Traditional museums are built like mausoleums. They house dead objects under thick glass, frozen in time, indifferent to your presence. You look at a painting, the painting does nothing back, and you move on.
That model just broke. Don't forget to check out our earlier article on this related article.
Downtown Los Angeles just welcomed Dataland at Frank Gehry’s landmark development, The Grand LA. Billed as the world's first museum dedicated entirely to artificial intelligence arts, the venue kicked off its public run with an inaugural exhibition titled Machine Dreams: Rainforest.
[Dataland Overview]
- Location: The Grand LA, Downtown Los Angeles
- Space: 25,000 square feet across 5 distinct galleries
- Core Tech: Large Nature Model (LNM) & NVIDIA Connectome
- Hardware: 1.5 billion total pixels, 84 Epson 4K projectors, 1,577 custom 10K LED panels
Co-founded by media artist Refik Anadol and Efsun Erkılıç, this space isn't just showing digital art on walls. It uses live human biodata and massive ecological datasets to construct an interactive environment that literally shifts based on your vital signs. If you want more about the history of this, ZDNet offers an informative summary.
The Math Behind the Metaphor
Most immersive digital exhibits are glorified projection mappings—essentially massive video loops playing on a timer. If you stand in the room for twenty minutes, you see the exact same loop twice.
Dataland runs on a completely different engine. The entire 25,000-square-foot facility is driven by a centralized high-performance computing system built with NVIDIA hardware, appropriately named Connectome. The brain behind the visuals is the Large Nature Model (LNM). Instead of scraping random images from the open web, Anadol’s studio partnered with major institutions including the Smithsonian, the Cornell Lab of Ornithology, and London's Natural History Museum to feed the system over 500 million ethically sourced images of plants, fungi, and wildlife.
The machine takes those half a billion data points and generates reality in real time. It calculates 1.5 billion pixels concurrently across 84 Epson 4K projectors and 1,577 custom 10K LED panels. Nothing is pre-rendered.
[500M+ Natural Data Points]
(Smithsonian, Cornell, Natural History Museum)
│
▼
[Large Nature Model (LNM)]
│
▼
[NVIDIA Connectome Core] ◄─── [Live Visitor Biodata]
│ (Heart Rate, Temp, Galvanic Skin)
▼
[Real-Time Rendering: 1.5B Pixels]
When Art Feels You Back
When you enter, you don't just buy a ticket; you get micro-tracked. Visitors receive an Empatica wristband, a medical-grade biosensing device that tracks heart rate, skin temperature, and galvanic skin response (sweat gland activity).
As you walk through the five interconnected galleries, wall-mounted sensors track your physical movement while the wristband beams your stress, excitement, or calm directly into the Connectome system. If a crowd gathers and their heart rates spike, the digital flora on the walls reacts, changing colors, morphing shapes, or accelerating the visual pace.
The experience hits your nose, too. Dataland partnered with L'Oréal Luxe to distribute 12 distinct, AI-generated scents through a wearable device worn around the neck. You smell damp earth, crushed leaves, and tropical humidity, changing dynamically as you move through different climate zones generated by the model.
In the Infinity Room, you are entirely enclosed in a three-dimensional LED cube—walls, ceiling, and floor are screens. The space simulates a giant glass hummingbird taking flight, accompanied by a 250-speaker audio layout using L-Acoustics immersive sound tech. The audio blends original musical scores by Kerim Karaoğlu with actual acoustic field recordings from 16 real-world rainforests.
The Eco-Footprint Question
Running massive AI models alongside thousands of square feet of high-definition LED displays sounds like an environmental nightmare, especially for an exhibition celebrating nature. The creators had to account for this irony.
The Large Nature Model itself doesn't run on local servers in Los Angeles. It is hosted on Google Cloud infrastructure in a specialized, low-CO2 compute zone in Oregon, utilizing 87% carbon-free renewable energy. According to data provided by the museum's structural consultants, Arup, the computational power required to reiterate the artwork for a single visitor’s stay equates roughly to charging a standard smartphone once.
The Disconnect Between Tech and Taste
Despite the engineering triumph, the exhibition highlights a growing divide in the art community. Critics point out that Dataland behaves less like a traditional art institution and more like a high-tech theme park. With ticket prices ranging from $49 for general admission up to $129 for priority access passes, it mimics the monetization models of Universal Studios or Disneyland rather than the nearby Museum of Contemporary Art (MOCA).
There is also the question of substance over spectacle. When art becomes entirely ephemeral and dependent on data loops, what does the viewer actually take away? The exhibit attempts to solve this at the very end with "Data.Chocolate"—a four-piece edible sculpture created with Valerie Confections where the flavor profiles are algorithmically determined by the day's collective data. You can also print custom t-shirts featuring a visual snapshot of the specific "machine dream" your presence helped generate.
Once you walk out the door, the system deletes your biometric data. The dream vanishes.
Practical Steps for Visiting Dataland
If you plan to navigate this digital jungle yourself, skip the tourist traps and approach the experience strategically.
- Book the morning slots: The real-time generation engine behaves differently when the galleries are empty. A crowded room muddies the biosensing data, turning the visual reactions into an averaged-out blur. Going early lets the system isolate your individual biometrics more accurately.
- Wear light clothing: The biosensing wristbands require direct, firm contact with the skin. Bulky sleeves or heavy jackets interfere with the sensors, and the immersive galleries can run warm due to the sheer volume of operating electronics.
- Pay attention to the transitions: The most impressive tech isn't the static imagery; it's how the model bridges different concepts. Watch the edges of the screens in the Latent Gallery where the "thinking brush" translates your touch inputs into neural network structures.
The exhibition runs at The Grand LA through January 31, 2027. Secure your tickets online well in advance, as the timed-entry slots limit capacity strictly to prevent sensor interference from overcrowded rooms.
