Opera Workshop Assignment: Use Artificial Intelligence to Help Write a New Opera. Then Assess.

Usually, the students enrolled in WP’s Opera Workshop course spend the semester memorizing and staging selections from classical operas by the masters, such as Mozart, Verdi, or Puccini.

Not this time. This fall, the students in the class are writing music for a brand-new work. And their assignment has a twist: they have to use artificial intelligence (AI) programs to help write the lyrics and music.

“Artificial intelligence is obviously a big issue in academia these days,” says Payton MacDonald, professor and chair of the music department. “These programs are here, and this project gives us an opportunity to explore how they do or don’t work as part of the creative process.”

MacDonald approached his music faculty colleague, Christopher Herbert, with the idea of integrating AI into a music course this year. Herbert enthusiastically agreed, and MacDonald developed a three-act operatic synopsis to guide the project. Set in 2034, the plot focuses on two friends whose higher education journeys are impacted by issues of access and affordability, a topic that resonates with the University’s student population, and also includes a chatbot as a major character—imagining a future in which AI is inextricably a part of everyday life.

MacDonald and Herbert, along with a third music professor, David Weisberg, are collaborating with the students in the course to write music for five to six scenes. Half the class has been assigned to write lyrics for specific pieces using ChatGPT, while the other half is composing the music using a variety of music generator programs.

A reading of the work-in-progress will be presented on Tuesday, December 5 at 7:30 p.m. in Shea Recital Hall 101 on campus, along with a discussion of how it all came together. Admission is free.

AI2_1000x700.jpgChelsea McBride (left), a master's degree student, shares some ideas with music professors David Weisberg (center) and Christopher Herbert

On a recent Tuesday night, the class is gathered in a conference room in Hunziker Hall on campus, where the atmosphere is more like a writer’s room. On the agenda: a discussion of one of the songs they are writing. With the score projected on a large screen, the group sings through it; then, the critiquing begins, with students and faculty offering thoughts on both the lyrics and the music.

The process of creating the works using AI has been a challenge. “Everyone is using it differently,” says Chelsea McBride, a master’s degree student from Toronto, Canada, who has been working on the music with the class. “Sometimes you have to go back and forth a lot. It’s about learning how to shape it to get us to the right point.”

Another student working on the music, senior Jason Braun from Howell, says that many of the music generators have been programmed with pop and hip hop tracks, not classical ones. “I’ve been using it as more of a springboard for ideas,” he says. “It’s a nice tool for brainstorming and experimenting.”

AI1_1000x700.jpg Senior Jason Braun works discusses his work writing music for the opera

“As musicians, we have ideas,” adds Austin Sidito of Holbrook, New York, who has been writing lyrics. “ChatGPT can get the ball rolling but it can be very annoying to actually get it to give you what you are thinking.”

“As a composer myself, this has been a really enjoyable experience,” says Weisberg, who has been working with the students to help shape the compositions. “It’s exciting to help them realize their vision.”

To help them get a feel for writing opera, the students also participated in a masterclass with Timothy Long, artistic and music director of opera at the Eastman School of Music at the University of Rochester and attended a performance of X: The Life and Times of Malcolm X, a contemporary opera being staged this season at the Metropolitan Opera in New York.

The professors are looking forward to presenting what the class has produced. “I am really loving that we are just exploring,” says Herbert, who has been leading the student team writing the lyrics.

MacDonald concurs. “I’m less concerned about the quality than the process,” he adds. “Successful people are constantly making things. Who cares if it’s not perfect? This gives all of us a chance to create. And that’s cool.”