2023–
Teaching writing when the machines started writing
When ChatGPT arrived in November 2022, I was in my fourth year teaching writing for the University of Maryland Global Campus: research writing, business writing, technical writing and first-year composition, online and face to face at Pearl City and on Oʻahu military installations. Few disciplines felt the shock sooner than writing instruction. The first institutional reflex everywhere was the one Lodge, de Barba and Broadbent describe: a fixation on cheating and on “securing” assessment. I shared the concern, but I had also spent twenty years arguing that new technology is unfairly blamed for problems that are really about design. I did not want to spend my last teaching years as a detection officer.
Learning it properly first
My rule since the streaming-media days has been “if I can do it, you can do it too,” which means I have to be able to do it. In 2024 I completed two Vanderbilt University specializations through Coursera: Prompt Engineering, and Prompt Engineering for Educators. I learned positive prompting strategies related to ethical use of AI in writing workflow, content mastery for students, and efficiency in assessment.
As with my past evaluation of technologies, I engage with the full range of frontier applications via my Perplexity account, and process privacy sensitive content behind the University’s network logged into M365 and CoPilot. I monitor functionality of tools like Research Assistant and Elicit, and I also use these tools daily in my own writing and research practice. The difference between a teacher who has read about AI and one who has argued with it at midnight over a comma is obvious to students within a week.
Teaching AI literacy, not AI avoidance
Two findings from recent research shape how I frame AI for writing students.
The first comes from Lee and colleagues’ 2025 CHI study of 319 knowledge workers and 936 real examples of AI-assisted tasks. The headline is uncomfortable for educators: the more people trusted the AI, the less critical thinking they did; the more confident they were in their own ability on the task, the more critical thinking they did. The nature of the thinking also shifted, away from producing content and toward verifying information, integrating responses and stewarding the task. That is, almost exactly, the skill set of a research writer. So I teach AI the way I teach sources: the tool is a confident, fluent, occasionally wrong informant, and your job is to interrogate it.
The second comes from Lodge (a former colleague at Griffith), de Barba and Broadbent (2023), who argue that AI literacy and critical thinking, while necessary, are not sufficient. What students need is self-regulated learning within what they call a “network of co-regulation,” with the learner’s own agency at the center of the human–machine relationship. That framing matches what I found in my own doctoral research on online learning self-efficacy two decades earlier: the students who thrive with a new technology are the ones who believe they can direct it. The goal is not students who can use AI, but students who stay in charge of their own learning while they do.
In practice, that looks like:
- Transparency as the default. I must enforce University policies, but I am a teacher, not a policeman. I admit my own usage and put forward strategies for using AI as a study buddy to master content, and regarding it as a willful junior staff writer to manage as their own human workload shifts from text generation to editorial.
- Verification as an assignment. I hold them responsible for any submissions with their name on it, and remind students that AI is trained on all writing, good and bad, so that AI text is just as likely to be returned for revision as non AI text.
- Process over product. Course designs include statements of if/how and what AI tools are used in writing workflow to be submitted with each assignment. I model prompts and use cases in my responses to their discussion post work, and the final assignment is a brief reflection essay.
Sources
- Lee, H.-P., Sarkar, A., Tankelevitch, L., Drosos, I., Rintel, S., Banks, R., & Wilson, N. (2025). The impact of generative AI on critical thinking: Self-reported reductions in cognitive effort and confidence effects from a survey of knowledge workers. CHI ’25. https://doi.org/10.1145/3706598.3713778
- Lodge, J. M., de Barba, P., & Broadbent, J. (2023). Learning with generative artificial intelligence within a network of co-regulation. Journal of University Teaching & Learning Practice, 20(7). https://doi.org/10.53761/1.20.7.02
- Fletcher, K. M. (2005). Self-efficacy as an evaluation measure for programs in support of online learning literacies for undergraduates. The Internet and Higher Education, 8, 307–322. https://doi.org/10.1016/j.iheduc.2005.09.004
Previous chapter: New Adventures (2018– )
