In Search of Generative Art History

To celebrate the launch of “Ex-Machina: A History of Generative Art” at Phillips auction house, Benjamin Kandler speaks to a group of leading artists and curators about the historical importance of generative art. The exhibition — on view at Phillips, London until 5 August — threads together multiple histories, including pioneers of computer art such as Herbert W. Franke, Vera Molnar, and Harold Cohen alongside generative photographers Gottfried Jäger, Hein Gravenhorst, and Vladimir Bonačić. Coupled with artists like Casey Reas, who co-founded Processing, and emerging stars Dmitri Cherniak, Tyler Hobbs, and Rhea Myers, the exhibition makes explicit the rich traditions of machine-automated art, while confronting ignorant histories of abstraction.

Whether or not generative art would have garnered its due without the NFT is an open question. What is not in doubt is the overwhelming importance of these works to the future of digital art. The following conversation foregrounds key definitions while surfacing the overlooked histories — and latent spaces — of human-machine collaboration. It reminds us that current developments are not born in an NFT silo, but in decades of experimentation.

‍‍

Benjamin Kandler: What is generative art to you, and how would you define it?

‍Georg Bak: Generative art is based on generative aesthetics, which was defined in the 1960s by Max Bense, a German philosopher teaching at [the University of] Stuttgart. He taught Georg Nees and Frieder Nake, who became the first artists to use generative methods. Nees defined generative aesthetics as a form of art that is automated by an algorithm created by the artist. Initially, they used analog methods to create this kind of art, and slowly they transitioned to using computing systems. 

‍Anna Ridler: I think of it in similar terms. I’m working in multiple different ways when I’m making generative art — creating a data set, creating an algorithm, and then working with the output. In each of those processes, I’m following different rules but I’m also allowing something to come out of them. 

For me, there’s a very strong parallel between working with machines to create generative art and working with nature. I often compare it to gardening, where although you might have an idea of what you’re going to produce, you can never exactly tell what’s going to flower. There’s this tension between what you plan and what actually comes out. ‍

‍Douglas Dodds: I agree. I also think of it as a set of rules within which the artist operates, and there’s an element of control in what goes into it, and an element of chance in what comes out. I like your analogy of the garden. Sometimes you get a few weeds that you pull up and throw away, but among them there are beautiful things that come out as well. We tend to think of it as being a computer program that an artist uses, but it could be something else entirely.

‍Sofia Garcia: Generative art is a systems-based approach to creating art. Yes, there is an element of surprise about what’s going to come out but it is up to artists to make a clear set of rules to put into the system. I also think generative art is more of an umbrella term. What we’re focusing on right now is computer-generated art. In the show, you’ll see algorithmic art using computer-generated algorithms, computer programming, AI, and data sets with Anna’s work. So generative art is a term that can be used to describe various methods of creation. Ultimately, however, there must be a sense of autonomy in the system that’s being initiated by the artist.

‍‍

BK: Sofia, you’re known for your work with generative artists like Matt DesLauriers. How do you pick your artists and what do you look for? 

‍SG: It’s important to know that I have a background in computer science, engineering, and in the arts. [For artists] it’s not enough to just have great technical skill, there has to be a conversation between technical skill and artistic ability — their understanding of color and composition, for example. I think their ability to tell you who they are with their work is also very important.

‍BK: Georg, we’ve been working together for eight months to put together this show. How did you choose some of the works?

‍GB: We initially had the idea to do a show where we present the pioneers of computer art for the first time. For me, it was important to have major pieces. The title “Ex-Machina” actually comes from the history of computer art. An essay by Abraham A. Moles was called “Ars Ex Machina” and later an exhibition planned by Herbert W. Franke — one of the earliest pioneers of computer art — was intended to be displayed in Vienna. What was going to be a wonderful show sadly didn’t happen because they couldn’t fund it. But then somebody contacted him and said: “Okay, let’s bring this show to Linz.” But it was more than just a show, it was a kind of festival. And that’s how Ars Electronica was founded. 

I thought it would be a good title for a show — to do something similar to what Franke had envisaged, and put together an overview of the most talented artists of our time, to contextualize them against some of the pioneers of computer art. ‍

We have works by Herbert W. Franke, Frieder Nake, Georg Nees, [and] pioneering generative photography from artists like Gottfried Jäger. We also have some major artists from the New Tendencies movement, like Tomislav Mikulić and Vladimir Bonačić. We also have some interesting artists from the net art movement like UBERMORGEN and a few contemporary artists that are quite relevant. 

‍BK: Douglas, you’re known for helping to build one of the world’s largest collections of generative art at the V&A [Victoria and Albert Museum]. How did that come about and how did you build it?

‍DD: Well, it wasn’t just me. In fact, I can’t claim credit at all for the first works that came into the V&A’s collection. Digital art didn’t just start ten or 15 years ago, you can find it as far back as the 1960s and arguably even earlier than that. There was an iconic exhibition held in London in 1968 called “Cybernetic Serendipity” — one of several that took place that year — so there was a kind of zeitgeist. Amazingly, the V&A collected its first computer-generated images as a result of that show. In 1969, the museum acquired a set of seven prints that were published in conjunction with “Cybernetic Serendipity” at the ICA [Institute of Contemporary Arts]. 

Some of the curators at the time weren’t that keen on them. There was a big argument in the museum — I could see the official correspondence about it — with some saying: “Why on earth would anybody want to collect this computer-generated stuff?” But eventually they did. ‍

After that, not a lot happened for a few decades — we collected a few works by Manfred Mohr and Vera Molnar, and one or two others. [...] It seemed obvious to me that we were experiencing a digital evolution, and that museums ought to be collecting this kind of thing. [...] People started giving things to the V&A because they were desperate to receive some sort of public recognition at a time when practically nobody was interested in it. It started with the acquisition of two major collections — one from a private collector called Patric Prince, and the other from the archives of The Computer Arts Society (CAS) that was founded as a result of the “Cybernetic Serendipity” exhibition. The V&A suddenly had five or six hundred artworks from those two sources alone, and it just grew from there. 

‍‍

BK: Anna, we speak a lot about very tangible artworks, which are somewhat easier to collect compared to data, AI, and GANs (generative adversarial networks) which are notoriously hard to control. How do you work with collectors to ensure that your work can be collected and conserved — what’s the process?

‍AR: It depends on the project I’m working on and what the end result is. One interesting aspect of working within the lineage of generative art is that it can be incredibly conceptual. I’ve started doing some NFT-driven works where the work is the smart contract itself. [...] Other times the work can be more tangible: printed data sets or thousands of photographs that are all hand labeled, or it can be a moving-image piece. 

That’s why I think generative art is very interesting, because it’s not one thing. It’s this umbrella term that can include moving image, prints, code, and even photography. So, for me, it’s important when I’m selling that collectors understand how a single project can actually have multiple different outcomes, some of which are tangible and very saleable, and others are less tangible, and more esoteric. [...] I have institutional collectors and crypto collectors. It’s all over the place. 

‍BK: Sofia, how does the split of both crypto and traditional collectors work for you? 

‍SG: Well, there’s the archetypal crypto collector who enters the space more for financial reasons. But you find a lot of these collectors become increasingly engaged with the art history side of things, trying to figure out: “What is this that I’m looking at?” [...] Then when it comes to traditional art collectors, I think most of them are interested in the dynamic between human expression and technology. 

Still, there is a lot of education that needs to happen around generative art on the blockchain, especially now in this blooming crypto space. NFTs and generative art work very well together, and have grown into something that we never would have expected. For collectors, the code itself is part of their acquisition. A few years ago, there were some works on [sale at] auction that would allow you to acquire the work as well as a printout of the code that produced it. Now, you’re acquiring the performative act of running it, rather than looking at a print or another output or video representation of it. It is a performance. Getting that message across to collectors is always really fun, especially in the traditional art world.‍

‍BK: Georg, why was it so important to include contemporary on-chain works and contextualize them [alongside] 70 years worth of photographic prints and sculpture?

‍GB: Well, I think a lot of these ideas have been around for decades, but nowadays technology has produced more advanced ways to display art. It’s quite spectacular. AI in the early days was very simple by comparison. Harold Cohen made AI pieces in the 1980s, and I think if you compare them to what’s happening now, we are almost witnessing a digital renaissance. First of all, this whole PFP thing that’s been happening in the last few years is amazing, how it has changed our whole entertainment industry, and by extension, society. You can reinvent marketing with PFPs. 

I also think the NFT movement has a lot of connections to art history, whether it’s the Situationists in the 1960s or Dadaism. It’s always interesting to see the context. And in this show, of course, we have the context of computer art. Still, there are other contexts that surround NFTs and digital art.

‍BK: [Opens up the panel to audience questions]

‍Anon: Benjamin, what is it like to be at an auction house now, at a time when DAOs are a substantial part of the collecting community?

‍BK: It’s fascinating because DAOs have a lot of buying power but they also present a brand new model for collecting. Traditionally, collectors have been quite fragmented. I think [DAOs] are a great way for art to become more accessible and [to allow] people to share and interact with each other. 

From the perspective of an auction house, the anonymity of DAOs can be a little bit tricky at times, but as they become more commonplace and more integrated with the traditional art world it will become a lot more normal. And we’ll see a lot more of them. ‍

‍Anon: How do you attribute value to emerging artists? Do you use more conventional standards? Or do you have a new set of parameters to look at, say, the code or the development potential of a piece?

‍SG: The code is really important, but for the most part, you don’t actually have to study it. It’s really interesting because a lot of these artists, especially in the NFT community, were hobbyists for the most part, who shared their work online and explored methods of expression using software. A lot of them were engineers. Seeing their talent is really important, but there’s a necessary technical skill that’s easy to overlook when you can’t see what’s happening behind the scenes. It’s a very human experience to write and produce the code, and it’s actually quite difficult. 

‍DD: Some of the same criteria still apply though: cultural significance, innovation, technical expertise, whether anybody is ever going to want to see it, etc. There’s no point in acquiring something if it’s just going to disappear without a trace. There’s a range of things that galleries take into account when they’re acquiring things. And quite often, you have to make the case internally within the institution as to why you should be acquiring one piece over another. The trick, of course, is to try and do it ahead of the curve. That way, things come into collections that are more affordable. That’s essentially how the public museum side of the art world operates. And where the interest often lies with digital art is that artists generally tend to push technical boundaries. 

‍SG: I think there’s an interesting tension occurring in this crypto community of wanting to reject the status quo — to reject the academy and the fine art world — but at the same time there’s an almost desperate urge to be a part of it. I don’t think anyone really expected the digital art market to explode as quickly as it did. 

‍Anon: Sofia, you mentioned selling the code itself as part of the work. What exactly is being sold? It’s not unheard of that someone can use their own original code and also incorporate code from other places. Does that complicate the ownership or partial ownership of the artwork itself?

‍SG: A great example of this was Christie’s first sale of an AI artwork in 2018. It made a record sale, but there was outrage over the fact that they had [used] a model that had been published online by another artist. 

A few years ago, there was this utopian view of open-source code, which was available for anyone to use. But now, since there’s money involved, it can get a little dicey when people put their work out there. [...] There are some artists in this space who say: “You can capitalize on this and use it for commercial purposes up to $100,000. After that, call me.”‍

‍‍

Anon: As technology and code develops, is there a fear that over the next 50 to 70 years that the same code won’t still be viewable on a screen? Will the art become obsolete? 

‍DD: Sometimes it’s important to document the piece in an archival sense. You might want to record it in some way so that you can still see how it looked when it was first made. The V&A, for example, acquired a work by Kyle McDonald, which is freely available on the Web at the moment, but presumably isn’t going to stay there forever. The museum has a copy of that work [and] can play the video recording and see the film as it was rather than as it might end up in five, ten, 15 years time. 

‍AR: To go back to my earlier analogy of gardening, I think of my works sometimes as flowers. You cut them, and then they disintegrate over time — they decay. I think there’s something quite nice about that. Because everyone thinks of digital art as this quick thing where you press a button and get stuff, when in fact it’s a very slow, long, laborious process. And I think that because it’s finite, it’s more beautiful.