What happens when we stop teaching people how to work?

AI is turning everyone into an instant expert. Need to write a strategy brief? Done. Build a financial model? Easy. Design a presentation? Already formatted.

But here's what's changing: we're skipping the part where people actually learn how to do those things. When you can get the answer immediately, why bother understanding how you got there?

The problem shows up later. If no one's learning the job anymore, who's going to know how to do it five years from now?

Look around your office. You might start noticing something shifting: the mix is changing. There are still plenty of younger people, sure—but the entry-level layer seems thinner than it used to be. The pipeline feels different.

In advertising and PR, juniors now make up just 6.5% of teams. Five years ago, it was over 10%. And it's not just those industries—this pattern is creeping across sectors. We're quietly dismantling the on-ramps where people used to learn by doing.

When you cut that bottom layer, you're not just trimming headcount. You're cutting off your future.

We're losing the bottom of the ladder

Since 2022, internships and entry-level jobs in creative fields have dropped hard. Summer hiring used to jump 125% compared to January. Now? Barely 40%. Junior positions are disappearing. Middle management? Still there.

We're building companies that are all bosses and no beginners.

This isn't just creative work. Consulting, tech, finance—same pattern everywhere. Lots of people making decisions. Not enough people learning how to make them.

Some leaders see it differently

Luis von Ahn at Duolingo is hiring more recent graduates while others pull back. Why? Young people see culture shifts faster. They bring energy. They ask questions that nobody else thinks to ask anymore.

Ravi Kumar at Cognizant thinks AI should create more entry-level jobs, not erase them. The difference is in how you design those roles—to teach judgment, not just tool use.

Jamie Dimon at JPMorgan is direct about it: "We don't have a people shortage. We have a skills shortage."

The skills we actually need—creativity, critical thinking, knowing what to do when there's no obvious answer—those still come from doing the work. Not from watching a machine do it for you.

AI is incredible at answers. Not so good at questions.

AI is amazing when you know exactly what to ask for. But most real work isn't about following instructions. It's about figuring out what the right question even is.

If your workplace only rewards people who do what they're told and never push back, AI will just make everything faster and more predictable. Innovation doesn't come from predictable. It comes from people who learned to think by trying things, failing, and trying differently.

The strange part? All those skills everyone says they want—adaptability, empathy, critical thinking—you learn them in the messy early years. Cut those years out, and you cut the whole learning process.

So what do we actually do?

This isn't about creating busywork just to give people something to do. It's about building real entry points—work that AI can help with, but can't do instead of you. Work that develops how you think, not just what's on your resume.

If you're leading people: stop optimizing only for speed. Create clear paths for people to grow while they work, not just through training courses they do on their own time.

If you're starting out: being new is actually useful. Ask "why" more than "how." Stay curious, even when people act like questions slow things down.

Because if nobody's learning the basics right now, who's going to solve the complex problems in five years?

You can automate tasks. You can't automate someone getting better at their job.

The companies that understand that difference? They're the ones who'll still be here—and still leading—ten years from now.