To ChatGPT or not to ChatGPT?

A content writer's guide to using GenAI without sounding like you use GenAI

By Joyce Wady

It’s hard to believe that two years have passed since ChatGPT burst onto the scene. I’m not sure I can remember the last time a technology both captured our imagination and terrified us of its possibilities. Now that we’ve all had time plenty of time to familarise ourselves with it, I thought I’d share my Dos and Don'ts of using AI tools for creating content.  

Do:

• Brainstorm ideas: On those days when writers block sets in and I’m struggling to find an opening line, I’ll ask ChatGPT to give me 200 words on the topic I’m about to write. I’ll be honest, it’s never good and rarely original, but more often than not there is one line in there that jumps out at me and sparks an idea of how to begin. When that happens, I’m off to the races!
• Ask how to improve your content: Recently, once I’ve finished writing something, I’ll drop it into ChatGPT and ask it how to improve the content. Of the five suggestions it churns out, there are usually two or three that spot something I didn’t notice and help me improve what I’ve written. Just beware, ChatGPT is an endless feedback loop – ask it once and you are fine, keep asking it and it will keep finding “improvements” whether you need them or not.
• Seek a little editing help: Sometimes I know what I want to say but the words just aren’t coming out right. Now, this could be the recovering perfectionist in me, but often it’s the byproduct of a distracted brain with too many plates spinning in the air. When that happens, I ask ChatGPT for help and nine times out of ten it delivers something that I’m happy with. One caveat – even though this is an original thought many of the AI checkers will still flag this as AI generated content. I’d be happy to argue that it’s not, but that won’t stop an editor from not publishing it.

Which brings me to my Don'ts:

• Blindly trust AI: Look, ChatGPT is a tool, and tools can help us be more productive and make our jobs a bit easier, but nothing it produces will ever sound fully human. You cannot pass off ChatGPT content as something original. It uses stock words and phrases like “pivotal” and “transformative” and it’s not always accurate. It also uses a stylised format for feedback and if you were to cut and paste that into an email, it will be really obvious. It needs human reasoning to sanity check what it has produced.
• Expect an editor to accept it: In the PR world, this is the penultimate don’t. Editors run content through AI checkers. Hundreds of pieces of content come across their desks every day. They are pros at this and professional writers to boot. They will not be fooled by your ChatGPT generated content. So just don’t. There is a reason most media requests these days come with a disclaimer of “no AI-generated content”.

This technology is only going to get better over time. Knowing how and when to use it as a PR pro is just as essential as a chef using his favourite pan or a doctor using her stethoscope. But just as those professions can’t create a meal or diagnose a patient with just one tool, we can’t build a content portfolio with just ChatGPT. The value you bring to the table lies in finding the right blend of technology and human creativity.