The Nomad Who’s Exploding the Internet Into Pieces

Could decentralizing online life make it more compatible with human life?

A robot bartender named Carl serves guests at the Robot Bar and Lounge in eastern Germany. (Fabrizio Bensch / Reuters)

Dominic Tarr is a computer programmer who grew up on a remote farm in New Zealand. Down in the antipodes, isolation is even more isolating. Getting goods, people, and information to and from Australasia for families like Tarr’s has always been difficult. Bad, unreliable internet service is a particular challenge. Australia and New Zealand are first-world countries with third-world latency.

Today, Tarr lives on a sailboat—another Kiwi staple, alongside sheep and distance. Connectivity is worse on the boat than on the farm, and even less reliable. But that’s by design rather than by misfortune. Tarr started living on the boat after burning out at a previous job and discovering that the peripatetic lifestyle suited him. Unreliable and sporadic internet connectivity became an interesting engineering challenge. What if isolation and disconnection could actually be desirable conditions for a computer network?

He built something called Secure Scuttlebutt, or SSB. It’s a decentralized system for sending messages to a specific community, rather than the global internet. It works by word of mouth. Instead of posting to an online service like Facebook or Twitter, Scuttlebutt applications hold onto their data locally. When a user runs into a friend, the system automatically synchronizes its stored updates with them via local-network transfer—or even by USB stick. Then the friend does likewise, and word spreads, slowly and deliberately.

For the contemporary internet user, it sounds like a bizarre proposition. Why make communication slower, inefficient, and reliant on random interactions between other people? But Tarr and others building SSB applications think it might solve many of the problems of today’s internet, giving people better and more granular control of their lives online and off.

* * *

The term “Scuttlebutt” comes from the original water-cooler gossip. Named for a water cask (a butt) that had been cut (or scuttled), early 19th-century sailors would dish dirt while drawing from it. Being a sailor, Tarr adopted the name thanks to its nautical provenance, an apt description of its behavior. On first blush, that might sound no different from Twitter and Facebook, where gossip reigns. Isn’t the internet decentralized already, for that matter: a network of servers distributed all around the globe?

Sort of. It has always been concentrated in some ways and dispersed in others. The internet’s precursor, ARPANET, was designed to withstand nuclear catastrophe. Geographically distributed servers could communicate with one another absent a central hub, thanks to the communication protocol TCP/IP. The ARPANET’s infrastructure was decentralized, but that design served a central authority: U.S. national defense.

When the web entered public use in the 1990s, it offered a publishing platform without intermediation, as commercial services like AOL had done for online access. And it worked, for a time, while the network and its user base were small. But the web quickly became unmanageable. Keeping a popular server running became too expensive for ordinary folk and too complicated for non-technical people. And so business dissolved the internet into commercial product offerings. Today, that authority rests in the hands of a handful of big companies that run services used by billions of people.

And those billions do indeed gossip online. But the services they use embrace gossip’s content rather than its form. Facebook and Twitter are only like water coolers if there were one, giant, global water cooler for all workplaces everywhere. That sounds empowering at first—people anywhere can see and spread news and ideas from anyone. But those users are entirely reliant on the service operator. Outages, bans, lack of connectivity, or state suppression might get in the way. More often, companies like Twitter, Facebook, and Google change their services’ behavior or the terms of their usage—especially the way customer data is gathered, stored, and used.

Proponents of decentralized services (which are sometimes abbreviated as “decents”) hope to overcome some of these limitations by scattering the software and data that run online services closer to their ultimate points of use. Tarr’s Secure Scuttlebutt isn’t a social network like Twitter or Facebook, nor is it an email client like Gmail. Instead, it’s a platform for encrypted, automated, and local replication of information. Atop this information, new, decentralized versions of services like Twitter—or anything else—can be built.

The key to Scuttlebutt’s operation is a simple approach to copying information between computer systems—a tricky problem due to ever-changing files across many systems. Instead of separate documents and images and other files, like the ones a computer might synchronize via Dropbox, Scuttlebutt treats all data as chunks of content added to the end of a list—like a new entry in a diary. A cryptographic key validates each new entry in the diary, and connects it with its author. This is a bit like how the Bitcoin blockchain works—a list of linked records in a chain of transactions, verified by their cryptographic relationship to the last item in the chain.

But Scuttlebutt doesn’t carry monetary transactions; it carries a payload of, well, gossip content. As it happens, most popular online services are just lists with new content appended. Twitter and Facebook are like that. So are Instagram and Soundcloud. A magazine like The Atlantic could be understood as an append-only list of articles and videos. Even email is, at base, just a pile of content.

So far, the SSB community has made social-network, messaging, music-sharing, and source-control management software that communicate by Scuttlebutt. But unlike Dropbox, Facebook, or every other Cloud service, Scuttlebutt doesn’t synchronize information by connecting to a central server. Instead, it distributes that data to the subscribers a user happens across. There’s no one Scuttlebutt, but as many as there are users.

Scuttlebutt-driven systems synchronize with one another via local networks—say, when a boat docks at port or a mountaineer descends to base camp. For more deliberate sharing, users can connect to an SSB island on the internet called a pub—as in public house; a virtual tavern where gossip can be shared more rapidly. Since no central server is required, no internet access is required either; a local network or data saved to a USB stick and handed to another user are both sufficient. It’s like having a series of private internets that still work like the online services popular on the commercial internet.

* * *

Decentralized software often accompanies political extremes. Bitcoin, a decentralized currency, is associated with the libertarian right, who distrust government. The anarchist left comprises another breed of decent fans. Collective resources and mutual aid are of concern to this group, who hope to replace both market- and state-run services with structures like cooperatives.

Tarr admits that some of the programmers creating and using Secure Scuttlebutt fall squarely into the anarchist camp. But he is more broad-minded about the project’s political aims. For Tarr, the philosophical underpinning of Secure Scuttlebutt is social relativism. Because Scuttlebutt is distributed, each user decides what to do with their network and how to do it. This means that the users of SSB-driven software must consciously deliberate about whom they want to interact with “online,” and where, and why.

Commercial online services, by contrast, regulate user behavior with software and legal controls. Even the way users are identified on a service like Twitter, Instagram, or WhatsApp must conform with the service provider’s wishes. A username is a globally unique ID. Otherwise, how would the service and the users tell one individual from another?

This is only a problem when a service is globalized, built to work for billions of people all at once. But real people deal with duplicate or similar names all the time. They do so by understanding the contexts and communities in which they live. On today’s internet, people don’t get a chance to ponder those circumstances. Instead, every context is the same context: the murky haze of the Cloud. Scuttlebutt doesn’t assume a replacement circumstance; instead it opens the door to many alternatives—the libertarians can have their markets, and the leftists can have their coops, and others can have anything in-between.

In that respect, Secure Scuttlebutt reveals some of the assumptions of the supposedly normal technologies people use. Isn’t it odd that every online service assumes it should be a global one, for example? Such a design benefits technology companies, of course, but there are obvious downsides. Security and abuse offer examples; those problems arise largely in software that insists on being both global and always-on. By offloading the work of synchronizing data to computers, a task computers do well, Tarr hopes Scuttlebutt can help people do what they do well: managing the real-world relationships that would inspire people to connect through software in the first place.

Some of those uses might entail democratic liberation—a fixation of internet activists, especially since the Arab Spring suggested that social media might help combat tyranny. But the central operation of those services also makes them easier for would-be authoritarians to control. Last year, for example, the Turkish government blocked access to Facebook, Twitter, and WhatsApp amidst protests related to the arrest of opposition-party leaders. Scuttlebutt could be used to organize people and disseminate information with less risk of state impediment. However, it also works more slowly and reaches a narrower community. In many cases, that might be fine—and it might avoid social media’s tendency to turn remote, political unrest into global entertainment.

But the likelier uses for SSB might end up being much more commonplace. For one, it works offline. In many parts of the world, access to reliable, affordable networking is a bigger challenge than access to computing. In India, for example, phones are ubiquitous, but network access is costly and slow. Technology companies have proposed solutions that double down on centralization. Google developed weather-balloon wifi to deliver access to Africa and Asia, and Facebook offered free internet on the subcontinent. Scuttlebutt might provide a simpler option with fewer strings and greater utility.

Connectivity loss also affects the first world, especially for those on the move. When in a subway or on a transcontinental flight—or even in a hotel room—networks are frequently unavailable or unreliable. Many services don’t work at all when a device is offline, even just to show what’s been downloaded since the last connection. They certainly don’t let you author new material offline. The cost and complexity of mobile roaming abroad also hampers always-on network usage. And even when accessible and affordable, constant connectivity has become a burden. Today, people often stay online not because they want to be there, but because there’s no way to avoid it.

Security and privacy offer further rationales for a system like SSB. Cyberattacks are common, and more organizations might want to decouple that data, even when encrypted, from the public internet. Decents might offer a solution. And when it comes to privacy, perhaps the best way to protect one’s personal information is to share it selectively for specific purposes. Services like Snapchat and Signal have already demonstrated a public preference for such behavior.

These rationales all derive from a bigger one: Centralized services are easy to use, but they offer one-size-fits all solutions. Why should a social network for a school or a family or a neighborhood work the same way as one meant for corporate advertisers, or governmental officials, or journalists? Even if Scuttlebutt never catches on, it shows that the future online might be far more customized and diverse than the present. And not just in its appearance, like MySpace or Geocities. But also in its functionality, its means of access, and its membership.

Today, Secure Scuttlebutt is both esoteric and unrefined. For those who aren’t already forking git repositories and hanging out in freenode on IRC, SSB will feel like a curiosity for eccentric technology geeks. But that’s also how the web once seemed, and Google, and Twitter, all the rest, even if it’s hard to remember when those systems were obscure rather than infrastructural.

* * *

In 19th-century Britain, the Church of England’s role as host of the official religion of the United Kingdom came under scrutiny. The Liberal Party’s drive to separate church and state had become viable, as industrialism, nationalism, and secularism rose to prominence. Decoupling the Anglican church from the state was called disestablishment, and its proponents were known as disestablishmentarians. In turn, conservative opposition to disestablishment was called antidisestablishmentarianism. Disestablishment was eventually achieved in Scotland, Ireland, and Wales, but not in England. In the process, antidisestablishmentarianism became the longest non-technical word in the English language—and a nebulous koan with which to shake a fist against the Man.

Lost in the modern misuse of antidisestablishmentarianism is the way it chains historical contingencies. The antidisestablishmentarians weren’t just proponents of the state church. They were opponents to those who hoped to decouple it from the state at a particular moment in time.

Secure Scuttlebutt exemplifies a similar principle, one that some fellow travelers in the decentralized software community have called counterantidisintermediation. The 20th century saw the rise of intermediation: centralized media systems run by corporations and governments. When the web became popular in the mid-1990s, it promised disintermediation—allowing individuals to reach one another directly, without middlemen. But harnessing disintermediation proved hard for ordinary people, and corporations like Google and Facebook discovered they could build huge wealth facilitating those interactions in aggregate. That’s antidisintermediation. Today, decentralized software projects oppose the centralized control of online media. That’s counterantidisintermediation.

The tech entrepreneur and activist Anil Dash has eulogized “the web we lost.” For Dash, that’s the disintermediationist 1990s. But the internet of that era couldn’t work today, even if the world wanted it back. History moves forward, and people must respond to present conditions. Whether via Secure Scuttlebutt or something else, counterantidisintermediationalism could become the driving political, economic, and technical worldview of the near future. If successful, it might find various political and economic implementations. Bitcoin-style anarcho-capitalism is one. Another is its opposite, leftist collectivist anarchism (Dmytri Kleiner, who helped popularize the term counterantidisintermediationalism, calls himself a “venture communist”).* Another is Dominic Tarr’s equal-opportunity, technical agnosticism, a centrist take that sheds the baggage of anarchy entirely.

Tarr’s pitch is appealing, and a poetic consequence of the counterantidisintermediationalist philosophy. Governments and corporations probably shouldn’t be trusted to contain and manage all of modern life. But neither should extremists, left or right, who happen to know how to program computers. A truly decentralized infrastructure wouldn’t just diversify control of its technical operation. It would also diversify political, economic, and cultural goals.

In an age awash with venture capitalists and billionaires, anarcho-capitalists and conspiracy theorists, oligarchs and neo-authoritarians, perhaps the most compelling vision of the technological future is also the most modest. Scuttlebutt offers one model of that humility. Diverse groups of people networked in equally diverse, and even mutually-contradictory ways—for profit, for community, for anarchy, for globalism, and for localism, among others. No revolution whatsoever. Just people of all stripes, in places of all kinds, who sometimes use computers together.

* This article originally misstated that Dmytri Kleiner coined the term “counterantidisintermediationalism.” We regret the error.