Dear CWCAB,

What are your thoughts on ChatGPT and other AI systems that have come out recently? Are writing centers facing an existential threat? Are there ways these technologies can help us?

C. Chatterly

When I first heard of ChatGPT, a chatbot that uses artificial intelligence to respond to user questions, not long after it was released in beta testing, I panicked a bit. In addition to writing center director, I serve as my campus’s WPA (writing program administrator). I wrote a long email to my dean and chair letting them know about it, in case they hadn’t seen the headlines. I ran some sample ideas through it, and it presented some boring, but plausible, first-year-student sounding writing, and it worried me. After sitting with it a bit, though, I think I overreacted. For now, until the AI gets better (and it will get better), the solution for this is the same as the solution for many plagiarism problems: write good, unique assignments for each semester.

My concern in my position as writing center director is different, but hard to put my finger on. How will we know if a student comes in with something they had written by a bot? What if a student wants to use it in a session? What if it’s a better writer than me? (This last one might not be strictly writing center related, but I bet I’m not alone in wondering that.)

I think, for now, the best course of action is to wait-and-see. There are a lot of people talking about this. Composition Twitter is throwing all kinds of responses out there. The higher education news industry is wringing its hands. What we don’t have, though, is any intensive research. There hasn’t been time. Writers of theoretical pieces haven’t had time to think deeply yet. Qualitative and quantitative researchers haven’t had time to get IRB approval, much less conduct a study.

Cultural norms and writing technologies have never not been changing and educators have never not been worrying about it. Plato worried that writing would damage students’ memories. Nietzsche’s writing changed when he switched to a typewriter. During the early days of the internet, educators worried students would stop doing their own writing, and there was at least one op-ed asking if Google was making people stupid. I’m currently worried that my students can’t read cursive.

Concerns over plagiarism are real, but I’m excited about some of the possible benefits that ChatGPT and the like might bring us. It could be a real help to people struggling with writers’ block, giving writers just the spark they need to get a project going. It might also help teach different genres. A student unfamiliar with a field could ask it to write an essay on a topic of their choice that they could use as a model for their work. The recent news that Microsoft has connected its search engine Bing to ChatGPT is a compelling development for conducting research.

When considering these positive directions, though, we need to pay very close attention to the ways the new technology may support and extend systemic injustice. For these technologies to work well, students may need to learn how to “talk” to them. Hegemonic and dominant linguistic systems are bound to be embedded in the systems that make the bots function. As with much of our work in writing centers, we will need to find a balance between resisting these linguistic structures and helping students learn how to interact with them. For example, the bots might have the same kinds of expectations as in academic writing as many to, like Standard Edited American English.

Students are not going to stop writing their own work because of AI. Classroom teachers will have to adjust their methods, but good teachers were already adjusting their methods with regularity. At the very least, learning to use the technology will be its own kind of writing. All of this could be a tempest in a teapot, and nothing interesting come of it at all. We could, though, be in the midst of a watershed moment in the production of human knowledge, akin to the Guttenberg press.

Note: I haven’t included a list of resources this time because the information is coming so quickly and changing so much that they would be out of date within just a few weeks.