Creative partner? Content machine?

The cursor blinked at me mockingly. I’d been staring at a blank document for twenty minutes, trying to articulate… Something.

The ideas were right there. Near to the front of my head. But trying to find a way to express them was… Hard.

So I did what any sensible person does when stuck: I started talking to my computer.

I opened up Claude (my choice – recommended by journalists and writers I follow) and began explaining the problem, and what I wanted to say, out loud, typing my scattered thoughts into a conversation.

Within minutes, I had three potential angles to explore.

This wasn’t AI replacing my thinking. It was AI helping me think better.

The assistant: Working through ideas together

Most people approach AI transactionaly. Fire off a question, get an answer, move on. But that misses the real value. AI works best as a thinking partner.

When I’m wrestling with a complex problem, I treat AI like that colleague who’s brilliant at asking the right questions. The one who listens to your rambling explanation and says, “Have you considered this angle?” or “What if we looked at it from the customer’s perspective instead?”

Take a recent piece I was writing about what I call “concentric circles of being interesting.” I had this half-formed idea about how the most engaging content creators operate in overlapping spheres—their core expertise at the centre, then related interests, then seemingly random obsessions that somehow make them more compelling.

I knew the concept was sound but I couldn’t articulate it clearly. Every time I tried to explain it, I either oversimplified or disappeared down a rabbit hole about why architecture matters to marketing.

So I started working through it with Claude. Not asking it to write the piece, but using it to test different explanations. “Does this metaphor work?” “What’s confusing about this definition?” “How would you explain this to someone who’s never thought about personal branding?”

The AI helped me work through dozens of angles I wouldn’t have considered alone. It became a collaborative process—me bringing the observed pattern and examples, AI offering systematic ways to structure and explain the idea.

The content machine: Writing to specification

Here’s where AI truly excels: creating content when you know exactly what you need but can’t be bothered with the mechanical bits.

Need a 500-word blog post about conversion optimisation with three specific SEO keywords and a conversational tone? AI can nail that in minutes. Want five different subject lines for an email campaign, each testing a different psychological trigger? Done and dusted.

This isn’t because AI is more creative than humans (it’s not). It’s because AI is pathologically good at following instructions. Where I might get distracted by a tangent about Victorian marketing techniques (fascinating stuff, honestly), AI stays relentlessly on brief.

I’ve used this approach for everything from social media captions to technical documentation. The process is simple: be specific about what you want, provide examples of the tone you’re after, and let AI handle the execution while you focus on strategy and quality control.

The unexpected benefit? It’s made me better at writing briefs. When you have to explain exactly what you want to a machine, you discover how often your creative briefs were actually quite vague. “Engaging” and “on-brand” mean nothing to AI. “Conversational tone like Malcolm Gladwell explaining something to a marketing manager over coffee” works perfectly.

The experiment lab: Testing wilder ideas

This is where things get interesting. AI has no shame, no reputation to protect, no fear of looking stupid. This makes it excellent for exploring ideas that might be terrible but could also be brilliant.

I’ve used AI to brainstorm campaign concepts that would never survive a traditional brainstorming session.

“What if we positioned our eCommerce tool like a financial advisor?”
“Could we create a marketing campaign based on the principles of brutalist architecture?”
“How would David Ogilvy approach TikTok advertising?”

Some of these experiments are genuinely awful. But buried in the mess are occasionally nuggets that transform how you think about a problem. AI gives you permission to be ridiculous, which is surprisingly valuable in a discipline that often takes itself too seriously.

The breakthrough moment came when I realised I could use AI to rapidly prototype content strategies. Instead of spending weeks developing a content plan only to discover it doesn’t work, I can test dozens of approaches in an afternoon.

AI helps me fail faster and cheaper, which means I can afford to try ideas that would otherwise never make it past the “that’s interesting but…” stage.

What AI can’t do (and why that matters)

Let me be clear about what AI isn’t: it’s not a replacement for understanding your customers, your market, or your business. It can’t tell you whether your strategy is fundamentally sound or whether you’re solving the right problem.

AI is good at execution and terrible at judgement. It can write compelling copy but can’t tell you if that copy will actually work with your audience. It can generate creative concepts but can’t evaluate whether they align with your brand values or market position.

This is why the human element remains crucial. AI amplifies human capability; it doesn’t replace human insight. The marketers who get the most value from AI are those who use it as a force multiplier for their existing skills, not as a substitute for developing those skills in the first place.

The practical approach

If you’re wondering how to start using AI more effectively, begin with the boring stuff. Use it for tasks where you know exactly what good looks like but don’t enjoy doing the work.

Content creation to specification? Perfect AI territory. Brainstorming alternatives when you’re stuck? Excellent use case. Making strategic decisions about budget allocation? Probably not.

The goal isn’t to automate everything—it’s to automate the bits that free you up for the work only you can do. The thinking. The connecting. The understanding of what actually matters to the humans you’re trying to reach.

Because in the end, that’s what marketing is about: understanding people. AI can help you explore that understanding more thoroughly and express it more effectively. But the understanding itself? That’s still yours to develop.

And honestly, that’s exactly how it should be.