In this episode, Ananya Bhargava interviews Lance Gharavi, an experimental artist and scholar, professor in the School of Music, Dance and Theatre, and Associate Director of ASU’s Interplanetary Initiative. The discussion tackles the complexity of defining creativity in the age of AI, the metaphors we use to make sense of intelligent machines, and how artists and storytellers can respond to this technological shift.
Transcript
[Intro Music]
Ananya Bhargava: Throughout history, technology has shaped the way art is created, shared, and experienced. Today, artificial intelligence is the latest force in that ongoing evolution, reshaping performance, design, and the creative tools artists rely on. As AI becomes more present in artistic spaces, it’s challenging long-standing ideas about what creativity is and how it happens.
I’m joined by Dr. Lance Gharavi, a professor in the School of Music, Dance and Theatre, and Associate Director of ASU’s Interplanetary Initiative. His work explores the intersections of performance, science, and technology, challenging traditional ideas about art.
In this episode, we’ll dig into the complexity of defining creativity in the age of AI, the metaphors we use to make sense of intelligent machines, and how artists and storytellers can respond to this technological shift. We’ll also look at how new technology has shaped art throughout history and what that means for the future of creative work.
Bhargava: When we talk about the more consumer facing AI, and we look at the things that it’s able to create, now, it really makes us question, like, what creativity is. So I would love to know how you would define creativity when it comes to AI.
Lance Gharavi: Words like creativity just don’t work that way. You can’t just decide that everybody else is wrong about creativity, because this is what creativity really is. One of the things that may be interesting about the intersection of AI and creativity is some social scientists have developed tests for creativity, and they’re fairly sophisticated tests, but they’re also kind of disappointing and inadequate, like a lot of tools we have around creativity and measuring it. But I will say this: currently, AI on these tests blows humans out of the water. It’s not close. One really standard fun game you can play is you take a common or even uncommon object, like, here’s a cloth napkin, and you play a game where you pass the object around, and you say, “Okay, come up with a use for this object.” And it can be normal, you know, ridiculous, whatever. And you keep passing it around, and the sort of low hanging fruit gets taken, like the obvious answers, like you could use this as a napkin, or you could use it as a towel, or the really obvious ones get taken early, but you keep going, and so after a while, it becomes really difficult to come up with the use that no one has thought of before. And this is really where your creativity comes in. If you asked an AI to play this game with you, you will not win, because it will keep generating ideas. If you ask it to make you an image that is a painting of some flowers, and you keep asking it to make new versions of it in new and surprising ways, it will do that all day. So at least with some very narrow definitions of creativity and some tools that we have developed to test how creative you are, AI can do that better than humans can.
A lot of people make other kinds of arguments that AI can’t really be creative because it’s just predicting the next token. I think those things often depend either in a misunderstanding of the history of art or a particular kind of humanistic “there’s something special about us” set of assumptions that are worthy of being questioned. I know I’m creative, so for me, AI becomes a kind of tool to use, to interact with. Here’s the way I sometimes think of it: AI can be a prosthetic for thinking and a prosthetic for creativity and imagination. To me, that’s a useful way of thinking about it, rather than thinking, “is the AI creative?”
Bhargava: A big debate surrounding AI is often, is it a tool, or is it something that’s going to replace us? And so usually, the argument making sure that it does not replace us is that it isn’t truly creative and it still needs human input and things like that. So I would like to ask you, since you sort of talk about it almost being like an extension of us, a prosthetic, would you say that it’s simply a tool that enhances human creativity? Or do you think that it will ever go a step further what a lot of people are afraid of, that it would replace human writers and actors and artists entirely?
Gharavi: I think that the metaphors that we use matter. They matter because they not only are used to describe our world, but in some significant ways, they produce the world that we live in. They shape the way we think. And so if we talk about, as I have here, as you have, if we talk about AI as a tool that prescribed how we think about it and what we can imagine doing with it. Because we know tools, we work with tools all the time. You know like, Photoshop is a tool and we use it in a certain way. A hammer is a tool and we use it in a certain way. And I think it’s useful to think about AI that way, in some contexts. But what happens if we change the metaphor and don’t think of AI as a tool, but we think of it as an assistant? You can tell it to do something, and it’ll go do that thing. An assistant has an agency to it that a tool does not. Now put both of those words aside and think of AI as a partner or a collaborator or a colleague, and suddenly the level of agency ramps way up because a colleague or collaborator is different than an assistant. An assistant is someone you tell what to do. A collaborator, a partner, a colleague, is someone you are in dialogue with. And the interaction between you and the colleague or collaborator produces outcomes that are richer than either of you could have developed on your own. So these words, these metaphors we use to talk about AI, that are not just in terms of appropriate descriptions of the technology or the platform, but they shape, they guide how we think about them, how we work with them.
You also asked, can it replace us? Obviously, currently, ChatGPT doesn’t do anything until I enter a prompt, but right now, they’re working on AI agents – that’s one of the big, buzzy terms – and these are AIs that can act in the world, right? And you still have to give it direction, but it can then go on and book your trip to Coachella or whatever it is that you tell it to do. And we’re still pretty much in the early days of these things, but one can imagine, in the not terribly distant future, that these agents, you could give it a much more complicated task, like create an ad campaign for my new product. It’s already happening. A lot of coding today, for example, is now done by AI, possibly by the end of this year or sometime next year, 90% of coding will be done by AI. You know, you may still need to have computer scientists, software engineers, be there to manage the team of AI that are doing the coding. Will it replace jobs? Yes, it will. You know, it’s hard to read the available data to tell, is it doing that now? And how much? You know, once it’s a better coder than 99.9% of humans, or 100%, what do we need the coders for, the human coders for anymore?
Bhargava: What would you think is the response to that? So especially we talked about coders, but I know that a lot of artists work in fields like marketing, graphic design, things like that, where they can take their skills from fine art and apply it to something that makes them a bit more money. So what advice do you have for them, what should the response to AI be from the side of human creatives?
Gharavi: Learn to use the technology. These technologies are still very new, and they’re advancing more rapidly than our ability to adapt to them. They’re advancing more rapidly than they are being adopted and because of that, the set of social and cultural norms and practices around these technologies are still in development. It’s not like we have no agency here. And so if you are an active part of that conversation and an active part of the set of practices around creative work and artificial intelligence, you can help shape how those sets of practices, policies, procedures, ways of doing things evolve. I say we have some agency. We get to decide, to a certain extent, how we’re going to use and not use these technologies.
There’s a sense, though, in which we feel a lack of control here. And that feeling is fairly accurate. It points to larger social forces, like capitalism, for example. But I think we still, within that context, have agency about how we do things there. I also think if we’re looking ahead, say, five years, where, let’s say we have advanced artificial intelligence that can do any cognitive or creative task that a human can do as well as any human. It may be that there are certain things we still like it when humans do that thing. And I think some things about the arts may be like that. We may know that a robot can play guitar better than any human possibly could, but we may prefer to go see a guitar player live than watch a robot play guitar. AI can already play chess better than any human being, but there are still chess clubs. People still like to play chess. People still compete. So I think certain things that depend on the hand of the artist and the visibility of the hand of the artist, especially things like live performance, live theater, live dance, live music, might start to take on a greater cultural cache, a greater value, because everything else is made by these wildly brilliant artificial intelligences.
Bhargava: We can see sort of that happening in history with certain technological innovations as well, where we see the general public relying more and more on technology, but that just creates space for, you can call it, more authentic art forms. So you’ve mentioned AI as a collaborator. You’ve mentioned how it can actually enhance creativity. And I know you have quite a bit of experience in this field, so I would love to hear about any specific examples of ways you’ve seen AI, or maybe even other technologies, enhance human creativity.
Gharavi: I suggested to a friend of mine that at some point in the future, maybe it’s near future, maybe it’s more distant, artists are going to seek out the early AI models that we have today that are kind of crappy. If you remember a couple of years ago, text to image models, you know, you type in a person interviewing someone over Zoom for a podcast, what we would get back is 12 fingers and kind of weird features and looks completely wrong and in various ways. And I said, artists are going to seek out those early models, because what they produce has this certain aesthetic quality that the newest AI models just don’t have, and that would be very, very difficult for a human to produce. And this is a very familiar idea to artists. Take musicians, for example, a lot of them will pay a lot of money or go to a lot of trouble to seek out old amplifiers or old mixing boards, or old guitar pedals or synthesizers that obviously are not nearly as sophisticated as what we have today, but because they produce a certain quality of sound that they want that is more interesting to them than much more sophisticated equipment that we have today.
What I’ve seen already with some of these artists is they’re doing that. They’re using AI not just “make me a perfect dolly shot of going through Times Square on New Year’s Eve.” Not just trying to make things that look like cinema made by human beings today, but rather to make something that’s unlike that that looks surprising and even shocking. And this is something very old in art, before photography, a lot of painters made their living by painting portraits. If you wanted an image of a loved one or of yourself, you had someone to paint your portrait. But once photographs became widely used, you didn’t need necessarily to have someone paint your image, because you could just have a photograph. And so a lot of work for those artists went away. Painting no longer had to focus on realistic representations of the world, because photography could do that as well or better than any painter could. And so painters had to figure out something else to do. And what happened was, painting just erupted in creativity in new ways of approaching the medium. And ultimately, photography did something very similar when photographers begin to realize, you know, we don’t just have to make a realistic depiction of the world, there’s all sorts of things we can do with this medium. I think it’s an exciting time to be an artist, to be a scholar, to be a learner, but it’s also really terrifying.
One of the terms that I’ve heard Ethan Malik use, he’s a professor, and he blogs a lot about AI. He talks about the jagged frontier of AI. And what he means by that is that AI is not progressing in a smooth line across all sectors and all disciplines. It might be really advanced when it comes to coding or solving mathematics problems, it might be quite good at writing research reports, but not so good at producing beautiful and consistent video or writing convincing narrative fiction. I think that capability is going to keep advancing. I was on another podcast a while back, and they were like, “Oh, people aren’t going to make art anymore because of AI.” And I’m like, “Of course they will! This is what we do as people.” We may not code anymore, we may not do stuff that we don’t really want to do, but people want to sing and perform and create beautiful, weird, ugly things. And that’s not going to stop. And even the, you know, the market forces of capitalism, even if they decide, oh, it’s much cheaper if we make Hollywood films using all AI, okay, great, but people are still going to want to pick up a camera and frame a shot to tell a story. My hope is that as AI advances in the next few years, it’ll make our lives better and richer and fuller with new possibilities.
[Outro Music]
