Precision: How I've begun handling AI written essays

Part One

Writing teachers are stuck in a bind with current generative AI (GenAI) tools. We don’t want students being reliant on these tools, but we can’t falsely accuse them of using the tools either. Sometimes, student writing might even read like GenAI text. More to the point, GenAI essays are difficult to catch, with some of the text being respectable. Despite respectable GenAI writing, these tools are not precise. Because they rely on math to calculate their output, their sentences must almost always be an average, which is by nature not precise. GenAI rarely uses examples. When it does, the examples are typically cliché.

My solution to this bind has been to orient my classroom approaches toward precision. I ignore my personal suspicions of whether the prose is GenAI or not—likely a combination of student and AI—and instead emphasize in my rubrics and writing lessons the value of precision. I design my assignments locally, asking students not only for examples in their writing but with local problems in mind. I ask too for proper names, places, and things. My students chuckle when I say that any people or places that get capitalized means they’re on the right track.

In my rubrics, I add precision as key category. In a writing assignment about local businesses or non-profits, I ask students to go learn about a local business in Champaign, IL (our town). This requirement gets them into the community while encouraging them to learn about being exact in their claims. ChatGPT doesn’t have this kind of local knowledge, especially given that many of these organizations have minimal digital footprints. It can’t talk about the businesses on Green Street, or the local boba tea place that recently closed down. It can’t bring up the good ice cream place (Jarlings). It can’t help decide which would be a better business to start: boba tea or ice cream.

Precise writing is good practice, regardless the presence of AI or not. As a writing teacher with nearly 20 years of experience, imprecise writing is nothing to me. When I taught at a community college Massachusetts, many of my novice working-class students told me they wanted to write “fancy.” They gussied up their prose with elaborate words and phrases that only served to jargonize the meaning of their writing. Even seasoned students make broad claims when they learn to write in formal settings—I remind them they’re attempting to adopt the “cloak” of elaborate academic discourse.

Broadly, though, if I am asking students to be precise, then I need to be precise about the purpose of their writing. Long gone are days where I ask students to write about famous texts, events, or things with a critical eye. I can’t ask them to write about social media in general anymore. Those essays were perhaps always a little boring. But now they’re obsolete. Instead, we focus on concrete situations and events with an eye toward purpose.

I end up needing to read and respond to student texts differently. I don’t focus as much on sentence structure, organization, or word choice. I instead end up asking them for examples in order to be precise. Three of my go-to comments are the following:

· “Add an example here. I don’t really follow the concept you’re writing about.”

· “What do you mean here? You’re too abstract. Bring us out of the stratosphere and add concrete details.”

· “I can’t follow your meaning. Your ideas aren’t wrong so much as they lack detail. I need you to make a concrete claim.”

I have found these comments allow me to bypass the problem of AI while addressing it. The main problem with AI writing isn’t that it’s bad. It’s vague and banal. The above comments (and those like them) help me to be a good writing teacher, regardless of AI writing or not.