Yes, Your Editor Can Tell You’ve Used ChatGPT.

If you’re an aspiring author or even an experienced one, you may be considering using artificial intelligence to help write your next book. ChatGPT, an AI chatbot, aims to provide its users with human-like text. However, relying on AI entirely to write your material has its downsides, including misinformation and repetition, to name a few. More often than not, your editor can easily tell if the material is AI-generated. Throughout this blog, we’ll be discussing the pros and cons of using AI on your writing journey.

AI can be an incredible tool for writers.

Unlike some of the publishing industry, we don’t think AI is the devil. In fact, we use it ourselves to research, brainstorm, or clarify ideas. It’s also a sanity-saver for when you can’t find a particular word (pro-tip: Write your sentence, use an underscore as a placeholder, and ask ChatGPT to fill in the blank with the correct term.)

That said, there’s a huge difference between using AI as a supplement to help with your writing and relying on it to churn out low-quality content. The former can boost quality and efficiency for writers. However, AI can also result in poor writing, which isn’t engaging for your audience.

The main issues we see with AI writing

As editors, when authors send us drafts of their manuscript, we can easily spot AI writing, whether it’s the full text or just parts of it. Many writers, especially new or aspiring ones, turn to AI-generated content, and it’s painfully obvious to us. In some cases, the content is repetitive, boring, and poorly written, not to mention factually incorrect.

Misinformation

Large language models like ChatGPT are primarily trained on publicly available text from cyberspace, which isn’t always accurate. Chatbots can also produce “hallucinations” or information that’s nonsensical and inaccurate. They may pull quotes, statistics, or ideas out of thin air or from incorrect sources, and unless you fact-check your work, you wont know if your information is correct.. We therefore recommend fact-checking the work that AI has produced and referring to the original sources. 

Repeated phrases

 Misinformation is the first red flag, but AI-generated content also contains the same phrases over and over again, which can make the reading experience less than ideal. It also makes it easy for us to spot when someone has used AI to generate their writing from scratch. Some examples of phrases that may make it obvious you have used ChatGPT are:

  • “Treasure trove”
  • “Intricate tapestry”
  • “While navigating the complexities of…”
  • “A testament to”
  • “In the world of”
  • “Let’s delve into”
  • “Look no further than”
  • “Whether you’re… or…”
  • “A plethora of”

Often, these phrases are not naturally how humans would write, particularly as AI tends to repeat words or phrases in close succession. For instance, we assessed a page of AI-generated text and found that it used the same “It’s not x; it’s y” structure four times in close proximity, which is not a ‘human’ way of writing.

But why would ChatGPT produce poor writing if language is supposed to be its speciality?


ChatGPT doesn’t “understand” language as humans do. It’s not sentient and can’t process emotions or meaning. Instead, it works using patterns, predicting what words, sentences, or text may be most likely to come next. This massively impacts how your writing reads, as it is unlikely to be original, as AI relies on online sources.


How humans write in comparison to AI 


Naturally, everyone has a unique writing style, which often includes the phrases listed above. The difference is that when a human uses these phrases, they are contextually correct and sporadic. Great writers will spend their time honing their own voice, by shaping a style distinct from other writers. Comparatively, AI will use its own voice, which is neither unique nor natural. While you may use ChatGPT for idea generation or research (ensuring you fact-check it), you should still attempt to rewrite the material using your unique style as this is what readers really want. 


As a hybrid publisher, we will consider accepting books written with AI, but our in-house human line editors will edit them to ensure they sound real and authentic to your voice. Alternatively, our ghostwriter can write your book for you, and we promise it’ll be much better than ChatGPT! Get in touch if you’d like to discuss how we can help you.