Last week, lunchtime, a meeting room at our place. Around thirty people from marketing, laptops open, sandwich on the side. No slide marathon, no «AI – an introduction». Instead: everyone's building. Each of them on their own little AI tool.
And somewhere between two soup bowls, something hit me that I haven't been able to shake since.
We almost always talk about AI in companies as a knowledge problem. «People need to understand it first.» So: another webinar. Another explainer video. Another deep-dive article. Another LinkedIn post by yet another AI expert. (Yes, I know. I'm writing this on LinkedIn right now. 🤷🏻♂️)
But honestly? That's the wrong question.
A year ago I wrote here about AI Information Overload. My point back then: don't drown in the noise, go deep on a few good sources. Today I'll take it one step further – and almost contradict myself:
Knowledge about AI isn't rare anymore. It's everywhere. One prompt away. Free, anytime, at any depth. What's rare is something else entirely. It's the doing. The threshold of putting your own hands on it. The moment you stop reading about AI and start building with it.
And what you build in that moment isn't knowledge.
It's judgment.
Nate B. Jones puts it beautifully: for code there's a compiler – it tells you whether it runs. For taste, there's none. AI hands you a draft in seconds. What's rare is you – the one who can tell whether the draft is any good. Whether the text actually sounds like you. Whether the analysis holds. Whether the image lands. Those last few percent, that «hm, smells off, again» – that's the real skill. And you won't pull it from any video. It only grows when you do it yourself. Again and again.
A little side note, because it's too good: one of the tools people built last week was a «Humanizer» – an assistant that strips the AI smell out of AI text. Translation: they built a tool that judges when language is good and when it isn't. Exactly that judgment. They built taste. 🤌
This is where the real lever is.
It's not about teaching people more. It's about taking away the hurdle of doing it themselves. Because the hurdle is rarely knowledge. It's respect. Uncertainty. «I'm not technical.» «What if I break something.» «Others can do this better anyway.»
That's exactly where we come in: a safe space, a bit of guidance, a real task from your own day – and suddenly someone who called themselves «not technical» that morning builds a working tool. Before lunch. We call it enablement. Honestly, the word doesn't matter to me. Coaching gets closer. Accompany. Bring people along. What counts is the step: from being informed to doing it yourself.
One number nails it for me: in our baseline survey, 74% already used AI daily. Daily! And still, average confidence sat at 6.6 out of 10. Read that again. These aren't beginners. These are people who already use the stuff – and still don't quite trust what it gives back. The gap isn't access. The gap is capability. And capability comes from doing, not from watching.
And if I zoom out for a second: this, to me, is the core of an «AI-first» organization. Not the tool stack. Not the number of licenses. But the question of who actually gets to build. In the old world, the idea traveled upward – into a brief, into an agency, into a queue. In an AI-first world, the person who knows the problem best builds the first version themselves – today, not in six weeks. That's not a tool advantage. It's an organizational one. And it compounds with every single person who's cleared that hurdle once.
If that's true, it changes what a marketing team has to do. Marketing has to move – every one of us, not «the AI team over there». And those of us driving it have, in my view, an obligation: not to inform and then hope. But to clear the path. To make the hurdle small enough that your own first click feels easy. And to stand next to you when it gets stuck.
And because this is allowed to be a bit of a love letter too: what Tanja and François pull off in these sessions is exactly that. They don't stand at the front and lecture. They walk the room. They get people building instead of listening. They turn a lunch slot into real capability. Quietly, concretely, session after session. Hats off.
I'm proud of what we're building as the Marketing Technology & Automation team. And almost prouder of the road we're on.
You don't learn AI by reading about it. You learn it by daring to do it – guided, in a safe space, with a real task.
Build together, learn together. We mean it. 🚀