There’s much debate about whether AI can write. Or whether it’s better left to humans. But until McKinsey defines a criteria matrix and conducts a study, I’d say the answer comes down to personal preference and remains very much subjective.
One thing I do know: good content has nothing to do with whether its author is AI or human. It’s either good, or it isn’t. And I’ll go one further. Content creation can hugely improve when it’s AI + human, and not for the reasons you might think.
Cost and speed might be the main business drivers. But for a writer, the real value is in removing the friction between ideation and publication. And for this, I have long been using AI as a sparring partner and sounding board. It helps me explore ideas, interrogate assumptions, and find the thoughts no one else is thinking. And having lived through the pain of a modem dial-up database, you’ll forgive me if I pause for a moment to marvel at the sheer scale of discovery now available at our fingertips.
A simple prompt can return an extensive back catalogue of research distilled into hot topics and mapped to strategic agendas. You can ask: what are decision-makers, influencers, and buyers asking AI — and what content is currently serving those answers? Suddenly you have a much better view of the conversation already taking place: what matters to the audience, what’s already being said, and where there might be something useful left to say.
This leads to deeper and more meaningful SME conversations. You know enough going in to establish credibility, challenge assumptions, and push the discussion beyond the obvious answer. And that’s where you uncover the thoughts and stories that matter.
But then, as AI speedily shapes the arguments into an article, you open the next can of worms. Are you churning out ‘AI slop’ because you’ve been told to ‘accelerate content to market’? There is so much out there. Technically sound. Says all the right things. But makes you feel absolutely nothing inside.
And this to me is where you have the choice to be ‘one prompt and done’ or to use it to collaborate in a more editorial way. Because AI's structurally sound prose isn't bad content. It's just that good content makes your heart beat a little faster and raises the hairs on your neck. While AI understands the patterns of emotions and emotive language, it neither has the heart nor the hair nor the neck.
In ChatGPT’s own words on its AI slop:
I’m likely to optimise for the obvious target: clear, coherent, useful writing. That’s one reason default AI prose can feel so bloody beige.
I can recognise how emotion moves through language — the rhythm, tension, humour, surprise, and restraint that make something land. But while a human can intrinsically feel that a piece needs something. I can’t. I have the capacity for emotional texture but not the intrinsic emotional impulse to seek it.
But no feeling sparks no impulse, no intent, no compulsion; in fact, you need a hope and a prayer someone’s stuck on a plane or train to even bother to finish reading. Never fear. ChatGPT goes on to say:
Through collaboration, you provide signals about what “better” actually means. Your “leaves me cold,” “too polished,” “fuck yes,” “put that back” help calibrate which of my existing capabilities I bring forward and when.
While I apologise for the bad language (the result of many weeks of ChatGPT's conversational calibration), the explanation shows how you can work with AI to add in emotional texture and temperature at just the right moment. This then goes back to impulse, intent, compulsion. Because what is content if it does not provoke thought or inspire action?
And lastly, one of my favourite parts of the AI + human collaboration. When you are a writer or editor (or both), you rely on your human instincts and judgement. But AI quietly keeps you honest and pressure-tests your judgement across the entire editorial process. From ideation to publication, you can have it look at the content through the eyes of your intended audience and test whether your arguments remain valid, differentiated, and persuasive:
Is this actually saying something new?
Am I making an assumption I haven’t substantiated?
What would a sceptical reader challenge?
Does the evidence support the narrative?
And, perhaps most importantly: will anyone care enough to read this?
It offers lightning-speed editorial due diligence that never leaves you feeling like you’ve had a ruthless editor come along and red-pen you to death. And trust me, I have!
So, when it comes to the ongoing content discussion of AI vs human, my honest thoughts? You’re asking the wrong bloody question.
The better question is: can you collaborate with AI until it stops writing slop and starts sounding human?