dmark/iStock/Getty Images Plus
A few months ago, I was let go from my side gig, writing content-driven advertising for a car services–related company. As a composition instructor and sometime creative writer, I admit to being stung by the email informing me “that we’re making some changes to our content priorities and production volume, and unfortunately, we won’t be able to offer you hours after this week.” After all, I had been so dutiful about my research and my writing. It didn’t help that the editor who fired me was only slightly older than most of my students.
Later, after a few email exchanges with her, I began to wonder if I’d been replaced by an artificial intelligence writing program. Grudgingly, I had to agree that an AI bot likely could do my job more efficiently. As my side gig ground on, my editor kept demanding that I churn out articles in about 20 minutes. I was being paid $25 an hour to write three short articles on topics related to car care and travel. Problem was, it was taking me more than 20 minutes to produce stories on topics such as “How can you replace an EGR valve?” and “Where’s a 2012 Mini Cooper’s battery location?” Perhaps I was too dutiful, but I always wanted to write accurate copy that read well while also staying on the right side of Google’s SEO rules. And not being a car mechanic, well, of course I could have been replaced by an AI bot. It made total sense.
I’m going to take a bit of a leap and assert that yes, AI writing programs like ChatGPT are the future of writing or at least the future of much of the writing people will be producing. That being said, what does this mean for teaching students to write well? I posit that what we will be focused on is helping students become good editors.
A good editor is someone who understands the rhetorical situation and what clearly written prose reads like. They also know how to go about checking for accuracy in terms of both content and conventions. Attention to detail is essential. However, does a good editor need to be a good writer themselves? In a 2016 article about Maxwell Perkins, legendary editor to Ernest Hemingway, F. Scott Fitzgerald and Thomas Wolfe, one of his biographers writes that “Perkins was unlikely for his profession: He was a terrible speller, his punctuation was idiosyncratic, and when it came to reading, he was by his own admission ‘slow as an ox.’” I know that Perkins’s letters to famous writers have been published, but other than those collections, I am not aware that he published anything of note.
So, while effective editors may or may not be exceptional writers, they must be great critical readers. And this is a skill that most writing instructors have been if not neglecting at least skimming over for years, mostly due to time constraints. A critical reader is one who engages with a piece of writing, recognizing both strengths and weaknesses in logic, content and style. Teaching students to be good critical readers takes time and requires instructors develop activities, such as social annotation assignments, that draw students’ attention to the details of a well-written text.
Of course, the underlying dilemma with AI-generated writing is that it is not the student’s own writing. But what do we mean by “own writing”? Before spell-check, my writing was “bordering on the illiterate,” as one of my English professors once wrote in the margins of an essay on Chaucer’s “The Knight’s Tale.” Until the advent of spell-check, I struggled, often clueless about how to look up a word in my well-thumbed Webster’s dictionary. Truthfully, my “own writing”—if I were to be put in a room today with a pad of paper, pencil and no other aids—would still likely alarm those of my colleagues who believe in spelling conventions.
Until recently, many of us in composition looked askance at such programs as the writing assistance app Grammarly and citation generators. Back in the early 2000s, I used to demonstrate to students how EasyBib often gets it wrong when it comes to MLA formatting. They were not impressed with my show-and-tell, the purpose of which was to instill in them the importance of understanding the basic concepts of MLA formatting so that they could create their own entries from scratch. Over time, I came to understand their resistance. Why not rely on help from technology? After all, the only thing more boring than formatting a works cited page is teaching other people to do it. Now, I direct students to use citation generators, and then act as editors, checking their work against a resource such as Purdue OWL.
When it comes to Grammarly (and other programs like it), I am a recent convert. After hearing rumors from other instructors that “it doesn’t work,” I never bothered to investigate. Then, a few semesters ago, one of my better students showed me how she was using Grammarly to improve her writing. Looking over her shoulder, I found myself agreeing with most of the editing suggestions, and we discussed how it could be used to help her correct some of the problems she was having with comma splices and sentence fragments.
Still, there’s an existential difference between spell-check and AI-generated writing. While computer programs can be leveraged to reduce the drudgery of proofreading for spelling, grammar and citation errors, these programs aren’t like ChatGPT, which produces coherent texts that students can hand in, with no revision, for a passing grade (at least some of the time). The text is being generated on behalf of the student and is being substituted for the student’s self-generated text. This use of AI is inherently dishonest.
Although discussing academic integrity will certainly be part of our conversations with students going forward, writing instructors would do well to showcase the limitations of the technology as a way to convince students to examine their use of it. Recent articles in the press have quoted instructors keen to disparage what ChatGPT is producing. A college history professor declared she would have given what the program spat out in response to an exam prompt “an F- if that’s possible.” Like my exercise with using EasyBib two decades ago, I can envision typing in an essay prompt in front of my first-year writing classes and having students socially annotate the paper, practicing their editing and fact-checking skills.
The debate over AI-generated writing’s role within academia will likely continue for years, but as my experience with writing ad copy demonstrates, it’s likely already a settled matter in the business world. This means educators would do well to find a way not only to live with the technology but also to incorporate it into our pedagogies. Training students to become good critical readers and effective editors is an ambitious goal, one that will require a major shift for many of us, but it also is a challenge we would do well to embrace sooner rather than later.