by At a reception for the first OpenAI developer conference in San Francisco last month, a crowd mingled, wine in hand, as withering critiques of art created with artificial intelligence appeared on a blue wall at the front of the room. “I’ve seen better-looking art on a malfunctioning printer,” said one reviewer. “The artistic equivalent of elevator music,” snorted another. “Harmless, unforgettable and terminally boring.”
It may seem like an odd strategy for OpenAI, the company behind widely used generative AI tools like ChatGPT and DALL-E, to promote disparagement of the art of AI, until you get the twist: AI itself wrote the critiques. Alexander Rebenthe MIT-educated artist behind the presentation combined his own custom code with GPT-4, a version of the large language model that powers the online chatbot ChatGPT.
Next month, Reben, 38, will become OpenAI’s first artist-in-residence. He weighs in as generative AI advances at a breakneck pace, with artists and writers trying to make sense of the changing possibilities and implications. Some consider artificial intelligence a powerful and innovative tool that can guide them in strange and wonderful directions. Others express outrage that AI is taking its work off the Internet to train systems without permission, compensation or credit.
At the end of November, a group of visual artists filed an amended copyright complaint against Stability AI, Midjourney and other AI tool makers after a federal judge dismissed parts of the original complaint, which accused the companies of misusing artists’ creations to train generative AI systems. Reben said he couldn’t speak to the specifics of AI and the law, “but as with any creative new technology, the law must catch up with the unpredictable future.”
(The New York Times sued OpenAI and Microsoft for copyright infringement on Wednesday.)
Technology companies included Google, Autodesk and Microsoft have hosted artists in residence. And over the past few years, artists have tested products like GPT and the DALL-E imager, offering insight into the tools’ creative potential before their public release. But the OpenAI residency, which is giving Reben a front-row view of the company’s work, is a first for the startup that is at the center of the debate over art and AI.
“Alex is one of the first people we shared our new models with,” said OpenAI spokesperson Natalie Summers.
Sam Altman, CEO of OpenAI, has long recognized that the technologies created by his company will change the nature of art. But he insists that no matter how good the technology is, artists (human artists) will always matter.
“There was a real scary moment where people asked, ‘Is this a tool we built or a creature we built?'” he said last month during an appearance in front of more than 300 artists and art lovers. crammed into an abandoned building. warehouse in downtown Oakland, California. “People now see these things as a new set of tools.”
After digital artist Android Jones said at the event that many artists were still very angry about the rise of AI image generators and the way they reduced the value of their own art, Altman said people would always look for art created by other people.
“It is clear that there will be more competition,” he said. “But, immersed in a sea of AI-generated art, that desire for human connection will increase, not decrease.”
Ge Wang, associate director of Stanford’s Institute for Human-Centered Artificial Intelligence and associate professor of music and computer science at the school’s Center for Computing Research in Music and Acoustics, wonders how receptive OpenAI will be to considering the tough questions. on the impact of AI on art. . What is the right balance between machine production and human healing? Will the instant results produced by companies like DALL-E discourage people from developing the kind of skills that require study and time?
“Asking these questions is bad for business, and OpenAI is a business,” said Dr. Wang. “You might have a wonderful artist in your residency asking questions. Are you willing to receive them?
But Dr. Wang, who is also a musician and designed two music-making apps, Ocarina and Magic Piano, for Apple’s iPhone, said he was encouraged that Mr. Reben was open to addressing questions about the impact of AI in the artistic community. .
Reben said that, as a technologist who had studied the impact of innovations like photography and recorded music on creativity, “I typically stay on the cautiously optimistic side.”
“But like any other technology of the past, there are both sides to the coin,” he added.
The New York native moved to Berkeley, California, a decade ago to become director of technology and research at Stochastic Labs, an incubator for creative scientists and engineers housed in a three-story 19th-century Victorian building. Mr. Reben’s highly conceptual art covers the walls of the main hallway and fills workspaces filled with printers, headphones, cables, capacitors, soldering supplies and other details.
On a rainy Thursday, Reben relaxed on a couch at Stochastic after a meeting at OpenAI to continue working out the details of what he’ll do during the three-month residency.
“If I come out of this and improve my art, or even come up with new questions or new directions to present to the world, that would be very valuable,” said Reben, who researched human-machine symbiosis as a graduate student at MIT Media. Lab, an interdisciplinary research center.
The residency overlaps with Mr. Reben’s first major retrospective, titled “Am I?” and on display through April at the Crocker Art Museum in Sacramento. DALL-E and other image generators like Halfway through the trip and Stability AI’s Stable Diffusion have taken the internet by storm by allowing anyone to instantly retrieve personalized visual images by simply typing a few words in a box. But while much AI-generated art exists in the form of pixels, Reben often manifests physical structures from ideas he refines with the help of artificial intelligence.
“I really like the absurdity and humor in my work, even if the underlying question is serious,” Reben said.
One sculpture in the exhibition features six toilet plungers lined up like a strange police formation. The AI-generated text on the wall poster explains that the work represents all that remains of the Plungers, an apocryphal art collective from the 1970s. Their fake artists adhered to “plungism,” a fictional philosophy “in the that the mind of an artist is in a state of flux and can be influenced by all things, even plungers.”
Plungism arose from Mr. Reben’s extensive back-and-forth with GPT-3: he would enter a message (an input intended to produce a desired response) and then tinker with his favorite responses, sometimes returning the edited language to the AI until it landed. in the correct wording.
Then there is “Dreams of the Cheese-Faced Knight,” which depicts a man whose face could be mistaken for a wheel of Swiss cheese. Reben worked with GPT-4 to find the right prompts to craft a compelling description of a painting and then fed the selected text into an image generator. He is not a painter, so he commissioned one to do the artwork.
Then, a large language model capable of assimilating both images and text studied the painting and described it in language that would fit in any museum. “The combination of psychedelic surrealism and fantasy gives the painting an air of playfulness, challenging the viewer to engage with the work’s complex layers of meaning,” the wall label reads.
Janisy Lagrue, the AI-imagined name for the real-life painter who produced the oil on canvas, explained: “I use cheese because it is a perfect symbol of the American dream. Cheese is a commodity, not a food. It’s totally artificial and delicious.”
The exhibit provokes more questions than answers, a reflection of Reben’s belief that as machines produce better results, humans need to ask better questions—about bias and propriety, among other things.
“Given how young this creative tool is, there is still a lot to solve, and addressing these issues falls on everyone involved, from its developers to its users,” Reben said. “The more people think about these questions, the better.”
Mr. Reben does not claim to speak for all artists as OpenAI’s first artist-in-residence. But he understands his concerns. Artists and writers worry that AI could steal their jobs, but Stanford’s Dr. Wang said the nervousness went beyond the possibility of losing their livelihoods.
The fear is that “not only will we be replaced as artists, but we will be replaced by something much more generic, much less interesting,” he said. “Maybe the generic will be enough to make a lot of money.”
Cade Metz contributed reports.
