AI Text Redactions (Mom, Patriot, Spiders, Valentine, Zilla), 2023

digital prints on aluminum
variable dimensions

I vividly remember learning how to write. I would painstakingly attempt to create letterforms as beautiful and precise as my teacher’s handwriting on the chalkboard. Later in life, I learned that my handwriting is not just about creating recognizable letterforms, but also about identity as our handwriting embodies a trace of who we are, carrying norms around sex, gender, and sexuality. 

Handwriting is now mediated by the new actor of artificial intelligence. Technologies such as Optical Character Recognition are driven by machine learning processes, setting new norms of machinic legibility based on the source material of the training models.

This series of digital prints explores the moments when I learned how to write. To create the works, I processed personal handwritten samples—mostly notes and cards written to my mother—through AI-based text-removal software. The resulting images are the marks that the software doesn’t recognize as valid language. Each print is named after a word that was removed by the text removal algorithm. 

These images evoke a sense of loss as we recognize these marks as belonging to language, but stripped of their semiotic significance by a technology that can only operate on sameness. But there is still is beauty in this loss of meaning, like looking at a language that we don’t understand. We don’t have to understand something for it to be meaningful.

Formalist and abstract, the remaining marks are the residue of a system derived from the corpus of language that was used to train the AI model. I’m interested in this excess of a system, the linguistic marks that artificial don’t recognize as legitimate, as source material for these formalist works that comment on the role of technology with digital and analog practices of reading and writing.

Artists often comment on how their work involves learning how “not to draw.” Perhaps I’m learning how not to write.