A few weeks ago, I was listening to The Daily. Rachel Abrams was talking with Steven Strogatz, a mathematics professor at Cornell, about AI going from struggling with basic math to helping solve some of math’s most famous open problems.

Strogatz described a town hall in his department. Students and faculty were all there, and, in his words, they were all freaking out. Then a student at the back of the room asked: What’s the point if AI can do math better than any human being?

Strogatz didn’t give a neat answer. He said he could sit there and learn from it, but he wants to discover things too. He wants that thrill. “If you take that away from me,” he asked, “what am I doing?” By the end of the conversation, he was in tears.

I couldn’t stop thinking about that student, because I hear the same question from designers now. And the industry’s answer has settled into a comforting line: AI handles the execution, and designers bring the judgment.

I agree with it. But as someone who hires and grows designers, I keep running into the question that line skips:

Where does judgment come from in the first place?

Nobody is born with it. Most of us earned it through years of doing exactly the work AI now does in seconds.

What the boring work quietly taught me

When I think back to my first design job, I spent months on work nobody would call important: resizing screens, building component variants, redrawing the same flow for the fifth time because a stakeholder changed their mind.

At the time, it felt like boring work. It was actually training.

Every one of those repetitive tasks taught me something no course could. I learned why a layout breaks at small sizes. I learned which edge cases engineering will always ask about. I learned what my senior designers removed from my work, and over time, I understood why they removed it.

Judgment isn’t a separate skill sitting on top of craft. It’s what builds up after you’ve made a thousand small decisions and seen what happened to each one.

AI didn’t take the job. It took the practice.

Today, a junior designer can generate ten polished variations of a screen before lunch. The output looks senior. The understanding behind it often isn’t.

That’s the quiet risk. When the tool makes the decisions, the person using it never feels the consequences of making them. They skip the struggle, and the struggle was where the learning happened.

I see it in portfolios already. The screens are cleaner than ever. But ask “why this and not that?” and the answer is often “it looked right” or “that’s what it generated.”

The paradox I see nobody is planning for

Here’s the uncomfortable math. AI makes experienced designers more valuable, because judgment is now the scarce part. At the same time, it removes the path that produced experienced designers.

If companies respond by hiring only seniors, the problem gets worse. In five years, who are those seniors? Every senior designer you’ll want to hire in 2031 is a junior somebody needs to train today.

This isn’t a problem the market will solve on its own. Every company would rather let someone else pay for training. So it falls to design leaders to treat it as a design problem.

What design leaders need to do: design how juniors learn

The old way of learning happened by accident. Juniors absorbed judgment because the work forced them to. Now we have to build it on purpose. Here are a few ways to do that:

  1. Review the reasoning, not just the output. A good critique asks a junior to explain one decision in depth, not to show ten options. Execution is easier than ever. The real challenge is setting the right direction, and without direction, more iterations are just more noise.
  2. Keep some work done by hand. Not everything, and not forever. But building a flow from scratch once teaches what a hundred generated ones can’t.
  3. Bring juniors close to real users. Usability sessions, support tickets, conversion data. Judgment grows fastest when you see a decision fail in front of a real user.
  4. Share the thinking, not just the result. Seniors used to pass on judgment just by sitting next to juniors. Now that thinking needs to be visible: written principles, explained trade-offs, the decisions behind the decisions.
  5. Keep hiring juniors. It’s tempting to build an AI-assisted team with seniors only. But a team with no juniors is a team with no future seniors.

A note to the junior designer reading this

Don’t wait for your company to solve this for you. A few habits will put you ahead of peers who let the tool think for them:

  • Ask “why” about every output you keep. If you can’t explain why a generated design works, you haven’t learned anything from it.
  • Study the decisions, not the screens. When you see great work, figure out the choices behind it.
  • Seek out feedback that stings. Critique from a senior who cares is the fastest judgment you’ll ever get.
  • Get close to users. The tool has seen millions of interfaces. It has never sat next to your user while they got confused.

Use AI to go faster. Just don’t let it do your learning for you.

Back to that student’s question

“Will AI take designers’ jobs?” is a fun meme, but it’s the wrong question.

The student asked: what’s the point if AI can do the work? For a junior designer, the point is doing the work yourself. That’s how you learn, and that’s how you become a designer AI can’t replace.

But that learning won’t happen on its own anymore. Design leads have to make it happen, and that may be the most important part of the job now.

And the question is still the same: Who trains the juniors when AI does the junior work?