A junior designer used to spend three days creating fifteen layout variations for a landing page. It was a slow, sometimes agonizing process, but each iteration taught them something. The third version revealed that the hierarchy wasn't working. The seventh exposed an inconsistency in the navigation pattern. By the fifteenth, an intuitive sense of what worked and why - "taste" was developing.

Today, that same designer can generate fifteen variations in fifteen minutes.

The output is faster, and the quality is often higher. But something essential is missing: the learning that used to come from the struggle. 

This is the central paradox of the AI era. The tools that make work easier simultaneously make growth harder. And most people don't realize it until they've spent a year producing impressive outputs while developing almost no judgment.

The Inefficiency Was the Point

The traditional entry to any profession was brutally repetitive. You wrote tickets, fixed bugs, created mockups, and updated documentation. It was boring.

But that repetition was building something invisible: pattern recognition. You saw problems solved in different ways. You learned which approaches worked in which contexts. You developed intuitions about what would break and what would hold.

These aren't about seniority in the traditional sense. They're about cognitive work. And in the age of AI, the gaps between the levels are compressing faster than ever.

AI has removed the inefficiency. And with it, the learning.

Fast Output, Slow Development

Consider what happens when a junior PM uses AI to draft a PRD in ten minutes instead of two days.

In the old model, those two days were painful. You'd write a section, realize it was vague, and rewrite it. You'd outline success metrics, discover they didn't measure what you wanted, and start over. Each micro-failure taught something.

With AI, the well-structured, comprehensive, and articulate PRD appears in minutes. But you didn't struggle with the logic. You didn't catch yourself making assumptions. You produced the output without doing the thinking that the output was supposed to develop.

The pattern appears everywhere. The researcher who uses AI to summarize fifty transcripts misses the deep familiarity that came from reading every word. The developer who asks AI to write a function skips the debugging that teaches how errors propagate. The designer who generates dozens of mockups instantly never develops the taste that comes from slowly exploring why one approach felt right.

In every case, output is faster. But learning is slower.

Why Learning Without Context Doesn't Stick

I’ve received countless pieces of good advice and read hundreds of insightful articles. Each time I think "yes, that's exactly what I should do," then a week later, I’m doing the same things I've always done.

My favourite metaphor for this is to think of professional development like a tree. 

The leaves are tips and best practices.
The trunk is the principles and mental models that make up the foundational understanding. The branches are patterns built through experience.

You can't skip to the leaves. Without the trunk and branches, advice has nothing to attach to. It sounds compelling, but doesn't stick.

AI accelerates this problem. It lets you produce impressive work (the leaves) without building the underlying structure. You look experienced from the outside but the foundation isn't there.

Building Judgment Deliberately

Since work no longer trains you automatically, you have to create learning experiences intentionally.

Study systems, not tasks. What matters now is understanding why things work, not just how to operate them. For researchers, this means learning evidence triage, not just interview techniques. For designers, constraint-based decision-making, not just the latest tool. For developers, it is system design patterns rather than frameworks. These are ways of thinking that apply regardless of which AI you're using next month.

Build things and reflect on them. Ship products and get them in front of real users. AI makes this easier than ever because you can ship in weeks what used to take months. But because AI handles execution, you need to be intentional about extracting lessons. After each project: What worked? What didn't? Where did AI sound right but miss something?

Seek out visible reasoning. You can't develop judgment in isolation. When a senior designer says, "This flow doesn't work," ask them to walk you through how they got there. Use AI to challenge your thinking and identify gaps. This keeps your brain in the driver's seat.

The trap is staying in your functional home. If you only ever see the design perspective or the engineering perspective, you'll only ever see part of the picture.

The Effort Is Still Required

The old model was passive. You showed up, did the work, and learned by accident.

The new model is active. It requires you to decide what to study, to create your own variety in projects, and to seek out mentors who will explain their reasoning rather than just their conclusions.

The bottom line is that while AI makes work easier, it makes growth harder. Be intentional about how you go about it.

This is the second article in a series exploring how AI is reshaping career progression. The next article examines the impact on the organisation chart as level 1 jobs are being automated, and the impact on hiring and onboarding. You can read the full details in our free ebook: Managing your Career in the Age of AI.