Dan Phiffer is a Brooklyn-based artist and programmer for the New Yorker. I first learned of Phiffer’s work during Occupy in New York, when he began building and testing a mobile and hyper-local network for protestors to communicate anonymously with one another, called Occupy.here. This network was exciting to me because it placed control of a communications infrastructure within an open-source project, instead of a large Internet service provider or a state-controlled technology.
Phiffer’s work is situated in a growing group of activists, technologists, and artists seeking to build alternative models for community that are focused on serving their users. Kansas City’s Freedom Network, the People’s Open Network in Oakland, the Free Network Foundation, and many more abroad are starting to build and run networks by and for the people. I conversed with Phiffer over email to get a better feel for what he is working on and what he thinks the future holds for mesh networks.
Ben Valentine: Why did you feel called to work with wireless and mesh technologies?
Dan Phiffer: An early interest in net art got me seeing the Web as a material that can be manipulated and subverted. I’ve started thinking of the network layer as part of that material, making it possible to situate Web-based projects in a particular place. Two examples of this are small towns where dial-up Internet is still the only option, or political demonstrations where mobile networks are being monitored or jammed. Mostly it’s about connecting people to their neighbors in ways the Internet isn’t able to—the model of the community garden or food co-op.
One pedantic thing I should mention here is that I don’t actually work with mesh networks. I am more interested in cheap commodity hardware and software, the existing tech in peoples’ pockets. That stuff still hasn’t come around to embracing mesh protocols. The projects I’m working on are more like offline local area networks (LAN) or intranets instead of building a true ad-hoc-style network.
BV: What are the unique qualities of mesh networks that differentiate them from the Internet?
DP: The Internet was a breakthrough success in large part because of its decentralized network topology. A message sent between two nodes will find the best path in response to which routes are possible. Mesh networks take this idea further; they use a distributed network topology, which is similar to a decentralized one except that it lacks a built-in hierarchy in its design. Nodes can join and leave much more fluidly, which means you need slightly more sophisticated software on your computer to move data through the network.
In my case, I’m relying on a single router that routes traffic in the traditional Internet style, except that my network doesn’t connect upstream to any Internet service provider. The only people who are able to use my websites are those physically close enough to connect to the Wi-Fi network. It’s like having your own web in miniature—a Little Internet, if you will—that is far less prone to problems like spam and trolling and mass surveillance. People can remain anonymous (or pseudonymous) without things instantly devolving into the comment flame wars that are so common on the Big Internet.
BV: What is the future of mesh networks?
DP: On one hand, in the United States there’s an increased awareness around issues like surveillance and network neutrality and a growing frustration with our oligopoly of network providers. That awareness makes alternatives like mesh networks more appealing. On the other hand, the Internet was built with abundant Cold-War–era military funding, and its only real competition was in mass broadcast communications. Ham radio, while it was distributed in its own network topology, wasn’t nearly as widely used as the Internet has now become. The big challenge to building alternative networks is the network effect: the value of the network is dependent on the number of participants, and everyone has already come to rely on our existing Internet.
But I think the underlying challenges to large-scale mesh networks are fundamentally social and economic. We must consider questions like: Who will you call when it breaks? I think the interesting answers are incompatible with American-style market capitalism. We need many more people to hear terms like sharing economy and the cloud as patently absurd, as a kind of technological newspeak. It’s the language of modern robber barons, who have a very limited and cynical idea of what terms like disruption or revolution might entail. I see a future of climate-driven disruption, where we’ll need solidarity and mutual aid more than we’ll need Netflix or Amazon.
The challenges of network alternatives are defined by their small scale in contrast to the Big Internet. This is another reason I’ve been focusing on an offline LAN model; it’s designed to work well even with a very small number of participants. Perhaps when it gets more popular, stitching the network islands together into a connected archipelago will be a more pressing need. Another model I’ve been considering is connecting the LANs together via the Big Internet by way of an encrypted darknet. That’s actually something I’ve been experimenting with a lot, recently.
BV: Could you say a little more about that?
DP: At my job at the New Yorker, one of my responsibilities is to look after something called Strongbox (based on SecureDrop, which is maintained by the Freedom of the Press Foundation). This system provides a web-based interface for sources to anonymously leave documents or messages that might inform or break a story. But in order to submit documents, one has to install the Tor anonymizing software; the Strongbox URL won’t work otherwise.
Along these lines, I’ve been thinking about how this same type of mechanism could connect individual people—more of a system for sharing funny animated GIFs rather than leaking sensitive documents. The darknet approach lacks the local broadcast component of offline Wi-Fi since it relies on the Internet as its data transport. I’m sure there are lots of things like this out there; I believe that’s how the black-market Silk Road was being run. But I’m interested in how this type of thing might offer a compelling alternative to the likes of Facebook and Twitter, not just provide a safe haven for disreputable activities.
BV: Right now the Internet risks becoming partitioned off, due to governments being upset about the reach of the National Security Agency (NSA) and other governmental entities. This worries me because a large part of my interest and love of the web stems from its knack for connecting like-minded people across vast distances. Do you think localized mesh networks can reinvent the web in a way that is both hyper-local as well as globally connected? How?
DP: If you live in China or Iran—or the United Kingdom, to a lesser degree—this isn’t just a risk but a current reality. An equally important problem is that, largely speaking, people living in different countries don’t have enough reason to care when the Internet does get partitioned off. The political conservatism that brought us the Patriot Act also seeks to keep out migrants who rely more heavily on global connectivity, to retain contact with their countries of origin. A related challenge is widespread xenophobia and racism. One of my Persian friends living in the States told me she’s become a kind of social media pariah, as her supposed friends have internalized the NSA’s multi-hop logic regarding the concept of foreignness. In cases like this, the chilling effect applies to social connectedness as much as to our right to free speech.
Progress, for me, looks like small-scale, bottom-up interventions—artistic, activist, or otherwise—over a long timeframe. I want to see hyper-local network alternatives contribute to political alternatives to the current model of a distant, corporate-funded, elected representative. This is the politics of the neighborhood, of the city, that eventually contribute to changing regional norms and making a national political impact. The nation-state and beyond maybe seems like the wrong scale at which to consider mesh networks, at least in the beginning.