As the current Core director, I oversee the first-year writing curriculum in ICC (Identity, Culture and Community), Academic Voice and Research and Writing. So of course when I heard about ChatGPT, one of the first things I did was to see how well it would write some of our first-year assignments. “Write an analysis of ‘Make Your Home Among Strangers,’” I commanded. 

There is a type of sustained thinking and deep learning that is inextricably linked with composition.

In less than a minute, it produced five paragraphs about the book that were focused, fluent and mostly on point, though somewhat formulaic and with a few inaccurate references. Still, it was a decent start. 

For those who haven’t been following this development, here is a quick summary of the tool: 

ChatGPT is a type of artificial intelligence (AI) that is trained to generate text. Essentially, it is a computer program that can understand and write text in a way that is similar to how humans do. The model is trained on a large dataset of text, such as books, articles and websites, which allows it to learn the patterns and structures of human language. People are using ChatGPT for purposes as varied as generating creative writing, automating repetitive tasks, creating chatbots and writing social media content.

So, what do we do with this? Do we call composition studies obsolete and accept that this tool will now write for us? No, I don’t think so. Do we ban it on campuses, call it academic dishonesty in all cases and steer students away from it? That’s not likely to be successful and it’s also shortsighted. I think we need to assess this as a new tool and think critically about when and how we use it. 

Let’s back up and think about why professors assign writing. It’s often for one of two reasons. In first-year writing, assignments further develop foundational skills such as making and sustaining an argument, interacting with source material and developing your voice and lens as a scholar. 

The second common purpose for college composition is to demonstrate critical thinking about or reflection on the subject of study. When you’re assigned an essay in an upper-level course, the hope is that you’ll generate more complex insights and more nuanced, cohesive arguments than when you simply speak conversationally about a topic. There is a type of sustained thinking and deep learning that is inextricably linked with composition. 

An AI essay that is simply generated and submitted by a student won’t address either of these purposes. However, if it’s part of a pre-writing process, or if it helps a student bypass writer’s block and get started, it could be a great tool. By bringing the tool into a class like Academic Voice, we can demonstrate traits like unity and fluency and try to emulate them. As students work on upper-level assignments, perhaps they use AI-generated sentences to find words for their more complex ideas or develop supplementary paragraphs, but the overall shape of the argument and the voice of the writer is the student’s. 

We already have programs claiming to catch the AI essay and they work reasonably well. I plugged my ICC essay into GPTZero along with an essay that was actually written by a first-year student. GPTZero quickly told me that the essay I had generated was likely composed by artificial intelligence and the latter essay by a human. My gut would have told me which was which; the tool simply gives me the data to confirm it.

Still, I hope that our collective response to ChatGPT is wiser than a cat-and-mouse game where students use the tool to completely bypass the hard work of writing and professors spend time trying to catch those not doing the work. We all have more important things to do with our minds and our time … like engaging new ideas, crafting meaningful learning experiences and putting words on a page with the authenticity and particularity that can only come from a human writer. 

 

*One paragraph in this piece was written by ChatGPT. I lightly edited it for flow. Can you tell which paragraph it is?