Ben Tarnoff on nostalgia, imagination, and deprivatizing the internet

"What if our experience of the internet could be a more collective one, and one that brought us into relationships of solidarity and mutual support with other people in our community?"

Ben Tarnoff is a tech worker, writer, and co-founder of Logic Magazine. His most recent book is Voices from the Valley: Tech Workers Talk About What They Do—and How They Do It, co-authored with Moira Weigel. He has written for the New York Times, the Guardian, the New Republic, and Jacobin.

I spoke to Ben about his new book, Internet for the People: The Fight for Our Digital Future. This interview has been edited for content and clarity.

What got you particularly focused on or interested in the internet?

I started writing about the internet more seriously in 2016, and I think the motivation was that I was trying to understand the origin story of all of the different crises that were starting to generate more public concern, often under the heading of so-called techlash. Since 2016, we've seen a growing recognition in the mainstream conversation that something has gone fundamentally wrong in our digital lives, and my instinct is always to historicize things to try to figure out how things came to be by looking into the history. That's what I wanted to do with the internet. Where did the internet come from? Why was it invented? How did it develop? How was it commercialized? And how did it come to generate the various social ills that we were beginning to talk more about after 2016?

The modern internet is very centralized in the hands of few companies, so many people imagine that decentralization is the solution to ameliorate that. What do you see as the problem with that idea, and what do you see as the truth in that idea?

You can't have a fully centralized internet anymore than you can have a fully decentralized internet. The question is always: what do you centralize and what do you decentralize? The internet exists at a number of different scales, and these scales co-produce one another. So the decentralized portions of the internet are often what make the centralized portions possible, and vice versa. The question of what we centralize and what we decentralize, ultimately has to be posed and answered at social, political and technical levels. Which is to say, if we were to decentralize certain aspects in the governance and the technical operation of the internet, what new kinds of decision-making might that make possible? The argument that I'm always trying to make is that decentralization, in certain instances, gives us the opportunity to embed face-to-face, in-person, democratic deliberation into how we govern the internet. Now decentralization is not always going to be the desirable or even feasible approach; there are aspects of the internet that will simply run better in a more centralized fashion. But again, I think it's important to emphasize that it's not a choice of either-or. And these days, decentralization is a term that is often used by people who are coming from the blockchain or the Web3 community, and what they mean by decentralization tends to be different than those of us who use the term to think about the opportunities for embedding forms of democratic decision-making into the internet.

In your book, you both compare internet distribution to and discuss the synergies with energy distribution and energy utilities. Can you explain that?

This is an interesting point that has come up simply in the course of how some local communities have built their broadband networks. One example I point to is called the Gig. This is probably the most famous municipally owned broadband network in the country. It's run by EPB, which is Chattanooga’s a city-owned telecom provider, which traces its origins to the New Deal. Chattanooga first built the Gig because they had created a so-called smart grid, which is an electric grid that has sensors that are embedded throughout it in order to help the grid operators detect outages and breakages and improve energy efficiency. In order to place these sensors throughout the system, you have to lay fiber optic cable so that the sensors can communicate. And in the course of laying fiber optic cable, you also give yourself the ability to start providing fiber broadband to people's homes. So, long story short, they realized that they had already created the infrastructure to provision broadband.

That's an interesting example of a situation in which you can deploy technology that both helps improve energy efficiency and is conceivably an important element in a green transition while also providing broadband internet access. Other folks sometimes make the argument that providing higher quality broadband service will encourage people to work from home, which will in turn cut down on their commuting, which will in turn cut down on their carbon footprint. When Jeremy Corbyn’s Labour Party proposed free broadband in their 2019 manifesto, they presented it along these lines as a decarbonization measure. That's another way to link those demands.

Sticking with the comparison with energy utilities, I'm very much familiar with and been involved in energy democracy fights, and I see a lot of similarities. Have you seen a lot of other successful examples?

Yeah, another area of overlap is rural electric cooperatives. Which, again, are primarily a legacy of New Deal-era efforts at rural electrification. These cooperatives were seeded with federal money, and to this day, serve enormous portions of rural America. Now many of these cooperatives have begun offering broadband service. One of the reasons that North Dakota has some of the best fiber broadband access in the country is because it's served by these rural electric cooperatives, which are owned and operated by their members. In order to retain their federal tax exemption, they have to return excess revenue to their membership and they have to hold regular democratic elections for their board. So these really are member-owned and -operated organizations, and in these rural parts of the country they've been quite successful in providing very good high-speed broadband service. There are more than 900 community networks in the United States; this refers to both cooperatively owned networks, such as those run by rural electric cooperatives, as well as municipally owned ventures like Chattanooga’s the Gig.

You write in your book that data is sometimes compared to oil but a better analogy might be coal because they both propelled not just the capitalist accumulation, but reorganization. Can you say more about that?

In the latter half of the 1990s, there is the dot-com boom, and we think of the dot-com boom today as an exercise in collective insanity. People just throwing money at very stupid ideas like Pets.com, Pets.com taking out a very expensive Super Bowl ad. The dot-com boom is remembered as somewhat of a punchline, and maybe rightly so. What I see the dot-com boom as representing at a deeper level is an attempt to push privatization up the stack. By the latter half of the 1990s, the privatization of the pipes of the internet—the physical infrastructure of the internet—had been accomplished, and now the question was how to push privatization up into the application layer of the internet—where the apps and the sites live. At the level of the pipes, the business model is clear: you can sell access to the internet, you can sell an internet subscription like Comcast today, for instance. But at the upper floors of the internet, the question is not how to monetize access, but how to monetize activity. Which is to say, not just selling tickets to the internet, but making money from what people did once they got there. This is what dot-com firms try and fail to do over the course of the latter half of the 1990s. And, as we know, the bubble pops in 2000 and 2001.

In the aftermath of the dot-com bust, a handful of tech firms begin to build these complex computational systems that are finally successful in pushing privatization up the stack. That is, they find a way to monetize activity and, in doing so, unlock the profit potential of the internet. Data is the essential ingredient for how they do that. If we think of these complex computational systems as a form of digital machinery, then data is their fuel. Data is their animating element, that it's by manufacturing immense quantities of data and finding various ways to monetize it that firms like Google, Facebook, Amazon, Uber, and others are able to become the leviathans of the modern internet.

The transformation of the internet in that way from more static web pages to applications, for example, increased the complexity and the computational intensity of the internet. And so much of the modern privatized internet is just straight-up wasteful; advertising being the most notable example. Crypto as well. How do you think we can deprivatize and democratize the internet in an ecosocialist way that doesn't require massive amounts of data centers? Or do you think that's something that we don't really need to worry about?

There's a couple elements to your question. The first is: what would deprivatization look like? And the second is: how might we imagine a more ecologically sane internet? I think we can connect those two questions as well. Deprivatization aims for an internet in which people and not profit rule, an internet in which people have the opportunity to participate in decisions that most affect them. They can't do that in a system in which the internet is owned by private firms and run for profit because in such a system immensely consequential decisions will be left in the hands of executives and investors and those decisions will always be bound by the parameters of the profit motive.

So what might a deprivatized internet look like? In order to create an internet in which people and not profit rule, you need to create spaces where the people can rule. That is, you have to create spaces that diminish the power of the profit motive, that shrink the space of the market, and encode practices of real democratic control. What those spaces look like, more concretely, is going to depend on what layer of the network we're talking about. At the level of the pipes, the level of the physical infrastructure of the internet, those community networks that we've been discussing—either municipally owned networks or cooperatively owned networks—provide the basic building blocks of the deprivatized internet. Because these are entities that can provide better service at lower cost and, crucially, can empower users and communities to determine how infrastructure is developed, designed, and deployed.

We'll have to develop new kinds of interventions for the deeper networks of the internet because “the pipes” is not purely the last mile of broadband coming to your house, it's also a complex set of deeper systems that ensure that the packets of data that compose the internet get to where they need to go. But community networks are the starting point for deprivatizing the pipes. Now as we move up the network, as we move to the application layer and we contemplate the so-called platforms, the situation becomes much more complex because platforms are not just more computationally intensive, they're also more diverse, they're more varied. They're entangled with our social, political, legal, and economic lives in more intricate ways. Facebook is a much more complicated beast than Comcast; accordingly, the strategies that we use to deprivatize the so-called platforms and that altitude of the internet will have to necessarily be more complex.

For inspiration, I look to experiments that are unfolding in some interesting communities, such as the decentralized web community and the platform cooperatives community. These are experiments with cooperatively run social media sites, worker-owned app-based services, particularly in ride-hailing. And I argue that these experiments, while small, give us some promising signs of what a deprivatized internet could look like. Now for these experiments to become more robust and more of a threat and more accessible to mainstream internet users, we need public investment, and we need a social movement that's capable of demanding that public investment. But they provide useful points of departure for envisioning an alternative internet. If we want to fulfill the promise of that alternative, however, we also need to create spaces of imagination; which is to say, spaces where ordinary internet users can get connected with technical resources in order to build the spaces and the structures that serve their everyday needs. I put a lot of emphasis in the book on imagination, and in particular a collective, materially supported, embodied form of imagination, because I think that's ultimately how we're going to get an internet for the people.

I feel this nostalgia for the internet that I grew up with, and I always wonder: was it actually better than it is now? Maybe, in some ways. You write that one way to think about that nostalgia is as a mourning of “possible futures that the internet might have had—futures that privatization, by programming the profit motive into every layer of the network, foreclosed. And maybe this nostalgia can help inspire the movement that will be needed to pry open a horizon that was shrunk by privatization, and recover a wider field of view.” Can you say more about looking to the past and using real or imagined paths as a way of looking forward to build a better future?

Many movements throughout history have looked to an imagined past, and indeed actively constructed an idea of the past, in order to envision a better future. And often there's a lot of political utility in making people feel as if they've lost something—a set of rights, a set of freedoms—that they now need to reclaim. Even if it's not entirely clear if those rights or freedoms existed. I refer to the Luddites, and in particular E. P. Thompson's discussion of the Luddites in The Making of the English Working Class, because I think it's a very useful illustration of how movements can apply nostalgia, which is a very powerful feeling toward the task of social mobilization. And internet nostalgia is having a bit of a moment right now; there's a real interest in the aesthetic of GeoCities and Yahoo. It's also important to point out that internet nostalgia is a constant of internet history. It's not just you and I who feel nostalgic for the era of GeoCities, but there are people younger than us that feel nostalgic for Myspace. And there's people who are older than us who feel nostalgic for even earlier eras of the internet.

You find internet nostalgia at every stage of the internet's history; it's not a recent phenomenon. So one could be a pain in the ass and tell people who feel nostalgic, “Well, that's not exactly accurate, you're not remembering that era of the internet correctly. In fact, the era of the internet that you're nostalgic for had all sorts of problems.” That’s one approach to nostalgia, to simply fact-check it. But that misses the point entirely because nostalgia is not really about facts, it's about feelings; say, even that it's a structure of feeling. There is a genuine insight that is embedded in that structure of feeling, which is: the internet is continuously being remade. But the problem is that the reason for its continuous remaking is because of the forces of privatization that are optimizing it for the profit motive. So in the book, I speculate as to the possibilities of harvesting this internet nostalgia for a higher political end, as the Luddites did. What if we could feel nostalgic not really for those previous eras of the internet that the onward march of privatization has obliterated—whether GeoCities or Myspace or even farther back—but what if we could feel nostalgic for the missed opportunities, for the forks in the road that could have gone a different way, for the the points in history in which privatization was deepened when the internet could have evolved in a different channel? Then perhaps nostalgia could be an aid to the social movements that will be necessary in order to deprivatize and democratize the internet.

That's such a good way of looking at it, because I think one of the biggest challenges is getting people to see that everything is contingent—the present is contingent, history is contingent—and to see that we could do something different, and the past could have been something different as well.

Yeah, exactly.

It can sometimes be hard to separate the negative properties of the modern internet that we see today with what might be possible with a deprivatized internet that's built around people instead of profit. What role do you think the internet could play as a positive force in connecting people and building a better world? 

In the book, I talk about an organization called the Equitable Internet Initiative in Detroit, where there's an effort to get low-income, primarily Black households in certain Detroit neighborhoods connected to high-speed internet. This service is provided either free or at much-reduced cost based on how much a household can pay. What's interesting about this model is that the technicians that go into people's homes to get them connected to and teach them how to use the internet—because in many cases, these are elderly folks who are not particularly literate with digital tools—is that they're not just technicians, they're also organizers. And they receive a political education with a curriculum that draws on folks like Grace Lee Boggs. When they go into people's homes, it's not just this act of service provision, it's an organizing effort. It's an effort to get people connected, not just to the internet, but to one another. To build a greater sense of connectivity within these neighborhoods so that they will be able to organize more effectively and achieve a greater degree of self-determination for the working-class communities of Detroit.

To me, that's an inspiring model for thinking about how to deprivatize the internet. Because when we talk about deprivatization, it's not simply the political economy that needs to be transformed, it's also how we experience the internet and how we interact with one another through the internet. Our experience of the internet in a privatized model is deeply solitary. We use the internet in the privacy of our bedrooms, or in the glow of our smartphones. What if our experience of the internet could be a more collective one, and one that brought us into relationships of solidarity and mutual support with other people in our community? So to that end, I think what the Equitable Internet Initiative is doing could provide a promising starting point for thinking about connecting differently through the internet.

That's a really cool example. What would you say to someone who reads the book and says, “Hell yeah, internet for the people, let's do it, let's fight for it.” What should they do? How do they get involved in the struggle?

Throughout the book, I make the point that to build a better internet, we need a social movement. There were often very good ideas about how to create a better internet throughout history, and the reason that those ideas were not enacted was the absence of a social movement that was powerful enough to demand them, and to demand them over industry opposition. But when I say that we need a social movement, I'm reluctant to say that we need a social movement that's solely devoted to deprivatizing and democratizing the internet. Because there are so many issues that urgently demand our attention at the moment—climate change above all—and organizers’ time and resources are so limited that I would be reluctant to try to make the case that the internet should somehow take priority, that it should be bumped to the very top of the list of the things that we need to organize around.

Rather, I’d encourage us to think about the internet as an integral part of any organizing campaign because the internet is entangled with everything. So whatever issue you're organizing around, there's an internet piece to it. You may not be interested in the internet, but the internet is interested in you. So if you're organizing around climate change, at some point you are going to have to reckon with the role that data centers, for instance, and the computationally intensive software that powers very profitable internet companies like Facebook, plays in climate change. These data centers, because the software they run is so complex, typically use a lot of energy. Much of that energy comes from carbon-intensive sources. And these data centers are only growing, particularly when one considers the role of very large machine learning models, so-called large language models, and their extraordinary levels of energy intensity. That's one example among many of why climate organizers have to think about the internet and have to think about how to transform the internet.

So I present the internet as something that people will have to engage with in the course of their normal organizing activity. And when they do, hopefully they can draw from some of the ideas that are developed in the book in order to have an analysis that's consistent with the rest of their politics. One of the things that I've noticed is that people on the Left will often have a very sophisticated critique of healthcare dysfunction in this country, or the capitalist roots of climate change, or the relationship between capitalism and white supremacy and the carceral state. But when it comes to the internet, when the internet surfaces in their organizing work, as it inevitably does for the reasons we discussed, they might default to a liberal anti-monopoly framework, or call for more regulation, or reach for certain ideas that aren't really native to the Left tradition. So, among other things, I'm trying to provide those folks with a set of analytical tools and a set of prescriptions that should feel closer to their political home.