Drivers Wanted

AI doesn't replace great marketing. It makes average work free.

Nobody wants to make average work. Yet most is. We're worried about AI stealing our jobs when the real problem is simpler: AI just made mediocre marketing easy, fast and almost free. Not bad marketing. Mediocre marketing. The decent campaigns. The "good enough" strategies. Anyone can now produce passable content. Reasonable copy. Decent analysis.

What used to differentiate marketers from the competition is now available by subscription.This first hit me last September (2024) when I started seriously experimenting with Claude on marketing campaign ideas. What we got to, was pretty decent. Was it better than the best people I've worked with, doing their best work? No. But it was better than a lot of stuff I've seen and done throughout the years.

Somehow, we went from being called marketers, creatives, researchers, to being labeled "humans in the loop". Nobody used to say "Human decision required". We just called it work. If we're not intentional about this, we'll end up the backup singers in a song written by software. "Don't worry, there's still a human involved" is the professional version of "We swear someone checked this." This is absurd. This framing will rewrite our value before we've had a say.

The goal isn't to protect our role from machines. It's to define it. Own it. Drive it. Because when you strip away the hype, the simple truth is AI is brilliant at doing. But it's still bad at deciding. It doesn't know what matters. It just knows what's next.

So the question isn't "How do we keep humans in the loop?" It's "What should never be outsourced?" That's your real value. The things that only a human, a sharp, strategic, culturally attuned human, can bring. But even "human value" is vague. What's yours? It could be taste. Timing. Ethics. Cultural nuance. Leadership. Emotional fluency. Be specific.

Two weeks ago, I watched this play out in real time. I put my phone on a client's desk, on speaker, and asked Claude to run a discovery call. We were physically in the client's office in Dublin. Two humans and one robot. Claude ran the meeting, and politely worked through a bunch of strategy-type questions, probing and pushing when not satisfied with the answers. Occasionally it didn't hear every reply, but overall, was pretty impressive.

"What's your core challenge?" "People don't know we exist." "How do current customers find you?" "Word-of-mouth mostly." "What's your customer acquisition cost for word-of-mouth versus paid channels?" "We don't really do paid channels." "Why not?" "Haven't needed to."

Claude kept drilling. Revenue per customer. Lifetime value. Market size. Competitive positioning. When answers got vague, it asked for specifics. Numbers. Examples. Did funnel conversion calculations on the spot. Even offered a decent diagnosis. Wrote up the entire meeting notes before I'd left the building.

Overall, both client and I were impressed with the quality of the meeting and the experience of this robot talking away with us. It's insane to me that it could do what it did. Claude was happy enough too apparently.

But it was also clear to me what it could not do. Claude needed to know which frameworks to use, when to push different directions, when to move on. AI doesn't replace the strategy work. It executes it. But someone still has to know what we should be doing, why, and how.

That's the role we must claim. Not a robot wrangler. Not a compliance officer. Driver.