AI in fashion education: if AI does the entry-level work, where will fashion’s next generation learn judgement?

There is a growing conversation about what career readiness means when AI can already perform some of the work we once expected students and junior employees to do.

Much of that conversation focuses on skills: which skills will still matter, which tasks will disappear, and how quickly education needs to adapt.

But I think there is another question underneath it.

If AI does more of the entry-level work, where will beginners develop the judgement that used to come from doing that work?

For much of my career in fashion, experience accumulated almost accidentally.

You researched the market. Prepared the figures. Sat in meetings. Made relatively small decisions before you were trusted with larger ones. You watched somebody more experienced notice something you had missed. Eventually you made the wrong call yourself and discovered, sometimes painfully, why it was wrong.

The task was producing an output. But underneath it, you were learning how to think.

That distinction matters now.

The work we automate may also be the work we learned from

An experienced professional can use AI to accelerate work because they already have a mental model against which to judge the result.

They know when something feels incomplete. They know which assumption needs questioning.

They can recognise that the technically correct answer is commercially useless.

A beginner does not have that advantage yet.

If the first draft, first piece of research, first analysis and first set of options are increasingly produced by AI, we may save people a great deal of repetitive work.

But some of that repetition was also practice.

This does not mean we should preserve inefficient tasks simply because previous generations learned through them. Work has changed many times before, and education has changed with it.

The more interesting question is: what was the old system teaching indirectly, and how can we teach that deliberately now?

The valuable skill may begin before the task

One idea in the current future-of-work conversation particularly interests me: the distinction between completing work someone else has defined and being able to recognise what is worth doing in the first place.

That starts to look less like traditional employability and more like entrepreneurial thinking.

Not entrepreneurship in the sense that everybody should start a company.

I mean the ability to notice a problem, an opportunity or a connection that is not already written into the brief.

Businesses are full of these moments.

Why is this product selling in one market but not another? Which customer are we overlooking?

What happens if we change the price? Is the problem really marketing, or is it positioning? Is this process inefficient because nobody has stopped to question it?

The people who become commercially valuable are often not simply the ones who execute instructions well. They are the ones who begin to see things other people have not yet framed as a problem.

That is extraordinarily difficult to teach through explanation alone.

But perhaps it can be practised.

What if AI created the problem instead of the answer?

This is where I think the conversation about AI in education can become more interesting.

We understandably worry about receiving thirty AI-written essays from thirty students.

But generative technology could also be used on the other side of the assignment.

Instead of asking every student to solve exactly the same fictional business problem, what if we could give them different ones?

At the same decision point, one learner might choose one market position, another a completely different one. From there, the businesses begin to diverge: different customers, different pricing logic, different distribution choices, different risks. The next problem each learner encounters can emerge from the decisions they have already made.

The learning objectives can remain the same.

The problems do not have to.

Perhaps one of the most useful roles for AI in education is not to give every student a better answer.

It is to give every student a different problem worth solving.

Then assess the consequences

That also opens up a different way to think about assessment.

In business, you are rarely judged because your answer matches the one at the back of the book.

You make a decision and something happens.

Sales rise but margin falls. A campaign brings customers who do not return. A product works in one market and fails in another. A premium positioning creates desirability but limits volume. A choice that looked reasonable produces an unexpected result.

The useful question then is not simply Were you right?

It is:

Why did this happen? What did you miss? What would you do differently now?

That is much closer to the commercial judgement I think fashion students need.

And it creates space for different students to succeed in different ways. A luxury business should not necessarily be judged against the same sales-volume target as a mass-market one. The assessment is partly the business consequences their decisions produce, and partly whether they can understand and explain those consequences.

I have written before about commercial awareness as the ability to connect decisions across money, timing, customers and risk, rather than simply knowing industry vocabulary. I have also argued that entrepreneurial thinking matters well beyond entrepreneurship itself: people working inside established fashion companies still make decisions with incomplete information, constraints and uncertain outcomes.

AI makes both questions more urgent.

Practising judgement before the stakes are real

This is increasingly how I think about the longer-term direction of Sim de la Mode.

The missions available today begin much earlier: learners first need to understand how the fashion industry works and start making decisions about the virtual label they are building.

But the direction is toward something more dynamic.

Different brands. Different decisions. Different problems emerging from those decisions. And eventually, commercial consequences that force the learner to understand not only what they chose, but what that choice did to the business.

Simulation cannot reproduce years inside a real fashion company.

It should not pretend to.

But if some of the junior work through which previous generations developed judgement is changing, perhaps education needs to create more places where that judgement can be practised deliberately.

AI may remove some of the work we once learned from.

It may also give us much better tools for recreating the part that mattered most: having to notice, decide, and live with what happens next.

---

Ekaterina Adamovich spent twenty years in fashion buying, wholesale, consulting and education before founding Sim de la Mode — an iOS learning game where players learn how the fashion industry works by building their own virtual fashion label. The first two learning missions are now live, and Sim de la Mode is ready to try in the classroom. Educators who would like to explore it with their students or discuss a small pilot can get in touch.