Timothy Laquintano, “Before Writing: LLMs as Reading Machines”

Several surveys show college students believe that using machine-generated text in their writing is cheating. However, they have a more liberal attitude toward using large language models (LLMs) for activities that are precursors to writing: brainstorming, outlining, and summarizing literature and readings. Although much attention has been paid to LLMs as writing machines, we have seen less focus on how LLMs operate as reading machines. But students have quietly been using LLMs for a variety of use cases related to reading--and these activities tend to be undetectable by instructors. This talk will address the state of undergraduate reading by focusing on LLMs as mediators of long-form text. It will look at the dynamic reading practices that LLMs make possible, and the errors they tend to make when summarizing and re-writing academic literature. After considering how LLMs might affect undergraduate reading and learning, the talk will conclude with a brief profile of some classroom activities designed to help students acquire a more critical understanding of LLMs.  

For disability access information or accommodation requests, call 413-585-2669. To request a sign language interpreter, email arc@smith.edu at least ten days before the event.

Timothy Laquintano holds a PhD in English with a specialty in writing and rhetoric from the University of Wisconsin-Madison. He is associate professor of English and director of the College Writing Program at Lafayette College. Laquintano uses qualitative research to study how ordinary writers adapt to new communications technologies. His book Mass Authorship and the Rise of Self-Publishing won the 2016 Computers and Writing Distinguished Book Award. He is co-editor of TextGenEd: Teaching with Text Generation Technologies, an open-access collection of resources to help faculty teach writing with large language models. He is currently working on a qualitative study of how workplace writers adopt (and sometimes abandon) AI tools.

This lecture is hosted as part of the Kahn Liberal Arts Institute short-term project, “Generative AI & Writing,“ organized by Julio Alves, Jacobson Center Director and English Language & Literature and MJ Wraga, Psychology.