The development of artificial intelligence (AI) has accelerated rapidly over just the past year. Generative AI algorithms can create images, develop scripts, make music and write poems. 

In response to the concerns raised by this potentially disruptive technology, Lehigh University Provost Nathan Urban issued a letter inviting anyone in the Lehigh community to participate in a contest that addressed the potential benefits of generative AI in education. 

The winners of the grant were Lyam Gabel, assistant professor of theatre, and Will Lowry, associate professor of theatre. The pair plan to use the funds to cover the cost to travel to a meeting or conference focused on educational innovation or education technology. 

In their entry, Gabel and Lowry addressed how they could use AI to enhance student learning at Lehigh, specifically proposing a co-taught first-year seminar in the College of Arts and Sciences that explores the intersection between theatre and AI in a pedagogical environment. 

That course, Can AI Make Art?, is a hands-on, interdisciplinary approach to exploring the question of not only whether AI can create art, but what role, if any, AI can play in theatre making, media making and storytelling. 

Throughout this fall semester, students have been creating original theatrical work with generative AI while examining the creative potential and limitations of this technology, as well as considering the ethical implications of this rapidly evolving field along the way.

“If you see the class as an experiment, we’re now coming to a place where we can look at the results for how students have engaged,” Lowry said. 

A New Pedagogy

Gabel and Lowry were inspired to respond to the provost’s contest because both have integrated AI into their coursework before, using platforms like ChatGPT and Midjourney to spur creative development in their students. 

“This is a field that’s moving rapidly, so our own experimentation is helping collaboratively figure out where we can explore and push students to engage in new things with minimal scaffolding,” Lowry said. “This course is both intersecting with how we’ve used it in our own scholarship as well as looking at how you learn using AI and how AI can be used for theatre making.” 

For Can AI Make Art?, students start out with a basic introduction to two popular generative AI platforms before moving on to more exploratory assignments.

“The first project was to write a script with ChatGPT,” Gabel said. “The second project was to create a series of images and then perform in front of them, using DALL-E to make the projections.”

The students then spent the next few weeks doing what Lowry and Gabel call “fast practices,” which is more of an ad hoc experimentation methodology for creating art, rather than a long-term, iterative creative process. 

“We just did our third fast practice, and with this one, we wanted to empower students to continue to explore on their own,” Lowry said. “We’ve used these two interfaces heavily throughout, and we’ve seen what they can do—but you also need to get out there and just play with them.” 

For this assignment, students were introduced to four generative AI programs they haven’t used yet to create a short film. The first platform is DALL-E 3, which is the next version that students have been using in class, so they can experience the differences between the two. 

The other three platforms are Runway ML, which is machine learning that can create video; ElevenLabs, which can do AI speech; and Stable Audio, which provides AI-generated sound effects. 

“Some of these students, if they haven’t done video editing before, will be, over the course of just a couple of days, using four new pieces of technology to make a brand-new piece of art, and that’s what’s exciting,” Lowry said.

Later on in the course, students will participate in Lowry and Gabel’s version of a Turing Test, a thought experiment made famous by British mathematician Alan Turing where a human judge engages in a conversation with both a machine and a human, but the judge is kept unaware of which one is which.

If the judge cannot reliably distinguish which participant is the machine and which is the human based on the conversation, then the machine is said to have passed the Turing Test because it demonstrated a level of artificial intelligence comparable to human intelligence.

Students will be given the option of either writing a monologue script themselves or writing it only using ChatGPT, and the class as a whole will try to figure out which ones were written by human students and which ones were written by AI.

Revolutionizing Theatre

One of the things Gabel particularly likes about the course is that, even though the first-year seminar is about AI and art—and its applications in the world of theatre, specifically—their students come from diverse backgrounds with a variety interests and will go on to major in a variety of disciplines. 

“So, it’s interesting discussing all of these ethical questions, knowing that our students will attack them from different angles, as users as well as makers,” Gabel said. “What are the ethics around it? What are the possibilities? And how do you develop a practice of interfacing with these new tools as they come out?”

Gabel said there are three primary ethical concerns that they’ve explored in class so far. One is the nature of creativity itself and what it means that AI is using other people’s work and then adapting it. The second concern is the impact of AI on the planet—will it solve our environmental issues, or is it creating new ones? 

“The other big one is, how does AI replicate, reinforce or disrupt existing biases and systems of oppression that are in our world?” Gabel said. “How is AI reflecting back to us our society’s racism, sexism, classism, homophobia, transphobia, etc.? Is it possible to use AI to break that, or do we need to change something about our world before AI can change its mind?”

One of the texts Lowry and Gabel use in the course to put these questions into context is More Than a Glitch by Meredith Broussard. 

“She talks a lot about how this stuff is built in and then the possibilities for it,” Gabel said, “because some of our students will be the people who will use it, and some of our students will be the people who might try to fix it.”

AI into the Classroom

Now that Lowry and Gabel are approaching the end of their experiment, both are already excited about the potential for teaching it again. And since they have developed it together, either one could offer it in the future, as another first-year seminar or as a regular part of their theatre studies curriculum. 

“We’re very excited about the enthusiasm about the course and the affirmation of what we’re trying to explore,” Lowry said. “And though these students might not take another course that’s specifically about AI, they will be interfacing with it in some way in their lives, as students and then also as professionals in an art field or outside of it.”