AI can write the caption. But who trains the marketer?

A business owner opens an AI app, types one line, and out comes a caption. A poster. A first draft of an ad. It took eleven seconds and cost nothing. So a fair question follows; one I hear more and more often: if the tool does the junior work, why hire, or become, a marketer at all?

I’ve spent 21 years answering some version of that question. This month, one of the most respected consulting firms in the world put it in a headline, and I couldn’t stop thinking about how squarely it lands on us here in Uganda.

The question everyone is quietly asking

McKinsey published a report titled “Building expertise in the age of AI: Who trains the next generation?” It’s written for global corporations, but it describes our market with uncomfortable accuracy.

The argument goes like this. For decades, the way a beginner became skilled was by doing the routine work: the research, the first drafts, social media posting, creating content calendars, and the basic graphics and web layouts. Nobody thought of those tasks as important, but they were the training ground. They were how instincts got built. And those are exactly the tasks that AI now absorbs in seconds. Take away the routine work, and you have to ask an awkward question: where does judgment come from now?

Here’s the part worth sitting with. McKinsey’s conclusion is not that skilled people are becoming unnecessary. It’s the opposite. The pipeline of talent isn’t disappearing — it has to be deliberately rebuilt. And the skill that now separates a hireable marketer from a hobbyist with an app is not tool fluency. It’s judgement.

The one thing AI can’t hand you

Anyone can generate a post now. That is precisely the problem. The AI tools give everybody the same average output. What it cannot give anyone is the judgement to know whether that output is any good, whether the hook actually stops the scroll, whether the offer fits the audience, whether a “clever” caption is quietly hurting the brand, and when to throw the whole thing out and start again.

On the platforms in 2026, an AI post with no human layer reads as exactly that, and audiences have learned to feel the difference in a heartbeat. Someone still has to be the human in the loop, the one steering the tool toward something that actually sells. That someone is the marketer worth paying.

“How do you think?” — not “What did you study?”

This is the line from the report I keep returning to. McKinsey says employers are changing the question they ask a candidate: from “What did you study?” to “How do you think?”

For Uganda, that shift is an open door. Most of the sharp, hungry people I meet don’t have a marketing degree. They have a phone, a hustle, and something to prove. The old system shut them out at the first hurdle. The new question is, can you think, can you judge, and can you apply this to a real business? The report even has a name for people who build real capability outside the traditional route: skilled through alternative routes. I’ve been training them for years. They become some of the best marketers I know, precisely because they learned by doing, under pressure, for real stakes.

So, how is judgement actually built?

If judgement is the skill, the obvious next question is how you build it, because you certainly don’t get it from watching a tutorial. The research points to something very specific, and it happens to be the thing I’ve built this entire school around.

You attempt the work yourself first. Then you compare your attempt against a stronger version. Then you sit with someone experienced who walks you through the gap between the two. Attempt, compare, and discuss. The researchers call it the answer-key model. I just call it how our classes run:

A live session, a real assignment on a real brand, and honest feedback from someone who does this work for a living. The gap between your first draft and where it needs to be — that gap, closing week after week — is judgement forming. You can watch it happen.

It’s also why a certificate from here means something to an employer: the judgment you build here is built on real business problems, not generic theory. It’s yours, and it travels with you.

What this means for you

If you’re looking for work: stop trying to out-type the AI. It will win that race every time. Become the person who can steer it — the one with the judgement to turn average output into something that actually moves a business. That person isn’t being replaced. That person is being hired.

If you run a business: the report’s warning applies to you too. Hiring someone who can only operate the tool solves nothing — you can operate the tool yourself. Hire for judgment. And the demand for exactly that is not theoretical:

McKinsey ends on a line I’ll happily borrow: the pipeline of skilled people isn’t disappearing — it’s being rebuilt. That rebuilding is the entire job of this school. The next generation of digital marketers won’t be trained by an algorithm. They’ll be trained by people who have done the work on the problems that actually matter.