Skip to content
Robin Schreiner

UX Management · Product Leadership

← All notes

People · 5 min read

Who becomes senior when AI does the junior work?

AI is changing the first steps into UX. What happens to the path from junior to senior, and how should someone entering the field learn now?

I keep coming back to an uncomfortable question: Who becomes a senior if we stop giving juniors room to learn?

AI can produce a first draft of a research plan, a flow, a prototype, or a presentation in minutes. That is useful. It also overlaps with work we used to give to people at the beginning of their careers. The brief arrives, someone experienced asks AI to do the first pass, and a junior never gets the chance to try.

The pressure on early careers is not just a feeling. Research from Stanford’s Digital Economy Lab finds a widening employment gap for young workers in AI-exposed occupations, driven largely by reduced hiring. The authors are careful: these are descriptive signals, not proof that AI alone caused the change. The data does not tell us exactly what is happening to UX roles. But the question it raises is familiar to anyone trying to enter the field.

The work changed faster than the learning path

We expect more, sooner. When a first draft takes minutes, a week spent on the same task can look slow. The quality bar rises as well: if AI can produce ten options, why did you bring only one? That sounds like progress until we forget how someone learns to tell a good option from a convincing bad one.

Senior people are increasingly asked to set boundaries for AI work: define the problem, give the right context, write useful prompts, inspect the output, and decide what should never have been delegated. That judgment is built through experience. It comes from doing work, making mistakes, seeing the consequences, and trying again.

If we automate every early attempt, we risk creating a learning plateau. Juniors become good at producing polished work before they understand the decisions behind it. Seniors keep carrying the judgment. Eventually some of those seniors leave, change roles, or retire. Where does the next group come from?

I do not think the answer is to make juniors work without AI. I do think we need to protect the parts of the work where learning happens.

Use AI like a colleague you need to brief

AI can be a way to practise delegation before you manage a team. You have to explain the goal, provide context, set constraints, and decide what a good result would look like. Then you review the work. What is missing? What is plausible but unsupported? Would you put your name on it?

For a junior, the most useful exercise may be to do a first pass yourself and then ask AI for another approach. Compare them. Ask it to challenge your assumptions. Test both against a real user or a real product constraint. Keep track of what you missed and what the tool missed.

Sometimes do the task without it. Run the interview. Sketch the flow. Write the awkward first draft. You need enough first-hand experience to know when the polished answer is wrong. The point is to use AI for your learning, as well as for speed.

Should you still study UX?

Yes, if you are curious about people, products, and why things work the way they do. I would be more careful about studying UX as a promise of a particular job title. The tools, titles, and hiring expectations are moving too quickly for that.

Learn how to talk to users and understand what they are trying to do. Learn how products make money or deliver a service. Learn enough about technology to understand what you are asking a team to build. Learn to explain a decision and show what changed because of it. Those skills travel across tools and job titles.

The World Economic Forum’s employer survey expects AI skills to grow in importance, alongside creative thinking, curiosity, and lifelong learning. That combination makes sense to me. Curiosity is still the thing that makes you ask whether the brief is right in the first place.

Specialist, generalist, or both?

I would not tell every junior to become a researcher because research sounds harder to automate. Nor would I tell them to become a generalist who can do a little of everything but cannot judge any of it well.

Larger organisations often have room for dedicated research, accessibility, content, or design system roles. Smaller companies may need one person who can research, design, prototype, and work with developers. AI can help you move across those tasks. It does not automatically give you depth in each one.

My advice is to build an anchor and learn the neighbouring work. Maybe your anchor is research, interaction design, visual craft, accessibility, or prototyping. Get good enough at it that you can recognise quality. Then use AI to help you explore the parts around it, without pretending the tool has made you an expert overnight. That gives you a way to contribute in both a specialised team and a smaller, broader role.

If you are starting now, I would focus on a few habits:

  1. Stay close to real people. AI can simulate feedback, but it cannot replace seeing someone struggle with your product.
  2. Show your reasoning. A portfolio of polished screens says less than a clear account of the problem, the trade-offs, and what you learned.
  3. Learn to brief and review AI work. Set boundaries, check sources, test outcomes, and take responsibility for the result.
  4. Understand the product. Know which customer need and which business goal your work is meant to serve.
  5. Seek feedback from people with experience. A good mentor can show you what an AI answer missed and why it matters.

Companies have a part in this too

It is easy to ask for senior judgment and stop hiring people who have not built it yet. It is harder to give juniors real work, review it properly, and let them make small mistakes where the cost is manageable. But if we want senior people in a few years, that work cannot be optional.

The path into UX may look different now. It should still be a path. AI can shorten the distance between an idea and a result. We have to make sure it does not remove the experiences that teach us whether the result is any good.

There is another question alongside the AI one: what do we mean by being qualified for UX work in the first place? I explore that in The UX shortcut that never existed.

Keep reading

UX has a seat at the table. Now what?