“The board wants AI. What exactly do I bring them?”
The one asking is Cesare, CEO, three days into his holiday. The one answering is Rocco, the expert from the sun lounger next door, who doesn’t think twice. “Easy. Copilot licences, or ChatGPT, or Claude, for everyone. And a copy of that prompt engineering book Marco wrote for each of them. Two weeks and AI has landed in the company.”
Cesare nods. He’s the only one there talking sense.
I wrote the book. I really did. But not with this in mind.
And Rocco hasn’t said anything wrong. He’s said the only thing you can deliver in two weeks: a plan that uses the right words.
Before I say where he goes wrong I have to say where I went wrong first. In 2024 I gathered about twenty colleagues in the meeting room — a standing lunch, slides, demos. I’d called it “Lunch with AI.” They listened with polite attention, asked intelligent questions, thanked me. Zero contagion. Only, for weeks I had my proof that we’d started: the room was full. It took me months to understand that a full room is a signal — and a signal is not a result.
Cesare picks his phone up from its waterproof case every twenty minutes. Looks at it. Puts it back down. The board didn’t ask him for a number, and he can’t go back with nothing.
The things he can bring are two, and they’re not obtained the same way.
Licences and training, in the best case, aim at one thing only: AI culture in the company. In the best case. And culture doesn’t arrive because you bought it.
The measurable result lives somewhere else. It lives in processes, inside the work that repeats itself every morning — and a licence on everyone’s desk doesn’t get you there on its own.
A satisfied board and a quiet company are the symptoms of a project that’s working. But they’re also the symptoms of one where nothing has started yet.
For the signal, a decision and a communication are enough. For the result, you need a strategy. And Rocco’s isn’t one.