Maybe it’s time for content writers to start leveraging AI and stop fearing it.

When technology first started replacing tollbooth operators and grocery store cashiers, who would have ever imagined that we, as writers, could be next. Yet, here we are, staring down the same fate, with an AI approximation of ourselves so good that 52% of business leaders are already using AI tools for content generation.

Amazon is littered with AI-generated books on complex topics; the Associated Press is now producing around 40,000 automatically-generated stories a year; and Hollywood writers went on strike, in part, due to the fear of being replaced by AI. But is it all as scary as it sounds, or can writers harness the power of AI – without losing our livelihoods in the process?

One writer’s experiment with ChatGPT

I have to admit, I was probably one of the last holdouts when it came to experimenting with ChatGPT. After all, how could a bot replicate my human emotion, intuition, and lived experience? Not to mention that the idea of using chatbot-generated text to help me in my work felt an awful lot like cheating. 

But, a couple of months ago, I was tasked with writing a particularly technical blog post, and my curiosity got the better of me. If I was going to Google the topic anyway, why not just query ChatGPT? I knew I wouldn’t use the auto-generated text verbatim. Want to threaten my livelihood, chatbot? OK, give me the best you’ve got!

And it did; ChatGPT immediately knocked me down a peg or two. The text was decent, good even. I couldn’t find a single grammatical error or even a misplaced comma, which I had been gleefully looking forward to up until that point. Had I been a tad less ethical and whole lot less fearful of AI detectors, I could have turned that post in to my editor, and I probably would have earned praise. But it soon became clear where the technology thrives and where it falters. 

The limitations of ChatGPT in content creation

Based on my quick experimentation with ChatGPT and a little research, I realized that my job is safe – at least for now. Here’s why.

SEO guidelines favor “helpful” content

Google is no longer penalizing AI-generated content, but their latest guidelines favor content they consider to be “helpful” and “people-first.” What does that mean exactly, and why does it matter? Well, for one, they name originality as a criteria for helpfulness. While ChatGPT is great at regurgitating information, it’s humans who convey original ideas and insights. It seems our humanity still has an edge in the eyes of Google, and brands looking to boost their SEO will hopefully take heed.

ChatGPT isn’t very clever

My query provided me with plenty of information in a well-structured article, but the text was exceedingly dry. There were no playful alliterations or clever turns of phrase, and while some brands will gladly trade a clever blog post for a free one, I can’t imagine they’d feel the same about short copy, like taglines. When I asked ChatGPT to suggest names for my own content writing business, for example, the suggestions were abysmal. Marissa’s PromoProse Hub? Yaar’s Strategic Scribbles? Thanks, but no thanks.

AI’s accuracy is questionable

As humans and bots face off, you can find people on every corner of the internet relishing in the fact that ChatGPT got something wrong, whether it’s coding questions or the color of the Royal Marines’ uniforms. And the evidence isn’t just anecdotal: A recent study from Stanford and UC Berkeley found that GPT-4’s accuracy has gotten worse over time. AI is only as good as its training data, and none is foolproof, so trusting AI at its word is a dangerous game.

How writers can benefit from AI

Despite its flaws, AI still has its place in marketing and communications. And, as a writer, I’ve found that place to be in structuring content. Creating outlines is the bane of my existence, and ChatGPT can create one effortlessly with the right inputs, leaving me with a solid base to work from. Conversely, it can take my own words and break them down into bullets to make long text more readable. 

AI also comes in handy for research purposes. Ask Google, and I’m left with pages of results to parse. Ask ChatGPT, and everything is condensed into a nice, neat article that I can use as a springboard for my own work. Again, though, I would have to triple-check any facts and figures before using them.

So, have I taken full advantage of AI’s capabilities? Well, no, not really, apart from that bit of experimenting. But I now feel secure in my knowledge that my work won’t dry up overnight. Instead of viewing ChatGPT as a threat, I’m looking at it as a tool to keep in my back pocket and pull out when I’m feeling especially stuck. Used carefully and in small doses, AI is set to make writing more efficient, freeing us up to do our best creative work. Take that, bots!