The Union Between Computers and Theatre in Annie Dorsen’s Yesterday Tomorrow

Laura Gaines, Ohio State second-year, Statistics, minors in Dance and Political Science

Mar 02, 2020

A man stands with his back to the camera on a darkened stage with individuals sitting on either side of him as he looks up at screens showing the notes in a music composition in a scene from Annie Dorsen's performance work Yesterday Tomorrow

I spoke with Annie Dorsen this month about her upcoming work at the Wex, Yesterday Tomorrow, as well as her creative processes. Our conversation introduced me to a novel intersection of science, technology, and the humanities: using computer algorithms as collaborators on theatrical pieces. As a student majoring in statistics and minoring in dance, this idea intrigued me; my discussion with Dorsen did not disappoint.

I asked Dorsen to define the term she coined for her style of composing pieces, “algorithmic theatre.” She said she uses it to differentiate between her works and those of multimedia performance, in that computers and computer algorithms serve as her “collaborators.” That means she works with computer programmers to design and write code that “performs in a live theatrical situation.” The computer generates material for the performance in real time and the algorithm won’t generate the same output each time, so each performance is “one of a vast number of possible performances,” and most likely will never be seen again. 

Dorsen had very limited knowledge of computer programming going into this creative endeavor. The idea came from an essay she read by Alan Turing entitled “On Computing Machinery and Intelligence” (1950) which, she says, is “a foundational text for the field of artificial intelligence, and particularly for natural language programming, [which is] the side of computer science that’s trying to get computers to be able to create plausible speech or text the way a human would.” From this essay, and an idea from a televised philosophical debate, Dorsen created her 2010 Hello Hi There, her first algorithmic theatre piece revolving around “a very old-fashioned kind of chatbot system.” She has since created several algorithmic theatre works, including Yesterday Tomorrow.

Although they are all in the same genre, Dorsen’s works come from different sources. Unlike the Turing essay that inspired her first work, Yesterday Tomorrow was a completely unplanned idea she came up with when a friend explained programming genetic algorithms to her. Dorsen recalled the moment:

“While we were talking, I just sort of said, ‘Oh I see, so you could, for example, use a genetic algorithm to turn one thing into another, you could take the song Yesterday and turn it into Tomorrow’ … and then the next day I thought, well actually, that’s an interesting idea. It would be fun to try that.”

"The computer generates material for the performance in real time and the algorithm won’t generate the same output each time, so each performance is 'one of a vast number of possible performances,' and most likely will never be seen again."

Dorsen frequently comes up with her concepts by reading, doing research, and speaking with people, and the next idea “pops out of her mouth in conversation.” 

She also noted that her pieces have evolved through time. In her earlier works, she intended to “demystify” computer science for her audiences, particularly since her audiences had limited knowledge about computers and were not necessarily “technically-minded.” Dorsen used relatively simple programs that were easy to understand so as to prevent her audiences responding only to the “whiz-bang effects”. Instead, she hoped the audience would think alongside the piece. Additionally, her earlier pieces had “very emotional… almost biographical” sources. Today, Dorsen says she is moving away from simpler code and more personal pieces and heading toward more technically complicated machine learning algorithms and the cultural, social, and political questions they raise. Dorsen mentioned that the algorithms have “some interesting aspects to [them] in terms of very old aspects of theatre… [including] plausibility, verisimilitude, [and] representation.” As such, Dorsen’s works are shifting from this unmasking of computer algorithms to more intense machine learning questions.

Since she doesn’t want the “whiz-bang effect,” I asked if she hoped for a particular reaction from her audiences. She responded that she didn’t have a particular message she wants to communicate, but wants her audience to approach the performance “in the spirit of curiosity… I like the notion that the audience [is] very active in their minds trying to make sense of it… [there are] moments of beauty, moments of humor, there are moments that are very surprising, also to me, because I don’t know what the programs will produce each night.”

My conversation with Dorsen acquainted me with a brand-new (at least to me) intersection of science, technology, and the arts, and I greatly look forward to seeing the unique performance. I plan to bring an open and curious mind, and I can’t wait to see what Dorsen’s algorithmic collaborator shows us. 

This selection is part of Writing about the Performing Arts at Ohio State, an interdisciplinary student-led project supported by the Ronald and Deborah Ratner Distinguished Teaching Award. Students from departments across the university composed responses to the center’s 2019–20 Performing Arts season under the direction of award recipient and Department of Dance Professor Karen Eliot and Manager of Public University Programs Alana Ryder with support from Performing Arts Director Lane Czaplinski.