Over more than 24 months of AI on the market, I’ve had the chance to meet many companies and many teams — and I’ve had to navigate the introduction and evolution of AI inside IdeaTech itself (and it’s not over yet…).
At first I supported every approach to AI that came my way, proposing projects, solutions, ideas. Over time I understood that not every approach has real prospects — in fact, only one points toward likely success.
I started to identify two coordinates.
The first coordinate: you
The first dimension is individual. It’s the single professional, employee, or collaborator who learns to work with AI — not to “use it” the way you use software, but to think with it. To ask questions they wouldn’t have asked before. To explore angles they wouldn’t have had time to explore before. To produce at a level that used to require a team.
As happened with tools like Excel, it’s a skill that silently — but clearly — separates those who have it from those who don’t. Like Excel, the gap is invisible from the outside: the colleague who uses AI well has no aura, wears no badge. They simply deliver more, better, with less visible effort. And you think they’re talented. You don’t think they’re using a tool you haven’t yet learned to master.
Like Excel, it’s not a phase. It’s not the course you take before moving on to serious things. It’s a permanent skill that will accompany every professional throughout their career. Twenty years on, people are still discovering macros. With AI it will be the same — multiplied by a hundred.
But unlike Excel, the first coordinate isn’t just a question of personal productivity. It’s something more: it’s the cultural foundation without which everything else fails to take root.
The second coordinate: the system
The second dimension is organisational. And it’s a completely different game.
It means rethinking how the organisation works — knowing that AI exists. Redesigning flows, consolidating functions, automating steps, transforming the way things get done. And it’s not about giving AI to people. It’s not the sum of individual skills — it’s organisational architecture with a completely new paradigm at its disposal.
The difference between these two coordinates is profound. With the first coordinate, the process stays — you move through it better. With the second, the process changes shape.
Take any weekly report produced inside a company. With the first coordinate you generate it in five seconds instead of two hours. With the second you ask: what decision should this flow of information be feeding? Who needs it, in what form, at what frequency? And you rebuild from scratch. Maybe the report becomes an automated alert that fires only when there’s a deviation. Maybe it merges with three other flows. Maybe the recipient changes entirely.
This isn’t optimisation. It’s design.
The second coordinate is the one that scales. It’s where an organisation stops doing the same things faster and starts doing different things. But here comes the knot that almost everyone underestimates.
The knot: why the second needs the first
The temptation is to jump straight to the second coordinate. Buy the platforms, redesign the processes, launch the pilot. The first coordinate — the individual one — looks like a minor step, something to handle with a training course and an email with a link to the platform. Tick the box and move on.
That’s exactly the mistake that costs the most.
The second coordinate needs the first — but not in the way you’d think. It’s not that people who know how to use AI become the ones who redesign the processes. The prerequisite isn’t operational — it’s cultural.
An organisation where people haven’t touched AI hands-on is an organisation that doesn’t know what’s possible. And when you don’t know what’s possible, two things happen. Those designing the transformation from the top work in a vacuum — they draw up processes that look good on paper but that people don’t know how to inhabit, don’t understand, and end up passively undermining. Those inside the processes can’t contribute — they don’t have the language, the experience, the intuition for what’s feasible and what’s an illusion.
The transformation gets decided at the top and endured at the bottom. Six months later, the results haven’t arrived.
Culture can’t be taught
And here’s the most subtle point of all. The first coordinate can’t be covered by a course. It can’t be solved with a manual. It’s not a tick in the mandatory training box.
Going back to the Excel analogy: I’m fairly certain that whoever truly learned to use pivot tables didn’t learn them by sitting in a course. Maybe they heard the term mentioned there, but they really understood them the day they recognised a problem, sensed that pivot tables might work for it, tried, discovered they did, and from that moment took ownership of the concept.
With AI it’s the same. The skill grows through practice, experimentation, the permission to explore. It needs to be nurtured, encouraged, cultivated over time. An organisation that wants to build the first coordinate has to create conditions for people to experiment — not deliver content and measure attendance.
The cost of reversing the order
Those who reverse the sequence pay a price that goes beyond wasted budget.
The real price is cynicism. When an AI transformation fails — and it always fails when the cultural foundation is missing — a conviction crystallises in people’s minds: “AI doesn’t work for us.” That phrase is slow poison. The next time someone proposes something involving AI, they’ll hit walls. Eyes rolling. Passive resistance. The cynicism born from a badly managed failure costs more than any failed investment, because it feeds itself. Every subsequent attempt starts already hampered by the memory of the last one.
The well is poisoned.
The map
If we plot the two coordinates (individual AI and organisational AI) on a graph, four quadrants emerge.
Low individual AI, low organisational AI. Ground zero. No skills, no transformation. It’s not a tragedy — it’s simply the starting point. The risk is staying there too long without noticing, because the gap, as we’ve seen, is invisible — but real.
High individual AI, low organisational AI. The launchpad. People have built the muscle. The organisation hasn’t yet redesigned anything. There’s compressed energy, unexpressed potential. But the ground is fertile — from here, transformation can take root.
Low individual AI, high organisational AI. The castle on sand. The company has bought the platforms, redesigned the processes, launched the pilot. But the people aren’t there. They don’t understand, they don’t participate, they endure. This is the cynicism quadrant — the one where the well gets poisoned.
High individual AI, high organisational AI. Real scale. Widespread culture and redesigned processes. People understand what’s possible; the organisation knows where it’s going. Everything multiplies.
The healthy path is clear: ground zero → fertile ground → real scale.
The individual coordinate is cultivated, not taught. The organisational coordinate is designed, not bought.
And the order is non-negotiable.