This article is part of the Fine Arts & Exhibits special section on how creativity can inspire in challenging times.
Words, in a variety of fonts and colors, slide over a bright 25-foot-square screen glowing from millions of light-emitting diodes in the Agnes Gund Garden Lobby at the Museum of Modern Art in New York. Refreshing every minute, the words are by turns seemingly banal — “Just Keep Breathing” — and then beautiful and ephemeral:
Trace this poem
in mud
so rain dissolves it.
Carve it in wind on stone,
a script no hand can hold, not tightly.
The work, “A Living Poem,” showing simultaneously in South Korea and running until spring at MoMA, is generated by an artificial intelligence program trained on the artist Sasha Stiles’ writing, as well as data from MoMA’s collection. (“Keep breathing” refers to Yoko Ono’s “1st Day Breathe,” a 1965 instruction-based artwork). The text goes on forever, never repeating itself exactly.
“Because each 60-minute performance is generated live by code, what viewers experience is not a fixed recording, not a video loop, but a living manuscript that’s always in motion, always becoming,” Stiles wrote in an email, explaining how the technology works. “Improvisation happens here at algorithmic scale, so no two tellings are ever the same.”
With the rise of artificial intelligence, machines are increasingly engaged in the creative process, raising the question: What is art if the human hand is no longer involved? It can be baffling for viewers in an age when A.I.-generated text and imagery already permeate the culture.
In response, major art museums are wrestling with what, within the kaleidoscope, counts as art, and are trying to teach the public to see beyond spectacle. In so doing, they are creating a canon that artists, collectors and audiences will treat as a guide to what matters and why.
Curators argue A.I. art is less about aesthetics — some of it is, in a word, ugly — than about intent and process.
“The technology is not the art,” said Paola Antonelli, a senior curator at MoMA and director of the museum’s research and development. “In the end, it’s about the person’s creativity and vision.”
Melanie Lenz, curator of digital art at the Victoria and Albert Museum in London, echoes that point: “Audiences can quickly tire of works that rely solely on technological novelty.”
That tension between novelty and substance is central for curators. Christiane Paul, curator of digital art at the Whitney Museum of American Art in New York, said, “There seems to be a general assumption that the term A.I. art simply means visuals created by means of a text prompt, which ignores artists’ deep engagement with the medium.” In fact, it’s a wide and varied field.
Even off-the-shelf tools can demand conceptual rigor, Paul said, pointing to the work of Bennett Miller, who uses publicly available A.I. image generators. But curators seem most interested in artists who work in the computer code of A.I. systems. Many renowned artists working with A.I. today spend months or even years accumulating and curating data for bespoke A.I. models that they have a hand in building themselves.
Artists have long experimented with computers, from Vera Molnar’s IBM mainframe drawings in 1968 to Harold Cohen’s painting system AARON in the 1970s. A.I.-generated art punched into the public consciousness in 2018 with the $432,500 sale of “Portrait of Edmond de Belamy” by the Parisian art collective Obvious at a Christie’s auction. Since then, artists have begun building their own art-generating A.I.s.
The Whitney’s Paul said that what matters is the sophistication of both the medium’s use and the concept itself. The curator’s role is to help people understand where the art lies.
William Latham, an artist who organized the ongoing Creative Machine Exhibition a decade ago in London, said things work best when the machine is a creative partner, “like Picasso with his assistants.” He and a collaborator, Stephen Todd, are working with Google DeepMind on an evolutionary system that creates images from text prompts, breeding increasingly ornamental variants based on mathematical formulas. Humans can pick which variant to breed further, or they can let the machine choose.
“It’s sort of like survival of the aesthetic,” Latham said by telephone. In the new version, “Evolution and Foundation,” on view at the Oxo Gallery in London until Oct. 26, the machine is making all the decisions.
That raises questions about autonomy and human input.
“A world in which machines write our music and make our art is a world we could lean into if we wanted to,” said Maya Ackerman, professor, A.I. musician and author of “Creative Machines: A.I., Art and Us.” “But I really think it’s probably a net loss for humanity to disengage the arts to such a complete degree.”
But for many artists working in the medium, the act of creation is in writing the code.
“I don’t really believe in the artist’s hand,” said Casey Reas, a pioneering artist and computer programmer in the space. “I believe in the artist’s point of view.”
Refik Anadol, a Turkish-born artist who helped design MoMA’s LED screen on which Stiles’ work is displayed, has embarked on an ambitious plan to build an A.I. art museum in Los Angeles from scratch, equipped with the infrastructure needed to display A.I.-generated art. A.I. runs on a lot of high-end computer chips and electricity, so many artists work closely with technology companies.
Dataland, as Anadol’s museum is called, will have its own A.I. model for use by artists, built with support from Google and Nvidia and trained on half a billion images, all collected with permission of the owners. Dataland is planned for The Grand LA, a Frank Gehry-designed restaurant and entertainment complex in Los Angeles. He said traditional museums don’t have the resources to display the full breadth of A.I. art being created.
“Without museums, the art forms are lonely,” said Anadol. “To me, the museums are the campfire for humanity.”
Reas said the biggest question facing A.I. and art is whether to integrate that work into existing canons or to break away and build new institutions around a new canon. “I personally believe both should be happening,” he said. He put that vision into practice with Feral File, an online platform dedicated to digital art.
But not all museums are ready to embrace the art form. Many are watching and waiting.
“We didn’t show any NFTs when they first emerged, and that seems to have kind of fizzled out,” said Marshall N. Price, chief curator at the Nasher Museum of Art at Duke University, referring to digital art works known as nonfungible tokens. “We’re taking a similar tack with A.I.-generated art.” His team famously experimented with having ChatGPT curate an exhibition a few years ago with subpar results.
Yet as some museums hesitate, others lean forward. “The role of A.I.s in generating images is a cultural milestone that cannot be ignored,” said Valentina Ravaglia, curator of the recently closed Electric Dreams exhibition at London’s Tate Modern.
Other current exhibitions of A.I. art include “Subject to Change” at Gazelli Art House in London and “The Age of Creative Intelligence” at the Goodman Center Lobby at the Fashion Institute of Technology in New York City.
At the New York Armory show in 1913, American audiences were unsettled by cubism, particularly Marcel Duchamp’s “Nude Descending a Staircase.” Today’s A.I.-generated works provoke a similar unease, challenging audiences to decide whether new tools expand or dilute the meaning of art. Of course, Duchamp and the cubists are now canonical figures in art history.
“Whenever there’s a new technology, there’s that moment of drunkenness,” said MoMA’s Antonelli. “Then you sober up, you establish base lines.”
At some point, she added, audiences will start having their own critical sense, and they will vote for the institutions that truly matter in a very organic way.
The post Amid the A.I. Deluge, What Counts as Art? Ask the Curators. appeared first on New York Times.
