‍

Julia Kaganskiy on Generativity’s Deeper Consideration

Julia Kaganskiy curated Infinite Images: The Art of Algorithms at the Toledo Museum of Art and spoke with Peter Bauman (Monk Antony) about its focus on generativity—spanning systems art, AI and randomness. Kaganskiy reflects on curatorial decisions shaped by the museum’s collection, its embrace of NFTs and the challenges of showing digital work in dialogue with tradition.‍

‍Peter Bauman: Your show Infinite Images at the Toledo Museum of Art opens July 12, 2025, and celebrates generativity’s various forms. Why is the show relevant now? What commentary are you making on the current moment?‍

‍Julia Kaganskiy: From the outset, I really wanted to focus on this term “generative,” which I feel has become more important and widely used but still feels ill-defined. From some of your writing on Le Random, I feel like you might agree that it's a term I think is very much in flux.

There are broader definitions that speak more to the idea of an open system—artists working with systems in ways that can be analog, digital or even biological. I’m thinking of the Galanter definition. But today, when people hear the word generative, they typically think of generative AI. Then in the Web3 community, it means something very specific.
‍

I wanted to give it deeper consideration within a museum context.

That felt relevant for a number of reasons. First, thinking about the way that we design and live with systems—algorithmic and otherwise—and how these systems tend to have an element of randomness, of emergence arising from constraints. Ultimately, that means it's something never completely within our control. There is always this element of dancing with the unknown, as I've come to think of it.

Artists have worked with randomness, chance and mathematical principles for centuries. I wanted to highlight these in the exhibition as a way to ground this practice in an art historical conversation.

I have a long-term interest in how artists engage with automation. This show is an opportunity to show how artists are using automation with intent in a way where they are designing their own tools, building their own systems, writing their own algorithms and training their own models. Automation is not necessarily new but I think what is new and different today is the extreme concentration of power, whether that is power in the form of data, compute or access.

I’m looking at the ways that artists working with automation give us an understanding of a different approach, a different ethics and a different sense of possibility.

‍Peter Bauman: With generativity as your guiding thesis, how did you go about choosing artists for the program? With so many living artists, how did you pinpoint, without the benefit of retrospect, who today is really making a lasting impact?‍

‍Julia Kaganskiy: It is always nice to have constraints. In this case, I was working primarily with the TMA's collection and two private collections the TMA has a relationship with.

That provided a great starting point for material to make an excellent show. I was thinking about what would give the best mapping of the space—about showing the breadth of generative processes and generative outputs. When we think of generative art, we tend to associate it with geometric abstraction but work using GANs tends to be more figurative.
‍

‍

The TMA hasn't done many digital-focused exhibitions, meaning a lot of this material is going to be new for the Toledo audience. Even for a lot of the staff, digital art is something rather unfamiliar. So as much as possible, I wanted to incorporate an understanding of what it means for an artist to work with code.
‍

What is the materiality of code that differentiates it from working with paint or marble? How does that help us understand systems and the generative process? It shows that art can be something alive, in flux and evolving—and how an artist might work with that.

‍
‍Peter Bauman: Was this broad understanding of generativity in mind when you chose artists that traditionally haven't emphasized generative systems in their work, like Sam Spratt?‍

‍Julia Kaganskiy: Sam’s work in the show is the Masquerade, which is a new body of work created using an entirely new process for Sam. In collaboration with Nifty Gateway, he created his own generative AI engine that was trained on his archive of digital paintings—down to the brush stroke.

To me, this represented the most contemporary version of how an artist might work with a generative system. I think what's unique about Sam is that because he's an artist who has worked as a digital painter for many years, the data set can be trained down to the brush stroke. It's a remarkable data set for an artist to work with. Of course, Sam’s not relinquishing control 100% to the generative process. It’s something he's in dialogue with, as it is for many artists. He's using it to create 3D models that are extrapolated from drawings that he uses to do a life drawing process from.

He's using it to test out different compositions for his densely illustrated, large-scale paintings. It allows him to continue experimenting up until the very last moment without having this compositional lock-in, where you can realize, “I've spent months laboring on this particular composition when it would be better if I changed it in this way.” The process that he underwent to get to this system was infinitely more complex and labor-intensive than if he had just painted the works in his usual style. But now that he has the system in place, it gives him a lot more flexibility—opportunity to experiment.

‍Peter Bauman: How does this kind of system align with the broader breakdown of the show?‍

‍Julia Kaganskiy: There are four main sections in the show. The first is the Imaginary Machine. It's the historical section that looks at Vera Molnár alongside contemporaries like Sol LeWitt, Max Bill and Joseph and Anni Albers. It makes connections between her algorithmic process—both analog and digital—and some of the ways that artists of the 20th century were similarly using rule-based systems, mathematical principles and geometric abstraction.

The second section is Chance and Control, looking at the role of randomness—thinking about the way that artists have used chance operations. With generative art, specifically, it explores the role of the random number generator in computing and the way that in long-form generative art the hash informs what the work ultimately becomes. Neither the artist nor the collector knows what the work is going to look like until the moment it's minted.

The third is focused on Digital Materiality, which foregrounds the digital’s unique qualities. It’s thinking about code as a process of simulation but also exploring the specific material constraints of hardware and software—the way that artists exploit these in things like glitch.

Then the last section is called Coded Nature. It is thinking about how generative art both borrows from and, in some ways, tries to mirror the generative processes in nature. Mutation, emergence, evolution in living systems—these are all things that a lot of artists explore with generative systems.

‍Peter Bauman: You mentioned TMA, like the Tate before Electric Dreams, hadn't engaged much with digital art or its history. TMA has moved in the even more radical direction of embracing NFTs and blockchains. Usually with exhibitions like this, museums acquire work to pad out their collection as well as the show. Is that what’s happening here?‍

‍Julia Kaganskiy: Yes, the museum has recently acquired some NFTs, several of which are in the exhibition. Also, in the last two years, TMA has been hosting a digital art residency, which has been featuring primarily artists working with the blockchain and NFTs. The first one was Osinachi. The second one was Yatreda. Currently, it's Emily Xie.
‍

‍

This interest in digital art was really spearheaded by the museum's director, Adam Levine, who is a very young, bold, visionary museum director and is interested in experimenting and pushing the museum in new directions.
‍

Levine really thinks of Toledo and the TMA as a place that—because of its long history and rich collection—could also be a site for imagining what the museum of the future might look like. Embracing digital technologies and digital art is very much a part of that. 

Not just taking cues from the major museums on the coasts or places like London and Paris but a place like Toledo, Ohio, being a site of original thinking and developing new standards for the field. Last year, the museum set up TMA Labs, which is a small R&D department within the museum tasked with thinking about the role of technology within the institution at a broad level.

It's a great initiative for a museum like the TMA, which honestly has not had a major role in collecting time-based media. So unlike the Tate, for instance, which has a very impressive and extensive time-based media collection, or somewhere like Buffalo AKG, which also was very early to this work.
‍

I think the TMA is starting afresh here. With a space like Web3, which is new, they have the opportunity to be at the vanguard of something.

‍
‍Peter Bauman: It’s an example of a younger generation of museum leadership and how they can reflect changing tastes. What challenges arose from presenting numerous NFTs and what were the broader challenges of the show?‍

‍Julia Kaganskiy: What was really nice was that the museum already had a wallet and already understood what it would mean to purchase an NFT. Some of the questions we had were trying to think more strategically about what it means to borrow this work. For instance, in this show, almost everything we're showing is a display copy. We're not transferring NFTs from the collector's wallet into the museum's wallet. We're not necessarily showing things on chain because we don't need to. There were those kinds of considerations—what does it mean to show an NFT that's being borrowed from an external lender? What is the most important aspect of the work?

For the exhibition broadly, one of the biggest challenges for this work is just how technically intensive it is: the amount of screens, projectors and media players it requires; the AV support that it demands; the networking; and the power infrastructure, especially for a building that was built in the early 1900s. It's not something that a museum could do without buy-in from leadership saying, “Hey, this is something that is a priority for us and we're going to invest in it appropriately.”

‍Peter Bauman: Was it intentional to emphasize the screen?‍

‍Julia Kaganskiy: There are a handful of interactive works, as well as sculptural ones. I've tried to vary the media and display formats as much as possible. Some works just demand a screen, maybe because they are audiovisual or they have motion or what have you.

In many cases, we've shown a series of prints and really emphasized seriality as being a core byproduct of generative work. For instance, we're displaying prints of twenty-five Ringers by Dmitri Cherniak, twelve Fidenza by Tyler Hobbs and twelve CryptoPunks. In some cases, they are accompanied by a digital representation of that work. For the CryptoPunks, we're going to have a custom eighty-minute GIF that Matt and John created for us, showing all 10,000 Punks, each for half a second. I'm excited about some of the display things that we're doing here and the ways that artists have also chosen to partner with us to bring an element of physicality and tactility to some of these works.

I also have to give a lot of credit to our exhibition design partner, TheGreenEyl, and particularly the creative director, Richard The, who has been an invaluable thought partner in thinking through the logistics of display, design and visitor experience of this exhibition.

‍Peter Bauman: This show demonstrates how throughout your career you’ve had a foot in traditional and digital spaces, advocating for nuanced discussion around emerging technologies. What do you think traditional art spaces can learn from the digital? And what can artists with more technical backgrounds learn from traditional art spaces?‍

‍Julia Kaganskiy: That's a big question. I’ve encountered many people throughout my career in museums and in more traditional fine arts programs that, almost as a point of pride, disavow technology, hold it at arm's length—really try to distance themselves from it. There's a part of me that’s sympathetic to that because I do think there are many things about technology as it has been presented to us by tech corporations that are intentionally disruptive, destructive, extractive and exploitative. All of these things are true. The more we know about how these corporations work, what their incentive structures are and what they're doing behind the scenes to manipulate people, the more that feeling intensifies and is proven correct.
‍

But I am someone who has always felt that just because you dislike something doesn't mean that you can ignore it. Just because you don't understand something doesn't mean that you should avoid it. 

I think the opposite is true and that's the approach I've always taken—and the approach a lot of digital artists I tend to work with take. It's a critical investigation of technology and its socio-political, environmental and economic effects.
‍

The aim is to reveal these processes and how these systems influence and shape our world. 

But it’s also to search for spaces of possibility—the cracks that we see the light through. I wish that more of my colleagues coming from the traditional side of things would also see the opportunity of working in that way.

On the flip side, I do think that a lot of artists working digitally—historically and still today—are folks that have not come through traditional fine art circles and education. Many have come from engineering and computer science backgrounds or fields like architecture and design. Honestly, digital art programs haven't typically existed in arts universities so that just wasn't a possible path.

But I do think there is a tendency to disregard art history, even the history of technology in art, in a lot of these circles. The lack of curiosity and interest in understanding what came before and trying to situate yourself and your practice within a broader lineage can be frustrating.
‍

I think it’s crucial to think about how what's happening now is actually part of a much longer trajectory that was set in motion decades ago. 

It's always so humbling to me when I go back and I look at texts written in the '60s and '70s and think about how little has actually changed in terms of what we're concerned with—what the potential issues, threats and opportunities are.

I just wish more artists working with new tools were not only enamored by the new and the next and being the first and whatever, but also considered how maybe this isn't that new, actually.
‍

‍
‍
‍Peter Bauman: Here’s some more evidence. Curators in the past have mentioned to me the importance of artists, without that art school background, still speaking the language of contemporary art. And it may be a case of needing to learn that language—taking a year abroad.‍

‍Julia Kaganskiy: What you're speaking to is true. A lot of curators who have gone through the training of an art history or a curatorial program do tend to speak this rarefied language of International Art English. They look for certain signifiers that an artist is able to do a contextualization of themselves in their work. It's a useful tool for an artist to be aware of—to think how they can harness.

But I'm not even talking about that, about playing the capital “A” art game. I’m talking about improving your own practice, the caliber of your own work as a thinking creative person by engaging with the history of art, the history of new media, the history of technology. That's what I'm interested in for myself as much as for anyone else as someone who didn't formally study this when I first started working in the field.

‍Peter Bauman: It’s a good point: the language should come naturally. Speaking French because you love France and have immersed yourself in the culture will sound more fluent than someone learning it for work.‍

‍Julia Kaganskiy: It reflects my own journey. I went to grad school when I was 32, in part because I thought,
‍

“I've done as much as I can learning by doing. I need to go read some theory now—read some history.”

‍
‍Peter Bauman: An interest in theory and history, to me, just shows that you care; that it’s in your blood. Many artists display this sensitivity and I’m disappointed when those engaging with deep learning and generativity automatically have their work dismissed as lacking a certain depth. I’m thinking of Joanna Zylinska referring to all generative AI work as Candy Crush art, even GANs. With all the criticism of AI, how did you, as a curator, balance those with the needs and message of the show?‍

‍Julia Kaganskiy: That is partly a product of working with the materials at hand, with the artists at hand who tend to have more formalist interests in their work. That being said, I do think that my choices highlight artists who build their own tools, their own data sets or train their own models.

It does speak to a different ethics of how artists—and people more generally—could engage with generative tools in ways that feel less extractive and exploitative of other people's labor. I also gave quite a prominent space to two works that I feel do a good job of peeling back the curtain and exposing the underlying processes, labor and decision-making—showing the potential for embedded bias that can make their way into these systems.

Sarah Meyohas’s Cloud of Petals and Infinite Petals have quite a large footprint in one of the main galleries, as does Operator’s Human Unreadable. These two works help manifest the human side that is ultimately powering these systems and informing how they are designed and what they're optimized for—what is prioritized.

Most critics tackle tech corporations and the tools that they produce. But artists like Sarah Meyohas or Holly Herndon and Mat Dryhurst are engaging with the technologies as a kind of foil. They are making their own tools in a way that’s not the same as OpenAI or Google—that often points to alternatives. The fact that we tend to lump everything in together under the undifferentiated terms of “AI” or “generative art” is also a problem.

At a moment when there is such intense backlash against generative AI, I almost want to temper that a tiny bit. I think much of it is warranted and these are discussions and debates that absolutely need to happen. But sometimes it feels like we're throwing the baby out with the bathwater. Is there another way?



---



Julia Kaganskiy is a curator and cultural strategist working across art, science and technology. Her show Infinite Images: The Art of Algorithms at the Toledo Museum of Art opens on July 12, 2025.

‍‍Peter Bauman (Monk Antony) is Le Random's editor in chief.