A gallery in downtown Los Angeles is trying to prove that art generated by artificial intelligence can be more than a five-second clip stitched together from a text prompt. Dataland, billed as the world's first "museum of AI arts," opened its doors on June 20 and drew more than 10,000 visitors within its first two weeks, according to cofounder Refik Anadol, the artist behind the gallery's technological installations exploring the bond between humans and machines. Anadol built the space with his studio partner Efsun Erkılıç, and its flagship exhibit is designed to answer, once and for all, whether a machine can feel the people standing in front of it. Dataland positions itself as an answer to a broader argument raging across the creative world about whether artificial intelligence belongs in art at all, and Anadol wants his gallery to be proof that a serious answer is possible.
An immersive rainforest built from five petabytes of raw data
The centerpiece of Dataland's opening show is Machine Dreams: Rainforest, an immersive architectural installation that Anadol calls his most ambitious work to date. Interactive digital displays across the space respond directly to visitors' movements and to biometric data collected through wearable devices, generating constantly shifting images and soundscapes. All of it is drawn from Anadol's Large Nature Model, an AI system trained on natural science archives supplied by respected research institutions, including the Smithsonian.
Building that model took three years of work from scratch, with Anadol and his team traveling to the Amazon and other rainforests to gather raw material that would later feed the system's hallucinated versions of those landscapes. "For three years, we started from scratch and trained our own AI models, and we worked with our own data sets," Anadol says. The result, he says, is "5 petabytes worth of raw data that we collected by ourselves." He is especially proud that this trove was gathered with the consent and participation of researchers, a deliberate contrast to the backlash and lawsuits major Silicon Valley AI firms have faced over what many creators call unlicensed, extractive use of their work as training data.
Anadol adds that Google DeepMind gave Dataland access to "experimental low-energy" resources, letting the gallery run on Google Cloud while maintaining what he calls "sustainable compute." That relationship traces back to 2016, when Anadol became the first person awarded the Google Artists and Machine Intelligence Artist Residency. The point matters because the enormous computing power behind modern AI systems has drawn its own criticism over energy use, and Anadol is keen to show that an ambitious, always-on installation like this one doesn't have to come at that cost.
Answering the skeptics who call AI art "slop"
Anadol knows exactly what he is up against. The phrase "AI art" is a nonstarter for many creatives and critics who see it as shorthand for the generative "slop" that has spread across visual media. He doesn't dismiss that reaction. "I mean, 100 percent, the majority is right," he says, adding that when people hear the term "AI art," "their first assumption is like, prompt engineering, or a bunch of eight-second clips." His answer isn't to argue with that perception but to build something that sits entirely outside it. "The reason for Dataland is to understand and explain and explore, and tell the world that there is not only one option," he says.
A waiver, a smartwatch and the smell of trees
Machine Dreams: Rainforest makes its case for AI as a tool for entirely new kinds of artistic engagement, and Anadol says it's the first of his projects where "it is impossible to record what it feels like." That claim holds up. Entry involves a somewhat fiddly process: signing a waiver, using a specialized app, then receiving a smartwatch and a U-shaped plastic collar worn across the shoulders. As the devices calibrate, the shoulder harness releases the exhibit's first striking scent, the unmistakable smell of trees, filling a hypermodern downtown building with the impression of a forest that shouldn't exist inside it.
The largest gallery in the exhibit is otherwise empty aside from a couple of pillars, with ultravivid fusions of nature scenes and computer-chip textures moving across the walls and floor in a 40-minute cycle of shifting effects. Those effects respond, in part, to the movement of everyone standing in the room. Walking in during a sequence simulating heavy rain and thunder reveals a watery circle trailing each visitor's feet, which stretches into an aqueous streak when they walk faster. Waving a hand in front of the falling droplets bends their paths, and the biosensor releases the clean scent of a summer storm to match.
Choosing to be tracked, or choosing to be a ghost
Visitors can move through the installation as a "ghost," skipping the biosensors entirely, but Anadol says the wearables, modified from medical-grade devices, let people leave a temporary fingerprint on the work. "For almost 5,000 years, we as humanity, we look at artworks and we feel something," he says. "At Dataland, as a laboratory of imagination, our first deep question is: Can artwork feel us back?"
Asked whether the devices tracked a trip to the restroom, Anadol laughs it off. "No, no, no, no, the restroom has zero connection," he says. He adds that the museum "forgets" a visitor's data the moment they leave, although it remains accessible to that visitor through a personal token handed out at the exit. "Data is a form of memory," Anadol says, and Dataland tries to treat it with that kind of respect. "So it's the opposite," he says, "of what we have in the whole world," where invasive surveillance has become routine.
A hummingbird flight and a look behind the algorithm's curtain
Further into the museum sits the Infinity Room, where visitors follow a glittery hummingbird soaring over a fantastical neon forest, an aerial sequence reminiscent of the flight scenes in the Avatar films, disorienting enough at points that the shifting point of view can throw off your balance. Next door, the Latent Gallery pulls back the curtain on the Large Nature Model itself. At one console, visitors can scroll through the model's training data by category. Selecting "Frogs" turns the wall ahead into a massive grid of frog photographs pulled directly from the system's dataset, a deliberate demonstration of how much verifiable material sits underneath the dreamlike images shown elsewhere in the museum.
Anadol says he wanted visitors to glimpse the sheer scale of information behind the installation's surreal, often impossible recombinations of real nature. The Smithsonian's Encyclopedia of Life alone supplied data on more than 2 million species. "These are the places where we demystify our algorithms, demystify our data sets, demystify our training experiments," he says, adding that he wanted to see "what happens if we go next-level and tell the audience and the visitors exactly the reason for that machine dream or hallucination."
Hallucination as the point, not a flaw
That word, hallucination, is deliberate, and it's also what separates Dataland's debut show from the ultra-realistic AI-generated images and deepfakes feeding today's misinformation problem. "It's about the idea of a machine that falls in love with nature, that's the rainforest, and we enter the machine dreams as a concept where we smell the machine's hallucinations, we taste the machine dreams, we hear millions of bird songs that are impossible to hear in a human world," Anadol says. The Large Nature Model may draw on real scientific libraries and real-time weather data, but the environment it constructs from those building blocks remains, in almost every respect, entirely alien.
The Sanctuary: an artwork that exists once and disappears
The exhibit closes with the Sanctuary, a concept Anadol has been developing for several years. As a group of visitors enters the room, the biometric information collected from each person, including heart rate, skin temperature, the path they took through Dataland and the pace of their walk, is combined to generate a swirling, abstract, three-dimensional representation of the room's collective energy. That representation is never seen again once it forms, making every visit to the Sanctuary a one-time event that can't be repeated or recreated.
Anadol says the gallery treats "emotion as an input" throughout the exhibit, and he's been struck by how often the biosensors detect goosebumps on visitors, which he calls an "incredibly relevant" sign that the sensory experience is landing. "The artwork can feel this information," he says. "I saw people with tears, I saw people with joy, I saw people with excitement," he adds. "And I think, if that's not art, what is art?"
"This is all about being human, not about AI"
That kind of visitor reaction is exactly what Anadol needs to win over skeptics who doubt artificial intelligence has anything to offer the creative arts. He doesn't frame the technology as a production shortcut but as a way to rediscover something essentially human. "This is all about being human at the end, not about AI," he says of Dataland and Machine Dreams: Rainforest. "It's just an incredible tool, but the messaging and the context and the meaning is still all about being human."
