Another Sunday reflection. This follows on from my recent article on AI agents, the changing nature of work and the advantage this gives senior professionals.
What happens to junior jobs when AI can already do some of the work we used to give people at the start of their careers? And how can they gain experience if they are no longer given the opportunity to learn their trade? We are told that young people need AI training. But is that really the answer? I think we first need to ask what work will look like tomorrow, and where each person’s real competitive advantage will lie, before teaching them how to write prompts.
AI and employment: why senior professionals have an advantage
Why does this give them an advantage? Because they hold the keys to the system. They have the experience and perspective to move things forward. A junior employee still needs a few years to learn their trade before their thinking can begin to shape their work or their company.
In software development, this is something we can see in hiring figures: recruitment of senior developers is rising while recruitment of juniors is falling. I recently spoke to a managing director who told me he was actually planning to replace junior staff with senior people. Honestly, I do not think we have seen anything quite like this before.
In the digital sector, many roles and tasks can already be handled, more or less, by AI under a manager’s supervision.
Does time saved in one place get lost somewhere else?
But we should not assume that the equation balances out.
What we are seeing increasingly is managers overwhelmed by AI-generated work. They spend more and more time checking, correcting and improving what the systems produce. Time saved in one place may be lost somewhere else. Some businesses are learning this the hard way, and some are hiring again to make up for it.
Yet here, too, senior professionals have the advantage. Junior staff do not yet have the experience and know-how to assess the work AI produces.
So what do we do about graduates entering a labour market that seems to have little use for them? I have a very concrete example of this at home.
What if we simply talked to our software?
I firmly believe that most young professionals still need to master the tools of their trade, whether they work in design, development or another field. But let us be honest: they may be the last generation to need to do so in quite this way.
As someone whose posts I enjoy recently suggested on X, it is quite possible that, in the coming months, we will stop using our production software directly and work through a chat interface instead. To tell you the truth, that is already how I work.
Excel is becoming almost irrelevant to me as a tool I operate myself. It is not that I no longer use it. AI uses it on my behalf, and I no longer need to know every step it takes inside the application. I think the same will happen with the rest of our software. We will simply talk to it as we would to another person, and that will feel natural.
And please, spare me the obsession with “prompting” and “knowing how to prompt”. We are talking about AI systems that can understand most of our instructions without needing an elaborate explanation of the context every time. If you have not tried ChatGPT Work, I would encourage you to give it a go. In my experience, it can do an astonishing range of things.
Thinking, inventing and solving problems: the skills employers will need
The labour market will increasingly look for people who can think, devise solutions and solve problems. That poses an enormous challenge. Not everyone will find that easy, and I am not convinced that schools and universities are ready for it. That is my impression, at least: I do not claim to know what every institution in France and Europe is teaching.
A few months ago, I would not have believed it. Now I think we are genuinely approaching a major shift in service-sector work and what I loosely call “office jobs”, for want of a better term.
As so often happens, those who are best equipped intellectually are likely to fare best.
Experience and broader knowledge as competitive advantages
Experience is going to become a huge advantage. Good news for senior professionals, I hope.
As I have argued before, broader knowledge will matter too. By that I do not just mean cinema, theatre or literature, although a little more of those would do us no harm. I mean an understanding of economics, engineering, communication, marketing, sales and entrepreneurship.
In short, I suspect that, with a few exceptions, the people who benefit most from this revolution will be the most versatile: those who can connect ideas and knowledge across different disciplines.
AI and junior jobs: which skills should young professionals develop?
So I still have not answered the question: what should we do about junior professionals?
If I were speaking to someone just finishing a software development course, I would encourage them to take an interest in design, UX, marketing, communication and business. I would tell them to understand how a company works: its moving parts, its priorities and the way it operates.
I would encourage them to develop skills in neighbouring disciplines seriously. They do not need to become specialists in everything, but they should aim to become very good generalists.
I may be wrong. Still, I remain convinced that the more people add related knowledge to their core expertise, the better placed they will be to adapt and find new opportunities.
There are no insurmountable barriers to this, but we need to recognise its importance. When I hear people insisting that the answer is to train thousands of young people in AI, I want to tell them they are missing the point entirely. Unless they mean engineering or research, which I suspect they do not.
We need young people who know how to think, not simply how to prompt. Those are two very different things.