Public Internet

Nearly a decade ago, Facebook pioneered the concept of social media. However, it wasn’t until recently— as the platform boasts 2.1 billion users— that the full consequences of such an ubiquitous network have burst into the public eye.¹ Now, Zuckerberg’s famed motto, “move fast and break things,” may have in fact contributed to a broken democracy. In his recent testimony before Congress, it also became evident that our elected representatives' lack of digital literacy only serves to exacerbate the situation.²

Technology has had an ever more intimate relationship with politics— and I'm appropriating the term technopolitics to describe this entanglement. It's a purposefully broad term, a hyperobject (to borrow a term from Timothy Morton) for reconciling disparate processes into discrete events and behaviors.³ Technopolitics could be i.e., using the Internet to influence political campaigns. It could be psychographically curated information on social media, policy and regulation changes (GDPR), net neutrality, and much more.⁴ My optimism is that "technopolitics" will also promote deliberation, especially in answering the following questions:

How have wireless information networks supported democracy and shared experiences? How does network topology affect its performance in this capacity?

Facebook's continuous and relentless immersion in technopolitics deserves attention. In some ways, the service has become a digital privately-owned public space (POPS). In the past, determinedly public spaces have been legally protected as arenas of deliberative practice and free speech. The growing privatization of these spaces, however, wicks away some allowance of free expression— bolstering the legal defense of any imposed regulations by situating itself as borrowed, rather than public.

Ownership and authority are counter-intuitive to public deliberation. Historically, the commons— or village green— was shared among multiple owners and would functionally transition to suit each of the user's unique needs. Despite the challenges of POPS, they have often served as popular locations for shared experience and the exchange of information or news.

For example, the public house (a "pub", or tavern) was once a prime hub; as was the modern mall— capitalism's ideal intersection with democratic deliberation. Malls were once vibrant centers of shared experience among a diverse demographic. Since 2002, however, the so-called "retail apocalypse" has erased 480,000 jobs at these social marketplaces.⁵ With Amazon, the commercial function has become obsolete, but what about its social function? If people are adapting social media to simulate these interactions, how does Facebook differ from the mall, the park, or the plaza?

Although Facebook might provide an adequate platform for organization and deliberation, it is not the property of its users. Additionally, Facebook in some ways resists appropriation by providing a standard for communication in ways that physical space remains ambiguous. Furthermore, Facebook's complex algorithms organizing information have been found to avoid serving anything "disagreeable" to the tastes and opinions of you, or your psychographic model.⁶ The lack of ideologically oppositional content has re-enforced, and polarized, filtered “echo chambers”— facilitating partisan conflict and distrust.

As of 2017, some 67% of Americans "get at least some of their news" from social media. Therefore, an in-depth understanding of how these networks behave is critical. From 1990 to 2016, the number of people working in journalism has fallen by about 250,000.⁷ Many of these jobs have simply disappeared— disabled by the hybrid media environment brought about by the Internet.⁸

In a talk at Harvard's HKS school, The Washington Post media reporter Margaret Sullivan cited the death of local news as the greatest threat to democracy.⁹ Local news has largely been subsumed by the Facebook “feed”, Twitter and other social media sources— attracting users with information aggregation and constant, rapid availability and convenience.

The trend that I’d like to address in this paper is the adoption of small-scale, distributed "intranets". Several initiatives, usually in larger cities, are already underway (the Redhook Initiative, Detroit Community Technology Project, NYC Mesh, Hyperboria, Project Byzantium, and LibreMesh among others in Europe) to bring affordable— even free— Internet-access, usually to neglected and poor communities within the US.

Essentially, these networks distribute wireless connections from a series of inter-connected hubs, breaking up the direct Internet service-provider (ISP) to consumer relationship in favor of a cheaper community-driven alternative. At present, these networks are merely re-distributing a full-fledged Internet connection; however, Cuba's historic underground SNET, or “street network”, presents an intriguing example for how these mesh networks could evolve with internal grassroots support.

A case study of the Cuban SNET— its network protocols, infrastructural assembly, and grassroots culture— could lend valuable insight into how communities in the United States might restructure their own information pipelines in support of political and social agonism and democracy. Additionally, I hypothesize that such a network applied en masse— but centered around local information ecologies— could revitalize news (providing accountability against misinformation, or “fake news”), draw power away from big tech, and encourage vigorous cultural growth online and off.

Recently, I've seen renewed interest in revising the models for how we exchange data and information across the Web. Juan Benet, most notably, introduced the idea of an Interplanetary File System (IPFS) that would dynamically record its "state"— an encrypted snapshot of its collective data.¹⁰

This "soft" network has already begun to enlist ordinary computers for maintaining this cryptographic "ledger" of the Internet. Through this model, it creates a transforming digital image of the Internet. Additionally, this system could altogether prevent the loss of data by archiving the Web into discrete, retrievable packets.

Another interesting development— this time in the consumer market— are companies like goTenna.¹¹ Their mesh network dongles are essentially Internet-age walkie-talkies. Instead of sound, goTenna's devices connect to phones or laptops— expanding the types of media and information that can be shared without a central routing authority. Additionally, apps— like Dr. Gardner-Stephen's Serval— are starting to turn phones into ad hoc "nodes" in self-assembling networks.¹²

In 2015, another app— FireChat— was tested along with goTenna's devices at Burning Man. Similar to Serval, FireChat generated a self-assembling, peer-to-peer network that allowed attendees to connect with one another— without needing to proxy through a data-center or commercial carrier.¹³

Now that the Internet fantasy has overwhelmed the skeptics and made its way into 80% of American households, people are beginning to re-evaluate it's historical function and how it can better adapt to modern media, culture, and politics.¹⁴ The concept of a wireless Internet— as well as the ethereal "Cloud"— promotes somewhat of a farce: that the Internet does not take up physical space.

In reality, data are stored on servers, and on your own personal computers. The Internet describes these facilities, but beyond that it points to the infrastructure that makes the exchange of data possible between those stores. There is also a distinction here from the World-Wide Web (WWW)— which describes the digital organization of the data the Internet carries and distributes.

In Albert-László Barabási and Réka Albert's Emergence and Scaling in Random Networks, the topography of a scale-free network— like the WWW— follows a unique set of observable laws, or behaviors.¹⁵ One such behavior recalls the "rich-get-richer" phenomena, or preferential attachment.

This property models the way in which new nodes assemble into a network, showing preference toward nodes with high degree (connectivity). As a result, well-established node become increasingly vertical. Franklin Foer, in his book World Without Mind, is critical of the Frightful Five (Google, Amazon, Apple, Microsoft, and Facebook) and the way in which they've capitalized on this concentration of activity and power.¹⁶

In a study conducted in Changsha, China— Degree and connectivity of the Internet's scale-free topology— the researchers demanded:

In order to better understand the dynamic behaviors of the Internet, to accurately depict the topological structure of the Internet and to availably control the efficient operation of the Internet there should be further research on the statistical characteristics of Internet topology.¹⁷

In the same study, the three researchers built a mathematical model, not of the WWW, but of the Internet— similar to that of Barabási and Réka, but "at the autonomous systems (AS) level."¹⁸ Asserting that the average degree— the connectivity of a node (e.g., a server or wireless base-station)— is one of the most important variables at indicating other global network behaviors and properties.

The connectivity of nodes on the Internet has been found to follow a similar power-law distribution, with the most connected nodes being very few, but reaching a majority of the remainder. Likewise, the least connected nodes make up the majority, but have a smaller average degree.

A high-degree node in isolation might be said to have more of a star topology— one origin serving and connecting many adjacent nodes— which carries risk by concentrating data and traffic over a single node. Hundreds of thousands of computers might rely on a single hub. Because of its high degree, it will also act as an intermediary for exchanges between other nodes. As a result, if the hub were disabled, those computers might also lose contact with one another.

Speaking to polarization, highly connected nodes that become compromised— perhaps promoting a single ideology, or censoring opposing information— might also contribute to a radicalization of opinion. In an ideal distributed network, each node is able to reach every other node in the network. The distribution of the network— as well as data redundancies— could significantly mitigate this risk.

While the star model— a single hub with high connectivity— deals with heavy traffic by "widening the pipes" (amassing more storage, electricity, and computing power), the modern "mesh" network handles traffic in a different way. In a sense, they are "self-assembling"— i.e. dynamic— using metrics like physical proximity, signal strength, and shortest walk (fewest "hops") to determine which nodes should be activated.

How our devices efficiently navigate this network is a difficult problem, and it has proven to be a significant obstacle for popular adoption. An ad hoc network or routing protocol— rather than relying on static connections— reacts to changing conditions and persistent connections despite mobile nodes. In this way, a user on the network might walk from source A to source B, and as they cross the threshold from one signal to the other the signal switches without disconnecting.

Several other open-source projects have developed alternative protocols to suit a new distributed infrastructure. Like the ham radio craze before it, hundreds of people are moving to these networks and discussing their implementation. I'd also like to note that across different forums, I've often seen the term "mesh" networks. For the most part, I think they are thinking decentralized.

Mesh networks seem to describe a more perfect grid of routing devices and servers. I don't think this is what they're hoping for, and it's not the way networks will want to naturally behave. A decentralized network, however, does calls for a flattening of degrees— more homogeneity across local nodes and repeaters.

The Cuban SNET, or "street network", was first conceived by the online gaming community in Havana when teenagers started stringing LAN cables across alleys to play World of Warcraft, Call of Duty and other titles. This also allowed for IRC (an early form of Internet chatroom) and file-sharing, which drew more popular support. A more organized push brought these small-scale LAN networks together, and created for the first time a digitally connected Havana.

For a long time, government workers and tourists were the only ones given access to the WWW. Now, Cubans can purchase scratch-cards for timed connections at public hotspots situated in Havana, and other cities. They are expensive, and the connection is crowded and slow.

It would be illegal for anyone to get Internet access without permission from ETECSA, the state comms company. However, they've largely turned a blind eye toward the SNET, and it has been allowed to grow (and similar networks of smaller size have popped up elsewhere on the island)— yet its moderators strictly disallow religious or political speech, sexually explicit material, or commercial promotion in order to ensure its continuity.

As a result of a complicated history of isolation— mostly as a result of the aggressive diplomacy of the United States— Cuba developed without a digital link to the world. The information quarantine, as well as the thousands of jobless professionals, created a cultural precondition for the internal growth of these networks.

In Havana, the SNET has established itself as a functioning alternative to the global Web— in a country with only 5.6% household Internet penetration.¹⁹ The only reason a global connection is ever used is generally to check Facebook or facetime a family member abroad— an actual geographic separation to which the SNET remains dismembered.

The government hasn't made it an easy task. Purchasing routers or other wireless tech in the 2.4GHz or 5GHz range also requires permission from ETECSA, and can't be brought into the country. As a result, these social and technical entrepreneurs must be careful not to overstep the law— or, not to get caught doing so...

A recent (2017) study shows that the SNET is centered around several hubs, or pillars:

While an "individual node may connect up to 200 users... a pillar may connect tens of regional nodes." Each pillar "has a distinctive name, sometime reflecting geography or the SNET's origin in gamer networks. The pillars' regions sometimes overlap, but together they cover a significant portion of the Havana metropolitan area."²⁰

Many of the movies, TV shows, and music popular in the United States are physically distributed via hard drive— what's known as El Paquete Semanal, the weekly packet. The report notes that heavy-bandwidth activity is limited outside of the hours of 3AM to noon, during which time El Paquete can be beamed across the pillars.

The paper marks 2011 as the beginning of the SNET, but perhaps that is the actual deployment of the ".snet" top level domain (e.g., .com, .org, .net). I've read in different places that the gamer networks started as early as 1998.

The study also mentions that the SNET uses an open shortest path first (OSPF) routing protocol— a variant of Dijkstra's shortest walk algorithm used to search for the quickest path across an interconnected web of nodes. Another interesting note in the report added that, "each pillar is connected to at least two others."

As with the SNET, local network initiatives in the US are using a powerful set of access points and repeaters to stretch signals from more powerful hubs. In reality, it is geographically designating invisible zones of intranet coverage with unique top-level domains.

The relationship of ISP-to-consumer gives only a single option: the Internet. I wonder if the developing decentralized networks might begin to shape an all-wireless, community-assembled Public Internet— a parallel surface to the commercial Internet that can be switched on or off.

Imagining, for example, Brooklyn as a container for a series of community intranets— disconnected from other community networks. Not completely disconnected, but having the ability to bridge, filter, or query the contents of this Public Internet.

By mapping this kind of digital commons to physical space and limiting each cluster's exposure to high-degree nodes, this separation might increase the average degree of nodes within each intranet. In this way, it might resemble a series of digital archipelagos.

The way distinct cultures, or institutions, develop around these digital spaces, and the way they could be appropriated for deliberative democracy is also intriguing considering the possibility of higher average degree within each digital commons and how that proceeds to connect these districts.

Lastly, it seems that popular participation in the assembly of these networks might be vital in determining their success. Internal maintenance, development, apps and services that make this digital public space unique should be an important value that these communities build. Altogether, I believe it could also foster continued digital literacy, learning, information equality and community connectivity.