Can Human Writers Compete with Humans Using A.I. To Churn Out Material?

By John A. Tures, Professor of Political Science, LaGrange College

One of the biggest challenges A.I. faces is dominating human-based writing. Artificial Intelligence creators have sunk plenty into this conquest. How well are their tools performing? Do humans still have a chance? And what should readers, and writers, be doing?

The Test

Kevin Roose and Stuart A. Thompson with The New York Times made news on March 9, 2026, with their article “Who’s a Better Writer: A.I. or Humans? Take Our Quiz.” You can take their test as well.

Roose and Thompson write “Skeptics have argued that A.I. can never be truly creative, because it lacks the kind of worldly experiences humans have. But several recent studies have suggested that, in blind tests, many readers prefer A.I.-generated writing to human-authored works.”

I took their test, looking at five paired short paragraphs of literary fiction. Half were written by humans like Cormac McCarthy, Ursula LeGuin, and Carl Sagan. The other half were generated by A.I. I got 60 percent right (I’m not good at picking poetry and I whiffed on one of the four prose pieces). The test informed me that humans usually pick the A.I. piece over the human writing.

But is it an accurate test?

“We asked A.I. to choose an existing piece of strong writing and then craft its own version using its own voice,” the authors wrote. Therefore, each of the passages is made of dazzling language that simulates the real thing, much like a mockingbird mimicking the call of another bird. Bravo.

Could A.I. keep it up over an entire novel? A runner who can run a 100-yard dash may not sustain that over a marathon, much less a 5k.

“AI leaders boast about their models’ superhuman technical abilities,” writes Jasmine Sun with The Atlantic. “The technology can predict protein structures, create realistic videos, and build apps with a single prompt. But these executives and researchers also readily admit that they have not yet released a model that writes well. OpenAI CEO Sam Altman has predicted that large language models will soon be capable of ‘fixing the climate, establishing a space colony, and the discovery of all of physics,’ but in an October interview with the economist Tyler Cowen, he guessed that even future models—an eventual GPT-6 or GPT-7—might be able to extrude only something equivalent to ‘a real poet’s okay poem.’”

Improvement Over Time In Writing

As Alyssa Matesic, an independent book development editor and publishing consultant, points out, there’s always at least one person who changes from writing your book: you. I discovered this when I wrote my first novel, Branded, during the pandemic. While A.I. proponents of the world tell you that they’ll stop the dreaded “blank page,” and end the lengthy editing process, I found those to be useful. When you’re stuck, it forces you to think about where your book is going. The editing process also gives you a chance for improving your work, and learning too.

For example, I spent much of the summer of 2020 plotting the novel, the characters, the story, the settings and other elements. My biggest struggle was how to begin. I initially began the first chapter with the villains and their evil product placement scheme, as well as a murder, and didn’t introduce the protagonist. I then chose to introduce a terrifying episode set 2/3 of the way through the novel, then used other chapters to show how we got to the tense moment, which confused readers. Finally, in interactions with other humans in my writing group, and with my publisher and his team at Huntsville Independent Press, I opted for an opening which introduced the main character, threw him into a disastrous pre-tenure hearing, raising the stakes and setting up the inciting incident: the need for a key publication or award to keep his job.

The point is that it was an iterative process, with the chance for bettering the book at a number of junctures. And that’s just one of several problems which required time, and thought, to improve. Books written in 45 minutes lack any of that.

Human Authors vs. Humans Using A.I.

The is a battle, but not between humans and machines. Rather, it’s more about humans versus humans with performance-enhancing tools. A.I. supporters are dazzled by “Coral Hart,” the pseudonym for a romance writer who uses artificial intelligence programs to churn out 200 books a year, claiming a six-figure income from tens of thousands of sales. “She” also markets a plan to do the same. It’s a bit odd that someone with a successful plan would need to make sales to follow “her” formula. Mike Pearl calls it A.I. Slop in the title of his article about “her.”

“Hart, the story says, ‘requested anonymity’ for what only sound like reasons related to professional expediency. She apparently works as some kind of coach, and has some unnamed role in publishing—work she performs under her real name. But she ‘fears that revealing her A.I. use would damage her business for that work.’”

That’s because a lot of readers and writers really dislike those who use artificial intelligence for shortcuts in churning out material for publication. On the Facebook page “Authors Supporting Authors,” those employing A.I. decry the “discrimination” they receive, and typically demand recognition, even respect, from those who do their own work. They claim their A.I. tool is little more than using a calculator to do math, or spell-check for editing. They insist they don’t have time to write, or learn how to write, and want the same opportunities as others seem to have who actually pen their own works.

This battle played out over a national writing conference, according to Jack Cox with The Writing Collective. Last year the “National Novel Writing Month (NaNoWriMo) shut down last week after 25 years of supporting the writing community.” Participants would spend November coming up with a 50,000 word novel in that month. “Last year, however, the organization behind NaNoWriMo burned all of that goodwill to the ground by embracing AI-generated slop novels,” Cox adds. “‘Writers’ could submit 50,000 words they ‘wrote’ by prompting generative AI and still ‘win’ the challenge. It wasn’t a good November for the organization, and it’s no surprise they decided to call it quits.”

It’s not just NaNoWriMo and novels that are part of the battlefield. Nearly every short story I submit to has a policy about not using artificial intelligence to write stories. And yet these users flood the market with such material, which means at least some possibly lying when asked. An editor told me a horror story of having to sift through thousands of slop stories, including 12 versions of the same story, with only slight variations, all involving A.I. There used to be hundreds of submissions for a few spots. Now there are thousands of these stories competing for those same limited spaces. Despite more opportunities for publication, there seem to be less spots available for humans using their own hard work and skills to churn out such stories.

If A.I. has an advantage, it may be overwhelming humans and their craft by sheer numbers, along with the decision of a number of artificial intelligence users with inauthentic responses to questions about tools used. In war, it’s called a “human wave” attack. In this arena, it’s an A.I. wave, which may simply crowd humans and their own craft. If readers, writers, editors and publishers still value human work, they have to continue to reject such material that is not human written.

John A. Tures is a professor of political science at LaGrange Collegein LaGrange, Georgia. His views are his own. He can be reached at jtures@lagrange.eduor on “X” at @johntures2. His first book “Branded” a thriller novel about corporate malfeasance, media manipulation, and academic intrigue, has been published by the Huntsville Independent Press (https://www.huntsvilleindependent.com/product-page/branded).