TO SPEAK IS TO LOVE: TECHNOLOGY AS A BODY

PICKLE AI EXPERIMENT

Technology as a body. 


In my written essay “To Speak Is To Love,” I explore my relationship to the Korean language as well as themes of bodily failure, digital communication, and memory. Using Pickle AI, a software that allows you to create an AI body double of yourself to use for virtual meetings such as Zoom, I wanted to create a short film/video essay that explores technology and artificial intelligence as a body in the way that it consumes and outputs information. I thought about the ways that Pickle AI could be used to alleviate the burden of having difficult conversations with others— a contemporary pressure I explore in my written essay. With this in mind, I used my deeply personal essay as a script for my body double to read and used auto-generated captions to juxtapose technology’s failures with the linguistic failures I grapple with in the essay. The AI’s lack of emotion, unwavering composure, off-kilter lip synching, and inaccurate captions in some ways emulate a very human struggle to communicate intent— another major theme that surfaces in the essay. In order to get the AI “Lydia” closer to the original essay, I manually corrected each auto-generated caption for hours with strikethroughs and blue text to mark my changes to the punctuation and words. Even though my voice is speaking, technology could not capture or understand the stylistic difference between a semicolon and a period through merely the tone of my voice. Is this my own failure or technology’s? It was particularly interesting to me in that even with AI and time-saving technological processes, I was laboriously altering the colors of each character so that I could convey my true intent. With many references to author Theresa Cha’s written work, this process was also a way to engage in the visual form that her work often takes—known to be difficult to decipher. Through images and uncommon formatting in her published work, Cha freed herself from many of the constraints of grammar and language. In nonsense, perhaps the burden of being understood is lifted. However, through this effort to make myself understood, I found that it was difficult to preserve the words and form of the cited authors in my essay. With each visual and sonic output of my film, whether it was the glitchy audio of my voice or the inaccurate captions, my original writing and theirs were muddled in this game of telephone with my AI self that was made to be free from some of the physical and linguistic pressures that make communication so difficult.