“Most Art Today Is Digital Art”. Trevor Paglen on AI, curation, and machine vision

As artificial intelligence reshapes public debates around images, authorship, and cultural production, Trevor Paglen‘s ideas have taken on renewed urgency. Bringing together a decade of essays, his new book How to See Like a Machine argues that today’s AI systems are part of a much longer history of machine vision, surveillance, and computational ways of seeing. At the same time, Paglen recently co-curated Zero 10 at Art Basel, making the case that digital art belongs within the broader history of contemporary art rather than at its margins.

In conversation with Anika Meier for SLEEK, Paglen reflects on the changing status of digital art, the role of curation as a form of storytelling, and why art continues to operate according to different logics than the algorithmic systems that increasingly shape everyday life.

Anika Meier The essays in your just-published book How to See Like a Machine were written over the course of a decade that saw AI become a dominant cultural preoccupation. Did bringing them together change the way you read your own work? Did you discover a through-line you hadn’t seen before?

Trevor Paglen It was remarkable to see that it all still worked. I was actually surprised because I hadn’t reread some of the older essays until I started putting the book together. When I did, I realized that while some of the examples had become dated, the ideas still felt relevant. I was surprised by how naturally the essays formed an arc from the past to the present, and by how well they still held together.

AM You’ve been writing about machine vision and operational images for over a decade. Why do you think these ideas are only reaching a broader audience now?

TP AI has simply become much more ubiquitous. People are much more aware of it, and much more concerned about it. It’s become much more present in everyday life, whether in the workplace, at home, or through social media and the technologies our kids use. So I think the book is coming at the right time because it’s been thinking through these questions over a much longer period. When some of the essays were first published, AI wasn’t part of everyday life in the way it is now.

AM Digital art is suddenly also something people seem to be much more interested in after the NFT boom brought it to a mainstream audience. You recently co-curated Zero 10 at Art Basel, a curated initiative for digital art. It’s an art fair, so it’s inevitably a commercial setting, but it also seems to make a serious curatorial argument for digital art as part of the history of contemporary art. Was that one of the ambitions of the project, or am I reading too much into it?

TP I don’t think that’s taking it too far. The argument of the exhibition is that most art today is digital art. We’ve just been reluctant to name it as such. Gallerists, curators, and institutions have tended to keep “digital art” as a separate category, probably because in the late 1990s and early 2000s it was associated with places like Ars Electronica, which existed outside the mainstream contemporary art world. Our argument is that if you look at what artists are actually doing, almost every artist is making digital work in one way or another.

AM There seem to be two very different ways of talking about digital art. In the contemporary art world, institutions, galleries, and collectors often don’t use the term at all. They simply call it contemporary art or, as collector Julia Stoschek does, time-based media. At the same time, online communities explicitly embrace the label “digital art.” You’re part of both worlds. You’re represented by Pace, but you also work closely with Fellowship. Why do you think there’s still such a divide over the label?

TP What’s going on is that people are trying to create different frameworks through which to view the work. That’s how I interpret it. You’re prompting your audience to look at a particular artist or artwork through a certain lens, and that’s fine. People have always done that. There’s nothing metaphysical about it. It’s really just a way of telling a story. Zero 10 in Basel was meant to show that people have been thinking about these questions and working in these ways for much longer than the NFT boom. None of this was invented during COVID. There’s a much longer, richer, and more interesting history of artists foregrounding digital tools in their practice.

For me, that history also includes systems thinking and cybernetics. I originally wanted to include some of Hans Haacke’s systems works, as well as works by Hanne Darboven. We weren’t able to secure those loans, but I wanted the exhibition to reflect that broader, almost cybernetic sensibility as well.

AM It sounds like you already have an institutional exhibition in mind. Is that the next project?

TP No, definitely not. (laughs)

AM Does that mean your exhibition invites us to read the work of artists like Andreas Gursky, Hito Steyerl, and Avery Singer within the history of digital art, even if they usually aren’t labeled as digital artists?

TP I wouldn’t put it that way. I wasn’t trying to label artists. I was talking about the concept for the exhibition. I hope it didn’t come across that crudely.

AM By your definition, Andreas Gursky, Hito Steyerl, and Avery Singer are digital artists.

TP Sure, why not? They’ve all been making digital work.

AM If curating is a form of storytelling, where does that leave the role of the artist? Mat Dryhurst has argued that the creators of protocols increasingly shape culture, he calls himself a creator of protocols.

TP I still call myself an artist. Old school, I guess. (laughs) Mat Dryhurst is right to point out that protocols shape what becomes visible. Whatever the Instagram algorithm is doing will largely determine what kinds of work get seen and circulated. You could say museums are analogous in that sense. There are certain criteria a work has to meet in order to become legible within an institutional context. That said, I wouldn’t go as far as Mat Dryhurst does in arguing that the real artist is the person designing the attention algorithm. That’s certainly part of the story, and it’s an interesting thesis that’s worth taking seriously, but I don’t think it’s the whole story. Art is many things.

AM In your book, you argue that most images today are made for machines, seen by machines, and ranked by machines. What does that mean for artists? Can they still be visible without playing into the attention economy?

TP You can just make art for humans. No problem. I make a distinction between artworks and what we might call visual culture, or simply images. It’s true that the overwhelming majority of images in the world are made by machines for other machines, and that humans often have to do extra work just to see them. But I don’t think an Avery Singer painting, for example, is made for machines. It’s made for humans. It’s operating according to a different protocol. Artworks function differently from something like a facial recognition algorithm. They simply exist in the world differently.

Of course, you can make art that’s intended for machines, or art that’s made by machines for other machines. That’s perfectly possible. But art is a much more flexible category than that. You can take many different positions, address many different audiences, and work with very different assumptions. That’s what makes it exciting to me. As an artist, you don’t have to make work for the TikTok algorithm. You can if you want to, but you don’t have to.

AM But then how do you get seen? That visibility has to come from somewhere. If it doesn’t come from institutions, doesn’t it inevitably depend on how you behave online and what you share?

TP Or from the gallery show you put together in your basement for your friends. There are many different ways to do it. It would be a very bad thing if all art had to pass through the choke point of the Instagram algorithm. Take Avery Singer’s work. If you see it on Instagram, that’s not really the work. The same is true for a lot of my own work. I get frustrated putting it online because it isn’t designed to be seen that way. It’s meant to be experienced in person and at scale. There are aspects of the work you simply can’t see online.

AM In your book, you write about the “society of the psyop,” drawing on Guy Debord’s idea of the society of the spectacle. How do the two relate? After all, Debord’s theory was also fundamentally about capitalism.

TP The move I’m trying to make is this: in The Society of the Spectacle, the basic model is that humans look at images, and those images mediate relationships between people. As Debord famously put it, the spectacle is “a social relation among people, mediated by images.” The “society of the psyop” modifies that model. You look at an image, but the image also looks back at you. You’re in a feedback loop with it. It’s constantly evolving in order to capture your attention or influence your behavior. It’s much more directed, much more active, and fundamentally recursive. And it all happens very quickly. That’s the key difference between the present moment and the society of the spectacle that Debord described.

AM This is perhaps a clumsy bridge, but the word “spectacle” also comes up all the time in the art world, where exhibitions are increasingly expected to produce spectacular, shareable moments. With Zero 10, it felt as though you deliberately resisted that. Was that part of the curatorial strategy?

TP With Zero 10, I specifically didn’t want to have one thing at the center that everybody was Instagramming. I wanted something that felt more nuanced and exploratory, something that asked more of the audience. I wanted different modulations, without a single focal point that everything revolved around.

AM Zero 10 was your first time curating an exhibition. Did the experience change the way you think about your own practice? Could you imagine curating again?

TP It was the first exhibition I’ve ever curated, apart from putting together little shows with friends in graduate school. I’d happily do it again, but curating is a lot of work. I already have a full-time job making art. It’s not just selecting artists. It’s all the emotional labor, the convincing, the problem-solving, and dealing with everyone’s different needs.

AM Were you also responsible for how the works were placed?

TP It was much more classically curated than a typical art fair. We thought carefully about sightlines and relationships between works. You might see Vera Molnár, Harold Cohen, and William Latham together, and through that sequence you literally see a historical trajectory unfold.

AM One of the promises of NFTs was that artists no longer needed so-called gatekeepers or curators. They could simply make their work available for sale online and connect directly with collectors.

TP Online spaces have, in some ways, taken over the role that basement galleries once had. They’ve made it much easier to share work, and I think that’s great. The downside is that they privilege the kinds of work that are easily shared online, which is only a small fraction of what’s possible in art. I value curation because it’s a form of storytelling. It creates a framework through which audiences encounter artworks. If you place 0xDEAFBEEF next to Rafael Lozano-Hemmer’s and Vera Molnár’s work, you see the 0xDEAFBEEF work differently.