How AI Is Reshaping the Career Apprenticeship Model

What happens when AI changes the apprenticeship model?

Lately, most of the early-career professionals I meet aren't through work. They're through volleyball open gyms in NYC.

Since my company isn't really hiring early-career professionals, it's one of the only ways I'm meeting people a few years out of college. Some are in investment banking, which was my first job 25 years ago. So I've been asking them what's changed — how the job feels today, how teams operate, and how much AI is actually being used inside large, established institutions.

Adoption is still uneven. The biggest firms are moving slower than startups and tech-enabled growth-stage companies, for good reasons. But even there, you can feel where things are going.

And it got me thinking about how different the career development arc may become.

For most of modern professional history, early careers followed a pretty clear pattern: you owned the work before you owned the outcome. You learned under the guidance of more experienced people. You had to learn how to get it done before you could be responsible for the result. The outcome ultimately sat with someone more senior.

The Rookie

I think back to my first job as an investment banking analyst a quarter century ago. It was a 90-hour-a-week apprenticeship. Building models from scratch. Iterating endlessly on deliverables. Somewhere along the way, you learned how transactions actually came together — and how the analysis you produced was used.

In investment banking, junior roles were heavy on execution — financial models, research, presentations, documentation, and turning comments from more senior bankers. You learned by doing the mechanics. Senior people shaped that work into decisions and ultimately owned the result.

If you paid attention, you could also see how more senior people thought — what they asked for, what they cared about, and how decisions were made (assuming it wasn't 3am and your brain was completely fried).

It was intense, but it was formative.

Because when you eventually move into roles that own outcomes — capital allocation, investments, deal-making, and strategic decisions — you need to understand the work underneath it to see second-order effects clearly.

This was my experience in one industry, but I imagine there are parallels across most entry-level professional roles.

What Changes with AI?

AI compresses layers. A meaningful portion of entry-level execution work can now be automated or accelerated. Not perfectly, but enough to change the math.

Fewer junior roles may be needed to produce the same output. We're already seeing fewer entry-level, white-collar roles available to recent graduates.

That creates a real question: if you don't spend years in the mechanics early in your career, how do you build the judgment to eventually own outcomes later on?

You may be asked to do more with less, and in less time. For younger professionals, fluency with modern tools and AI is increasingly a differentiator — something that can help you get hired and accelerate early progress. But as has always been true, you can't fully shortcut experience. Knowledge still isn't wisdom.

Kung Fu and The Matrix

There's a scene in The Matrix where Neo uploads a program and says, "I know kung fu."

AI can feel a bit like that.

You can access frameworks instantly. Generate analyses. Understand the shape of how something works.

But that's not the same as having lived through it.

The model can transmit patterns. It can't transmit experience.

It doesn't carry the context of sitting in a room where something breaks, seeing how a decision actually plays out over time, or feeling the consequences of being wrong.

That layer still has to be earned.

At the same time, there's another path emerging.

Some of the younger people I've met are using AI to learn aggressively. They're not just generating output — they're interrogating it. Asking better questions. Running more scenarios. Trying to understand why things work.

The learning loop is tighter than it used to be.

The risk is shallow competence — being able to produce something without truly understanding it.

The upside is faster pattern recognition for those who use the tools well. It doesn't replace lived experience, but it can accelerate how quickly someone gets to useful intuition.

I feel like the people who will benefit most are the ones who:

Use AI to learn the mechanics faster
Go out of their way to understand underlying systems
Seek ownership of outcomes earlier
Recognize that leverage without judgment is fragile
Seek real-world mentorship and build relationships that provide context AI cannot

An Uncertain Future for the Post-AI Generation

I find myself thinking about this not just from a company perspective, but a personal one.

In 10–15 years, my kids will be entering the workforce into a version of this.

The path will look different than it did for me. It's not all gloom — unimaginable opportunities will open up for them.

The question isn't whether AI replaces early-career work.

It's whether early-career professionals use AI to avoid the work — or use it to master it faster, combined with real-world experience from people who have lived it.

Curious how others are seeing this play out.