A common misconception is that plugging your work into a language assistance software and applying its suggestions is sufficient in order for a piece to be published. Many believe that for new English writers, these tools are enough to catch all of our mistakes and get our essays ready for the world. Because few free alternatives can provide much better guidance than Grammarly, though, it’s an application that editors are often quick to both use and recommend.
But the truth is that using Grammarly isn’t simple. If I were to employ each edit that Grammarly proposed when I plugged in my work, I’d end up obscuring my meaning. I’d introduce errors where previously none were. Lines would lose their cogency.
Many of Grammarly’s suggestions are useful and applicable; it frequently still catches me in mistakes I’ve overlooked. Yet other corrections it points toward make it as clear as day that these grammar assistance programs are imperfect. It’s a hard truth for many up-and-coming writers that it demands a fairly deep understanding of the English language to even know which edits make sense to apply.
As an editor myself, oftentimes the best I can do is to direct people toward these grammar aid programs. But it’s hard to explain to people that they’re no substitute for the years and years of experience that it requires to discern whether an edit is worth making.
With the following passage from my own writing, for example:
“Being an only child has meant different things at different ages. As a child, it was never needing to share.”
Grammarly advises changing “needing” to “needed.” Why exactly it would be wrong to replace the present progressive form of the word with the simple past completed cousin, I can’t easily articulate. There’s just something amiss with the alternative that I’ve learned to identify through living with this odd language for my entire life.
The next correction Grammarly puts forward for the very same piece recommends that in the sentence:
“In one moment, I gained a sudden glimpse into the life full of meaningful moments that I could never share with a brother or sister at my side.”
that I should change “the life” to “a life.” Why exactly this substitution doesn’t work for me, I can’t quite explain either. There’s just a rhetorical value to the first that feels diminished by the edit Grammarly provides.
For the line, “If I’d always be the only one — the only existing combination of he and my mother,” Grammarly indicates that I should change “he” to “him.” While I see that change wouldn’t be incorrect to make, it sacrifices a certain punch that I feel is better achieved through the route I initially decided on. And Grammarly has no appreciation for this sort of subtlety.
But artificial intelligence does.
With AI, though, the gamble is even greater. In many regards, it’s the same Grammarly conundrum only amplified.
Tools like ChatGPT, Google Bard, and Microsoft Bing can be a colossal aid to writing. But they’re tools with an even higher bar of entry. They analyze work on a much deeper level and offer a deeper level of critique.
Anyone at all can go to the OpenAI website and prompt ChatGPT to “Craft a story in the style of Ernest Hemingway.” But it takes someone well-versed in English to see the AI drivel that emerges for what exactly it is. For anyone less experienced or who’s writing English as a second or third language, though, it’s far easier to be fooled into buying those hollow recitations about “weaving tapestries” and “diving into intricate webs.” The irregularities and forced tones are a challenge to spot for almost anyone unfamiliar with these programs.
The potential is there for AI to be a groundbreaking aid to the creative process. It can be a terrific tool for gradually learning the language, evaluating tone, and finding synonyms. It can effectively unpack the distinctions between different approaches to the same idea.
AI can converse at length about whether sentences would be best served by em dashes, commas, or enclosed parentheticals. It can talk about the scenarios in which words would be best used, and the slightly different situations in which their close synonyms might be better fits.
AI can be a much more analytical editor than the standard grammar-checking application or website. It can write ad nauseam about those places where the cracks in our use of language show and explain how best to repair them. It can show us sources to help us prove our hypotheses in the same way that any research would. But it still falls on us to craft those ideas into our own voices.
On a few occasions where I’ve directed writers toward AI as a means of refining their work, I’ve ultimately felt as though I’ve done them a disservice.
Afterward, their work will suddenly be error-free, but display all the hallmarks of robot writing. The sentences riddled throughout their pieces will be verbatim transplants from ChatGPT.
I can’t fully blame the writers using AI in inappropriate ways. These judgments are still subjective. And the allure of these automatic essay-writing machines is just too tempting for many to use sparingly. There are no established rules or standards dictating when using these software rises to something that could be called cheating. No alarm goes off when AI has been abused to a certain degree.
This strange new terrain is something we’re all navigating together. We can’t yet be expected to know the correct ways of using these technologies much more than we could expect cavemen to use computers.
If you’re still learning and hoping to one day make a living through writing, it’s discouraging to see the sort of work ChatGPT can effortlessly spit out. If I didn’t have a foundation in English, I might run from the craft entirely if I were discovering the world of generative AI for the first time today. Even now I sometimes fear whether a future in writing is possible.
AI is a powerful tool. I’m not sure whether I can advocate new writers using it at all. It’s a cybernetic sword with two sides.
On one hand, I don’t think there’s a writer alive who couldn’t occasionally learn something from a dialogue with one of these text-generation programs. But they’re not instruments that are easy to wield responsibly without certain prior foundations.
For those with a grounding in the world of professional writing, ChatGPT can be that apparatus to help push writing to the next level. It can catch tonal inconsistencies another program might miss. It can advise edits that less sophisticated software wouldn’t. It can rationalize its reasoning. It can be the next generation of Grammarly. But it can’t do our work for us and the ideas aren’t its own. It can’t democratize expression.
AI can’t grant us access to avenues of communication that we haven’t gained access to on our own. It can’t give us the ability to wonder about each word in the way that writers must. It can’t help us to feel what’s implicitly buried in each sentence — to glean the nuance between two near-identical routes for getting the same idea across.
While Grammarly and AI can be helpful ways of improving as writers, they’re tools that need to be approached with caution. And while they can help most of us, they’re not substitutes for the years and years of education and experience it requires for us to grapple with the fine details of this boisterously erratic language.
Ben, I am not a believer in AI. Having spent a large period of my life working on the very problems that the generative folk claim have been solved I consider that snake oil. Be that as it may, I would like to suggest that you include examples (specific) that show when you do x, you will get y. You have described some examples but I have found some is not enough. The more examples you have, the more it becomes a learning opportunity for the reader and I take it that is your goal here. Just a suggestion from an old guy. Randy.
My experience with Grammarly is pretty much exactly equivalent to yours. It catches a lot of my genuine mistakes but when I’m trying for a certain style - my own - it has difficulty following my meaning. In particular I love wordplay and cultural references and Grammarly is all at sea tracking my train of thought.
I ignore a lot of its suggestions.
As an editor, my rule of finger is four errors before I return a piece to an author with advice to run it through Grammarly or similar. I’m not going to spend my day correcting typos.
Here’s an experiment, though. Get a current AI model to generate some text and run *that* through Grammarly. There will be no mistakes. Indeed, it is one of the ways people pick AI writing. No mistakes and none of the “style” that generates words that might be seen as mistakes and their awkward corrections.
Run Shakespeare or Steinbeck through Grammarly and you'll see a sea of incarnadine.
Compare AI to the best of human writers and you'll find it falls short. Compare AI to the vast bulk of humanity - execrable stylists and banal with blandishments - and AI is clearly a better communicator than most. Especially those who have never learnt English.
AI is improving all the time. It’s evolving so fast that any story you publish is out of date. The twin drivers of massive investment and massive feedback are galloping it ahead.
As an editor, if someone submits AI that I can pick, I suggest they submit it to our specialist publication that handles AI. If I can’t pick it, I’m not going to do that. I’m going to publish it. It might read a little awkward but so what? A lot of human writing reads awkward. Am I really going to try to identify someone's racial or ethnic ability and suggest to them that one of their sort can’t write English proper?
Right now, AI writing - at least the stuff I can pick - is an accent. Too much and it’s obvious. Too little, and maybe the writer has crafted their prompt well, maybe they have massaged the output, and maybe it’s just a fallible human being who don't write good to start.
Or maybe it’s an AI model I can’t pick.
I don’t really care. If it’s good enough to pass as human, to communicate the intended message, and not contain any howling hallucinations, it does the job and it’s up to human readers to decide if they are entertained or informed enough to give it a thumbs up or down.
AI is in a journey and its potential to evolve is greater than ours. It’s evolving right now and I’m watching it with combined wonder and worry for what lies ahead.
Britni