The Sign Had to Be Real
On judgment, creative instinct, and why knowing which tool to reach for is the skill nobody’s talking about.
There’s a piece in Fortune this week that stuck with me. Matti Yahav, CMO of Fiverr, wrote about the moment he watched a thirty-foot sign go up on a hillside along the 101 in Los Angeles, saw the city lose its collective mind over it, felt the elation of a campaign landing exactly the way he’d designed it to land, and then felt something else entirely. A stone in the stomach. The quiet realisation that he may have just spent months building the most persuasive argument for why his own role was about to be fundamentally rewritten.
The sign spelled out the name Billy Boman. An AI video director working largely alone out of Stockholm. Clients including Google and Klarna. Commercials that compete with those produced by the largest advertising agencies in the world. No massive crews. No seven-figure budgets. Just a laptop and a point of view.
Yahav writes that Boman isn’t valuable because he uses AI. Plenty of people use AI. He’s valuable because he brings a visual sensibility, a creative instinct, and a point of view that no prompt can generate. The technology expands what he can execute. It doesn’t do the thinking for him.
The gap between someone who can operate AI tools and someone with genuine judgment is the one that will define who thrives and who gets replaced.
In the creative industries especially, that distinction matters enormously right now, because the tools have never been more democratized, and democratized tools have a way of making everyone look the same.
I’ve spent over twenty years in factual television. Across the BBC, Channel 4, HBO Max, History Channel, Discovery. In that time, I’ve learned that the job title on the call sheet has never been what makes someone irreplaceable. What makes someone irreplaceable is instinct, taste, that they know what a thing should feel like before it exists. They know how to read a room, a subject, an archive, a silence and see what others can’t. They know, and this is the part that’s genuinely difficult to articulate and almost impossible to automate, when to push and when to leave it alone.
None of that is workflow. It can’t be prompted.
Which brings me to the other thing in Yahav’s piece that I can’t stop thinking about. The irony at the heart of the whole campaign. They could have rendered that Hollywood hillside in AI. Indistinguishable from the real thing.
But it wouldn’t have worked. Nobody would have pulled over on the 101 to film it on their phone. Nobody would have posted it on Reddit. No news crew would have shown up. Tom Sandoval from Vanderpump Rules wouldn’t have been worried about his view.
The organic chaos, the thing that made the campaign actually work, required something that existed in the physical world. Something with weight and dimension and presence. Something people could drive past and argue about and wonder at.
The judgment call wasn’t which software to use. It was whether to use software at all.
And there’s something almost perfectly ironic about it: a campaign promoting an AI-powered creative, during a time when conversations are ongoing about whether AI is replacing human practitioners, achieved all of that traction precisely because it was built around something real.
The thing that made it land wasn’t the AI. It was wood and steel bolted to a hillside, catching the morning light on the 101.
That’s the part of the conversation we keep sliding past. Everyone’s talking about who can use AI best, who’s building the most efficient prompt, who’s automating the most steps. And those are real questions worth asking. But they’re not the only questions.
The deeper question, the one that will separate the people who navigate this moment from the ones it swallows, is whether you know what the job actually requires.
Whether you understand the outcome well enough to choose the right method or tool for the job. Whether you have the judgment to know when the real thing still matters. Can your judgment and taste connect with an audience.
I grew up in the transition from analogue to digital. I make linocut prints, shoot on film and read physical books. Not out of nostalgia, but because I understand something about the difference between a thing that exists and a simulation of it.
Both have value. They have different value. Confusing the two, reaching for a render when you need something people can touch, is exactly the kind of mistake that judgment is supposed to prevent.
The tools are extraordinary right now. I use AI tools. I’ll keep using them. But the tools are not the point of view. The tools don’t know what the project needs. The tools don’t know when the sign has to be real.
That’s still on you.
You can read the full Fortune article HERE.