Art with a Life of Its Own: A Conversation with Gary Hustwit

By Eric Skillman

Interviews—

Jun 18, 2026

Art with a Life of Its Own: A Conversation with Gary Hustwit
Share

Over the course of his first three documentaries—Helvetica (2007), Objectified (2009), and Urbanized (2011)—Gary Hustwit established a clean and clear cinematic language that he used to describe the complex and often contradictory systems of thinking that designers use to shape the world around us. His latest film, Eno (2024), a kaleidoscopic portrait of the legendary musician, artist, and producer Brian Eno, strips that language down to its basic grammar, then allows it to rebuild itself. Eno is an ambitious formal experiment in what Hustwit calls “generative” filmmaking. The term is fraught these days, but there’s no AI slop in Hustwit’s approach. Instead, Hustwit and his collaborators have designed software to edit and assemble real footage on the fly, surrendering direct control of the final edit to foreground the very human choices happening at the structural level. To celebrate the arrival of his films on the Criterion Channel, I sat down with Hustwit to discuss his approach to designing films.

Let me start with the most basic question: What, in this case, is a generative film?

It’s a film that’s created dynamically in software, according to rules created by human beings. It’s pulling from this much larger pool of material—edited scenes, music, raw footage—and it’s constructing a film from that material that’s different every time. But the system knows how to put together a narrative arc, regardless of what the individual pieces in each iteration are, so that it still tells a story about Brian Eno. You’re just getting a different story about him every time you watch it.

There’s this idea that the essence of art is in the artist’s choices. “You need to cut away the parts of the marble that aren’t David.” But you’re kind of avoiding that here. You’ve left at least those final editing choices to the software.

I think you have to take a bigger-picture view of it, because I’m still making all those choices in the sense that I’m curating all the material that is available for the system to choose from, and I’m designing all the ways that those things can interact. If you look at it from a conventional filmmaking lens, then yeah, I am not controlling the content of each iteration of the film. But I’ve designed it—with my team, an incredible group of people, who put thousands of hours of work into making it possible—so that when I click “generate,” it’s going to work every time. And there’s a freedom to that, as a creator. I get to be surprised by my own film every time.

 

People have asked me, “Are there versions of the film that come out that you don’t like, or that you wouldn’t normally have made?” And there definitely are! There are times where the system will come up with something at a screening, and I’ll scratch my head and go, “God, I might have maybe put that scene in a different place. Maybe this isn’t my favorite version of the movie.” And then somebody will come up to me afterward and say, “This was my favorite version of the movie so far!” So I came to this realization that it’s all subjective, and I don’t need to be so precious about each individual iteration.

Can you give me an example of what kind of parameters you’re defining? For example, in the version I saw, we had a couple scenes early on where Eno asks, in the middle of a conversation, “Is it lunchtime yet?” kind of urgently. And then toward the end—again, of the version I saw—we get this explanation of his new philosophy of skipping breakfast. And I laughed out loud at it; it was this great punchline that had been set up the whole way through. But there’s probably a version where the punchline comes before the joke and it plays very differently. Is there a comedy parameter?

No, and sometimes those things just happen by chance. But there is a thematic parameter to the film. If a scene is selected, the system is more likely to place it in a part of the film where it thinks it’s going to make sense thematically and then figure out things that could come before and after it that might work well. It is very—intentionally—chaotic sometimes. But we establish early on that, scene-to-scene, you could be anywhere. You could be in the 1970s, you could be with Brian in the present, you could be with a band that he’s working with. And I think there’s an excitement about that too. I like films where I don’t know what’s going to happen next. That’s part of the reason that we let the software show itself in the film. There are certain points where you’ll see code on-screen or you’ll see it scrubbing through different files in the data set, as a signal to the viewer that something’s happening now. In reality, most of those choices that the system is making, it makes in like a thousandth of a second at the beginning of the output. But there were places where we wanted it to show its process a little bit.

Documentaries by Gary Hustwit
Watch Now On
Top of page and above: Eno

At one point, toward the beginning, there was something that looked like a serial number in the corner?

Oh yeah! That is a real number. Yeah, each iteration has a generation number in the opening titles.

So could you, in theory, generate that version again from that number?

No, it would never be the same. Even if we forced it to use the same footage, it would still not be the same because there are unique things—the transitions, and some whole scenes—that the system is making in real time for that version alone.

 

There was no precedent for how to make a film that could change every time. How much should it change? Or why should it change? A lot of our development process was about trial and error, too. There were outputs early on that just were horrible, that did not work, that were too random. And ones that were too structured. We just kept experimenting until we found what we thought was a good medium for it. That process of inventing the software, digitizing hundreds of hours of Brian’s archival footage, and filming all the new interviews took about four years.

Which are the chicken and egg here? Did you have the idea for this type of filmmaking and thought Eno would be a great subject for it, or did you want to make an Eno film and this grew from that?

Initially, I was getting bored with the constraints of filmmaking—or mostly film exhibiting. You know, going out on screening tours for my design films. I’d be going to sixty or seventy cities for each film, pressing play, and then going to a bar or something for ninety minutes and then coming back for the Q and A. Because I can’t watch the film anymore, I’m completely sick of watching it! My background was in music before I got involved in film. So it felt like I was a band, but for my tour I was just going onstage and pushing play on my record and walking off. So I was dreaming of a way of showing a film that was more like a performance, more like music. Not just the same thing over and over again. That was the initial impetus.

 

Brian did the soundtrack to Rams (2018), my previous film. At that point I had approached him about making a film about him, and he said no. He’d turned down many filmmakers, because he thought bio documentaries were horrible, they were just one person telling their version of another person’s story. But I have this incredible friend, Brendan Dawes, who’s a digital artist and coder and just a super creative individual. I asked him, “Hey, I want to make a film that could be assembled in software and be different every time it’s shown, but still feel like a cinematic documentary. Do you think we could do that?” And he thought we could. We started working on an early demo of the platform, using the raw material from Rams, including some of Brian’s music. After six months of doing that, we were already making an Eno generative movie, in a sense, by using that as our raw material. So I decided to approach Brian about the idea, and showed him the demo. He loved it and agreed to participate. I still don’t think he wanted to have a film made about himself, but he wanted to be part of this experiment. I knew he was thinking about these kinds of ideas, different ways of trying to make art that could have a life of its own.

Rams

You’ve been working on this project for many years. I imagine when you started, you wouldn’t have predicted that you’d be releasing it into a world overwhelmed by AI discourse. But it seems to me that the thing you’ve built is actually kind of the opposite of what we think of as generative AI. I think of AI as basically built to elide the whole artistic process. Just press a button and here’s your art. Whereas your film is largely about its process, right?

It’s also organic to Brian’s process. But I think that the main difference is that this is all our own material. We’re dealing with real footage; the system is not making anything up. The algorithms are written by us as the artists. There is so much human creativity in this—I would say more than even a conventional movie. The reasons that we designed the system were not to make a movie cheaper or faster, or to cut people out of the process. This was a capability that we didn’t have as filmmakers, to make a film that could change and could tell different stories every time. And so we invented the software to get to that capability. It wasn’t the other way around. Sometimes I hesitate to even use the word generative in what we do because people automatically assume it’s generative AI. But there’s been generative software for decades now. I mean, Eno was writing generative music software back in the ’90s. It’s not something that just happened with ChatGPT. I think there’s this whole universe of new cinema that our technology can enable that I’m excited about exploring. Those are the projects that we’re working on now, in addition to continuing to evolve Eno. We’ve added hours of footage to the film since it premiered. I just added another thirty minutes last month.

I found that watching Eno recontextualized some of your earlier films for me in interesting ways, in terms of narrative structure especially. Helvetica is pretty chronological, right? You’re telling a history and a relatively straightforward chronology. Then by the time you get to Urbanized, you’ve already expanded into a more kaleidoscopic style of storytelling. Can you talk a little bit about that progression?

When I first started thinking about Helvetica, I wasn’t thinking, “I want to make a film about this one font, let’s go!” I’d been a design geek and I had played around with typography and graphic design for fifteen-plus years before I got involved in making films. So I really wanted to watch a film about type in our environment and all the words that we see every day. But one didn’t exist. Originally the film was going to be more expansive and a bit more like a tone poem. And it was around that time that Helvetica had a resurgence, and it was very polarizing. So I thought, “Oh, this is a kind of interesting mix here. Maybe I could do all those things that I wanted to do with the film, but I could do it through the lens of this one typeface.”

 

But in the subsequent films, it was more about all these different things that connect. They were more like surveys of these areas of design that I was interested in, but that nobody had made a documentary about. That’s always been my big factor in my creative choices. All the films that I’ve made are things that there was no documentary about and I was just like, “God, I really wanna watch the documentary about urban planning!” They’re seemingly boring subjects that I was personally obsessed with and couldn’t believe there weren’t films about already.

Urbanized

In those early films, your subject matter is very visual, obviously. It’s typography. It’s design. But then you get to Eno, and while Eno’s work has some visual components to it, it’s primarily audio. Did that affect your planning or thinking at all?

It didn’t, because all the other films are still about the creative process, and that’s something I’m always fascinated by. I want to try to decode what these people do and how they do it and why they do it. I think Brian is like the pinnacle of creative experimentation: the work that he’s done with his own music, with all the bands that he’s produced, and then these tools like the Oblique Strategies cards and other games that he created to disrupt his creative flow or inspire others.

Eno’s form of creativity is an interesting contrast with the methodology of a lot of the designers and planners that you’ve talked to in the previous films, who have a tendency toward rule structures and toward creating universal principles, like Dieter Rams’s ten principles for good design. Then Eno is on this other end of the spectrum where he doesn’t have rules, he has strategies. Or, maybe they are rules, but they’re only valuable insofar as they’re useful. On that spectrum of rules to strategies, where would you place yourself?

Obviously with this movie and the approach that we took, it’s the complete opposite of when I started making the design films. There was very much a sense of construction and obsessing over sequence: controlling the story and controlling the audience’s experience. But after I’d made the Design Trilogy films, I was burning out. I felt like I was making almost the same film, but with different subjects. I wanted to disrupt that. And it was interesting because around that time, I went and talked to Massimo Vignelli, who is in Helvetica. He’s an incredible, legendary designer, he did the whole New York subway system signage and map design, and so many other projects. I was trying to tell him that I felt like I was making the same film. And he said, “You invented a new language with Helvetica. Don’t be afraid to keep speaking it.” And I thought that was really great, wise advice . . . but I still felt like formally I needed to change things up.

Helvetica

I love that quote because it defines everything that the modernists were trying to do, right? They had this idea that there’s a universal best design, that there is some universal truth of design. And then Eno is, in some ways, the opposite.

Well, like Vignelli said, designers can make everything from a spoon to a city. You have a point of view, you have an aesthetic, you have a strategy. You have a design philosophy. It doesn’t matter if you’re designing a phone or a table or an education system; you’re going to use the same principles and the same methodology for every project. And what’s fascinating about Eno is he does the opposite, intentionally. He reinvents his process with every project, even if it’s the same type of project, or if he’s producing yet another record. He’ll intentionally force himself to change his process to not get stuck in a creative merry-go-round.

I notice there’s a throughline in the Design Trilogy, I think especially in Objectified, about finding the hidden humanity and intention underneath modern manufactured life, right? That everything has a designer or a human that was behind it.

Yep.

And your films show the stories of those people who are lucky enough to have been able to have this measurable impact on the world. But what do you think about people who maybe don’t command industrial resources, or who don’t have the Braun brothers sponsoring them? Is there a way to have a creative impact in a smaller way, on a smaller scale? Or is it a symptom of modern life that everything has got to be—

—has got to be commodified? Yeah, obviously, everything we do after we open our eyes every day is some sort of creative act, you know? How we arrange our home and the clothes we wear and what our haircut is. It’s all a form of self-expression. And the things that we make, even if they’re of a very modest scale, are that too. I think it’s all creativity and it’s all valuable and important for us to do. It helps us process what’s happening in our world. I think art has always been that for us. It’s a way of feeling like we’re part of something bigger. So yeah, I think we’re all designers. We’re all creatives. The scale doesn’t matter. We’re all creating all the time and it’s all valuable.

Eric Skillman is a designer and art director at the Criterion Collection.

More: Interviews

Charting the Rise of Trans Filmmaking with Caden Mark Gardner and Willow Maclay
Charting the Rise of Trans Filmmaking with Caden Mark Gardner and Willow Maclay

The curators of a showcase of trans directors now playing on the Criterion Channel discuss the work of these trailblazing artists, who have brought new layers of nuance and insight to cinematic depictions of their community.

By Andrew Chan

On Film / Interviews — Jun 5, 2026
Writing About Cinema: A Conversation with Peter Cowie
Writing About Cinema: A Conversation with Peter Cowie

The longtime Criterion Collection and Janus Films contributor reflects on his role in the golden age of art-house cinema and his life as a film critic, historian, publisher, and festivalgoer.

By Liz Helfgott

On Film / Interviews — May 28, 2026
Lumière, le cinéma!: A Conversation with Thierry Frémaux
Lumière, le cinéma!: A Conversation with Thierry Frémaux

The Lumière brothers were not just inventors who brought about major advancements in motion-picture technology, they were also cinema’s first artists, capturing modern life with dynamic compositions and ingeniously choreographed action.

By David Schwartz

On Film / Interviews — May 19, 2026
On Restoration and Repair: A Conversation with Ja’Tovia Gary
On Restoration and Repair: A Conversation with Ja’Tovia Gary

The shape-shifting artist and director of The Giverny Document talks about the Black feminist tradition, her approach to direct animation, and the influence of Toni Morrison and Soviet montage theory on her work.

By Beandrea July

On Film / Interviews — Apr 22, 2026

You have no items in your shopping cart

View Cart Check Out
Close Menu

SEARCH

For streaming, search the  Criterion Channel