There is something rather dizzying about the times we are living through.
Every day, I discover possibilities that would have seemed almost absurd just two or three years ago. Complex things become accessible. Technical barriers disappear. Tasks that once required rare skills, a great deal of time or substantial resources can now be completed in a few minutes.
And yet, when I look around me, I sometimes feel that nothing is really changing.
That is probably the paradox that strikes me most today.
Technology is advancing at an extraordinary pace. Our ability to adopt it, to change ourselves and to question our habits is moving much more slowly.
We still too often confuse the availability of a tool with our ability to do something useful with it.
Giving everyone access to artificial intelligence does not automatically give them the curiosity to explore it, the judgement to assess its answers, or the patience to try again when it fails.
Access to technology is becoming more widespread. Mastery of it, much less so.
But I think the real issue lies elsewhere.
When a technology becomes capable of doing part of our work, it does more than change our methods. It also touches something much more personal: the place we occupy, the value we believe we bring and the way we see ourselves.
To varying degrees, we have all built our legitimacy around skills, knowledge and experience, sometimes acquired over decades.
Being the person who knows. The person others consult. The person who holds information, understands a process or can solve a problem that others do not understand.
And suddenly, a tool makes some of that knowledge available to everyone.
How could we not understand that this might be unsettling?
A complicated procedure can be inefficient for an organisation while remaining reassuring to the person who has mastered it. Information that is hard to obtain can be a real source of power. Expertise can become an identity that is painful to relinquish, even in part.
This is where the subject becomes fascinating.
AI does more than transform our relationship with work. It questions our relationship with usefulness, status and power.
And I think it would be too easy to dismiss resistance as nothing more than a lack of open-mindedness.
Some fears are entirely rational.
When we ask someone to share knowledge built over twenty years, without explaining what their place will become once that knowledge is available to everyone, we should not be surprised if they hesitate.
Likewise, promising productivity gains without saying who will benefit leaves an essential question unanswered: will this new efficiency help us work better and learn more, or will it simply mean asking for ever more from fewer people?
These are legitimate questions. Ignoring them would be a mistake.
I must also acknowledge my own impatience with this situation.
When you experiment with these tools every day, when you can see in concrete terms what they make possible, it sometimes becomes difficult to understand why adoption is not moving faster.
You want to demonstrate, persuade, accelerate.
But a successful demonstration does not change a culture. A high-performing application does not spontaneously alter long-established habits. And a promise of efficiency is not enough to dispel the fear of becoming less necessary.
We must also accept that some objections are well founded: a system can be impressive while still being unreliable, poorly integrated or unable to account for the complexity of real work.
This is probably one of the most important lessons I take from this period: being right about a technology’s possibilities does not mean understanding the conditions for its adoption.
Transformation cannot simply be decreed. It is built with the people concerned, through an understanding of their actual work, constraints, interests and concerns.
Above all, it rests on something no technical demonstration can produce instantly: trust.
Trust does not emerge from a compelling presentation. It is built over time, through consistency between words and actions, through the ability to acknowledge a mistake, honour a commitment and listen to an objection without immediately trying to brush it aside.
It also requires a kind of honesty: being willing to say what we do not yet know, what the tool cannot do, and what might genuinely change for those who use it.
We talk a great deal about models, performance, automation and autonomous agents.
Perhaps we still talk too little about what it means, on a human level, to accept that we may no longer be indispensable in the same way.
Behind this technological revolution lies a question that reaches far beyond AI:
Can we change our understanding of our own value without experiencing every advance as a threat?
Perhaps we will gradually need to learn to draw less recognition from possessing an answer, and more from our ability to ask the right questions, exercise judgement, pass on knowledge, make decisions and help others progress.
Yet here too, I would be wary of turning that into another certainty. Nothing guarantees that these abilities will remain forever beyond the reach of machines.
What seems more solid to me is to distinguish between the ability to act, the responsibility to decide and the trust needed to act together.
An organisation can have the best tools in the world and still be unable to make full use of them if everyone feels threatened by what others might now be able to accomplish.
Conversely, a culture in which knowledge circulates, experimentation is encouraged and people can question their own methods without losing their legitimacy already has a considerable advantage.
I am deeply convinced of AI’s potential. But increasingly unconvinced that its main difficulty will be purely technological.
We will probably continue to see extraordinary capabilities emerge at a pace we will struggle to get used to.
The hardest part may be learning to live and work with them, without confusing our value with being irreplaceable.
Real transformation will not begin when machines become capable of doing more. It will begin when we agree to change what we do, what we share and the space we leave for others.
And that could take much longer than expected.