I am currently writing my doctoral thesis on AI and employment. Its scope requires choices, and one important issue will remain outside it. Yet I believe it deserves urgent attention.
Much of the debate about AI and employment focuses on numbers: how many jobs will disappear, how many will be created, and how productivity will change.
A less discussed risk is more structural.
What happens when young people can no longer begin the journey that turns inexperience into expertise?
The Barbell Company
In my view, the ideal company has something of a barbell shape.
At one end are young people: inexperienced, sometimes naive, full of energy and eager to learn. They may not yet possess refined judgement, but they contribute effort, curiosity and fresh questions. Their value is not limited to what they know today. It also lies in what they can become.
At the other end are highly experienced professionals. They remain equally committed, but their contribution is different. They recognise patterns, anticipate consequences and navigate problems with a calmness earned over many years. They have learned to surf uncertainty rather than be overwhelmed by every wave.
Between these two ends are people travelling from one side to the other.
That journey is how companies renew themselves. Juniors become managers. Managers become experts. Knowledge is accumulated, tested and eventually passed to the next generation.
AI could interrupt this movement.
When Efficiency Breaks the Journey
Many tasks traditionally assigned to junior employees are precisely the tasks AI can perform most easily: basic coding, preliminary research, document preparation, routine analysis or administrative coordination.
Automating them may appear entirely rational. But a task can have limited value as output while still having enormous value as education.
A recent study of US software vacancies found a 14–15% relative decline in junior versus senior developer vacancies following the release of ChatGPT. It also found that the remaining junior positions increasingly demanded problem-solving and communication capabilities.1 PwC’s 2026 analysis identified a similar paradox: highly AI-exposed entry-level roles were seven times more likely to require skills traditionally associated with seniority, including judgement and leadership.2
We may be asking young people to arrive with judgement while removing the experiences through which judgement is formed.
The danger is not only that fewer juniors will enter the labour market. Those who do enter may travel a shorter and less complete road towards seniority. In ten years, organisations could discover that they have plenty of AI-generated answers but too few people capable of recognising when those answers are wrong.
From Informal Learning to Deliberate Apprenticeship
This challenge is particularly visible in the two sectors I know best: tourism and technology.
In technology, expertise grows through debugging, failed deployments, difficult clients and imperfect systems. In tourism, it develops through live operations: an unhappy guest, an absent colleague, a disrupted journey or a hotel operating under pressure.
AI can suggest a solution. It cannot give a young professional the memory of having lived through the problem.
This does not mean companies should preserve every repetitive task. It means they must replace accidental learning with deliberate learning.
Businesses will need intensive in-house programmes that combine mentoring, rotations, supervised responsibility, exposure to real decisions and the disciplined critique of AI-generated work. Senior employees must be recognised not only for solving problems, but also for teaching others how to think through them.
Learning can no longer be treated as an optional HR benefit. It must become a strategic capability discussed by CEOs and boards.
A conscious leader asks more than, “What can we automate this year?”
The deeper question is: “Who are we preparing to lead this company ten years from now?”
The organisations that answer that question well will not have to choose between technology and people. They will use technology to accelerate the journey from energy to expertise, while ensuring that the journey still exists.
Conscious Leadership Tip #36:
When you automate a process, think beyond the tasks you are eliminating and consider the skills your team may stop developing as a result. Identify the capabilities that could disappear with automation and create a learn-by-doing pathway, supported by traditional training, so that critical knowledge remains inside the organization. A skill can appear unnecessary today and prove valuable later—much like studying Latin may seem irrelevant to a student, yet provide a useful foundation for understanding law and the origins of a legal system.