Member Spotlight: X.A Li

Olivia Sherman
Who are you and where are you located?

X. A. Li
My name is X. A. Li. I am an artist and computer scientist living in Chicago.

Olivia Sherman
What brought you to Chicago?

X. A. Li
I attended the University of Chicago for undergrad, and I've been here ever since. I think it's a fun, wonderful city. It's just big enough to have everything you might want, especially culturally, but it's also affordable and unpretentious. It's a really exciting community and laboratory to test down new ideas and try new things.

Olivia Sherman
What does new media mean to you?

X. A. Li
That’s a really interesting question. I was actually thinking about this recently, because a friend of mine recommended this book by Friedrich Kittler called Gramophone, Film, Typewriter. The book is all about how these “new media” technologies, at the turn of the century, were transformative in relation to the printed text beforehand. It's funny, because when we hear these words now, we don't tend to think of gramophones or typewriters as being “new media”. But I think it sets up this interesting contrast where the idea of new media is ever-changing. You have these established notions of what traditional media might be. People have been painting, sculpting, and so on, for millennia, and these are well-established media for creation. But new media tends to arise from these technological advances and innovations that are not necessarily brought into the world just for art making. They might be brought into the world for mass communication, for entertainment, for education, for something else. I think new media art is really interesting because it's both a way of engaging with the technologies of your time to unlock new aesthetic possibilities, but then also in in those aesthetic possibilities, commenting on embodying something interesting about these technological forces that are co-existing all around us and shaping all of our day-to-day lives.

Olivia Sherman
What brought you to the New Media Caucus?

X. A. Li
I've been making art in some form my entire life, but it was about 2018-2019 when I started really working with what you might consider new media arts and making videos using algorithms and technology.
New Media Caucus actually organized one of the very first shows I ever showed work in. This was an exhibition called Border Control that was hosted at the Stamps Gallery at the University of Michigan, and I contributed a work to that show. Since I live in Chicago, which is not that far from Ann Arbor, I was able to go out and see the show in person, meet some of the other artists. It's really wonderful and really exciting, especially that early in my journey in new media art, to be in this kind of environment where your work is situated in the appropriate context, and you get to see all of these exciting pieces across video and sound and installation and technology. I've been familiar with New Media Caucus ever since, and it's really wonderful to be able to receive this award now, you know, many years after I first encountered this organization.

Olivia Sherman
What does receiving this award mean to you?

X. A. Li
I'm very thankful, very grateful. I think it's nice to be intellectually and aesthetically validated, but also to have the material support to spend on resources and equipment and stuff for new projects- that's much appreciated as well. So on both levels, it's very nice.

Olivia Sherman
Would you like to talk a little bit more about the projects you submitted?

X. A. Li
Yes, definitely. The video work that is included in the Header/Footer show is actually a piece that I've only shown once before this. It was really nice to be able to include it in this exhibition. It's a video called Opening and Closing that juxtaposes the banality of the New York Stock Exchange against the inevitable externalities caused by unchecked capitalism.
This project originated from kind of a funny discovery, which is, of course, on the Internet- there’s so many artifacts and things that kind of go unseen. Some of these things are uploaded for lots and lots of people to see. Other things are uploaded just kind of incidentally to have a record, and you're not even really sure why it's up there. When I started working on this project, I found the New York Stock Exchange's YouTube channel where they upload these daily videos of the opening and closing bells. When the stock market opens, they invite a company or a nonprofit to come and ring the bell, and vice versa when it closes. And these videos are so strange because nobody ever talks in them. The back is completely fixed. There's always a group of people who are kind of standing there, like the representatives who are invited and they're clapping, they're cheering, they ring the bell. But that's it. It's very abrupt. It's very blunt. It's very non-contextualized.
When I stumbled upon this YouTube channel, it was in the thick of COVID. So, in a lot of instances, there would just be one person who's physically there, but then there would be a big Zoom call in the background with10 other people who are cheering and clapping.
I found this so fascinating because it's not really something that has entertainment value necessarily, or it's not entirely clear why it's up there because it's very unprocessed. It forms this massive catalog of every single organization or company they've invited, and they all look the same. There's this haunting quality to it for me, especially with how uniform the sense of complete cheer and complete optimism is. It’s not a very nuanced scene. It's a very absolute, “this is the best thing that could ever be happening when we're so happy and we're so excited” kind of mood.
In creating the work, I processed many, many of these videos, including organizing the sections of each according to how many people were present. The idea is that the final video stitches together hundreds, if not thousands of these original clips and it proceeds from one person at the beginning to increasingly adding more and more and more figures until the very end you have these very large crowds of people and they're all doing the same thing. The only sort of audio element is the cheering and the clapping and the bell. And the bell is so loud. It's so blaring, it's so insistent, it's so sharp. But eventually, these scenes fade away a bit. They're interpolated with footage from the Deepwater Horizon oil spill, which caused a lot of horrible environmental desolation to a very wide geographic area. This video is kind of overlaid on top of the clips of people cheering and celebrating at the New York Stock Exchange. Crucially, the people never fully go away, right? The way that the video ends is, the viewer is hovering above the water. You're watching the oil slicks spread and you can still see kind of the faint outlines of some of these people, and they're still cheering. They're still clapping. They never stop. And it's meant to be a bit of a meditation, kind of unearthing the tension between the absolute reality that's contained within that one room versus the much more complicated realities that exist and are directly connected to what happens in that one room.

Olivia Sherman
I know using AI in art can be a pretty controversial topic, especially right now. Could you talk a little bit about how you use AI in your in your work?

X. A. Li
That’s a really great question. For me, it's very interesting because I am formally trained as an AI researcher. I have a master's degree in computer science. I've worked in R&D labs. I've been training generative AI models for a long time now.
So I've had a very interesting kind of in-depth view of how the field has advanced over the years and kind of what the implications of using AI in different forms might be. To get the obvious things out of the way, I think large tech companies training generative AI models on artists’ original work without their consent is horrible. It's bad. It shouldn't happen. I think unfortunately that's going to require a much more complicated, societal governmental policy-oriented solution to enforce that. Because we generally can't expect companies to enforce themselves. But I think from an artist's perspective, I'm much less interested in using AI to generate kind of final artifacts, and I'm much more interested in misapplying AI in ways that reveals its own limitations.
The project that I'm showing as part of this New Media Caucus show, it doesn't actually use AI that much. In fact, it uses some custom image processing algorithms that I implemented, but the use of AI is primarily almost as a behind the scenes tool. I'm using AI to basically detect how many people are in each of these frames and I'm using that AI generated count to structure the sequence of the frames in the final video. That's a very boring, utilitarian use of AI- because I can either have an AI model count how many people are in each of one of these frames, or I can count myself over thousands of videos. But practically speaking, it makes a lot more sense to use the model than for me to sit down and just count every single frame, right? But I do have other more recent projects where the idea is very much, using AI to do something, but it looks weird, it looks bad, it makes you feel bad. And that's meant to highlight some of the tensions that are inherent in the way that we have fully internalized and widely deployed predictive technologies kind of across the board.
I have one recent project for example, where I took a video of my hands typing this government taxonomy of body parts and I processed the video using a couple of separate AI models where first I used what's called in search to take every frame of my hands typing and have the AI try to return one of the other frames in the video that's most similar. So it's trying to predict what the next frame might be. And in practice it doesn't work that well, so you get these weird jittery effects where my hands are moving in ways that.
the human hand obviously doesn't move in, but then I'm further interpolating between these frames using a video interpolation networks. The idea for that is that it's meant to seamlessly fill in frames between frames that exist. You're not meant to really notice the final object. It's supposed to just look like the video is smoother.
But again, in practice, when you apply this to my video, it's weird. There's kind of all of these telltale AI artifacts where you'll see briefly, I have 6 fingers in one hand all of a sudden and then it goes back to five fingers. And all of these are kind of unintended use cases of these technologies. This is not what these models were invented for. But for me, it's very interesting to apply these very dominant techniques, very ubiquitous frameworks to something as simple as the human form and have it be made very evident that these models don't understand what is human. These mathematical matrices can only do so much to represent the infinite, complex human realities that we inhabit. So, the AI is not doing a good job in a lot of ways, but that's kind of the point, and it's kind of meant to be a little bit unexpected or jarring or even bad to kind of make you think about- well then how does the ubiquity of these technologies around us affect my day-to-day life? How does this control the reality that I am engaging with as a person myself?

Olivia Sherman
Do you have any recent projects that you would like to share with us?

X. A. Li
In terms of upcoming projects, I am really interested these days in repurposing these advanced algorithmic systems to their own self-critique in a way to kind of reveal these sort of tensions and instabilities in surfaces that might normally seem seamless and inevitable, but are actually pretty strange if you really think about it.
That includes not just looking at AI in isolation, but AI as predictive technologies, as part of this regime of data and optimization and prediction that affects the global economy. The way that we're able to form social relationships affects the politics and ideologies that become dominant, that gain power. I have a couple of projects that I'm working on right now exploring topics ranging from, how do these kinds of predictive algorithms and advanced AI networks change the interpretation of reality? How do they change how events are perceived and interpreted and codified into history?
I have another project that's looking at kind of the illogic of applying frameworks of resource extraction to the natural world. To to give you an example, why is it that if we're going out to take a hike, we might look up the reviews of the hike that we go on? We want to go on this hike because it's 4.9 stars and that other hike is, say, 4.2 stars. Why do we evaluate the natural world according to either recreational value in a quantitative sense or the mineral resources that can be extracted and their economic worth?
And then I have a third project that I'm working on that is looking at how agentic A.I. systems might embody fundamental American values and then reproduce those in ways that maybe would resonate with us, or maybe would horrify us in a historic simulation. Lots of interesting ideas I'm exploring these days and I'm very excited about.

Olivia Sherman
Do you have a website that I can link in this interview where readers can view your work?

X. A. Li
xa-li.com