A few months ago I wrote about the growing trend of using AI-generated art in books and what that future looks like for illustrators and artists. Now I want to take a look at the ways in which AI will shape the future of the writing industry as a whole.

There is some concern that soon we’ll see novels “written” by AI taking the place of those produced by living, breathing humans—but I think that concern is overblown at this point in time. For starters, the programs we have available to us now (like ChatGPT and Rytr) simply regurgitate content and writing styles found on the internet.

While ChatGPT in particular has recently undergone a massive upgrade—going from being able to source only information from 2021 and earlier to now being able to access present-day material due to its partnership with Google Search and Microsoft’s Bing—the fact remains that AI in its current form is still recycling content that has been created by people.

There is, at present, no new information or insight that these programs can produce. This means there’s no way for AI programs to give opinions about anything either, since anything it presents is really just someone else’s opinion that it found on the internet. This in and of itself makes the application somewhat limited.

Style is another area where AI-produced writing doesn’t quite cut it. An author’s unique voice is sometimes just as important as the plot and characters, and AI writing (understandably) lacks that innate personality that makes certain pieces so compelling. In fact, once you read a few AI-generated articles or blogs, for example, it becomes rather obvious that you’re reading something produced by a machine.

It tends to fall back on slight variations of the old-school model that we were all taught in elementary school: one introduction paragraph, three body paragraphs, and one conclusion paragraph, each at least three to five sentences long. The sentences tend to be simple, stiff, and short in length—in other words, not very interesting to read.

While using machines to assist with grammar and punctuation is an even older concept than AI (spell-check, anyone?), the improved editorial services provided by this newer technology still can’t compare to the complexities and nuances that skilled (human) editors and proofreaders deal with on a daily basis.

More than just checking grammar and punctuation, these roles involve reading and understanding a body of work as a whole, then making suggested changes from there. This involves an intimate knowledge of the plot, characters, and tone of the work that can only come from the unique sense of human empathy.

All of this isn’t to say that AI-generated writing can’t be extremely useful to writers in their line of work. While grammar and spell-checks are by no means foolproof, they can certainly give our tired eyes a little help after staring at the same manuscript for hours. Beyond that, though, AI can help in a big way with conducting research needed for your novel or paper. Historical fiction or nonfiction writers may find that using an AI program as an assistant can help cut down on hours of research needed to produce a historically accurate text.

Those in the world of academia, on the other hand, can consider AI a more in-depth Google, helping find more relevant results for your specific area of study. This is because it doesn’t just gather a handful of links for you to look over but also helps summarize the findings and present them in an easy, digestible way.

AI is also particularly handy at flagging plagiarism (largely because it is itself based on material that is already publicly available). This is useful not only for those who grade essays but also those who want to make sure that their writing doesn’t too closely resemble something that’s already been published.

One of the biggest cautions that writers should be aware of when employing AI, however, is its tendency to show unwitting bias through the massive amounts of data that it currently uses to spit out relevant information in response to the questions or prompts that humans enter. How does this happen? Harvard Business Review reports that “bias can creep into algorithms in several ways. AI systems learn to make decisions based on training data, which can include biased human decisions or reflect historical or social inequities.”

While this may not seem like a particularly relevant or troubling problem for an author using AI to check grammar issues in their manuscript, the problem really shows itself for anyone writing in the public sphere. This can include those working in law, politics, social services, or any other person-facing career. While all information learned from the internet should be taken with a grain of salt until you’re sure it comes from a trusted, reliable source, AI’s results should be treated with a little extra dose of skepticism until these kinks have been worked out.

Overall, it looks as though AI is at the beginning stages of revolutionizing certain aspects of the writing industry. But for now, authors should rest assured that it’s their humanness that infuses meaning into their work for both themselves and their readers.

Andrea Moran lives outside of Nashville with her husband and two kids. She’s a professional copywriter and editor who loves all things books. Find her on LinkedIn.