6 Artists Who Were Using Artificial Intelligence Before ChatGPT
Linda Dounia, still from Spannungsbogen, 2022. Courtesy of the artist.
To an extent greater than most of us are even aware of, the convenience of our everyday lives is underpinned by a technology that has been in development for decades: artificial intelligence (AI). Now, suddenly, that subject is on everyone’s lips, including its present and potential future darker sides, especially since Google engineer Blake Lemoine claimed a chatbot he was working on had become sentient—shortly before Sydney, Bing’s chatbot, told New York Times reporter Kevin Roose that it wanted to become human, and counseled him to leave his wife for the chatbot itself.
Potential sentience aside, questions about AI abound. Is the technology a threat to artists? Can AI truly create art? What are the copyright and intellectual property implications of image-producing AI programs that are partially based on artists’ existing works? These questions remain to be worked out, on computer servers, in artists’ studios, and in the courts.
But already, artworks made with artificial intelligence have reached the highest level of recognition. Most notably, New York’s Museum of Modern Art has an ongoing display of Refik Anadol’s project “Unsupervised,” which allows an AI to “hallucinate” and “dream” new visions, based on the inputs of all the existing works in the museum’s collection.
But many other artists have been exploring AI since long before New York Times reporters chatted with bots that professed their love for them. Here are six artists who have been in the trenches for several years, working on themes such as identity, language, and human–machine collaboration.
Memo Akten
B. 1975, Istanbul. Lives and works in Los Angeles.
Memo Akten, Distributed Consciousness, 2021. Courtesy of the artist.
If artificial intelligence might seem like the kind of field where an artist would want a PhD in order to truly grasp the ways to work with the medium, Memo Akten is just such an artist. His 2021 dissertation at Goldsmiths, University of London, “Deep Visual Instruments: Realtime Continuous, Meaningful Human Control over Deep Neural Networks for Creative Expression,” investigated deep learning models as an artistic medium.
As an example, his piece Distributed Consciousness (2021) explores the cognition of the octopus, whose neurons are not centralized like those of humans but rather spread across their bodies, as a model to explore what he calls the the “synthetic alien intelligences” now being developed in the form of AI. The project exists as two NFT collections.
Memo Akten, still from All Watched Over by Machines of Loving Grace, 2021. Courtesy of the artist.
The video work All Watched Over by Machines of Loving Grace (2021), meanwhile, takes its title from a 1967 Richard Brautigan poem that cries out for cybernetic meadows and forests, “where we are free of our labors and joined back to nature.” The trippy video illustrates these images literally, and in the end it’s a little hard to know if the image is appealing or appalling. Created using custom software based on machine learning, the work seems to bring out a modern, cynical take on a hopeful 1960s vision.
Akten has exhibited his work globally, including at Unit London (2023), ZKM Center for Art and Media (2022), Stedelijk Museum Schiedam (2022), Haus der Kunst, Munich (2020), and the Mori Art Museum in Tokyo (2019).
Sougwen Chung
B. 1985, Toronto. Lives and works in London, New York, and Hong Kong.
Chinese Canadian artist, researcher, and coder Sougwen Chung is no stranger to the meeting of art and tech: Her father was an opera singer; her mother a computer programmer. As a child, she studied violin and began to code websites while still in grade school.
During a research fellowship at MIT, she discovered robotics. “I was interested in the physical embodiment, and what it would feel like to evolve my own drawing practice,” she told the Washington Post, “and I hadn’t seen robots used collaboratively at that time. I wanted to try something less about robots executing an existing code and more about working together.”
What resulted was several generations of robots she calls Doug, which stands for Drawing Operations Unit Generation (appended with consecutive numbers beginning with 1). She has built and programmed these robots, which are driven by AI, employing recurrent neural networks to learn how to draw in the artist’s own style. With their sleek forms, they can be considered artworks in their own right.
“What would it be like,” she told the Post she wanted to know, “to have a drawing collaborator that was a nonhuman machine entity?” Whether live, or, during coronavirus lockdowns, streaming on video, she has performed live with these robots, and she calls her work “embodied AI.”
Among other accomplishments, she has had her work acquired by the Victoria and Albert Museum in London, and has exhibited at the Vancouver Art Gallery, Art Basel in Miami Beach, and the New Museum in New York.
Linda Dounia
B. Senegal. Lives and works in Dakar.
Linda Dounia, still from Dust is hard to breathe, 2022. Courtesy of the artist.
Linda Dounia, still from Dust is hard to breathe, 2022. Courtesy of the artist.
Senegalese artist Linda Dounia combines generative adversarial networks—a model for deep learning in which one computer creates an image and another evaluates it—with old-fashioned materials like ink and pastel. While many developers (like Midjourney and Stable Diffusion) have created image-making programs that use images scraped from the internet to train their AI—and generally to create figurative images—Dounia has been training models in abstract painting based on her own works, as she describes it, “to explore the extent to which AI art can convey meaning and feel as spontaneous as analog art-making.”
Dounia’s NFT Dust is hard to breathe (2022), auctioned on Artsy in 2022, is an animation of curated outputs of a generative adversarial network (GAN) model trained on the artist’s own paintings, all in response to a work by artist Alma Thomas. In what is likely the first large scale AI drop by an African woman, in 2022 she published Spannungsbogen online, some 2,000 images based on her own acrylic paintings. This time, they were inspired not by a fine artist but by science-fiction characters: the Fremen, of Frank Herbert’s classic novel Dune. The Fremen are able to ride the massive sandworms in their resistance to colonizing forces; Dounia’s project was created in response to issues of racial discrimination in face recognition technology, calling on a similar spirit of resistance to that of the Fremen, but harnessing the most modern means.
Linda Dounia, still from Spannungsbogen, 2022. Courtesy of the artist.
Linda Dounia, still from Spannungsbogen, 2022. Courtesy of the artist.
To that end, at Art Dubai 2022, she launched Cyber Baat, a platform to promote African-descended digital artists on the blockchain.
Dounia’s efforts have been shown worldwide, from Art Basel in Miami Beach to Art X Lagos and the Dakar Biennale; she’s curated exhibitions, such as “Black*Rare” at NFT marketplace SuperRare; and she’s been invited to lecture at such august venues as New York University.
Jake Elwes
B. 1993, London. Lives and works in London.
Jake Elwes, installation view of “The Zizi Show” at the Victoria and Albert Museum, 2023. Courtesy of the Victoria and Albert Museum.
“You’re born naked,” says RuPaul, “and the rest is drag.” But what if the drag is digital? New media video artist Jake Elwes explores the AI implications of drag and masquerade in a show on view through 2024 at the renowned Victoria and Albert Museum, in London.
“The Zizi Show” is “a deepfake drag cabaret” that probes the ethical problems presented by artificial intelligence, in the hopes of demystifying this technology, which, for most of us, is very much in a black box. While we may think of AI as all-powerful, Elwes has discovered that computers “have difficulty recognising trans, queer, and other marginalized identities.” In his work, he hopes to “give the power back” to such communities.
In the beginning, his interest in autonomous systems, used by artists like Nam June Paik and John Cage, led him to wonder if computers could ever be truly creative. But instead he decided, in order to expose their limitations, to dig into political issues such as bias in the ways that computer systems are trained.
His video piece Zizi - Queering the Dataset (2019) questioned existing facial recognition datasets, which can leave out people who are not white and male, such as women of color or trans people. He inserted the faces of drag performers into an existing set of images that had been used to train such software. “It started to create fluid identities, these other, queer, much less recognizable faces,” he said in an interview with the Edinburgh International Festival.
Expressing optimism for the future, Elwes has said that “artificial intelligence is at its best when used as a tool to complement human creativity, and it can actually teach us a lot about the working of our own brain and our society and structures.”
In addition to the V&A, his work will be on view this year at the Max Ernst Museum in Brühl, Germany. He has recently exhibited at ZKM Center for Art and Media in Karlsruhe, Germany; the Zabludowicz Collection in London; and the Yuz Museum in Shanghai.
Anna Ridler
B. 1985, London. Lives and works in London.
Anna Ridler, installation view of Bloemenveiling, 2019, in “DYOR” at Kunsthalle Zurich, 2022. Photo by Julien Gremaud. Courtesy of the artist.
When NFTs exploded in popularity, especially after the earth-shaking $69 million sale of Beeple (a.k.a. Mike Winklemann)’s Everydays - The First 5000 Days in 2021, skeptics invoked the 17th-century Dutch tulip mania, in which exotic breeds of the flower fetched inflated prices, to call NFT trading a bubble (which seems, just a few years later, to have popped).
Artist Anna Ridler was referring to that historical phenomenon two years earlier with Bloemenveiling (2019), a critical commentary on the 2017 Bitcoin craze and crash of 2018. Bloemenveiling supplied an online exchange, complete with bots to help drive up prices, for auctioning AI-generated digital tulips that, like nature’s originals, wilt a few days after they bloom.
Ridler doesn’t use out-of-the-box technology; having earned a master’s degree in information experience design at Royal College of Art in London, she aims to steer clear of big tech, developing her own data sets (her own drawings and paintings, that is) to train AI, and programming her own generative adversarial networks.
She has exhibited all over the world in just the last few years, from the Tokyo Photographic Art Museum and London’s Barbican Centre, to the ZKM Center for Art and Media in Karlsruhe, Germany.
Jenna Sutela
B. 1983, Turku, Finland. Lives and works in Berlin.
Jenna Sutela, still from nimiia cétiï, 2018. Courtesy of the artist.
Finnish artist Jenna Sutela often brings together the futuristic and the ancient, the evidently sophisticated and what might be considered primitive, to introduce a vision of symbiotic relationships between the organic and the human-made.
In a 2018 residency at Google Arts & Culture, she produced a characteristically ambitious work, nimiia cétiï, based on her study of the 19th-century French medium Hélène Smith, who wrote about séances in which she claimed to channel the speech of Martians. Sutela says this was one of the earliest known examples of glossolalia, or speaking in tongues. The artist used these texts, and the movements of the nattō bacteria, long held to be a secret to long life among the Japanese, as a teaching material for an AI to reconstruct the Martian language.
The artist’s work has appeared at venues including the Guggenheim Bilbao, the Serpentine Galleries in London, Castello di Rivoli, and National Museum of Modern and Contemporary Art Korea, along with major exhibitions such as the Shanghai Biennale and Liverpool Biennial. She has also spent time as a visiting artist at Massachusetts Institute of Technology.
