‍

Dmitri Cherniak on Strictly for Art

‍Dmitri Cherniak is participating in the group show Infinite Images at the Toledo Museum of Art. He discusses the show in a conversation with Peter Bauman (Monk Antony), reflecting on the growing presence of digital and code-based art in museums. They also cover the changing nature of automation and the role of engineering in contemporary digital art practice.‍

‍Peter Bauman: You’ve mentioned that a long-term goal of yours is to have more generative art in museums. In the last 12 months, we’ve seen a significant increase in just that: Electric Op, Electric Dreams, Radical Software, PST and more. Now with Infinite Images, we have a show even more explicitly focused on generativity. How do you contextualize the importance of this show and this moment? To what extent do you see it as a turning point for digital art?‍

‍Dmitri Cherniak: It makes a lot of sense because in the past few years the popularity of generative art has increased by an order of magnitude—initially by technological enthusiasts. Now also it's broadened out to a larger group of people so that when I'm talking with museum curators, they are starting to piece this together as a line from Sol LeWitt and Constructivism to where we are now.

It seems like a lot of them, at least more forward-thinking ones, get it and understand its place in twenty-first-century art history. I think that’s really important because this is a continuation of that history; it is not just a niche community, especially compared to what it used to be. Since I was living in New York—now for almost 15 years—I've been going to new media and code art shows, seeing people like Danny Rozin when he was doing his mirrors early on. I always felt the art, the quality and the conceptual work were there, but it was just very much catered to a smaller audience. I'm really glad now that more people will have the chance to appreciate it.
‍

I still think there's definitely a divide between the actual technology that gets used and a more general audience. 

But as more and more people are forced to engage with technology—in the workplace, at school, in their everyday lives—I think they really have to contend with it.

I'm glad that it's finally being canonized in these institutions. They also are looking to the future in a way that perhaps they haven't before. From a demographic perspective or the idea of digital art being a movement, they're much more receptive and trying to showcase that because they realize they might have to appeal to a younger audience now as well.
‍

‍

‍Peter Bauman: Yeah, museums are dealing with their own post-COVID and, especially American ones, political realities with funding. This all does seem to have made them rethink strategies that may not be working with a newer generation of museum visitor. How do you think we can keep positive momentum going for contemporary digital art?

Dmitri Cherniak: Putting on good shows and bringing in new people to learn about it, having good curations, having explanations. I know that some people are of the mindset that you shouldn't have to explain your art and I'm totally okay with that. I'm very happy to explain my art because I feel that there is a divide between generations and the tools that they use.
‍

It’s worth giving context into what creativity means for a technologically more engaged generation, building these bridges and trying to explain.

Many times I've given talks at institutions and people have come up to me afterwards and said, “Oh, I thought that this blockchain thing was just social investing and monkey pictures. I didn't realize there was conceptual art that was such a core element to this movement.”

So it can help more people understand and appreciate it. Along these lines, I’m really impressed with what Beeple has built with his studio. He made an amazing point, which is that these institutions are putting up these large screens now. This gives them the opportunity to show so much art. Even if you're showing the main exhibition five, six hours a day, they can show other art at other times and highlight many different artists.

There's so much more possible engagement; it is actually quite powerful, his thinking, which is basically, “I'm going to try to build the best version of a new type of institution. I'll get to be at the forefront of understanding what these museums are thinking about with installations or these digital rooms.”

That's quite clever. That is the way I see a lot of museums going. Whether that ultimately works out for them, I actually don't know.But we have seen these all-encompassing exhibitions can be quite powerful when done well, for example, Ryoji Ikeda’s data-verse trilogy.

‍Peter Bauman: There does seem to be a new energy in digital art in parts of the States.‍

‍Dmitri Cherniak: Yeah, the reality is how many digital galleries have opened in the past few months in New York. That is a good sign.
‍

We don't want the new media, code and generative art landscape to be some big fish in a small pond.

I would much rather we are smaller fish in a much bigger world, helping people appreciate the art in the context that it exists.

Over the past several years, that has started happening. There's an order of magnitude more people employed in the industry, which is a good sign.

‍Peter Bauman: We can thank digital curators throughout the world who’ve been doing this work for decades. People like Tina Rivers Ryan, Christiane Paul and many more. I think people also look to you for that context because you’ve become a prominent part of digital art’s progression in the twenty-first century. Since you made your first work as a college student in 2009, what major turning points in digital art have you witnessed—even taken part in?‍

‍Dmitri Cherniak: Another curator to mention is Michelle Kuo at the MoMA, who is now chief curator. It's amazing that someone who is in that position is so deeply engaged with the art form. That’s not something that I think we've had before.

The Christiane Paul show Programmed at the Whitney in 2019 was a huge moment because I felt like it demonstrated the potential for what I knew and what I loved. It was in a significant museum in New York and curated so well. You had a Sol LeWitt wall drawing next to Casey Reas work, and you can just see the connections. I even said to my wife when we went together:
‍

“This is inspiring. And not only that, I feel like my work could be here one day.” 

Then I do think working with blockchains was quite important because it gave collectors a way to engage with the software directly, especially with the on-chain work. That was so different from what I had done before, which was essentially building a system and printing one-offs of screenshots. You have to imagine at the time, too, these works were selling for a few thousand dollars. The idea of printing and framing a hundred works for something that you barely had any wall space for and had no idea who was interested in it. That wasn't something you could really ask of artists or galleries. Rune Madsen or Casey Reas would have works, multiple versions, but maybe they were pinned-up prints as opposed to something properly framed.

Blockchains were such an important moment, personally, just for people being able to collect these works in a way that made sense and was native to the actual code.
‍

It just felt for the first time that I was not encumbered by the economics of producing physical representations of what was my core work. 

It wasn't called long-form generative art at the time but I think that was what I was hoping to accomplish—to highlight the artist giving up control but also being able to show off the diversity of the algorithm.

For me at the time, that was such a core idea of what I thought a good creative coder or generative artist could do.

‍Peter Bauman: Zach Lieberman a couple of years ago also mentioned to me the importance of him seeing a show at the Whitney—for him it was BitStreams—that made him feel like code art belonged in museums. It shows how powerful these shows can be.‍

‍You mentioned earlier this conversation with art history and your practice perfectly demonstrates this with your connection to the László Moholy-Nagy estate.‍

‍Moholy-Nagy brought ideas of Constructivism to the Bauhaus, which saw engineering and artistry as integrated, not distinct. They aimed to connect technology and art. You’ve described your work as a demonstration that engineering is a creative practice. How do you see the relationship between artist and engineer today? Do you see your practice as extending this Constructivist legacy of engineer-artists?‍

‍Dmitri Cherniak: While I have seen a lot of engineers focus on code reusability, legibility and program structure as a form of artistry, unfortunately in practice I have seen quite a big divide between the engineering mindset and art for art’s purpose. Maybe that wall is something people have built up or it's just a cultural blockage because I actually see them as being naturally quite close.

‍
But the divide goes both ways. There are extremely talented artists who have reached out to me wanting to learn more about code and generative art. I'll spend the time going through something with them, trying to teach them something. Then they just put it into ChatGPT and say, “Oh, yeah, I just made the changes this way.”

They didn't really learn from it. It was just a platform to do one thing. Then they left it and didn't follow through.

But for me, as an engineer, I was put in this position where I was told from, let's say, the business side of things that, “Okay, we need to accomplish this nebulous problem. We need to understand all of this unstructured data” before machine learning was in use in most businesses.

There was a lot of creativity in how that would work. There was also a lot of scrappiness. There was sometimes cutting-age technology and sometimes you had to do it yourself. But ultimately, it was never something black and white.

I was just mad, to be honest, that other people couldn't appreciate it. That made me think a lot about artistry—even to the point that when I was the CTO, when people asked what I did, sometimes I would say, “I'm an artist.”

I couldn't think of who else besides artists straddled that boundary where you were solving these open-ended problems for some benefit.
‍
‍Peter Bauman: Is that why you’ve long said that automation is your artistic medium?‍

‍Dmitri Cherniak: As a kid, I loved to draw. Art was one of the things that I just loved to do but I was good at math. So society was like, “You better do math because that's how you have a chance for social mobility or success” or whatever that means. Ultimately, the toolset that I learned was still used for art: for drawing, for sketching. It was just a different toolset.

Automation, for me, was this creative process where I was solving these problems and doing it in a way that required this balance of having enough knowledge and understanding, but then also giving up some of the freedom. Giving up some of that freedom is where the creativity and the nuance really comes in.

‍Peter Bauman: Working with automation as your artistic medium, you’ve developed a deeply nuanced understanding of it. I wonder how you’ve observed automation changing over the last fifteen years.‍

‍Dmitri Cherniak: A lot of people are thinking, “Now with LLMs and GPTs, what's going to happen to automation?” I feel they're quite overrated. I don't know exactly what's going to happen. In my experience, I've played around with using some of them for some code. One of the things that people say is this is going to replace coding entirely.

But as an exercise, I was like, “Okay, let me try to do this using the GitHub Copilot.” And LLMs can do basic things, sometimes quite nicely. But as soon as you start to get into anything a little more complicated, it doesn’t deal with all of the actual issues that you might need to do.

For some of the easier tasks, I feel like engineers could have already automated some of that in the first place.

‍Peter Bauman: I’d also like to connect ideas of automation to your relationship with Moholy-Nagy’s estate. He once said, “The human construct is the synthesis of all its functional apparatuses, i.e. man will be most perfect in his own time if the functional apparatuses of which he is composed his cells as well as the most sophisticated organs are conscious and trained to the limit of their capacity.”‍

‍To what extent do you see automation as one of those apparatuses? In the way that Donna Haraway speaks of the human as cyborg, does your work consciously build from that lineage?‍

‍Dmitri Cherniak: I definitely agree with his idea in the quotation. I highly recommend Vision in Motion, a posthumously released manifesto by Moholy-Nagy. One of the things that he really speaks about is the idea that we have to use technology; we have to use industrial materials as artists.
‍

We have to use these things that are created in our time strictly for art, strictly for creative purposes, because if we don't, they're only going to get used for economic and political purposes. That is going to be the biggest problem of our time. 

If you look at what's going on now, I would argue that is a huge problem and a disconnect. It's one of the reasons that I really appreciate not just the aesthetics of his work, but the thinking behind it and how he went about teaching people. That is where I found myself, obviously not as elevated and not as experienced as him, but in my own little pocket. I was feeling the same way where there was just a disconnect between this, and I felt it was important for people, not just as a cog, but for people.

Some people are much more into the idea of being a cyborg and that integrating into you as a person. I guess I see them more just as tools. To me, that doesn't cross the boundary into actual personhood; it's a tool for you to use and engage with.

‍Peter Bauman: I’m interested in your first digital work, which was Taylor Swift ASCII Art from 2009. It’s AI, machine learning art from 2009, when not many people in the world were making those connections. It also seems quite prescient because today, the idea of AI work that's competitive, networked and simulated like a game—just like your Taylor Swift ASCII Art—is becoming more popular.‍

‍Do you see that project as anticipating some of these major trends—not only the rise of AI art, but automated, agentic AI art?
‍‍

Dmitri Cherniak: Looking back, Taylor Swift ASCII Art was just me trying to embed some emotion, or feeling—just my own personality—into this super fast hyper-networked competition that we didn't really have much of a say in. 

We built this agent and then they would play against each other. So obviously, anthropomorphizing ideas or concepts has been around; that's not something new. But for me, it was just a friendly competition with my friend. He was Miley Cyrus and I was Taylor Swift. The idea was they were competing for who's number one. Definitely part of it was the competition but then also I wanted something visual.

You couldn’t actually get to watch them play because it happened so quickly. But you got a chance to look through the logs of the competition.

‍Peter Bauman: What was the actual competition? A variation of Go?‍

‍Dmitri Cherniak: Yeah, the game was a modified version of Go. It was called Philosopher's Football. Go at the time was the game that even the AIs weren't close to the human champions. Then not that much later we got AlphaGo and saw that just completely crumble.

For the work, I wanted people to see, when they looked through the logs, Taylor Swift blowing them a kiss when they got beaten. It was just a little cute thing. It was in the logs so needed to be character based and that's where I came into the ASCII art.
‍

‍

I also did a stylized version for the term cover paper. It was still grayscale but a little more nuanced than just the plain characters in the terminal. That was how it started.
‍

‍
Years later, it turned into the show at Heron Arts because I shared that term paper online and immediately ten people asked me, “Could I get a print of that or something?"

That’s when I started to really dig into it more. It was, let's say, my escape from managing people as opposed to building things. It was a chance to have time to code again and not just manage or help other engineers with their problems.

‍Peter Bauman: Fascinatingly, that project started your digital art career twice. You’ve mentioned how Taylor Swift ASCII Art, Ringers and Light Years were all multi-year projects that still managed to feel cohesive. Do you see them linked through any single trajectory or more as separate inquiries?‍

‍Dmitri Cherniak: I think of them mostly as separate inquiries. I guess you could say they're very separate to me. A lot of my work is definitely iterative exploration, working on tools, building on things. But then with those three projects in particular, they are the projects that I feel the most proud about. They're most cohesive.
‍

When I think about releasing work now, what's important to me is that I feel proud of that work myself—that I feel like I’ve touched on something I feel is important; then I'm trying to share that.

I guess you could say they're on a similar trajectory but they're project-oriented inquiries for sure. I use some of the same tools. I build up my toolkit to help me advance. In that way, it's shared but they're definitely all conceptually different.

‍Peter Bauman: I wonder if there’s a way to link not only that work but, more fundamentally, your ideas of creative exploration. You’ve said, “I am empowered by randomness, speed and reproducibility, which lets me dig into the visual ‘space’ of an aesthetic concept in a way that you can’t when working in a strictly physical medium.”‍

‍Can we unpack that? Does working digitally shift your visual thinking from execution to navigation—more like searching than making? How do you think differently about exploring a pictorial, output or even latent space?‍

‍Dmitri Cherniak: I'll try to give you where I feel my work sits versus AI when you're talking about exploring the latent space versus the visual space. A lot of my work is definitely an iterative process. I have the tools where I can reproduce things very quickly. I have the seed, which basically serves as the DNA that you're about to express through the software. I can go back and forth in my code.

If you imagine Mondrian doing his rectangular subdivision in 2D space, he has probably sketched hundreds, thousands of these. He understands the form quite well. But if you were to give an artist software that could do a rectangular subdivision where you could manipulate lots of parameters—maybe include things like rotation—you could see thousands of those in a single day.
‍

You see so much and the opportunity to notice something and dive in is just so much greater.

So in some ways—just by brute force—you have the ability to see so many options. You can really understand or see something much more than someone who has this initial spark and works on it only once.

It really empowers you as an artist. If you're making one painting, it takes you X amount of time to make one painting. When you think of Warhol or some of this factory-era art that exists, you can see more but you're still limited and you still are dependent on labor.

With generative methods, you basically no longer are dependent on that aspect. You can take the same seed and rerun something with different code or go back to the same code if you don't like it.
‍

You're so empowered to explore the visual space. 

My experience with AI or latent space, is that it feels much more like navigating through a multidimensional matrix as opposed to seeing explicit feedback and running with it. My work, at least, gets generated so quickly. The iteration loop is so fast.

Whereas with exploring through a latent space, it's not as one-to-one connected. I actually think that there's a lot of interesting art that happens with AI but a lot of the creativity does not come through exploring the latent space. It comes through the artist conceptualizing or using that to achieve a goal or purpose.

In my work, a lot of the creativity and artistry comes through this iterative process of seeing all these outputs and being able to explicitly go back and dive down the space, as opposed to a more blackbox exploration.

To be fair, that hyper-connected, feedback-driven blackbox is kind of the special sauce of machine learning in the first place. While I like the ability to logic my way through my code to directly impact my artwork, the actual mechanism and reason why I am able to do that is more akin to how AI works.
‍

‍
----
‍


Dmitri Cherniak is an artist that creates hand-coded goods with automation as his artistic medium.
‍
‍‍Peter Bauman (Monk Antony) is Le Random's editor in chief.