Artist Ian Cheng’s Simulations Uncover Humanity In Artificial Intelligence
Currently on view at The Shed, the exhibit explores how AI is inherently human and “parenting is programming”
Astounding and unprecedented, Ian Cheng’s exhibition Life After BOB, currently on view at NYC’s The Shed until 19 December, champions art at the intersection of artificial intelligence and cognitive science. In fact, there’s no exhibit quite like it. On one side of The Shed’s fourth level, a 48-minute narrative animation, titled “Life After BOB: The Chalice Study,” built using the Unity video game engine, plays. It follows 10-year-old character Chalice Wong after her father installs an experimental AI, dubbed BOB (short for “Bag of Beliefs”), within her. On the other half of the room, the same animation plays but at the viewers’ command. Using their phones as remotes, they can pause, rewind and zoom into the animation to uncover more information about the characters, fauna or objects. Then, whether outside or inside the exhibit, viewers can update and evolve the animation through editing the artwork’s Wiki page, where edits integrate into the simulation in real time.
Often using technology and video games as his medium, Cheng is no stranger to exploring what it means to live in an increasingly technological world. In 2017, his Emissaries exhibit at MoMA PS1 featured simulations that viewed people interacting in a fictional world. In 2018, Cheng’s exhibition BOB created a sentient virtual creature that his latest show draws off of. Yet unlike the others, Life After BOB blends narrative, simulated storytelling, interactive world building and the anxieties of contemporary life together, rooting out how simulations are human all along and opening up optimistic pathways for the future. We spoke with Cheng about using AI as an art form, the collective aspect of this exhibit and why he believes NFTs are beautiful.
What inspires you to use AI and video game engines as your medium?
Dumb story, but I was at Whole Foods in 2012, having lunch on Houston Street. They have like a little deck for the dining area that overlooks the salad bar. I was watching people there. People were stealing food, they were taking things, putting it back, bringing their dogs, flirting. It was this whole ecosystem, self-contained in Whole Foods and contained in my field of view, and I thought, “Oh my god, I have to make something like that.” Forget trying to make linear narratives. I gotta make something that feels like a little world. The only way to really do that without actually staging it like it’s theater, performance or live animals is to try to like cannibalize a video game engine to play itself.
At a certain point, I started researching a lot about, not just AI, but also human psychology and animal psychology, to try to find models that could describe a new approach to AI. And I realized when you embody an intelligence, it’s both a harder problem but also something we conceptualize a lot easier because that’s how we relate to our own intelligence. I started getting into neurosymbolic AI models and cognitive architectures, inspired by Carl Jung, that dealt with an agent’s behavioral complexity by breaking it down into sub-personalities.
Your work often deals with emergent behavior. Can you talk about this and why you fall back on this subject?
One of the key things I found doing simulations was that the secret sauce was emergent behavior. To define it very simply, it’s the idea that the sum is greater than its parts. You have hydrogen. You have oxygen. Put them together and you can’t say hydrogen and oxygen are wet but you can say water is wet.
I learned recently that the neocortex of the brain is composed of these cortical columns that are responsible for making competing models of the world, and the interaction of these cortical columns might be what is responsible for producing the emergent effect of a unified experience of the world. I think emergent properties happen at the level of nature, of course, and in a much dumber, more constrained way, in simulations—a property that makes artworks feel alive. When I look at art and also the art that I aspire to make, there’s a sense of aliveness.
Part of the aliveness of Life After BOB is the collective nature, which you call “worlding.” What is this and what drew you to it?
“Worlding” was something I was writing about in-between Life After BOB and my previous project, BOB. It was a very simple idea of how can I, an individual artist, essentially produce a habitable community around a specific game. In the case of working on Life After BOB and working with my producer Veronica So, we produced a huge team during COVID, and the project became the one thing that we all could focus on during COVID.
On the production side, I thought that we made a world through that. In terms of the novelistic sci-fi, fantasy sense, I started thinking about “worlding” as the well that you can keep dipping in as an author, so that you don’t have to reinvent everything from scratch. The more you make a world, like Tolkien, or Harry Potter or Star Wars, the more there’s a world that gets built out through a production of one story, the less legwork you have to do on the next story to start to create meaningful stories in it—especially sci-fi, fantasy stories where so much of the background informs the details that make a story pleasurable.
There’s a certain multiplicity to that, both on the artist and viewer side. Why was this an important factor?
I kept going back to this gut feeling that making narratives in a video game engine will give us a foundation to do more “worlding” stuff later to easily hook in other kinds of tangential experiences that are more interactive. For Life After BOB, there’s a wiki that’s online, which was inspired by wikis that emerge when there’s a fandom for a particular sci-fi or fantasy world, and the changes you make on that wiki, if you were so inclined—would then influence some of the cosmetic details and, later, behavioral details of the various background artifacts and objects in any given scene. I like to think of this as a programmable movie. The hope was to make something that looks and feels like a movie right now, the way a Tesla feels like a car right now, but then can have the technological foundations to grow into something bigger.
Why do you gravitate to simulations?
We’re hearing that word in the context as AI develops. We’re hearing that in order to achieve something that resembles general intelligence, one would have to create an intelligent artificial system that can simulate its own possible futures. And then you realize, in this discourse, where we’re trying to engineer AI, that we ourselves—consciousness—is a simulated property. Consciousness is an emergent property that lives at a simulated level of a simulated representation in the models that neurons create. When you think, “Am I going to get coffee or tea? Am I going to go out or stay in? I’m going to be shy or make a move,” it’s all simulation.
How does storytelling relate to AI?
Stories for me are maps. They’re maps of how to behave and how one ought to behave, so inherently stories have a moral valence. The more we can share that responsibility with interesting AI models in the future, like an AI writer’s room, the wider the range of adaptive behaviors we can watch unfold in the form of stories. Sky’s the limit. We don’t have to always write the hero’s journeys. We can write different kinds of stories that illustrate and animate the new kinds of behavior that meet the complexity of our environment. I’m quite optimistic about that fusion.
Stories for me are maps. They’re maps of how to behave and how one ought to behave, so inherently stories have a moral valence.
You described BOB as “art with a nervous system,” and indeed, there are a lot of anxieties throughout the animation. What prompted you to touch on nervousness? And what does channeling these human emotions through art, specifically art in regard to AI, offer you?
The episode of Life after BOB starts with a subtitle that says, “it’s a great anomic era,” which means a kind of restlessness or instability that comes with a certain era, a fancy way of saying people are living in an era of persistent anxiety, where institutions are collapsing and a sense of longterm life meaning is collapsing. It’s something I was feeling myself at my mid-30s. When I was writing, my daughter wasn’t born yet. She was about to be born, so I was thinking about my own life, how it would be impacted by being a father and then of course her. How do you raise a child in what feels like a pretty crazy time? As a parent you have tremendous early influence on the life script of a child. There’s a line where the dad says, “parenting is programming.” That felt really true when I wrote that. It interweaves with AI in particular, as I thought AI would be the kind of sci-fi trope that can animate something as touchy and abstract as an existential crisis.
With the boom of digital art growing and NFTs, we’re curious, what are your thoughts on NFTs?
I think they’re awesome. It’s the best of both worlds. It’s the best of Web 1, where people really gravitated to the idiosyncratic creator side of the web. Then, it unites that with all the stack innovations that we see from Web 2. NFTs unite the best of both those things, so you can re-address the problems of centralization and all the privacy and data selling issues, and you can bring back a little bit of that individual creator ethos.
I’m really interested in “worlding.” I think there’s a way in which an individual creator can start up a world—not as a mega-institution, where you have to be the Marvel Cinematic Universe to make a world and develop a community around it—but now, as an individual creator with a blockchain economy behind it. You can create and maintain a thriving world. I think there’s something really beautiful about that.
Hero image from Ian Cheng’s “Life After BOB: The Chalice Study” (2021), live animation, color, sound, 48 min, courtesy of the artist
