Maybe it’s time to build a new internet

We’ve done it before and we can do it again

mehan jayasuriya
5 min readJun 18, 2014


Last month, the U.S. Federal Communications Commission voted to advance a plan that would strike down one of the core tenets of net neutrality, the principle that asserts that all internet traffic be treated equally. This act marked the latest move in a decade-long fight between internet users and large service providers like Comcast, Verizon and AT&T. Playing the role of referee, the FCC has been advancing both slowly and clumsily toward enshrining some definition of net neutrality in the law—it’s the nature of that definition that is hotly contested. The plan proposed last month would reportedly allow service providers to sell access to “fast lanes” to moneyed interests—effectively transforming the internet from a level playing field into a two-tiered system. While many were surprised by the speed and resolve with which the FCC unveiled this plan, careful observers noted that the agency’s new chairman, Tom Wheeler had previously lobbied on behalf of virtually every large cable and mobile service provider.

Meanwhile, Comcast, America’s largest provider of internet services, is seeking to absorb Time Warner Cable, its only real competitor on a national scale. Having become the world’s largest media and communications company following its purchase of NBC Universal in 2009, Comcast now seeks to reshape the landscape of internet service providers in its image. If its acquisition of Time Warner is approved, Comcast will control roughly a third of all U.S. internet connections and will be the dominant player in nearly all of the nation’s top markets. Despite the clear threat to competition (not to mention the additional weight this would lend Comcast in negotiations with infrastructure and device providers), regulatory opposition is anything but guaranteed. Comcast is among the top spenders on lobbying in Washington, making large campaign contributions on both sides of the aisle; its CEO is known to play golf with the President.

And then there’s the pervasive surveillance of internet communications by the U.S. and its allies, new facets of which are still being revealed on a monthly basis. As the Snowden documents have shown, many of these surveillance programs are made possible by the National Security Agency’s ability to intercept and decrypt data in transit, a technical feat previously thought to be near impossible. Given the nature of these programs, online privacy now seems unattainable for all but the most sophisticated cryptographers. What’s more, it’s likely that consolidation among service providers will only make conducting mass surveillance easier.

Thoroughly commercialized, consolidated and surveilled: this is the state of the network in 2014. Once hailed for its revolutionary potential, the historically open, decentralized, egalitarian internet seems to inch closer toward centralized control with each passing day. But what can be done?

Here’s a modest proposal: perhaps it’s time to build a new internet.

This might seem like a radical idea but it’s not without precedent. As Columbia Law School professor Tim Wu demonstrated in his 2010 book, The Master Switch, the history of American communications technology has followed the rough arc of a pendulum swing. In their infancies, all of our communications networks were largely open, decentralized and free of corporate or governmental control. After demonstrating their economic value, these networks were slowly captured by large communications firms, who exerted greater control over each network until a new, more open medium arrived to usurp its place. So a closed radio network begat an open phone network; a closed phone network begat an open television network; a closed television network begat an open cable network; and a closed cable network begat the open internet. Might we be approaching the point at which the internet’s centralization begs for a technological solution?

The thought of building a new global network from scratch might seem herculean but similarly ambitious projects have been successfully undertaken in the past by hobbyists. Take FidoNet, a global, decentralized communications network built by the hobbyist Tom Jennings during a few weeks of downtime between jobs in 1984. FidoNet allowed computers attached to phone lines to send messages that could be received by users on the other side of the globe. Using a “hub and spoke” model that was in many ways even more distributed than that of today’s internet, the network connected 50,000 machines at its peak, spread across North America, Europe, Asia and Australia.

Similarly, the Usenet networking protocol, developed by a small group of North Carolina graduate students in 1980, offered public access to a worldwide computing network far in advance of the internet (ARPANET, the internet’s direct predecessor, was only available to academic and military institutions). At its peak, Usenet was far larger than even FidoNet, with a userbase in the millions. Computer historian Howard Rheingold once described Usenet as, “[A]n anarchic, unkillable, censorship-resistant, aggressively noncommercial, voraciously growing conversation among millions of people in dozens of countries.” When Tim Berners-Lee decided to announce the launch of the World Wide Web to the public, he did so on Usenet. Like FidoNet, Usenet was wholly decentralized and built and maintained by a small group of volunteers.

Remarkably, both FidoNet and Usenet were crafted using technology that now seems downright antediluvian: phone lines, dial-up-modems and computers less powerful than modern calculators. The situation today is, of course, very different. Powerful, WiFi-enabled devices can be found in the majority of American homes, be they laptops, smartphones, tablets or routers. This provides massive potential for so-called “mesh networks”—networks that connect devices directly to each other, forming a sort of daisy-chained connection that requires no central access point. Historically, mesh networks have been used mostly to connect machines to the existing internet, though a modern equivalent of FidoNet or Usenet could easily leverage this technique. Because a mesh topology allows computers to connect with no intermediary, mesh networks are inherently resistant to surveillance, censorship and centralization. There are a number of mesh networking projects currently in progress that could potentially form the basis for a new global network—this piece by Adrienne LaFrance in The Atlantic describes some of them.

It’s possible that somewhere out there, a hobbyist is already hard at work, building the robust, next-generation network that will supersede the internet. Whether or not it takes hold will likely depend on the outcomes of current political battles. Following a public outcry, the FCC has decided to accept public comments on its proposed net neutrality plan. As a result, the agency was flooded by more than 45,000 comments in May, crashing its servers. Congress has also been spurred to action, with Democrats introducing a bill yesterday that would force the FCC to explicitly ban “fast lane” agreements. It’s difficult to imagine that the FCC will capitulate to service providers if the public remains engaged on the issue. But rest assured that if the FCC fails to save the internet’s soul, technologists will.



mehan jayasuriya

writer and technologist in portland, OR // as seen in Pitchfork, The Guardian, Gizmodo, etc. //