This story is free to read because readers choose to support LAist. If you find value in independent local reporting, make a donation to power our newsroom today.
Yes, I enjoyed LA's new AI Museum. Here's why I still didn't like it
Dataland, the world’s first museum dedicated to art generated with artificial intelligence, opens Saturday. Located at the Frank Gehry designed Grand LA building in downtown, it’s positioning itself among institutions like MOCA, The Broad and the Disney Concert Hall.
The inaugural exhibit, Machine Dreams: Rainforests, is a project of museum co-founder and artist Refik Anadol and his studio’s Large Nature Model (LNM), which Anadol describes as “ethically sourced image archives transforming into real-time artworks” that span five “multi-sensory” galleries.
What does that all mean exactly? And would it all be worth the lowest ticket price of $49? I attended a recent press preview to find out.
Something you should know about me is that I love art. A lot. I want artists to thrive, and I don’t want AI to threaten that. At the same time, I believe in creative freedom. Who am I to say what is or isn’t a valid tool of an artist?
So I went into the museum as a curious skeptic. And in the end, while it was an enjoyable experience, I didn’t like it.
Here’s why.
Gearing up: Biosensor watches, scent gadgets and shoe coverings
First, the press event felt more like a tech start-up launch than a museum opening, which made sense once learning the museum is partnered with tech giants like Google Cloud and NVIDIA.
Before going in, we were given shoe coverings, making me feel like we were entering a sterile laboratory.
Inside the lobby, a large screen wall illuminated preview images of the exhibit, framed by scrolling environmental data of different rainforests. Loud ambient music –- what I can best describe as what you would hear in a futuristic spa — permeated the space, and a similar AI-generated sound design scored the whole exhibit.
The first official room of the exhibit was the “Discovery Portal,” with more data and dazzling graphics conceptualizing the environmental data captured to feed the LNM. We were each instructed to use a QR code to unlock a box, which revealed a wearable biosensor that looked like a blank digital watch, and another device to wear around our necks that emitted various scents designed by L'Oréal Luxe.
Psychedelic sensory overload
We walked down a hallway and descended an escalator into what was arguably the most impressive gallery: the “Data Pavilion.” It's a large room with curved edges, and mirrored columns and ceilings.
Evolving images inspired by rainforest data moved quickly along the walls and floors: pink-purple lightscapes merged into yellow-blue polka-dots, and gave way to watery blue ripples, with circles of light that followed my footsteps. The exhibit is designed to react to its audience and I enjoyed seeing the images shift based on everyone’s movements, as well the sweet earthy smell released by the scent neck device. The Large Nature Model generating these experiences was trained on environmental data sets, including imagery, sound and weather data from 16 different rainforests.
“ As an artist, I believe the best way of creating work is machine-human collaboration,” said Anadol, who told me that every installation in the museum will be “human-made”. For him, the benefit of using AI is in how it can supersede human capacity.
“I will never remember a half billion images, but there's an installation here that can allow me to let visitors feel connected and listen to 50 million bird songs,” said Anadol “That is impossible to do by traditional tools.”
So did the resulting experience help me connect with nature in a radical new way?
Here’s where I have to disclose that I have sensory sensitivities. And like a lot of neurodivergent people, I crave and revel in sensory stimulation but can also find it overwhelming, and even physically painful in extreme instances. It would turn out, both experiences were true for me. After some time in the “Data Pavilion,” I found it dizzying, especially to walk while looking at a fast moving image beneath you, and my colleague was even more nauseous. (We are in our 30’s, which may also have something to do with it!)
On the bright side, I thought it was the closest experience to being on a psychedelic you can get without consuming anything.
Drawings, chocolates and environmental questions
Next, was the “Latent Room” which is mostly centered on interactive stations.I tested out a touch screen panel where visitors can make a drawing and then choose a nature model to interpret it into a piece of AI art.
Frankly, I had trouble using it. I did not find the interactive stations very intuitive, and maybe it's because I’m an aging millennial.
Visitors here can also choose from one of four chocolates created by Valerie Confections, based in Glendale. The one I tried was very delicious but the connection to the rest of the exhibit was lost on me.
This is where the scent started to be too much and I took it off and kind of held it at arm’s length.
At this point, I had started to wonder about the museum’s promises of creating environmentally friendly artificial intelligence, claiming the energy required to render the visuals for each visitor equals the energy used to charge a cell phone. This is based on Google Cloud’s low-CO2 compute zone which operates on carbon-free energy. Most A.I. isn’t structured this way, consuming exorbitant amounts of energy and water. But I wondered if masking all the chips and wires behind the exhibit's walls erases the fact that it does take machinery and energy to create artificial intelligence.
To Anadol, though, hiding the tech is the point: “I didn't want to get excited about some cables and computers. It's not about truly that, but it's more about becoming part of a new art form.”
Personally, I was hoping for something more as important as the environmental impact of AI in an exhibit about connecting to rainforests.
67 percent emotionality
The penultimate experience is called the “Infinity Room,” which was truly the most immersive animation I have seen, also the most dizzying.
You stand in a mirrored cube space, watching an eight-minute long semi-AI generated film with video-game like nature graphics, inspired by an Amazonian folktale about a glass hummingbird and wisdom tree. The story had a moral and felt like a call to action, but to what, was unclear to me based on the exhibit alone.
The final room, called “The Sanctuary,” is where your biometric data — like your temperature, heart rate, and skin conductivity — is displayed on the wall, before transforming into a visualization of the collective visitors’ experiences, a colorful abstraction of swirling sand.
While my data was only identifiable through the initials on my biometric device, and Dataland says visitor biometrics are not saved, nor tied to a specific person, I started to feel funny about having my data recorded at all.
I was also puzzled by how my biometrics were being interpreted into engagement and emotionality (at one point, mine was displayed as 67 percent).
“The percentages are not medical or diagnostic measurements,” said a museum representative, when we followed up over email. “They are artistic interpretations generated from the visitor’s interaction with the exhibition. The system may respond to signals such as movement, duration of the visitor’s stay, and physiological patterns gathered through the wearable experience.These signals are translated into a poetic profile of the visitor’s journey through the exhibition.”
But it was the very concept of poeticising artificial intelligence that was nagging at me the whole time.
Virtual trees and touching actual grass
After the press preview, I drove from DTLA up to the Arroyo Seco near the LAist office without much thinking. By that point, I was so overstimulated by the press tour that I just needed some peace and a little actual nature.
I sat in my car, with the window rolled down, on a quiet street, staring at the rows of trees along the narrow canyon that provides urban Pasadena with a bastion of green space.
I did get the impression Anadol and his team at Dataland are interested in critical discourse and artistic evolution, and I’m curious to see if future exhibits may more deeply explore the complexity of incorporating A.I. into art.
But as I decompressed in the quiet of the Arroyo, I wondered — if the exhibit’s purpose is to connect visitors to the natural world, what's the advantage of machine driven visualizations, as opposed to being in nature?
When asked over email, a rep replied that “nothing can replace being in nature.”
“We ask whether data can help us perceive aspects of nature that are usually invisible to human senses [...] Dataland is not an alternative to nature. It is an invitation to experience nature differently. People may leave the exhibition with a renewed desire to visit, study, listen to, or preserve the natural world.”
And seek nature, I did!
Dataland opens June 20 in Downtown Los Angeles.
Natalie Chudnovsky contributed to the reporting of this piece.