It can be a nifty helper, but it’s essential to be keenly aware of its limitations. And no, this post wasn’t written by AI.
When OpenAI launched ChatGPT to the public late last year, my initial reaction was one of hysteria. It’s a generative artificial intelligence (GenAI) that threatens to strip away jobs for writers like me, so I’m pretty sure my knee-jerk reaction was valid. I believe it still is, but I soon realized that constantly worrying about how it can displace my career wouldn’t get me anywhere. Hence, I decided to first get to know my enemy so I could figure out how to move forward and deal with it the best way I can. I logged on the OpenAI’s website and reluctantly signed on. I’d like to say, “the rest is history,” but I won’t, because this is, in no way, a panegyric to GenAI.
Once the initial shock and disbelief and despair wore off, I got to work. I took my time figuring out how to use ChatGPT: what it can and cannot do, and how I can incorporate it in my writing and editing process.
Before I begin stringing words together, I first draw up an outline. It allows me to keep my momentum going even on days when I’m uninspired to write. But there are days when I can barely lift a finger to come up with a decent working outline. This is especially true with technical writeups I need to produce for Pushpins, a GIS company I run with my husband. When I get stuck in this part of my writing process, I log on to ChatGPT and ask it for help. I give it a prompt that details what the topic is, and in less than a minute, it already has an outline. Well, sort of.
What it churns out is rough around the edges, so I’d strike out some ideas and rearrange the others until I have an outline that I could work on. ChatGPT tends to repeat subtopics. It also comes up with a predictable flow that I usually forgo to avoid sounding cliché or boring. Despite having to make these changes, I’m still able to save time. Instead of staring at my computer screen waiting for an idea, I now have a tool that can help jumpstart things for me.
ChatGPT is also a handy word bank. While I have online dictionaries bookmarked for easy access, they’re not always helpful when I’m looking for a specific word or phrase that has slipped my mind. When an idiomatic expression is right at the tip of my tongue but I can’t spit it out, I let AI beef up my vocabulary.
GenAI also assists me with captions for social media posts. But I simply do not copy-paste what it gives me. Much like how I use it for drafting a story outline, I make sure to tweak the copy it produces according to how I would write it. Otherwise, I’ll only end up with cookie-cutter content.
ChatGPT, Bing Chat, and Google Bard also lend me a hand in research. It’s essential, however, to always keep in mind that the results these tools come up with should be taken with a grain of salt. The free version of ChatGPT, which I use, has been trained on information only until 2021; asking it about anything that happened beyond that year is practically futile. If I want an up-to-date context, I head over to Bing Chat or Bard. Then again, updated content doesn’t always mean factual, so I still need to make sure the information I get is correct.
So why use AI in the first place and not just stick to Google Search? Because GenAI is better at collating information and fleshing out ideas that would otherwise be cumbersome to do with a simple Google query. For instance, if I want to know about the applications of drones in agriculture, I’ll have to wade through pages after pages of Google search results just to come up with a list. It’s a process that will take me hours to accomplish. With GenAI, I need only ask it for examples, and it’ll come up with a list that has a brief explanation for each item. From there, I narrow the list down into the exact details I want, verify those details for accuracy, and do more research when necessary.
I can’t stress it enough, though: using GenAI comes with extreme caution. Plagiarism and factual errors are two of the things I always need to watch out for. I definitely don’t want to end up like CNET, whose “AI journalist” ended up copying other works and committing grave factual errors. This is why I don’t simply copy what ChatGPT gives me, unless I want to settle for mediocre, bland, lifeless, and possibly stolen content. Neither do I feed it with information I don’t want it to train on, like articles I need to edit (or this blog post, for that matter). My clients pay me to write and/or edit; they hire me for my skills. I’m the one who’s supposed to do the heavy lifting. Having someone—or something—else do the work for me is an insult not only to my clients who trust my abilities; it’s also an affront to myself.
Do I think GenAI can replace writers and editors like me? Yes and no.
If I don’t make an effort to stay at the top of my game, if I’m okay turning in work that is merely on par with what GenAI can do, then my career—my passion—will definitely be on the line. ChatGPT can produce a 2,000-word article on just about any topic within minutes, an endeavor that takes at least two days for human writers. But does it produce mesmerizing pieces that grab and hold the attention of readers? No.
And yet, it’s ignorant to dismiss that the threat is not real. I’d like to believe the writers who have lost their jobs to ChatGPT aren’t lousy at what they do (or did). I also believe the Hollywood writers’ demand for AI regulations hold water.
Equally important to remember: Using GenAI is, by no means, a reason to undermine the work we do, nor is it a reason to haggle for a lower writing or editing fee. (Rewriting AI-generated content is never an easier task compared with writing something from scratch.)
So, what can wordsmiths and editors do? I honestly have no answers. Right now, all I can do is be on my toes and turn in my best work for every single assignment. A mentor once told me that I’m only as good as my last story, and this is something I’ve kept in mind since even before GenAI took the world by storm. And as naive as this may sound, I know there’ll always be clients who’ll put a premium on quality over quantity. After all, the proof is in the pudding, and GenAI’s pudding isn’t that great. Yet.