Abstract
This paper discusses the history of experiments in literature based in procedure and algorithm–works which have challenged literary concepts of authorship, originality, and intentionality. By looking to the past, I hope to situate contemporary poetry and prose created with LLMs within a genealogy of older experimental texts. Computational language-art has been around a long time, but the use of Computer-Generated Text and Machine-Writing has lead to new writing practices and processes. I hope to show that while new technologies challenge the basis of literary concepts, writers are using these technologies to reassert human control and domain in the realm of the literary by marking computer-generated text with the trace of the human. These attempts harken back to historical experimental movements, as when the Oulipo group generated algorithmic texts, while inserting a chaotic element which they termed a “clinamen,” upending the inexorable and rigid logic of the algorithm. In contemporary digital poems by Nick Montfort, the poet elevates a computer language (PERL) to the status of poetry, while the coding language itself generates a series of poems. Montfort reverses the experimental process: the code IS the poem; the execution of the code produces computer-generated writing. Students now often exhibit a similar impulse when they send ChatGPT-assisted essays through “humanizer” software that deliberately misspells words and breaks sentences into fragments. My aim is to raise questions about the knotty entanglement of humans and writing machines, by raising aesthetic and ethical concerns as to our ability to control computer-generated texts.
Presenters
Dimitri AnastasopoulosAssociate Professor, Director of Creative Writing, English, University at Buffalo, NY, United States
Details
Presentation Type
Paper Presentation in a Themed Session
Theme
New Media, Technology and the Arts
KEYWORDS
Experimental Literature, Computational Language-Art, Computer Generated Text, LLM, Machine-Writing, OULIPO
