They must be in agony since the two black-clad individuals are still as they kneel on the ground. Oversized, slick gold masks cover their features, making it impossible to tell if they are grimacing—it looks like they’ve buried their faces in half of an Easter egg.
They resemble sculptures because of their stillness, and the only way to tell they are human is to listen for the little rise and fall of their chests. Which is extremely appropriate considering that they aren’t entirely human. These are “Idioms,” creatures of human and machine that were made by French artist Pierre Huyghe for Liminal, his biggest-ever show at the Punta della Dogana in Venice.
The show, which runs from March to November, features roving idioms. Their masks contain sensors that track the rooms they sit in and people they meet. Artificial intelligence will progressively translate this data into a new language. like instance, the masks of the Idioms will gradually generate words like “door,” “humans,” and “writing,” creating a language that will eventually allow them to converse with one another. Huyghe is curious about what they could be able to say in 20 years since their knowledge will grow every day.
A large black box suspended from the ceiling, a “self-generating instrument” that is also loaded with environmental sensors and produces ambient music and crisscrossing beams of light, is seen across from two kneeling Idioms in a darkened room on a crisp March day, just before the exhibition opens to the public. The LED screens on their foreheads shine gold, and it seems as though the Idioms could only muster a few syllables in reaction to the artwork before them, repeated over and over again. They speak in a whispery, hissing voice. It has a similar sound to “What’s this?”
It’s reasonable to inquire. The challenge for any artist attempting to work on a topic as revolutionary and defining as artificial intelligence is that the actual magic frequently occurs behind the scenes on some hard drives. Huyghe acknowledged at a press conference three days prior to Liminal’s opening that it might be difficult for the average visitor to understand that the language coming from the Idioms’ masks is artificial intelligence (AI)-generated. He was concerned that visitors would assume that the people wearing the masks are the ones whispering, even though there is a blinking server on display.
With ChatGPT’s launch in 2022, buzzy technology has quickly revolutionized everything from homework to journalism, and modern artists are under clear pressure to address and engage with it.
Like Huyghe, several artists have employed AI to create or improve their works, including British conceptualist Gillian Wearing and German filmmaker Hito Steyerl. An apparently “fully AI-driven” multimedia installation of historical works by French artist Philippe Parreno will premiere at Munich’s Haus der Kunst shortly after Liminal’s first run ends.
It’s not always clear if artists are utilizing technology in an intriguing and challenging way or are just trying to cash in on the excitement. It’s not evident from a draft press release for the Munich exhibition precisely which components of Parreno’s display would be artificially intelligent, but it’s simple to understand how AI could be cynically applied to an exhibition like a slick veneer to make old work appear new.
Artificial intelligence (AI) is already present everywhere; it can autocomplete our emails, recommend a new Netflix series, and recite the weather report in the voice of Alexa on Amazon. Writing has been revolutionized in recent years by chatbots that compose plays, poetry, essays, cover letters, and code in response to prompts. Text-to-image models like DALL.E and Midjourney enable anyone to produce “art” with just a few lines typed.
However, as AI becomes more commonplace in our daily lives, artists’ use of it runs the risk of seeming cliched. According to reports, Turkish artist Refik Anadol’s “live paintings,” which are presently on show at the Serpentine Gallery in London, have “transfixed for an hour or more” large crowds. AI was given images of coral reefs and rainforests to create Echoes of the Earth: Living Archive, an immersive exhibition by Anadol that allows visitors to explore “artificial realities.” Although some claim that Anadol’s prior AI-generated work was overhyped, people may still be enthralled.
Everything about it resembles a gigantic technological lava lamp. Jerry Saltz of New York Magazine talked about Anadol’s Unsupervised, a 24 foot screen at the Museum of Modern Art that utilized artificial intelligence to create visuals continuously between 2022 and 2023. Saltz thought the work was dull and insignificant, capable of briefly amusing you but ultimately “not disturbing anything inside you.” Put simply, he thought the piece said nothing.
According to Saltz, “AI will have to provide its own vision and vocabulary if it is to create meaningful art.” In a literal sense, Huyghe’s Idioms accomplish just this. It’s fascinating to watch them because, rather of presenting the audience with a static state of “artificial intelligence,” they present an ongoing process of “artificial learning.”
It’s intriguing that Huyghe is using AI to take charge of the art in this instance, even though there’s a chance something could go wrong. It’s possible for the Idioms to either produce no language at all or a language that sounds discordant and unpleasant to us. They might be rebelling in some way, or they might be unduly swayed by boisterous exhibition attendees, repeatedly saying the same phrases.
Seeing the Idioms’ reactions to the surrounding art on a daily basis would surely be entertaining. Even though my immediate impression was, “I bet their knees hurt from all that kneeling,” Huyghe intended for these bizarre veiled individuals to inspire reflection about the relationship between the human and the non-human.
The usage of AI in his work, Camata, is less thought-provoking. In one of the harshest deserts on Earth, a skeleton is surrounded by robotic limbs carrying out an enigmatic ritual. Despite not being live, the film is altered in real time, with artificially intelligent “editors” obtaining information from a big brass sensor that resembles a telephone pole close to the exhibition opening. This sensor keeps track of everything, including the number of patrons in the gallery and the outdoor temperature. The Camata film is then edited appropriately.
However, curator Anne Stenne makes it clear that this isn’t a straightforward case of “x” leading to “y”; for instance, the AI editors wouldn’t chose film taken at night automatically if there was only one person in the show. This means that, although the never-ending editing process is exciting (you could, after all, sit through the entire exhibition and never see the same sequence twice), it’s difficult for the average person to see why artificial intelligence was required. Should the editing be done at random, how would the task change? In casual viewing, it’s really difficult to tell.
Attendees of these displays must, in fact, have faith that extraordinary things are happening behind the scenes. Although Huyghe’s sensors are on display throughout the show, the artist is reticent to discuss the specifics of the software that interprets this data and how it operates. “Pierre doesn’t want to focus on the technical parameters of his works,” a spokesman claims. His goal is to focus on the experience of the guest. In a world where firms have been shown to be deploying “pseudo-AI” that is actually managed by hidden people behind the scenes, audiences may find this alarming.
As with Huyghe’s self-generating language, AI art functions best when it accomplishes a task that the artist alone was unable to. Anything less runs the danger of being, at best, gimmicky and, at worst, ridiculous. Whatever the case, the AI movement will keep sweeping across galleries, and before long, doubting the tool will be akin to doubting a pen or pencil.
“Computer art” appeared in exhibitions all over the world in the 1960s, from London to Stuttgart to Zagreb to Las Vegas. “Perhaps a computer will never produce a painting all by itself,” a modern writer speculated, adding a warning that “at least one expert thinks such art represents a genuine new art form.” Talks about AI’s place in art will eventually sound like this, without a doubt.
