You may have been told that “the land is disappearing.” That no one wants these jobs anymore. That they would be less sexy, less bankable, almost condemned by tech.
You may have also been told that training field employees is risky: “If you train them too much, they will leave.”
These claims have a problem: they don't stand up to the test of facts.
In France, between 50 and 60% of employees work in an operational profession. At the European level, this represents nearly 120 million people. And these women and men are at the forefront of all the major transitions we are going through: energy, industrial, demographic, logistical, ecological, strategic. In other words, everything that is presented as a priority is based, very concretely, on skills in the field.
The paradox is there, and it's awkward. This population is at the same time a majority, essential... and historically underformed.
Initial training has long been seen as a secondary path. Continuing education has often remained anecdotal. Little time, few adapted tools, little recognition. The field was asked to carry out major transformations, without always giving it the means to do so.
So the real question is not “does the land still count?” It matters more than ever. The question is: Have we really given ourselves the means to train him up to the challenge?
What we tried. And why that wasn't enough.
For years, technology has been used to respond. Training has been digitized. It has been made mobile. It was cut into micro-modules. And it was essential. These developments have made it possible to open up access, to gain in flexibility, and to bring training closer to daily life.
But despite this progress, a gap persists. Why? Because the problem in the field has never been the lack of content. The field knows. What he needs is help when the situation really arises, not three weeks before, in a training room.
What AI is changing. Really.
What AI brings today is more profound than a new tool in the stack. In a few years, we have gone from an essentially conversational AI to agentic logics. From general tools to specialized systems by use. From great models who know how to say everything to more targeted, more sober, action-oriented models.
When AI systems solve complex scientific problems, it's not because they are better at formulating sentences. It's because they know act within a specific framework. Applied to the field, it opens up something new: we no longer disseminate knowledge, we support situations.
It is a fundamental difference. On the ground, AI is not a content engine. It can become a driver of situations : simulate, have them repeated, correct in real time, without judgment, at the exact moment when the action is being played. And that is precisely why it fits operational constraints: little time, pressure, decisions to be taken now.
The field has always learned differently. It is a strength, not a defect.
Do for the sake of learning, rather than learning for the sake of doing. Observe, repeat, adjust. Rely on a peer, a local manager, an itinerant trainer who knows the context. These models work. They have always worked.
The problem has never been their effectiveness. It's their ability to scale. Key people become bottlenecks. Time is running out. Situations are multiplying. Even social learning — which is extremely powerful — has sometimes been thought of as a side feature, a dedicated space, separate from everyday life. However, its strength comes precisely from where it is born: in action, in spontaneous exchange, in shared gesture. When it's too far removed from the real flow of work, it loses some of what makes it so effective.
Well-designed AI agents don't replace these human roles. They in extend support capacity, by encouraging practice, by correcting in real time, by adapting to the context, by remaining available where humans cannot always be, whether for time, geographic or linguistic reasons.
Train en masse, while remaining personalized and close. That is change.
The status of the training is changing.
It ceases to be a catalog of modules to become a Lever of execution.
Many organizations have a clear strategy: improving the customer experience, launching a product, securing quality. But transformation often gets stuck between the decision and its concrete implementation on the ground. This “last mile” is the most critical. And this is where many initiatives fail, not because of a lack of vision, but because of a lack of accompanied execution.
So the subject is not only about producing content. It's making sure they become practices. This means start with much more operational questions : what concrete behavior are we trying to change? What real situation do we want to secure? What problem are we really trying to solve?
A customer welcome that varies from one site to another. A sensitive business gesture. A critical procedure that is poorly applied under pressure. It is from these situations that everything must be designed, through observation of the field, the analysis of real practices, the support of reliable sources: procedures, product sheets, business standards. AI then makes it possible to formalize this knowledge in formats that can be activated much more quickly than before.
And the time saved in design is not used to produce more. It is used to support practice: to define situations, to work on postures and behaviors, until competence becomes a reflex. This is where operational excellence comes in.
What that says about organizations, basically.
This movement goes beyond training. It affects governance itself.
For decades, technology has especially strengthened the office: more tools to analyze, manage, and decide. The field was left relatively alone in the action. But there is something more profound to name here. For a long time, knowledge was seen as a competitive advantage... to be protected. The more analysis, strategy and management tools were concentrated at headquarters, the more centralized the power remained. And in the meantime, the terrain was working. Point.
What AI is doing is much more interesting than a simple debate about employment. It's not about massively destroying jobs. It's from rebalance the value between those who decide... and those who turn those decisions into reality. And so Give operational action back the place it deserves in value creation.
Tomorrow, there won't be a generic AI agent. There will be a constellation of specialized agents, aligned to specific business contexts. The value will no longer be in the accumulation of software, but in the ability to finely understand its realities on the ground, and to translate them into truly useful devices.
So the challenge is not to have more AI. The challenge is to no longer separate the strategic from the operational. The organizations that will succeed tomorrow will be those that have understood that performance is built in the exact place where decision meets action.
And that this place is called, quite simply, the land.
