My goal is to follow this topic long-term through the Spring 2023 semester. The topic of AI models that can generate text (articles, essays, letters, and even fiction) has become widespread, so there's a lot under discussion. For this first round, I'm focusing on AI text generators and the potential impact on education-- particularly writing assignments.
Five on AI-generated writing
"Eight ways to engage with AI writers in higher education," Lucinda McKnight for Times Higher Education, October 14, 2022
Among McKnight's ideas is to "[u]se different AI writers to produce different versions of text on the same topic, to compare and evaluate." This may be an intriguing exercise in leading students to note the nuances of academic or informative writing. Even when we all write about the same subject, we all do so differently.
"The End of High School English," Daniel Herman for The Atlantic, December 9, 2022
This piece captures some of the fear and uncertainty that rattles around the brain of a writing instructor who has encountered ChatGPT and feels rendered obsolete. It also reflects on why we require students to write, or read. It speaks from a personal perspective, and most instructors will identify with Herman's anxieties.
"AI bot ChatGPT writes smart essays-- should professors worry?" Chris Stokel-Walker for Nature, December 9, 2022
This brief explainer gives a few reasons why AI-generated text used to cheat may not be a long-term concern. They aren't ironclad, but they are worth mulling over.
"Enraged Worries That Generative AI ChatGPT Spurs Students to Cheat..." Lance Eliot for Forbes, December 18, 2022
Eliot, an AI expert and Stanford Fellow, ponders many different questions and scenarios that come to mind for all kinds of instructors. One of them includes thinking about ways to use ChatGPT as part of the writing process for students.
"A College Kid Built an App That Sniffs Out Text Penned by AI," Tony Ho Tran for The Daily Beast, January 4, 2023
A Princeton student appears to have developed a program, GPTZero that discerns whether text has been generated by an AI model. This piece introduces the delightful terms of "perplexity" and "burstiness," two characteristics that this student embedded as criteria for detection.
תגובות