Job

How AI changes the skills required and the training of future professionals

AI is compressing traditional learning times, eliminating the apprenticeship phase and demanding a new training model based on soft skills, critical thinking and continuous learning

by Luigi Riva*

19 March 2026

(Adobe Stock) VideoFlow - stock.adobe.com

3' min read

Translated by AI

Versione italiana

3' min read

Translated by AI

Versione italiana

A few weeks ago a young colleague asked me what the deadline was for a job. He proposed two days. I replied that I would not accept it for a week. "Why?" he asked me? 'Because of AI,' I replied. Not because technology is a problem in itself, but because it threatens to compress the time needed to learn.

According to the CEO of Microsoft IA, Mustafa Suleyman, in less than 18 months the world will reach full automation of clerical tasks. We will see what will happen in the labour market, in processes and organisations, but that something has changed in education and in particular in the training of professionals is already a certainty. Before AI, it was slower and sometimes a job was assembled and disassembled countless times, everything by hand, from the colour of the slides to checking for typos, which took hours. It was a method for internalising quality criteria and critical thinking, for learning to recognise fragile reasoning: in short, for professional growth. It was our gavetta. Which is no longer there, at least in that form.

This is not a matter of nostalgia, but of understanding what the training of tomorrow's professionals will be. A question that concerns managerial positions, specialists and technicians. Because if 92 million clerical positions will gradually disappear by 2030 (but not only, because HR, legal, media and other will also be involved), 170 million jobs will be created in new fields - specialists in Big Data and AI, fintech engineers, sustainability and ecological transition experts - with a positive balance of 78 million new positions (World Economic Forum data).

Human skills that cannot be replicated by AI

From a different perspective, Oxford University argues something similar: if by 2035 almost half of the traditional jobs could be replaced by AI, human skills such as empathy, creativity and strategic thinking will remain essential and difficult to be replicated by machines. We will see if the numbers are confirmed, but today AI can do many things; however, it has no strategic thinking, no vision, no imagination, no judgement, no culture, no ethics, it does not connect the dots, it does not think 'outside the box'. So, now that AI has eliminated the need for apprenticeships, how do we train the seniors of tomorrow?

Of course, competence building is a central and much discussed issue. If students use AI to the full - and they do - university faculties cannot turn a blind eye. They need to update and revise their study paths. It is also well known that companies and workers need to invest in reskilling and upskilling, in lifelong learning that supports growth paths and the development of transversal skills. And this must be done by favouring interdisciplinarity and lateral thinking. I could hazard a prediction, but I hypothesise that faculties of Philosophy will return very relevant in the selection paths of companies.

However, soft skills such as critical thinking, problem solving, emotional intelligence, collaboration or strategic vision are not always learned in books. Studying is essential, but having a teacher is sometimes even more so. That is why it is important that today's senior not only corrects results and provides guidance, but also raises doubts and reflections, asks for reasons, discarded alternatives, and criteria for choice. Hoping that today's junior will in turn be able to become tomorrow's senior teacher.

In any case, companies must be open, emotional, experimental, and it is therefore essential that they have an 'Agile' organisation, as has already been written on these pages. Today, AI is forcing us to redesign the perimeter of traditional roles, career expectations, growth stages, progression, but also the required skills and internal training. The old parameters must be abandoned. For example, in a logic of HR in the time of AI, ambition, motivation and learning skills can be more functional than a deep and rich track record. If implicit training is reduced, managerial growth is no longer an almost automatic effect of experience.

International analyses (OECD, WEF, etc.) agree that the central point is the formation of judgement. A generation accustomed to powerful tools can be very efficient in production, less trained in supporting a decision under conditions of uncertainty. Production becomes simpler; evaluation remains complex. If the first phase is drastically shortened, the second risks not receiving the necessary space. This is why one must be wary of any work delivered too quickly. And, as with any complex problem, action is needed on several levels (Agile organisation, pixelisation, continuous training, universities, new career paradigms). A multi-level approach that AI could not have.

*President Strategic Management Partners