How ChatGPT Turns You Into a Terrible Writer
ChatGPT is constantly evolving, so my judgment today as of late March, 2024, may not hold in the near future. However, a current assessment can be instructive regarding the current pitfalls of ChatGPT, how it ensnares students and writers alike, degrades their writing abilities, and creates zombified hollowed-out prose.
This became clear to me when I submitted ten different essay prompts for F. Scott Fitzgerald’s short story “Winter Dreams” to ChatGPT and discovered that regardless of the prompt, I got basically the same writing response.
Here are 7 things about ChatGPT that can turn you into a terrible writer:
One. The Reductionary Effect. Complex ideas are reduced to easy-to-explain soundbites, bromides, and self-evident observations. “The story is a poignant lesson in the self-destruction that occurs when we’re not mindful of our obsessions.”
Two. The Homogeneous Effect: Because ChatGPT can’t delve into nuance or playfully address contrary ideas, it tends toward the same kind of writing, preferring to present the same moral lesson over and over for a variety of essay prompts for the same short story, and it does so in the same anodyne, banal prose style.
Three: The Attenuation Effect: ChatGPT attenuates us to basic, dull, cliche-larded writing. Because it can generate prose at such a rapid volume and it is inundating business and academia with its hopelessly boring style, it is becoming the preferred currency, evidence of a culture that attenuated itself to writing mediocrity. Regardless of the prompt, ChatGPT gave me this obvious remark about the story “Winter Dreams”: As Dexter's journey unfolds, it serves as a cautionary tale, reminding us of the importance of weighing the costs and benefits of our choices and recognizing the true value of what we stand to gain or lose.” Telling us we need to weigh our options is too obvious to be worthy of writing, but the style is confident and fluent enough to sound sophisticated. This brings us to our next problem:
Four. The Tuxedo Effect: I told one of my colleagues that one of the things I'm doing to encourage my students to produce authentic writing is to compile a list of ChatGPT's stock phrases, cliches, and overused words as it does a good job of sounding sophisticated while saying nothing. My colleague responded brilliantly. She emailed me this: “The stock phrase list is a great idea...some of the writing ChatGPT produces is the equivalent to wearing a tuxedo to class. Completely inappropriate and totally sticks out.” Exactly. Using “big” words that amount to nothing, ChatGPT will make your writing both bland and ostentatious. You’re wearing sartorial gaucherie to the classroom and your professors don’t like it.
Five. The Word-Fixation Effect: ChatGPT is fixated on certain words and phrases. Here’s a small sample: “poignant,” “poignant reminder,” “unfurling,” “delve” “unspooling,” “fostering,” “ultimately,” “unwavering,” “validate,” “relentless pursuit,” “as the story unfolds,” “cautionary tale,” “importance of staying true to our values,” “highlighting,” “underscoring,” “reminding us of the importance of,” “sobering,” “fleeting,” and so on. These words stand out like a tuxedo and make professors suspicious of the essays’ origins. Reading the accretion of these words over time has a soul-crushing quality and speaks to the deleterious effects ChatGPT has on writing.
Six. The HMO Effect. As someone who once crawled into my HMO clinic with a herniated disc and was greeted by a grouchy doctor who told me to get out of his office while throwing a bottle of Ibuprofen at me, I know the degradation and futility of looking for high-quality medical care through an HMO. Just as HMOs too often provide inadequate healthcare, ChatGPT provides an abysmal substandard writing service. In contrast, real writing with authorial presence, in-text citations, specific examples, and nuance is the PPO of literary endeavors. ChatGPT is the low-rent HMO of writing. You’ll get fed a bunch of cliches and obvious observations about the human condition, but your writing--or I should say ChatGPT’s writing--will be a lackluster affair.
Seven. The Evidence-Deficit Effect: Being the HMO of writing, ChatGPT will not quote or paraphrase a given text; it won’t provide in-text citations, it won’t give salient examples to illustrate its claims. It’s bare bones all the way.
If you want a PPO, you’ll have to work for it. Be skeptical of the cheap, easy responses you get from ChatGPT. They deaden writing, crush the soul, and compromise anyone with pride who wants to present something of high quality.
Comments