

Twitter could assimilate into the Fediverse or work on solutions even more decentralized and privacy-focused than what already exists. But it seems that it won’t—presumably because doing so could destroy its business.

Dorsey’s true motivation may be to fend off regulatory action through self-regulation. For example, legislators are considering revisions to Section 230 of the Communications Decency Act, which limits intermediary liability for harmful content posted by users. A new standard that places content moderation into the hands of end-users, small communities, and recommendation algorithms could alleviate Twitter’s responsibility for defamation, hateful speech, and sensational content.

It could also be that Dorsey wants to appear hip to “radical new solutions” in times of growing discontent with social media surveillance and manipulation. For human rights advocates, decentralizing social networks is a noble goal, but difficult to scale up due to the entrenched dominance of platforms like Facebook and Twitter. It seems unlikely that Twitter would make this change without being forced to by the government.

In the meantime, Dorsey hasn’t addressed the elephant in the room: Can social networking giants generate huge profits in a decentralized ecosystem based on thousands of interoperable networks? And how can Twitter be trusted when its source of revenue is surveillance, consumer lock-in, and advertising? If Dorsey wants his proposal taken seriously, he needs to start answering these questions.

Future Tense is a partnership of Slate, New America, and Arizona State University that examines emerging technologies, public policy, and society.