If AI does more of the doing, how do inexperienced people become experienced?

AI creates an obvious productivity opportunity: experienced people can delegate more of the doing to increasingly capable systems.

That is valuable. But some of the work we are most tempted to automate was doing two jobs at once.

It produced an output, and it developed the person doing it.

Junior people learn partly through repetition, mistakes, debugging, awkward first attempts, feedback and exposure to work that more experienced colleagues can already do much faster. The task may look inefficient when judged only by today’s output, while quietly contributing to tomorrow’s capability.

Some of what we’ve been calling low-value work was also, quietly, apprenticeship.

AI makes that tension sharper because it can allow people to produce work requiring more expertise than they currently possess. A relatively inexperienced person may be able to create something sophisticated with AI without independently possessing all of the knowledge that would previously have been required to produce it.

That is an extraordinary opportunity for learning. It is also possible to skip parts of the process through which judgement used to develop.

The distinction matters because producing an answer and knowing whether it is a good answer are different capabilities.

As AI takes on more execution, people may need more judgement rather than less: the ability to frame the problem, direct the work, challenge an answer, recognise when something is subtly wrong, understand consequences and know when not to accept what the system has produced.

The AI’s ability to produce software is increasingly greater than my ability to independently produce the same software.

That is already visible in self-taught building. AI can enable someone to create systems beyond what they could have coded unaided. Used actively, that can accelerate learning because the person can interrogate the output, test ideas, explore alternatives and work at a level that would previously have been inaccessible.

But simply receiving a better output does not guarantee that the person develops the judgement behind it.

This turns an automation question into a workforce-design question.

If organisations remove large amounts of routine work, they cannot assume the old development path will somehow continue unchanged. They may need deliberate alternatives: structured exposure to difficult decisions, review of AI-generated work, simulation, mentoring, explanation, challenge and opportunities to take responsibility with appropriate support.

We need to replace the learning, not just remove the work.

There is a further organisational tension. The most productive way to complete today’s task may not always be the best way to develop the person who needs to handle tomorrow’s much harder task.

That does not mean preserving inefficient work for its own sake. It means recognising that organisations produce more than outputs. They also produce capability.

We can’t just keep hiring experienced people forever. Somebody has to make them.

This becomes more important as software and other forms of execution become abundant. If experienced people can orchestrate far more work through AI, their judgement becomes highly leveraged. But the organisation still needs a route by which new people acquire that judgement.

AI may be part of the answer as well as the cause of the disruption. A system that can do the work can also explain it, generate exercises, challenge reasoning, simulate scenarios, review decisions and adapt learning to an individual’s level. The opportunity is not to recreate every old apprenticeship task exactly as it was, but to understand what learning it provided and design better ways to provide that learning.

The important question is therefore not whether AI should do the routine work. Much of it probably will.

The question is what must replace the experience that disappears with it.

We’ve spent a lot of time asking how AI can make experienced people more productive. We may need to spend the next few years asking how inexperienced people become experienced in the first place.