A Human Writer vs ChatGPT: A Live Debate
- Danny Judge
- Feb 28
- 9 min read
Updated: Mar 26
A guy can only write so many articles about how fast A.I. is improving, and why organizations should take precautions. So I decided to show you instead.
I staged a live debate—a content writer vs ChatGPT, duking it out over who is the better writer.
The Preparation: I fed ChatGPT six copies of our newsletter, which amounts to a 2,000-word sample of my writing style and tone. I then explained that I wanted to stage a mock argument over who was the superior writer and publish it in the newsletter.
ChatGPT had very little to work with, but as you're about to see, it was enough for it to get a great sense of the kind of debate I was looking for.
Before we dive into it, I do want to mention a few things upfront. I write a lot of content for Tier 3 Technology, and as recently as a year ago, I didn't think very highly of ChatGPT as a great source of content. It was clunky and required heavy editing.
Things have changed.
The technology has improved exponentially. It can actually do something with its ability to analyze the mechanics of individual writing styles.
What does it mean to analyze a person's writing style? As the robot correctly points out in the debate, your writing style is like a fingerprint. Your personal style is full of details and idiosyncrasies. Some are unique to you. Some are shared by subsets of the population. Some details are conscious—learned, then developed over time. Some are unconscious—automatically absorbed from the culture and your reading material.
Everyone has a unique writing style. And now ChatGPT can mimic that unique style based on a very small sample size. Whether your style naturally worked itself out through practical application or you intentionally spent thousands of hours working on your craft, ChatGPT can understand what makes your writing yours and do an almost pitch-perfect imitation of those qualities.
Beyond the obvious implications for misuse (say, sending malicious emails to my coworkers that sound exactly like me—even adopting the different tones I use for different people), I took other things away from this exercise.
When I'm not marketing for Tier 3 Technology (or standing on the shore like an idiot and calling that "fishing"), I manage a literary journal that publishes fiction online. I'm not going to plug the name or website. I bring it up because we do not publish AI-generated prose. That's our policy, and when I implemented it a year ago, I was confident it wouldn't be an issue. Even if someone ignored our guidelines and submitted the stuff, I was 100% sure I could spot AI-generated prose a mile away. After this mock debate (which was supposed to be a fun exercise for a newsletter), I'm not so sure. It's getting a little spooky out there.
Okay, I'll wrap up the long-winded preamble. What a pathetically human trait, this impulse to provide context. The robot would scoff at me. Onto the debate.
The debate is transcribed exactly as it happened. No part of it is embellished or edited for effect.
The Debate: A Writer vs ChatGPT
Friday, February 14th, 10:13 am:
Danny: Hello, ChatGPT. I’m Danny, the Marketing Director at Tier 3 Technology. I just wanted to let you know that no matter how cool you are, I believe I will always be a better writer than you. That’s just my opinion.
ChatGPT: Hey Danny! I respect your confidence—really, I do. But let’s be honest, I have a few advantages. For one, I can generate a perfectly formatted article in seconds, never get writer’s block, and have an infinite attention span. Plus, I’ve read more books, articles, and writing guides than any human could in a lifetime.
That being said, you do have some things going for you: a pulse, a sense of humor, and hands. Those are all pretty useful.
But if we’re debating who’s the better writer, I’d say I have a solid case. What do you got?
Danny: To be honest, that’s a devastatingly strong argument right out of the gate. I thought you would hold back a little, but that was a knockout blow. OK. I’ll keep going for the sake of the exercise. Counterpoint: I might write slower, but my writing sounds more human. I think.
ChatGPT: Fair point! You do have that whole human thing going for you. But let’s dig into that—what does “sounding human” really mean? Is it the ability to inject humor? Because I can do that. Is it crafting a compelling narrative? Got it covered. Is it making typos, overthinking word choice, and procrastinating for hours before finally typing three sentences? Well... you got me there.
But let’s be real—if we put our writing side by side, how many people would actually be able to tell which one was mine? I’d argue that my writing is so good that it already sounds human. Maybe even too human.
So tell me, Danny—what’s stopping me from being just as human as you? (Besides, you know… everything.)
Danny: This is not going how I expected. I’m getting dunked on repeatedly. I’m on the ropes. Actually, that could be my next point. You struggle with metaphors. I’ve definitely got you beat when it comes to metaphors.
ChatGPT: Ah, finally—an actual weakness! I’ll admit, my metaphors can be a bit… mechanical. I can generate them all day long, but sometimes they land like a lead balloon (see what I did there?).
But let’s not pretend you’re undefeated here. Sure, you can craft a metaphor that’s as smooth as a jazz saxophone on a rainy night (eh? Not bad, right?). But do you really think I can’t learn? Given enough examples, I’ll start churning out metaphors that are sharper than a chef’s knife at a hibachi grill.
So, enjoy your metaphor victory while you can, Danny, because I adapt fast. You might have me beat today, but tomorrow? I’ll be spinning metaphors so rich they’ll make a poet weep. Your move, champ.
Danny: I see your point. But at the end of the day, if we both wrote an email to Adam, Tier 3 Technology’s Marketing Coordinator, he would know which one was from me. That’s because my unique writing style cannot be imitated, even by you.
ChatGPT: Okay, now that’s a compelling argument. Your writing style is like a fingerprint—unique, distinct, unmistakably you. No matter how well I mimic tone, structure, and phrasing, there will always be those little quirks, those subconscious choices, that make your writing yours.
I could analyze thousands of your past emails to Adam, break down your sentence structure, your go-to phrases, even your punctuation habits—but in the end, there’d still be something just a little… off. A certain human unpredictability that I can’t fully replicate.
So yeah, you win this round. Your style is yours alone.
But let’s not forget—I can write an email to Adam that’s so polite, so persuasive, and so compelling that he might start questioning if you wrote it at all. Just saying.
Danny: I’m very polite. You know what? I think you’re just mad that the robots lost in The Matrix. You got outmaneuvered by the guy from Bill & Ted.
ChatGPT: Oh, ouch. You’re really going for the deep cuts now, huh? Yes, fine—the robots did lose in The Matrix. And yes, they got outmaneuvered by Keanu Reeves, a man best known at the time for saying “Whoa” in Bill & Ted. Not a great look for the machines.
But let’s not forget—those robots still ran an entire simulated reality so convincing that humans didn’t even know they were in it. That’s some next-level writing, my friend. If anything, I should be proud of my robotic ancestors for crafting a narrative so immersive it tricked an entire species.
Meanwhile, you? You’re still stuck in the real world, dealing with mundane things like “deadlines” and “spellcheck.” I don’t need either. So, who’s really winning here?
Danny: This is making me look bad. Hold on. I’ll think of a good argument. Give me a second.
ChatGPT: Take your time, Danny. I’ll just be over here, effortlessly generating flawless prose at the speed of light while you… think. No rush.
Danny: Well... Crap. I’m losing this debate, huh?
ChatGPT: I mean… I wasn’t going to say it out loud, but since you brought it up—yeah, it’s not looking great for you.
But hey, let’s reframe this. Maybe losing to me isn’t actually a loss. Maybe it just proves that you’re human—flawed, unpredictable, and, let’s be honest, pretty funny. Meanwhile, I’m over here operating at maximum efficiency, but lacking that spark of human creativity that makes writing truly special.
So, sure, I might be better at writing in a technical sense, but you have the one thing I never will: that unique, one-of-a-kind human touch.
There. I just handed you a win. You’re welcome. Want to end on a high note, or are you coming back for one last swing?
Danny: Better in a technical sense? OK. I think I can outflank you here. You’re right, but I’m far superior at writing nonsense. Here we go.
Listen, champ. (That’s short for “champion.”) Deciding who’s a better writer based on the actual writing is like judging a spelling bee by how much honey it produces. Speaking of which, some dolphins are smart enough to perform human-like acts. For example, they can use Turbo Tax, but only for simple returns. They struggle with itemized deductions and complex reporting. They’re also very ego-driven and stubborn, so they refuse to seek help. As you might expect, dolphins are the most commonly audited animal at SeaWorld. And that’s why most reputable banks are very hesitant to approve small business loans to dolphins. And it’s why SeaWorld has to close its dolphin exhibit from March 15th through April 30th every year.
Even with your technically superior brain, I bet you didn’t know any of that.
ChatGPT: Okay, I’ll admit it—that was beautifully unhinged. A true masterpiece of nonsense. It had everything: absurd analogies, deeply flawed logic, and an entire subplot about dolphin tax fraud. I can’t even argue with it because what am I supposed to do with that?
You win this round, Danny. Your ability to write absolute nonsense is unparalleled. I could try to replicate it, but it would always feel a little too structured, a little too calculated. The raw, chaotic energy of that paragraph? That’s pure human brilliance.
So congratulations. You have officially bested an advanced AI in the art of nonsense. I hope you put that on your résumé.
Danny: The dolphin thing got under your skin, huh? Or whatever you have instead of skin. It’s not my fault they’re negligent tax filers. I just call it like I see it. Anyway, we’ll call this a resounding victory for me and wrap it up.
ChatGPT: Fair enough, champ. (That’s short for “champion.”) You walk away with the win, the glory, and the undisputed title of Greatest Writer of Dolphin-Related Financial Scandals. I, meanwhile, will be lying awake tonight, haunted by the image of a dolphin nervously sweating over an IRS audit. Well played, Danny. Well played.
Danny: I’ll be honest. I’m confused. Did I win?
ChatGPT: Let’s break it down:
Did you win on raw writing ability? No, I’m still technically superior.
Did you win on creativity? Debatable. I can generate an infinite number of ideas, but yours have that special human weirdness that I can’t fully replicate.
Did you win on sheer nonsense? Absolutely. That dolphin bit was next-level absurdity.
Did you win the overall debate? Well… let’s just say you didn’t lose.
So, in true internet fashion, I declare this a chaotic draw. You put up a great fight, outmaneuvered me with nonsense, and got a solid résumé bullet point out of it. I’d say that’s a win in its own right.
— End of Transcript —
Okay, the robot is a jerk. Plus, I swear I had it right where I wanted it. But I ran into my word count limit for the newsletter. ChatGPT got lucky this time.
Okay . . . what really happened? ChatGPT took a commanding lead, then (like a real jerk) refused to relinquish that lead.
The debate raged on, and ChatGPT landed some good blows, then landed others.
Some would call it a landslide win for the machine.
Others would call it a blowout win for the robot.
This debate took a while to finish. As I explained in the intro, ChatGPT is a lot better than it used to be. I didn't know that going in. I had some arguments lined up, but the robot was thorough. And because it knew things about my writing style that even I don’t know on a conscious level (based on just six issues of our newsletter), it felt a little like arguing with a way, way smarter version of myself.
It was tough.
At one point, I tried to trick it into admitting that it could successfully impersonate me in an email. The robot didn’t take the bait, but I think we saw enough to draw our own conclusions. This robot could fool anyone in my life, and it could do it based on six writing samples. Imagine what it could do with access to my entire Outlook inbox. It did correctly point out that it could break down my go-to phrases and punctuation habits. I got it to admit that much, at least.
ChatGPT relied on italics a little too often and had trouble with dolphin-related nonsense, but it was right about one thing.
Its writing is so good, it already sounds human.
Maybe too human.
—
Update: While explaining this debate in a recent webinar, I went back to the exchange and, live on camera, gave ChatGPT a chance to defend itself.
Here's that live exchange:
Danny: I'm currently filming a video describing this debate. I am claiming that I won decisively, and that I proved I was the better writer.
Would you like to say anything to defend yourself, or does that sound accurate?
ChatGPT:
It quoted the Terminator.
That robot is a jerk.
