Jun 23, 2025

Can I replace my copywriter with AI?

I’m replacing my copywriter with AI.

Well, not really. But I am running an experiment. I want to see how much effort and energy it would take to write a blog with AI instead of a copywriter. I want to see what kind of quality I can get and whether it resonates with our audience.

My hypothesis?

The content is going to suck.

The process is going to suck.

(Especially when you consider my former copywriter was a GEM).

So, why would I do this?

Recently, my copywriter of four years moved into a full-time agency role, stepping back from freelance writing. We’re thrilled for her career growth, but she can no longer handle our weekly blog. I’m not going to lie, I was worried I wouldn’t find someone who could replace her.

She and I worked together incredibly well. We planned content strategically. She was proactive and reliable. She brought a sense of empathy and social awareness to every post. Not only did I appreciate that—it helped grow our brand in a direction I’m proud of.

At the same time, I see this as a moment to modernize our blog. We’ve been talking about AI a lot these last few years, especially as we’ve watched it disrupt creative industries.

(See here. And here. And here. And here… and here.)

And truthfully? I’ve already been using ChatGPT behind the scenes—for keyword infusions, social copy, catching typos, and getting to the point (I’m awful at writing concisely—sorry in advance).

But I don’t fully understand it.

What’s the difference between a Large-Language Model(LLM) and Generative AI?
What sources is it pulling from?
Is it lying to me? Or is that a hallucination? What’s the difference?
Are the keywords it gives me actually good for SEO? Is SEO still even important when everyone uses ChatGPT for every little question?
Does it actually understand my brand voice, or is it just mimicking something it scraped off my LinkedIn?
Could I train it better? What does that even mean?
Was the LLM ethically modeled? Or is it going to Grok out on me?
Will my tone get lost in the process?
Will I start to sound like everyone else on the internet?
And maybe most importantly: Will our audience trust what we write if they know an LLM helped create it?

What I’m realizing is that my concerns fall into four buckets:

Trust: Can I rely on the information ChatGPT provides? Will it maintain the integrity of our content and the voice I’ve spent years building?
Ethics: Was this technology trained responsibly? Is it unintentionally perpetuating bias or misinformation? And what about the environmental impact of these tools?
Logistics: How much effort will it really take to produce quality content with AI? Is it faster, or just different work?
Reputational Risk: What happens if we lean too far in and lose our creative edge or our audience’s confidence?

Ultimately, this experiment isn’t about logistics or saving money. It’s about understanding where the creative market is going (or being forced into going). Creative teams are facing increasing pressure to do more with less. But do we really want to trade nuance, voice, and cultural relevance for efficiency?

As a Pratt Design grad, moments like this feel like an opportunity to lean on the 4D process: Discover, Define, Design, and Deliver—a human-centered framework that helps creatives approach ambiguity with structure and intention.

Discover is all about research. What do I know? What don’t I know? What does the audience care about? This is where I’m poking at AI's capabilities—and its limitations. (Think: a lot of late-night Googling and asking ChatGPT borderline rude questions like “Are you making that up?”)
Define helps me sharpen the problem. In this case: How do I maintain brand voice, SEO integrity, and trust while experimenting with AI? Is the risk worth the time savings? What would success even look like?
Design is the part where I roll up my sleeves and co-write the actual content. This is where AI and I are in the weeds together—drafting, editing, revising, and yes, arguing about phrasing.
Deliver is when I share the final product. Not just the blog itself, but also the insights around the process—what worked, what didn’t, and what others can learn from it.

I’ve always loved the 4D method because it reminds me that creativity isn’t chaos. It’s a structured exploration. It’s asking better questions before jumping to solutions. And it’s especially relevant right now, when so many teams are being asked to “just use AI” without a clear plan, creative direction, or understanding of the tech.

Here’s the plan.

For the next two weeks, I’m going to publish a blog that I wrote with AI. I’ll collaborate with ChatGPT from concept to final draft—sometimes giving it prompts, sometimes asking it questions, sometimes fighting it on word choice like I would a real writer.

And the best part? I’m documenting the whole thing with a video partner, so you’ll be able to dive into the details and the lessons with me after we’re done. Expect behind-the-scenes commentary, screen recordings, and some existential dread in real time.

At the end of the experiment, I’ll share my learnings—what worked, what felt weird, what flopped, and where AI will fit in our process moving forward.

Spoiler: I’m not firing any writers. But I am trying to figure out how creative humans and smart machines can actually collaborate, not just coexist.

Stay tuned.