Abstract

This paper argues that the threat Large Language Models (LLMs), such as ChatGPT, pose to writing instruction is both not entirely new and a welcome disruption to the way writing instruction is typically delivered. This new technology seems to be prompting many instructors to question whether essay responses to paper prompts reflect students’ own thinking and learning. This uneasiness is long overdue, and the hope is it leads instructors to explore evidence-based best practices familiar from the scholarship of teaching and learning. We’ve known for some time how to better teach our students to think and write. Perhaps the arrival of LLMs will get us to put these lessons into widespread practice.