AI does not begin by stealing jobs
For months, the public debate has revolved around one simple, dramatic and anxiety-inducing storyline: artificial intelligence will replace everyone.
It is an effective narrative because it is sharp, memorable and tied to an old fear: the fear of becoming useless.
Yet when we look at the available signals, the picture is more nuanced. Snowflake reports that 77% of surveyed organizations see AI-driven job creation, even though 46% also report job losses. The dominant movement is therefore not pure erasure. It looks more like a recomposition of roles, skills and priorities. (Snowflake)
In other words, AI does not first remove a box from the org chart. It first attacks something more intimate: our certainty. The certainty that expertise alone will protect us forever. The certainty that a job will remain stable. The certainty that accumulated experience will be enough to preserve our value.
That is where the topic becomes truly interesting.
The largest shock is not only economic. It is psychological, cultural and managerial.
What AI is really changing: the value of work
The World Economic Forum explains that this decade should eliminate some jobs while creating even more, with a positive net balance globally. Its 2025 report points to 170 million new jobs created and 92 million displaced by 2030. That does not mean everyone will benefit equally. It means the labor market is being reshaped more than erased. (World Economic Forum)
The key issue lies elsewhere: when tasks change, value changes too.
Repetitive, standardized, documentable and easily transferable tasks become more exposed. By contrast, everything related to judgment, arbitration, relationships, contextual thinking, supervision, applied creativity and decision-making gains weight.
The International Labour Organization also notes that the most likely effect of generative AI is not the outright disappearance of jobs, but job transformation through tasks, with different exposure levels across occupations, sectors and countries. (ILO)
This is a major shift.
For a long time, many professionals could monetize expertise as if it were stock: accumulated knowledge, stabilized know-how, protected by experience and recognized by the organization.
AI pushes toward another logic: expertise as flow.
A flow of learning.
A flow of adaptation.
A flow of permanent updating.
Those who cling to fixed expertise are entering a fragile zone. Those who turn expertise into adaptive capacity become more valuable.
Obsolescence is no longer a theoretical risk
One of AI’s strongest effects is making visible a fear that many organizations had kept hidden: the fear of becoming obsolete.
In my book, chapter 14, I explain that the arrival of artificial intelligence quickly activates a survival reading in people: “what does this mean for me? is my job at risk?” That mechanism matters as much as the technology itself.
That is exactly what is happening now.
When a company deploys a copilot, a conversational assistant, a domain-specific agent or advanced automations, employees are not just evaluating a tool. They are evaluating their future. They are trying to measure their relative value against a system that produces quickly, cheaply and without visible fatigue.
Fear then emerges for three reasons:
- it blurs established reference points;
- it weakens the sense of control;
- it challenges the link between seniority and legitimacy.
At that moment, the issue is no longer AI alone. The issue becomes the absence of a framework.
Why so many companies fail at AI adoption
Many organizations still believe that adopting AI means buying licenses, launching a pilot and communicating an ambition to transform.
That is far from enough.
The OECD points out that AI can improve productivity, job quality and some aspects of safety, while also creating real risks around automation, loss of agency, bias, surveillance and lack of transparency. (OECD)
In other words, success does not depend only on tool performance. It depends on how the company prepares the human environment around it.
When fear takes the wheel, teams tense up.
When teams tense up, they experiment less.
When they experiment less, they learn less.
When they learn less, adoption slows down.
And when adoption slows down, leaders too quickly conclude that “people resist change.”
In reality, people often resist uncertainty that has been poorly managed.
Microsoft also notes that the organizations moving fastest are the ones that clearly expect AI literacy for everyone, organize continuous learning and help employees embed AI into daily work. (Microsoft)
The lesson is straightforward: buy-in does not come from an announcement. It comes from clarity, meaning, training and permission to learn by doing.
The company that protects its people does not hide AI from them
Some leadership teams believe they are protecting employees by downplaying the issue.
They avoid candid conversations.
They postpone training.
They let everyone experiment in isolation.
They reassure vaguely.
They hope to buy time.
That is a strategic mistake.
A company protects its people when it gives them a clear reading of the moment:
- what is changing,
- what is not changing,
- what will be automated,
- what will be augmented,
- what must be learned,
- what will remain deeply human.
PwC shows that the jobs most exposed to AI are still growing, while the skills required in those jobs are changing much faster. The challenge is therefore not only to keep a role, but to remain relevant inside a role that is evolving. (PwC)
That is why the smartest organizations do not hide AI. They make it understandable. They explain it. They train people. They test. They support.
They replace diffuse fear with visible capability-building.
Tomorrow’s decisive capital: learning faster than uncertainty
The real divide will not simply be between people who use AI and people who do not.
It will be between:
- those who learn fast,
- those who wait to feel safe before moving.
IBM reports that executives expect large-scale reskilling needs tied to AI and automation in the coming years. (IBM)
That matters because it changes the role of management.
Training is no longer a nice HR benefit.
Training becomes a strategic responsibility.
And learning is no longer a peripheral effort.
Learning becomes a condition for maintaining value.
In my book, chapter 4, I argue that innovation is not reserved for a few heroic profiles: it is everyone’s business, without exception. That idea becomes even stronger with AI. Adoption cannot rest on an innovation team, an IT department or a few enthusiasts. It has to spread across the whole organization.
The winning company will not necessarily be the one that deployed the most tools first.
It will be the one that built the best collective learning system.
AI reveals strong organizations… and the others
AI acts like a revealer.
It reveals the companies that already know how to:
- share a vision,
- experiment without punishing,
- circulate knowledge,
- develop skills,
- decide faster,
- create psychological safety.
And it reveals those that were living on illusions:
- frozen expertise,
- comforting silos,
- control-based management,
- cosmetic innovation,
- peripheral learning,
- vague communication.
In those structures, AI does not create the problem. It accelerates it. It shines a light on weaknesses that already existed.
That is why the debate “will AI steal jobs?” is too narrow.
The deeper issue is this: does your organization know how to turn fear into capability?
Because in the end, AI first steals our certainties.
And that may be very good news for organizations ready to learn again.
Turning fear into value
Fear is not shameful. It is human. It can even be useful when it pushes us to look clearly at what must evolve.
The danger begins when fear is denied, poorly named or left without structure.
At that point:
- employees defend territory,
- managers slow things down without saying it,
- leaders overplay confidence,
- the organization loses time,
- competitors learn faster.
By contrast, when a company addresses the issue with maturity, AI can become a very concrete value lever:
- better execution quality,
- time gains,
- skill growth,
- role redesign,
- broader innovation capacity,
- better-prepared decisions.
AI does not only remove tasks.
It redraws the map of value.
And in that reshaping, the winners will not always be the most technical people.
They will often be the clearest thinkers, the most adaptable and the best supported.
👉 In your organization, is AI creating more fear… or more value?
That is exactly where my keynotes, advisory work and workshops can help: naming the fears, clarifying the stakes, accelerating useful adoption and turning diffuse anxiety into a path of progress.
References
(Snowflake) = https://www.snowflake.com/en/news/press-releases/snowflake-research-reveals-ai-driven-job-creation-outpaces-job-loss-with-77-percent-reporting-workforce-gains/
(World Economic Forum) = https://www.weforum.org/publications/the-future-of-jobs-report-2025/
(ILO) = https://www.ilo.org/publications/generative-ai-and-jobs-2025-update
(OECD) = https://www.oecd.org/en/topics/ai-and-work.html
(Microsoft) = https://www.microsoft.com/en-us/worklab/work-trend-index/2025-the-year-the-frontier-firm-is-born
(PwC) = https://www.pwc.com/gx/en/services/ai/ai-jobs-barometer.html
(IBM) = https://www.ibm.com/think/insights/new-ibm-study-reveals-how-ai-is-changing-work-and-what-hr-leaders-should-do-about-it
