First Published: 28 July, 2023
Instead, I’m an advocator for first looking at the faults of it, then navigating those faults in order to get the best out of it.
I initially faced ChatGPT with a healthy amount of skepticism, uncertainty, and fear.
Now, having had months of research and hands-on learning, I’m able to share the dirty secret behind ChatGPT.
In a keynote presentation, Dr Jonnie Pen, a Professor of AI Ethics and Society at the University of Cambridge, referred to ChatGBT as ‘mansplaining as a service’. As a copywriter (and a woman), I knew exactly what he meant.
I have received outputs of straight-up lies. I’ve had ChatGPT give me made-up quotes attributed to made-up people and put them in contexts that suggest they're real.
ChatGPT has also produced outputs, explaining things that no one needs explained to them.
For example, I recently received an EDM from my bank, letting me know they were now offering a new mobile banking feature. From the start, I knew they had used ChatGPT to write it.
Here’s what I received:
"In our ever-evolving digital era, mastering your finances with finesse has become paramount. To transform this vision into a delightful reality, we've meticulously crafted an innovative mobile banking feature that places your financial world right at your fingertips. With this technology, you can effortlessly conduct a myriad of banking operations, such as checking your account balance, transferring funds, and even paying bills, all from the comfort of wherever you might be.”
Instead of telling me about a new feature, I was essentially told how great mobile banking is.
I already know that it’s great.
Here’s what I would’ve wrote:
"Say hello to our new mobile banking feature! Keeping an eye on your finances just got simpler."
And that covers it: the what (mobile banking feature), why (simpler financial tracking), who (the bank), and then the instructions on how to use this feature follow because that’s all people need to know. Punchy, to the point, clear, and word-fluff-free. No mansplaining.
When I received that initial output, I was reminded of some of the AI images I’ve seen. Much of what’s been produced is incredible, and I’m proud to be a part of a team that contributes stunning and innovative graphics, but every so often I come across an image that, upon closer inspection, has a person with nine fingers.
It’s the same with copy. You need to look closer to see its flaws in order to rectify them.
However, there’s a real danger of that not happening.
There’s a misconception that ChatGPT does the work for you. All it takes is a user telling it what to do, then out pops a bunch of words and off you go. But producing effective and engaging copy isn’t just about writing.
Just because ChatGPT can write things quickly, it doesn’t negate the need for research, strategy, or editing. If anything, those things are more important than ever. Otherwise, we’ll drown in a sea of AI-produced sameness, with no targeting, no real direction, and ultimately not really achieving anything.
I started this blog pointing out all of ChatGPT’s pitfalls, not because I’m anti-AI, but because when working with anything powerful, we need to understand the impacts of what happens when it’s misdirected.
In truth, I use ChatGPT every day to help me either improve or streamline work.
However, the mechanics in the background have remained relatively unchanged. I still do thorough audience research, topic research, strategise with other departments to ensure content is aligned, and of course, editing. What also happens is daily improvements to my prompts, all based on what I’ve learned from previous outputs.
It has certainly saved time, allowing more budget to be allocated to other areas where it otherwise wouldn’t have been able to. I’ll even be the first to admit that it has produced great copy.
But none of that is solely because I told it to do something, and it did it.
The dirty secret of AI is this: you still have to work.
The controversy around ChatGPT has escalated into class action lawsuits, and demands for government policies to react, and even its own creators have called for leaders to act.
All of this sets out a relatively uncertain future for ChatGPT, which means any kind of heavy reliance on it is dangerous. Ultimately, if users are using ChatGPT to create content without much thought and then it goes away, they’re right back where they started.
If I lost ChatGPT tomorrow, I won’t deny it would be annoying. But, like any change, I would take my skills and my learnings and adapt. I would continue producing high-quality content using my writing skills and apply what I’ve learned about AI to one of the other emerging platforms, like Google’s Bard or Meta’s Llama.
Had I wasted the last few months rolling out content explaining why mobile banking is so great, adapting to change might be a little harder.
While we look towards the future of AI-assisted copywriting, marketers must remember our role as the architects, not just the users.
It's up to us to turn AI output into gold.
.svg "MA-Logo (1)")
.png)
