I wanted to start by talking about the basics. What exactly is the difference between AI-generated art and digital art? 

Digital art is a series of very carefully procedurally generated images. It’s just like how you would draw. You have a piece of paper that’s 8.5 x 11, and you’re drawing on that. If you can imagine, in digital art, we have cells that make up a plane just like a piece of paper, and those cells can be assigned different colors. We can create lines by saying, okay, cells from this point to this point can use a similar color field, to draw something that looks like a line. It’s not exactly the same, but there’s a lot of very careful work that’s done in that. That’s not unlike the rules in traditional art in which you decide to paint with certain colors and to use certain brushes to cover a canvas. 

The difference between that and AI is that in AI, you’re asking it to generate art based on a series of keyboard prompts. So, you’re no longer making really exacting decisions about what is placed where. A lot of that comes through this kind of serendipitous combination that is actually kind of exciting to people because they don’t know always what they’re going to get. They’re going to describe a thing and they’re going to see what the data set sends back to them. There’s a lot of people who like to take some visual improvisation and put that into the art making process. 

With AI there are going to be consistent results, especially with, say, the DALL-E system. Everything kind of looks exactly the same in DALL-E. They always have the same very illustrative style and it’s going towards a certain type of preference. Systems where you get more variety tend to be the ones where people are setting up their own AI datasets and trying to push things a little bit further. But it’s still images. It’s still a plane of values. And you can still do something with that image, even though you generated it in AI. 

AI and art have had a bit of a complicated relationship, especially with concerns over transparency. I’m thinking about the 2022 Colorado State Fair when the winning piece in the digital artwork category was found to have been the result of AI. There are also all these fears right now that AI will be used to steal artists’ work. I wondered, when you see the future of AI, what do you see? Is AI art a friend or a foe? 

I don’t think it’s that kind of binary. I think that AI has incredible potential and just like any new technology, transformative technologies generally scare people. People see it as a means from which their livelihoods would be taken away. But if you think about the creation of the industrial loom, for instance, you can actually say old school looms are the original automatic computers. People were terrified that they would mean that no one would ever make any textiles by hand again. And we know that that’s just not the case. In fact, the looms made textiles more accessible, just like AI will make images more accessible. 

But to really make truly transformative images, it takes a little bit of dedication, it takes a little bit of time. It’s not an easy thing. When we see AI images, they’re just searching through datasets. I’m a programmer and I’ve actually played a lot with AI. It can only synthesize what it has been given from those datasets, and it can only produce hybrids between established concepts or a set of criteria. It’s an average of an average, and average sometimes can be great, but average is never transformational. Art always has the potential to be transformational, but it’s not always transformational. So, I think that that’s going to continue. There’s going to be some good art made with AI. There’s going to be a lot of bad art made with AI Those things are going to consistently shift. 

There have been a lot of questions about the validity of AI-generated art and while I think we can all safely agree that ChatGPT isn’t about to produce the next great American novel, when it comes to visual art, the quality seems to be a lot higher. 

There’s a preference, especially in American culture, for representation and for certain types of esthetics that help make people look more heroic or focus on a certain standard of beauty that is always shifting. I think art is a lot more than that. When you say that AI is, quote unquote, getting better. It’s really kind of touching on the kind of values of fashion photography, for instance. 

When I look at a lot of AI art, I find it really boring. It’s always the same kind of lighting. It’s always the same kind of really light smoothing textures and tonality. It’s not anything that feels unusual or like, I haven’t seen it a million times before. It’s like the chicken sandwich craze. Everyone will love that chicken sandwich for a bit. But then you realize you need a little bit more variety in your life. That’s also the case in art. It gets exciting at first and like all things, those trends kind of move on and the things that really last, those are the ones to kind of look for. 

Going back to the history of technologies, the camera obscura — which was eventually a technique used in photography. But with the original camera obscura, you have a little pinhole right in a box and the world is reflected and projected through that pinhole upside down on the other side of that box. People were using that to bounce mirrors up and they would trace the representational world. There were people at that time wondering: Is this going to kill drawing? Is this going to kill painting because you can now directly trace what you’re seeing in the world. The answer was, no, it didn’t do that. But also, painting and drawing evolved after that. After people realized that they could represent the world perfectly, painting and drawing started to abstract the world. And then those abstractions became really huge. Artists always find a way. 

AI’s going to be the same way. Someone’s going to come along and be like, okay, how can we deconstruct AI? How can we actually smooth in a certain way or rearrange its esthetic biases, or its procedural biases and find something new to combine it, or how will we generate some AI? There’s a lot of great artists right now who are generating AI and then they’re deconstructing those images in another more controlled format so — again, everyone finds a way. 

Who are the artists who are using AI really well? 

Well, there’s one artist named Dawnia Darkstone. She’s an artist who works a lot in NFT‘s, which not everyone loves those things. But Dawnia finds ways to have AI talk about this idea of gender fluidity, of cultural fluidity, and takes some of those images and then doing processes where she’s glitching them. What that means is that she is purposefully making them fall apart and in many ways she’s able to get to some of those bigger ideas about AI as something that’s automating things, like we’re going to work with the assumptions of how the world works and we’re going to represent it based on those assumptions. She’s quite literally unraveling that. 

Andrew Deutsch has some amazing animations that he did when Austrian artist Herman Nitsch died. He used AI to generate imaginary ideas of Herman Nitsch performing certain art pieces and then reconstructed some of that in After Effects in a way that made the whole rhythm and flow of the video really quite fabulous. So, even though it is really in its beginning now, there’s a lot of artists like Karl Erickson who are training AI to come up with new creepy Muppet characters for his animations. He’s a professor at Rhodes College. And I’m always amazed how different people are using the basis of those things and creating their own agents using things like stable diffusion. It’s really quite exciting times. 

I’m guessing that people probably treated digital art the same way that they’re looking at AI right now, with some fear and uncertainty. But now, it’s become a major area of artwork. It’s a lot more accepted. So, do you think AI will be treated the same way? 

The early days of anything in any new art form are always met with a little bit of a conservative backlash. People for the longest time would not accept photography as an art form. Part of that is that we like the things that we set up, and when something challenges the time and attention away from that, people get upset. Ultimately, that’s a cultural value that we have to examine. 

I think that the world is better with more kinds of images. Just like the world is better with more kinds of people and more kinds of ideas. So, it’s going to take a little while, but I think because people are using it so quickly, I think it’ll be more accepted. Then honestly, the hype will die down. People don’t just want easy things. The really memorable parts in life are the struggles and the triumphs that we go through. When we really work hard on something and it finally works out even just a little bit, there’s this relief that happens. 

We’re going to find that the easiness of AI is going to get a little old pretty quick. Just like when social media was huge. We’re still using it, but Facebook these days looks like a wasteland because people got sick and tired of screaming into the void and always being online. Young people these days don’t do the whole Instagram thing as much anymore because they find it cringy to always be promoting yourself. These things have ups and downs and trends and go round and round. But I think eventually it’ll be a part of a larger conversation about what creativity really is. 

How will you handle AI in your classes? Are you concerned at all about students maybe using this as a quick fix? A cheat? 

That is something that we were a little bit worried about initially, until I realized that I’m always encouraging innovative solutions. The things that we do would be very difficult to create in AI. Like, we’re not asking students to create a poster for me for the next “Dune” movie. Students can really figure that out pretty quickly with AI. 

We’re asking them to critically think through the process. We are getting them to look at artists who have innovated things. For example, I want you to work in a photo montage, but I want you to include these types of elements. I want you to include scanned textiles that you find. Well, if you have to scan some textiles in as a part of your compositions, it can be kind of hard for AI to get that quite right. So, if they have one element of AI, but they have all different other elements and they’re printing it out and putting it into a giant multi-media canvas with paint and print and different things, even though they’re generates of those images digitally. 

I think what they’ll find is that art that’s really, truly transformational is going to accept many different disciplines and techniques. That’s what we really push, especially in our first-year foundations courses at CSU. We really put them through the ringer of trying to get them to experiment with a lot of different things. Art’s not always about production. When they get to the real world, it is. But when they’re learning about it, it’s about a journey. We’re not so interested in results. We’re interested in how they learn. That’s the big deal. 

With the fall semester beginning how are you going to talk to students about whether they should use AI and how they can use it ethically? 

In the art world, what is copyright is actually kind of a slippery subject. We have people who are modifying Nintendo games and making them into conceptual art pieces in ways that aren’t really ripping the game off. They’re kind of referencing the game. But then you have artists like Richard Prince, who literally will take an ad and crop it and reproduce it. So, we talk about appropriate appropriation and what does that mean. When we start to take those images and place them into a new context, is it actually starting a new idea or are you riding on the coattails of someone else’s idea? 

It’s a difficult and long conversation, but it’s something that all our art faculty talk about a lot. So, our students are really well trained in understanding where those lines are. But we’re also introducing them to so many techniques that things that are easy answers, like somebody may in the graphic design world may come up with a thousand logos for their boss by just doing stuff in AI. We’re not asking for things like that. We’re asking about process and about repeating different processes in different ways. So, it’s not even possible to do an AI-generated image, like when we’re literally scanning items or I’m literally teaching you how to edit a thing and we’re going through the programs. 

We talked a lot about some of these transformational methods that you mentioned, the loom, the camera, and how they changed the art world. How do you think AI will change how we look at art? 

Honestly, you can never really tell, but we can look to the past, like the camera obscura and, once you can get representation well — and the same thing in digital art. There’s a trend that people call “digital painting,” which I hate that term because it’s not painting. Let painting be painting and, paint has its own qualities. What they’re talking about is illustrative qualities, everyone wants to illustrate something, and they want to illustrate something representational because it seems hard. AI makes that direct representation of a thing really easy. 

I think the more difficult ideas or the things that are on the fringe of art right now will start coming into the foreground. Questions about whether we can get back to some sort of abstraction or not worry so much about representing the world we’re in, but instead a world that we wish to see. All different kinds of things like that I think will start to come to the foreground. 

It also could really spur more performance art and people wanting to see things in person or integrated into their everyday lives. There’s a bit of a distance until we become more cybernetic or whatever you want to call that, that we still feel this distance between certain digital properties and our own bodies. So, people may start to experiment with stuff like that. There’s also this idea of an avatar or trying to project an idea about yourself. I think people are starting to get over that. Like I mentioned social media and so what other ways can we deal with representation of who we are, what we are as a culture and stuff? I think we’ll start to evolve through that. 

So, it’s a long answer. It’s not definitive, but you know, any time that we see a certain push, especially by corporations that this is a thing, you will always see a backlash. And then it’ll keep going back and forth. In the immediate, we’ll definitely see a backlash against that. Through that, we also may see some people who are thinking in really interesting ways, and then those people’s ideas and art will come to the foreground.