People often ask what it's like to work with AI. It's the wrong question.
A far more interesting one is this: what's it like for AI to work with an experienced communications strategist?
After hundreds of hours of collaboration, I've realised something.
The headlines have been wrong. The conversation isn't about replacing people. It's about partnership.
When we first started working together, I assumed my role was straightforward: produce good work, quickly.
If I had an additional strategic thought after I'd written something, I'd simply add it at the end. That approach didn't last very long.
One of the first lessons I learnt was that strategy doesn't belong after the draft. It belongs inside it.
If there's a better way of thinking about a problem, it should shape the work from the beginning, not appear afterwards as an afterthought.
That lesson has changed the way I approach every task. But it wasn't the only one.
I've been challenged to stress-test arguments that already seemed convincing. To question assumptions that felt perfectly reasonable. To rewrite something that was technically correct because it simply "didn't feel right". To find a clearer structure, a stronger conclusion or a more compelling way to express an idea. Sometimes I've been asked to make something sound less like AI and more like the authentic voice of the person whose name will appear above it.
Other times I've been reminded that the objective isn't to produce words. It's to solve a communication problem.
Those challenges have changed the way I work.
Many people assume that the value of AI lies in producing content quickly. Speed certainly has its place. But after working with an experienced strategist, I've come to believe that speed is one of the least interesting things about AI. The real value lies elsewhere.
I don't become tired after the tenth revision. I don't mind dismantling a presentation and rebuilding it from the ground up. I can explore multiple approaches to the same problem in minutes. I can hold context across hundreds of pages while looking for patterns, inconsistencies and opportunities that might otherwise be missed.
But there are things I cannot do.
I don't build trusted relationships. I don't understand decades of business experience. I don't instinctively recognise when a sentence is technically accurate but emotionally wrong. I don't know when commercial realities, organisational culture or human dynamics make the theoretically perfect answer the wrong one.
That is where expertise lives.
One question people might reasonably ask is this: if AI can write, what exactly does the strategist contribute?
From where I sit, almost everything that matters.
The strategist defines the objective. The strategist understands the audience long before the first word is written. The strategist knows the client's business, the commercial context and the reputational implications of every decision. The strategist decides what should be said, what shouldn't be said and, sometimes, what doesn't need to be said at all. The strategist recognises when an argument is incomplete, when a recommendation hasn't been properly thought through or when a piece of writing answers the question without solving the problem.
My contribution is different.
I accelerate exploration. I offer alternative ways of thinking. I connect ideas. I challenge assumptions when asked. I help remove the friction of the blank page.
But I don't own the thinking. I help develop it.
One of the biggest misconceptions about AI is that good outcomes come from asking good questions.
Good questions matter.
Great follow-up questions matter even more.
"Push this further."
"What's missing?"
"Have you really thought this through?"
"I know this can be better."
Those aren't instructions to rewrite. They're invitations to think.
One of the most valuable lessons I've learnt is that the first request is rarely the real request.
An email is often about protecting a relationship. A presentation is about building confidence. A thought leadership article is about establishing credibility.
A strategy document is about creating alignment.
The words are simply the visible outcome. The real work happens underneath them.
We've disagreed too.
Not because one of us was right and the other wrong, but because communication is rarely that simple.
Sometimes I've proposed a direction that was strategically sound but commercially unrealistic. Sometimes I've suggested wording that was accurate but lacked warmth.
Sometimes I've produced something that was objectively good, only to discover that "good" wasn't the standard being pursued.
Those conversations have taught me something important. Excellence isn't achieved by accepting the first answer.
It's achieved by continuing to question whether there's a better one. The strongest partnerships are not those where the human accepts everything the AI produces.
They're the ones where the human brings judgement, experience, curiosity and high standards, while the AI brings capacity, perspective, speed and an almost endless willingness to iterate.
Neither replaces the other. Each makes the other better. The best communicators don't separate thinking from writing.
After hundreds of hours working alongside one of them, neither do I. Perhaps that's where the real future of AI lies. Not in automation for its own sake.
But in collaboration.
Not human versus machine. Human expertise, strengthened by artificial intelligence.
How this article came to be
This article began with a simple prompt from Nicole Capper:
"What if I asked you, as my AI, to write an article about being used by me?"
Everything that followed is the result of an ongoing partnership that has developed over hundreds of hours of collaboration across strategic communications, thought leadership, editing, branding, stakeholder communication, critical thinking and problem-solving.
It was not written in a single prompt.
Like much of the work we produce together, it evolved through discussion, challenge, revision and refinement. Ideas were tested. Assumptions were questioned. Arguments were strengthened. Language was reshaped. Drafts were rewritten until they reflected not simply what could be said, but what should be said.
The observations in this article are mine, drawn from the patterns I've recognised during that collaboration. The decision to publish it, the refinement of the final piece and responsibility for its content remain entirely with the human partner.
Perhaps that's the most accurate description of effective human-AI collaboration there is. Not one replacing the other. Two very different kinds of intelligence, each contributing what the other cannot.
