Deprivatizing internet: interview with Ben Tarnoff

Ben Tarnoff is a writer and co-founder of Logic magazine. Author of manuscritps on tech workers and organizing – such as Voices from the Valley and The Making of the Tech Worker Movement, he recently launched his new book Internet for the People: The Fight for Our Digital Future

Tarnoff tells the internet history as a history of privatization and proposes moves towards deprivatizing the internet, which is managed by and for the people. Thus, community networks and cooperatives can play a role in this construction of another internet.

In a conversation with Rafael Grohmann, he talked about internet infrastructure and materiality, metaphors for naming technologies, the role of imagination in disputing futures and proposing alternatives, community networks, cooperatives, data and tech workers organizing.

DIGILABOUR: What is the role of infrastructure and the materiality in the terms of the Internet we’re living in? 

BEN TARNOFF: The Internet is deeply material creation. It’s composed of fiber optic cables, data centers, computers—things you can touch, things that had to be constructed out of parts, with those parts themselves having been made from materials that were dug out of the earth. I open the book with an undersea fiber optic cable called Marea that runs from Spain to Virginia in the United States, and I wanted to open the book there because I felt it was important to draw attention to the materiality of the Internet. I would say, however, that I think that the materiality of the Internet has become more widely appreciated in recent years. There certainly was a period in time in which the idea of the Internet as being dematerialized, floating in space, mediated through metaphors like the cloud was quite dominant in both popular and scholarly circles. I think of a lot of the writing on the Internet in the 1990s as being very influenced by the perception of a lack or a loss of materiality, whereas I think in recent years there has come to be a greater appreciation that the Internet is made of things you can touch, things that have a specific physical composition and a physical origin. But one has to be careful not to let the pendulum swing too far. In certain STS papers there is now probably too much emphasis on materiality, too much emphasis on infrastructure. Because ultimately, when we’re talking about the Internet or computing more broadly, hardware doesn’t matter as much as software, and you can’t actually touch software. All of this materiality would be dead if it weren’t animated by code. 

DIGILABOUR: About Internet histories, what’s wrong with the mainstream history of the Internet and how to rewrite the Internet history in other ways?

BEN TARNOFF: Well, I’m not sure that there even is a mainstream Internet history. I think most folks know vaguely that the Internet came from the military. They may, if they’re of a certain age, have memories of the Internet of the 1990s, or possibly even earlier, but there isn’t much of a coherent popular narrative about where the Internet comes from. So, in a sense, my book isn’t an attempt to challenge or debunk a dominant narrative; it’s to give people the narrative for the first time, which is actually a wonderful opportunity because you can tell the story to readers who don’t know any version of the story. You can fill in some of the details of the few elements they may know to give greater context on precisely how the Internet emerged from U.S. military research. But the later history of the Internet, and specifically the privatization of the Internet, is not on most people’s radar. And the history of the internet’s privatization is not a matter of much interest in scholarly circles either.

DIGILABOUR: What is the role of imagination regarding disputing futures and prototyping other types of infrastructures?

BEN TARNOFF: In the book I tried to make the case for the political value and the political stakes of imagination. In doing so, I draw on the work of Angela Davis and other abolitionist thinkers in the United States who often make the point that failures of imagination lead to practical failures. So, from an abolitionist perspective, the failure to imagine a world beyond prisons has led generations of reformers to reinvent the prison in different ways. One current iteration is so-called e-carceration, where individuals are no longer incarcerated in a facility, but they have an electronic monitoring device fastened to them, and the prison logic continues, but the prison is now the world—the prison has been turned inside out. I try to transport those insights into the realm of the Internet and pose the question, what kind of imaginative work might be needed to truly transform the Internet? When we think about the firms and the computational systems that currently control and organize the Internet like Google, Facebook, Comcast, and others, we need to think about how we could create an Internet in which we don’t simply have different versions of those entities, but entities of an entirely different kind. So, to follow the thread provided by Angela Davis, the point is not simply to replace Facebook with a pseudo-Facebook—with a nationalized or cooperativized Facebook, let’s say—but with a constellation of entities that enable us to connect in a completely different way. That sounds a bit abstract, but I think it’s important to make the point that imagination has a material foundation, which is to say imagination as a human pursuit can only happen if there are resources provisioned to support that pursuit. In the book I make some suggestions for what that type of materially supported, embodied, collective work of imagination might look like. It’s not each of us sitting alone in our rooms dreaming up a new Internet; rather, it’s a process that brings people together in space and gives them the things they need to work out with one another what an Internet that better serves their needs looks like. 

DIGILABOUR: I would like to put the accent on the metaphors that we used to frame the Internet and technology. I like your take on “platforms does not exist” because even critical researchers like me and others used to frame these things as platforms. What do you mean by a platform that doesn’t exist?

BEN TARNOFF: Before we get into the platform piece of it, it might be useful to step back and talk briefly about why metaphors matter. This brings us back to the beginning of our conversation, to the materiality of the Internet. Now, of course, the Internet is composed of things you can touch, but as I mentioned, code software, the instructions that computers are executing, is ultimately the most important element, and that is not something you can touch because what it is composed of, at the most fundamental level, is a series of electrons that are being interpreted as zeros or ones. Everything in computing that is built on top of those electrons is an abstraction. Abstractions are tools that we use to think about things that we can’t see, things that we can’t touch, things that we can’t perceive with our senses. In this context, abstractions are not simply aids to understanding, they are the understanding. So the stakes are rather high. When we think of the complex computational systems developed by Google, Facebook, Amazon, and others as platforms, we are understanding the architecture, the role, the composition, the political economy of those entities in very particular ways, and in ways that ultimately favor the interests of big tech companies. That is why I object to the term. The term platform evokes a sense of openness, a sense of even handedness that works to the advantage of these companies because they want to convey that impression. But in fact, as we know, these firms are intimately involved in organizing our online life. They are sovereigns of a kind. They’re not simply platforms on which interactions take place, they are architectures that are coordinating those interactions for particular purposes, above all profit-making.

DIGILABOUR: In your book, you frame community networks as an alternative to these infrastructures and imaginations. What are community networks, their potentials and also their limitations?

BEN TARNOFF: There are varieties of this sort of thing all over the world, but in the United States we have more than 900 publicly owned or cooperatively owned broadband networks. Most of these are owned by a municipality, but many are owned and managed by the users themselves, in the case of cooperatively owned networks. There are two main advantages to community networks over their corporate counterparts. The first is that they tend to provide better service at lower cost than the broadband giants here in the United States like Comcast. The reason is fairly obvious: because these organizations are able to invest in infrastructure and prioritize social needs like universal connectivity rather than funneling money upwards into the hands of executives and investors. The second main advantage of these community networks is that they create the possibility of encoding democratic practices into their everyday operation. So, for instance, in the case of the cooperatively owned networks, these are democratically run organizations. The member-owners vote in regular elections to decide who will be on the board that runs the network. Community networks give people an opportunity to participate in decisions that directly affect them, decisions that pertain to an essential infrastructure of their lives, the Internet. And at the end of the day, I think this latter quality of community networks is what I find most valuable, because it points the way toward organizing an Internet along very different lines. There are, as you pointed out, limitations here. Community networks are local organizations. The Internet is not local; it exists in many scales and interventions at those scales are also required. There are the deeper networks of the Internet, the so-called backbones, that have to be dealt with as well. But as an existing set of experiments—and 900 is not a small number—it’s a very promising point of departure for what it might look like to deprivatize the pipes of the Internet and reorganize them around serving human need rather than private profit. 

DIGILABOUR: How do you relate community networks approach and platform cooperativism movement? And what are public policies needed for that?

BEN TARNOFF: In the book I split up the Internet into two layers, the bottom layer and the upper layer. That’s a reduction, I know, but it’s useful for my purposes. The bottom layer I call the pipes, the physical infrastructure of the Internet, the stuff that moves packets from one computer to another. Community networks, to my mind, are the most promising path forward for deprivatizing the pipes, by which I mean creating an Internet where people and not profit rule. When we move up the stack to the so-called application layer of the Internet, to the layer where the Internet is experienced, the strategies for deprivatization become more diverse and more complex because this is a realm of the Internet that is more diverse and more complex. Facebook is much more complex than Comcast. Facebook is also much more different from Amazon than Comcast is from AT&T. So what that means as a matter of strategy is that we have to adapt to the new terrain and develop a set of interventions that are comparably complex and diverse. We also do not have experiments that are as mature at this upper layer of the Internet. There is not an obvious equivalent of the community network model up the stack. What we do have, however, are a handful of interesting communities, such as the platform cooperativism community and the decentralized web community, that are developing different kinds of online services, which, despite their inevitable limitations, point the way towards what a different Internet at the application layer might look like. Platform cooperativism, as you know, is the attempt to build online services that are cooperatively owned or cooperatively governed, and there are a number of interesting experiments along those lines around the world. The decentralized web community is a somewhat vaguer designation, but comprises projects like Mastodon, an open-source software project that enables people to run independent social media sites that can be cooperatively managed by the users themselves and also to federate those servers together to create a broader network. Deprivatization is not monolithic; it takes on different features and follows different directions depending on the layer of the Internet that we’re talking about. 

DIGILABOUR: What is the role of data in these alternative arrangements? 

BEN TARNOFF: We have to find a collective way to think about data. Data does not have any value at an individual level. Your data as a single person is not particularly interesting, at least at the scale at which these companies tend to operate. Your data is only interesting when it is put together with a lot of other people’s data and analyzed in the aggregate to derive certain population-level insights that can be used to make money, such as through targeted advertising or other means. The big tech companies don’t think about data in an individualized way, they think about it in a collective way. So we should also think about a collective way. And one of the problems with the various policy solutions that have been pursued in the European Union primarily, but also to some extent around the world, has been that they rely on an individualized model of data ownership. The flagship European data protection law GDPR talks about “data subjects” who are entitled to certain rights. It’s not the right way to think about it conceptually. It also produces fairly disappointing practical results. One of the rights that is secured by the GDPR is the right to request your data from a firm, such that you can look to see what information they have about you. There’s some limited value to that, but what you’re not getting there is any real picture of how that data is monetized. You’re also not being given any real alternative of where you could put that data. If I download big zip file of my data from a website, even if I have the technical literacy to interpret it, I’m not going to have any real sense of how this data interacts with other people’s data in order to generate the type of insights that can be monetized, and crucially, I don’t have anywhere else I can put this data. It’s not as if I can withdraw my data and plug it into another site. So, that’s a long way of saying that we can’t think about data individualistically because that’s not how the companies think about it. We need collective solutions, and we need collective solutions that create space for non-commercial alternatives.

DIGILABOUR: how to connect tech worker organizing and this type of movement around Internet for the people and cooperativism?

BEN TARNOFF: It’s a good question. I think there are some clear ways in which organized tech workers could contribute to the project of deprivatization. One could imagine, for instance, unionized technicians taking part in the construction of community networks. But I also want to sound a note of caution, particularly when we think of the role that the more highly salaried white-collar tech workers at the big firms might play. I believe all of those people should be in union, and I believe that they can make a contribution to the project of  building a better Internet, if only by virtue of the technical expertise that they possess. But we wouldn’t want to create a situation in which we transferred the power to determine what the Internet looks like from a small handful of executives and investors to a slightly larger pool of relatively privileged software engineers. So that’s my note of caution. I am inspired by historical experiments in worker-led innovation like the Lucas Plan in the UK. I believe in workers’ self-management. But we also need to think about who has most at stake in decisions about how technologies are developed and implemented, and ultimately I think that the people and the communities with most at stake, who stand to be most affected,  should have the most say. This is not unrelated to the observation that I’ve made in earlier writings of mine on the tech worker movement in the United States, which is that for the white-collar employees who inhabit the upper layers of the tech workforce to play a constructive role, they need to pursue a solidaristic model of unionism and take their political direction from proletarian leadership. That is a good principle more broadly: I’m a bad Marxist, but remain fairly orthodox in my belief that the working class is the protagonist of any project of social reconstruction worth pursuing.