I recently interviewed Chris Thompson, curator of the Burlington City Arts' exhibition Co-Created: Artists in the Age of Intelligent Machines, and Donna Rizzo, Professor of Civil and Environmental Engineering and an advisor for the Vermont Artificial Intelligence Task Force. We talked about Vermonters’ concerns about AI and how an art exhibition like Co-Created engages the public in the critical conversations we need to have about how our community is impacted by technological advancement.
Jenn: In the spirit of wanting to know the provenance of things, Chris, we'll start with you. So where are you from, and where are you situated now? And in a few sentences, how did you get here?
Chris: I have a somewhat unusual background for an artist. My dad was a physicist at Princeton and worked on the fusion Energy Research project there. So I always assumed that every other kid had a connection to a big mainframe to play computer games on and a summer job designing PC boards and learning basic electronics for a giant government research project. Turns out, when I went out into the world to become an artist and musician, I realized nobody I met had actually had that background.
So everywhere I went to work, I was the person who was not terrified of computers and ended up running the computer systems at these different companies out of default. Until one day, I woke up and I was a Corporate CTO with a dozen engineers working for me, and I thought, this is not exactly what I had in mind. So I quit and went back to being an artist, and soon afterward, I was asked to be the curator for Burlington City Arts, and then, as you mentioned, after that I co-started the Generator Makerspace.
The intersection of art and technology is a unifying theme throughout all of my career. When I was a curator at BCA, I realized that contemporary art is about trying to help understand the complexities and contradictions of modern life. However, most of the artists that we were showing were still dabbing pigment on canvas with animal hair on sticks, I thought, hmmm, we need to do a little bit better than this to understand the world today. So, I started emphasizing shows with artists who use technology. This exhibition, CoCreated: Artists in the Age of Intelligent Machines, was my Pandemic project when I actually had time to dive into figuring out how to write machine learning algorithms and neural networks to use in my own art projects. I am interested in doing computer simulations of living things, and I wanted to have them able to learn and evolve for themselves.
Jenn: Nice, and so how did you get to Vermont?
Chris: Oh, my parents actually retired up here, and I made the mistake of coming to visit, and I was like, why am I in New Jersey? (Laughter) And so my wife and I just loaded up a truck and moved up here.
Jenn: Donna, how about you?
Donna: Geez, how do I beat that? I'm at the University of Vermont in Civil and Environmental engineering. I was officially trained as a groundwater hydrologist/modeler. Never in a million years did I think I'd become a faculty member. It's just not what I wanted to do. But for many reasons that I won't get into, I became very disillusioned with traditional engineering consulting. I went to California for my graduate studies primarily because I had wanted to get to California since birth. And so, I moved out there and studied engineering at the University of California, Irvine, because it was one of the few programs to have an environmental engineering program at the time, which was an oxymoron in a way. But my heart was into ensuring that everyone had access to clean water, and they also had one of the first programs in artificial intelligence (AI). So, I went there hoping that I could merge AI to solve environmental problems and do something useful.
When I graduated, I worked for one of the national labs on groundwater remediation problems. And again, my research for my Ph.D. was trying to apply artificial intelligence to help people make better decisions about how we manage the landscape and the Earth's surface with respect to providing clean water for people.
How I got back to Vermont? Professor George Pinder, who was Dean of the College of Engineering and Mathematical Sciences at the time, recruited my advisor as faculty here at UVM. And since I grew up on the East Coast, I wanted to return, so I took advantage of the fact that I was just in the stage of writing my dissertation and moved back here, finished my PhD, and started a small business. I did that for quite a while, and then taught a class here at UVM (Sustainable Engineering Economics). When a position opened up, I figured the worst thing that could happen was they turn me down… but they didn't. They hired me, and now I teach classes like Applied Artificial Intelligence and environmental engineering courses. And I love it.
Jenn: And UVM is so lucky to have you here! The context of our larger conversation today is the exhibition Co-Created, what led up to it, and the conversation we hope it sparks now that it's open. Donna, could you tell us about your involvement with the Vermont AI Task Force?
Donna: Happy to; it was a Senate Committee and people were appointed. I was appointed by Lieutenant Governor David Zuckerman, as a representative of the University of Vermont who worked in AI, and was happy to be a part of the Task Force. I worked with many amazing people on that committee. This was in October 2018; it was designed at the time to just make recommendations on the responsible growth of Vermont's emerging AI and technology markets. Vermont was unique at that time (after California) in wanting to explore how Vermonters really felt about this topic. I served for two years, and it was one that was the most fun committees for me. We had workshops and focus groups that took us to all parts of the state. The committee actually listened to people's fears, hopes, and dreams for AI technology, concerns for what AI might do to the labor market, and whether people thought the state should create laws/regulations concerning these kinds of technologies. It was a learning experience for everyone on the committee and a lot of fun. Eventually, I got to meet Chris, and you know, I figured, oh, my God, we're going to tie this to art! This is going to be amazing!
…It's human nature to want to create and invent things that have never been created. I don't think you're going to be able to regulate or stop people from being creative. It's what makes us human. The regulation really belongs on the intent and how you plan to use that tool.
Donna Rizzo, environmental engineer
Jenn: Could you speak to some of the recommendations the committee proposed?
Donna: There were two themes that everything really centered around because you have to imagine the advantages and risks associated with AI is very different depending on the field that you're in. So, the way we tackled it is was to focus our committee discussions on themes (specific applications of AI) for example, in the healthcare arena. Another focus group would be on applications of AI in precision agriculture, another in transportation, etc. You can imagine that many of the fears at the time stemmed from the introduction of self-driving cars and other types of technologies that might eliminate jobs for Vermonters. But it was really interesting to have individual workshops on focused areas because people didn't seem to mind these technologies being used in certain sectors of the marketplace. Whereas there was much more fear/concern in other sectors (e.g., law enforcement) and whether these technologies will take away jobs. In Vermont, when we only have 650,000 people in the state, it is one of the poorer states in the nation from a tax revenue point of view. So, job elimination is a substantial fear. So, listening to people from around the state and learning about their fears and what the state might do to keep businesses thriving and wanting to come to Vermont is important. Some of these AI technologies are exactly the kinds of companies that you'd want to attract to a state that doesn't want to attract assembly lines, more pollution, and perhaps offer higher-paying jobs. So, there’s a delicate balance; we did a lot of investigative work. The recommendation is to be very cautious before putting any kind of regulations on industry when it comes to these technologies. We spent the majority of our time talking about ethical issues, ensuring that AI technologies are transparent, and how to maintain the freedoms and liberties of Vermonters.
We actually provided testimony to the legislature and recommended creating an AI Division within the Agency of Digital Services (guided by an Advisory Council) to implement the Task Force recommendations, which I believe is underway right now. (I'm not a member of the second task force). We also spent a lot of time, I might say - stealing ethical... hmm…that doesn’t sound like the right combination of words…(laughter).. but Canada and the European Union (and California) all had recommendations and codes of ethics for AI technologies that we kind of stole different pieces from to provide some recommendations/cautions regarding the ethical use of AI that people might want to bear in mind when considering these kinds of technologies.
Jenn: I remember talking with you about your conversations with farmers. Could you explain how rural communities may have different concerns about AI than urban ones?
Donna: Precision agriculture, the idea of using AI technologies to produce higher crop yields while saving farmers money with respect to reducing the amount of irrigation water and fertilizers that are needed – applying them at the right times. There are sensors that use AI technology that are built into farmers' tractors, and they're now widely used, especially on large farms. One of the fears is that if smaller farms can't afford these same kinds of technologies, will they lose out in the marketplace? We have a couple of grants now, where we're hoping to use satellite imagery to provide some of these same recommendations that can be offered via sensors (images) for free to farmers so that you don't have to buy the latest technology to make these same advances.
But one of the things I learned in listening to Vermont farmers is that they have a very progressive mindset. Vermont farmers, obviously don't want to go out of business, but they care deeply about the land and environment, and if they can do something that's a win-win in the sense of, 'oh, I can apply less fertilizer, which is going to cost me less, and cause less groundwater contamination,' they are on-board. We also have a pilot program researching payment for ecosystem services. So, if farmers do things on their land that actually help the crop yield but at the same time reduce the amount of phosphorus or nitrogen that's leaching from a particular field, there is a means to be partially compensated for these efforts. Farmers are really on board with this. These kinds of technologies in certain areas can be really beneficial.
Jenn: When working with the Vermont AI Task Force, at what point did you meet Chris?
Donna: Chris approached us during the pandemic (after the first Task Force wrapped up) and asked if we might tie some of the ethical issues about AI into the art + AI endeavor that he was launching. Chris is a kindred spirit in that he (like me) tricks students and kids into wanting to learn computing while doing something fun…in his case creating art or music without having them even realize the pain they’ll suffer if they want to go that route (choose a field like computing).
Jenn: Chris, can you pick it up from there?
Chris: As I mentioned before, Co-Created started as a pandemic project for me, and I was just excited for the opportunity to be able to do more sophisticated things with my art projects, including learning how to program neural networks. As Donna was talking about, I thought I'd never use math after 12th grade. Yet, art was actually bringing me back into math. When I was trying to make little software creatures move around in certain ways in my simulations, I was like, oh, sine and cosine. OK, I gotta figure out how this works. I discovered that math was actually fun and fairly painless, quite unlike my educational experience.
As soon as I started getting into neural networks and machine learning, I realized a lot was going on in the Artworld while I wasn't looking. For the years while I was executive director, Generator, we had some great talks on machine learning as part of our Reckless Idea Speaker Series, but I had yet to actually dig into it. When I did, I realized that the whole generative AI machine learning revolution had snuck up on me. I started talking to - you, Jenn - and talking to other local artists, and I realized there was really something going on here. So I did what I always do when something is happening that I don't completely understand; I say it's time to do an art show about it!
…Artists tend to function like a sort of cultural antenna—they tend to be more in tune with what's changing in the world than other people are. So if you want to get at something that's on the edge, you almost inevitably can find some artists who are probing it. They do it in this wonderfully cross-disciplinary way, perhaps mixing color theory and physics, which probably wouldn't be acceptable in the scientific world but is totally OK in the arts.
Chris Thompson, artist and curator
The idea is that artists tend to function like a sort of cultural antenna—they tend to be more in tune with what's changing in the world than other people are. So if you want to get at something that's on the edge, you almost inevitably can find some artists who are probing it. They do it in this wonderfully cross-disciplinary way, perhaps mixing color theory and physics, which probably wouldn't be acceptable in the scientific world but is totally OK in the arts. So that was perfect.
I started talking to artists about it, and soon after talking to you and a few other folks, I spoke to John Cohn about the possibility of doing a project for the show. He said let me tell you about the Vermont AI Task Force I'm working on: Everybody we talk to seems terrified of artificial intelligence — there are actually so many good things happening, but people just shut down right away. We decided that in addition to doing a great show, perhaps an art exhibition would be a more effective way to engage the public in a discussion about AI.
One of my favorite phrases is "hiding the peas.” We did lots of youth education at Generator in which we would “hide the peas” integrating all sorts of science and math in fun art and creative projects, and the students wouldn't even realize that they had just been learning physics.
We thought the show would be an interesting and none threatening way to start some of these conversations that we thought Vermont and our country as a whole really needs to have about what we want AI to do. Conversations about what we want the limitations to be, what the advantages are, and all the exciting things that can happen. But also, how do we frame it without all the fear? Burlington City Arts Center was excited to do this show because we rarely had time to do such a complex exhibition when I was the curator there a number of years ago. I approached them about doing this as a guest curator, which gave me the time to do the sort of show I never had – I was always throwing shows together in a month or two. For this project, I had two years.
The additional time allowed us to think about doing a symposium, panel discussions, youth education and all these other bigger things that really add to a show and engage the community.
Jenn: Chris, I was just thinking about when you and I first talked about the show; it was outside my parents' house in the New North End, I think the spring of 2021, when I was still in a COVID pod with my family and a few neighbors. You were on your bike, headed for the bike path.
Chris: Recalling that time in COVID, when our conversations were so isolated, it seems in stark contrast to the stage of COVID we're in now. The opening reception for Co-Created was lively and well attended, the weekly newspaper Seven Days published a cover story about the show this week, and an audience is engaged in a conversation about ideas we were mulling around in COVID isolation.
I thought the show would be perceived as covering an obtuse topic, but two months before the show opened, the revolutionary generative text ai ChatGPT came out, and the media world exploded, and everybody freaked out. It turned out to be the best-timed show that I ever curated. And, of course, I wish we meant to do this, and we knew all this would happen ahead of time… But yes, it's turned out to be a much bigger show. BCA’s said it is the most popular show they've had in a number of years.
So, there's a lot of exactly the conversations happening that we hoped would come out of it. But, for two years, I was just researching with a lot of outdoor meetings with artists. Having the time I did during the COVID isolation was perfect for digging into something that was outside my comfort zone.
Jenn: Donna, in the past few years, have you observed a difference in how your students approach AI as an area of study? Do things feel different?
Donna: Oh, God, yeah. In fact, the timing of this show, I think, kind of saved me because when COVID hit just as the first task force was winding down, we had just put in recommendations. Everything got put on hold because of COVID.
Even teaching here at UVM, we were determined to have some hybrid classes where students would come in-person and attend remotely from home the next day. This would reduce the class size to half and allow students to be spaced apart. The students were absolutely amazing about abiding by rules - talk about a code of ethics!
But after two years of working on the task force, it felt like a letdown in the sense that recommendations were going to be put on hold indefinitely.
So, to have Chris appear via John Cohn, knocking at my door, to talk about co-creating a creative AI symposium/art exhibition that could happen a year or two down the line, with public discussions around ethical AI and art. Oh my God, it was extremely hopeful for me. The one thing students (and faculty) learned during COVID is we don't like learning online. Imagine all the hoops we make students jump through to get an engineering degree; you don't love all your classes. For a student to learn online and not have the camaraderie and typical social interactions… Honestly, Chris, you saved me by knocking on the door and saying you wanted to do this two years in the future, so thank you.
Jenn: Chris, two years ago was before text-to-image generators like Dall-E and bots like ChatGPT were part of popular culture. How did you make decisions about what work you wanted in the show?
Chris: I focused on artists treating machine learning like it's an artistic medium—the idea of somebody digging into it and exploring the medium. I focused on artists who were building their own data sets and experimenting with the code, hacking or tweaking it. While work that was coming out of the early generative programs looked pretty cool, I saw very little that was compelling in a conceptual way. Contemporary art is all about the idea first, and then you figure out how you will render it from there.
As I got into it, I became increasingly interested in how the relationship between artist and medium and audience and context is all changing. Machine learning messes it all up these relationships in wonderful ways. When selecting artists, I was also interested in artists who were thinking about how they're collaborating with this new tool and about the new relationships between artist and machine.
I was lucky that there were a number of people locally who, like you, Jane Adams, and Lapo Frati who, were doing really interesting stuff. Ultimately the exhibition became a mix of international and local people.
Jenn: I appreciate that your essay in the BCA gallery guide asks many questions and doesn't assume answers. You ask big questions that my research team has also grappled with, like what is authorship?
Chris: That's what we got right now—there's this chaos that we were in…the genie is out of the bottle now. What do we do? In the last two years, companies like Google, Microsoft, and Open AI, were being really circumspect around ethical concepts. 'We've got this machine learning model, and we're not releasing it. We're just testing it. We're being really careful about it.’ But as soon OpenAI released ChatGPT3, that all just went right out the window, and now all these big companies are terrified of being left behind; they are all competing, and all that careful ethical and incremental work is just gone now. So we're all in this moment of chaos; it's super exciting—if exciting is the right word.
Jenn: Right. Well, maybe we should unpack that? (laughter)
Chris: I'm stuck by the Domino effect of public announcements from social media platforms, all announcing new AI chatbots at a time when the general public seems really uneasy with them. What will we get from these features, and what do we potentially lose? The public response seems uncomfortable and conflicted.
Donna: I think it would be very challenging to put regulations on AI technologies. I don't believe that we should regulate AI tools themselves. They are just tools and are usually designed with a specific purpose in mind. Hopefully, that purpose is good. But once a tool is developed, it can be used for many other applications. One of the things I learned from working in the environmental arena is that when you design a tool or product, you really should include the cost of properly disposing of the byproducts properly when bringing the product to market. If the costs of doing things responsibly are too large, then certain products wouldn't make it their way to market.
I believe it's human nature to want to create and invent things that have never been created. I don't think you're going to be able to regulate or stop people from being creative. It’s what makes us human. The regulation really belongs on the intent and how you plan to use that tool. I mean, and that's what was so interesting when Chris started talking about using AI and art.
The exhibition Co-Curated Artists in the Age of Intelligent Machines is up at Burlington City Arts through May 6, 2023.
The Vermont Creative AI Symposium on Art and Artificial Intelligence is on Saturday, April 15, at Hula Lakeside.
This interview was conducted from the UVM Howe Library podcasting studio on Wednesday, March 6, 2023.