The following prompt (in bold) was input into OpenAI’s GPT-3 Beta Playground using the Davinci model. The text in italics was the output for the prompt completed by GPT-3.

A Description of a Performance Artwork:
From the moment that a performer enters the space, they are being watched. There is a camera positioned on the opposite side of the room. The performer can hear music playing from a speaker. If the performer looks at the camera or the speaker, they will see an image of themselves on the camera or hear their own voice from the speaker. As the performer moves around the room, they can hear their shadow on the floor and see their shadow projected on a large screen opposite the camera. They can hear their voice echo in the space. They can see a video of themselves projected on the screen. The two elements – voice and image – create a feedback loop between the performer and their environment. The performer’s voice, both in real time and as it echoes, is in conflict with the image of themselves. The performer’s voice and image are both competing for the performer’s attention. The performer is both smaller and larger than the image on the screen. The performer hears their voice echo on the floor, but the image of themselves on the screen is seeing them in real time. The performer’s body becomes a manifestation of the feedback loop. It is a physical manifestation of the verbal and visual feedback that the performer experiences. In this performance, the performer’s body is a manifestation of the feedback loop. The performer’s body is not only competing with the feedback loop, but also invokes a feedback loop between the performer’s body and the audience’s body. The performer’s body is competing with the feedback loop for the attention of the audience. The performer’s body is larger than the feedback loop that they are experiencing. The performer’s body is in conflict with the feedback loop. The performer’s body is both smaller and larger than the feedback loop. There is a feedback loop between the audience’s body and the performer’s body. The performer’s body is competing for the attention of the audience and competing with the feedback loop. The performer’s body is larger than the feedback loop. The performer’s body is also a manifestation of the feedback.

Performance:

Production Date: February 06-11, 2021. Performed by: Chelsea Coon

Statement:

This GPT-3 prompt was challenging to realize. I must have read and re-read the instruction more than a dozen times before deciding on steps to produce the composition. It was a process that had to unfold in its own time – meaning, no amount of sketch or pre-planning was as effective as feeling my way through the process to better understand what the opposing webcams, mirror, audio recording, and my body were able to do in the space of this performance.

In my experience, endurance came into this work through the amount of trial and errors needed to achieve this final outcome. While the video is only 7 minutes, 21 seconds, the time of all the elements in this work took a different amount of time to take form. My interpretation of the GPT-3 instructions towards the realization of real time (2021) was that this performance was meant to encompass multiple spaces, times, perspectives, audios, etc. It was a composition of contradiction and confusion—and significantly disrupted a sense of orientation to space and time. A critical detail that emerged in the unfolding of the work was the fingerprints that smudged the mirrors edges which accumulated over the duration of the work and depicted the traces of my body in the mediated digitalscape. These traces on the mirror represented my body pushed up against the boundary of the edges of this object/frame/space.

Further, in this performance the body was trans-temporal as evidenced in its presence in: video and audio pre-recordings, a “live” recording between two webcams and a mirror, as well as the final video for dissemination online. 

In my process, I filmed myself looking straight into my webcam while generating an offbeat metronome sound with my fingers against a plastic object that was out of frame. I repeatedly said “forward,” “backward” and occasionally “pause.” Because these verbal articulations were produced in their own time, i.e. set to no particular rhythm, I knew this would lead to confusion as the work developed. Later in the process, I incorporated another pre-recorded audio with verbal instructions, and in the final filming of the work spoke over the audios at play.