Why the traditional career ladder is broken—and what’s replacing it
Yesterday I wrote about “Who reviews the AI’s code when the senior leaves”. A friend pointed out the sustainability problem in AI-augmented development—what happens when the experienced person moves on?
Today I want to flip that question.
What about the juniors who never got the chance to become seniors in the first place?
Junior engineering roles are collapsing. Companies are hiring more seniors, fewer entry-level positions. The traditional path—junior → mid → senior—is breaking down.
And it’s not just developers. Junior designers, junior PMs, junior anyone in creative/technical fields—they’re all facing the same squeeze.
The work that used to teach you the craft? AI does it now.
Here’s how it used to work:
For developers:
For designers:
That grunt work wasn’t just busywork. It was training. You developed taste by doing the reps.
Now AI does the reps. So how do you develop taste?
Here’s what nobody wants to say out loud:
Pattern recognition ≠ judgment.
A 2027 junior will have seen more code patterns than a 2017 senior. They’ll have generated more design variations in a month than previous generations did in a year.
But seeing isn’t understanding. Generating isn’t judging.
The senior knows why something is good. The junior using AI knows that it looks good—but can’t tell you why, and can’t catch when it’s subtly wrong.
This is the gap. And it’s growing.
Experienced developers spent 10+ years learning specific tools, editors, workflows. They’re attached to their ways of doing things. When the landscape shifts, they have to unlearn before they can learn.
Juniors? No baggage. No attachment to “how we’ve always done it.”
When the iPhone came out, the developers who’d spent years building for Nokia and Blackberry were suddenly useless. Their experience became a liability. The fresh graduates who dove straight into iOS? They won.
We might be at that moment again.
Here’s the insight that reframed everything for me:
Cold applications are dead. There’s too much AI-generated slop. You need to find ways to build trust.
Qualifications don’t matter like they used to. Certifications? Nobody cares. A GitHub full of random projects with zero stars? Useless.
What matters is trust.
And how do you build trust? By being useful.
I heard a story about someone named Roy that perfectly illustrates this.
Roy wasn’t a developer. He was doing semi-technical customer support, learning to code on the side. He found a Discord community, had some questions, and started hanging out.
Then something shifted. Instead of just asking questions, he started answering them. He got good at it. Really good.
Another developer started DMing Roy for help. After a few conversations, he said:
“Holy shit, Roy is one of the most useful people I’ve ever talked to.”
Someone else responded: “Yeah, isn’t it crazy that he doesn’t have a dev job yet?”
“Wait, you’re not joking? I thought he was at least ten years my senior. What do you mean he doesn’t have a job?”
Roy now works at Clerk. He skipped the resume pile entirely. He got hired because people trusted him.
If you’re a junior—designer, developer, whatever—here’s the playbook:
Your degree, your certifications, your “perfect” portfolio—none of it matters as much as whether someone at the company can vouch for you.
Join Discord servers. Hang out in GitHub issues. Answer questions on forums. Be where the experienced people are, and be useful.
Every problem you solve, write down the two sentences that would have saved you time. Share them.
Curious which AI models are better at writing? Build a little app to test it.
Want to understand why Figma’s auto-layout works the way it does? Rebuild a simplified version yourself.
Don’t just consume answers. Build your way to understanding.
There’s a difference between:
When you hit a problem, don’t just let the AI solve it. Ask it to explain. Then try to solve a similar problem yourself. Build the muscle.
Anthropic just published research on exactly this—I’ll dig into it tomorrow.
Netflix’s CTO said they’re specifically hiring juniors for their “native AI familiarity.” It’s kind of bullshit, but it’s a lever you can use.
You grew up with these tools. Seniors are still figuring them out. That’s an actual advantage—use it.
Seniors have capability but lower energy. They’ve seen it all. They’re a bit jaded.
You have energy. You’re excited. You’ll try things they won’t bother with.
That energy is valuable. Don’t let rejection kill it.
Here’s what I keep thinking about:
What if the traditional junior → senior path just… doesn’t exist anymore?
What if the new path is:
It’s messier. Less predictable. But maybe more meritocratic?
The people who win won’t be the ones with the best degrees. They’ll be the ones who figured out how to be useful, how to build trust, and how to learn faster than the tools are changing.
If you’re reading this and you hire people:
Your interview process is probably broken.
If a candidate can cheat your technical interview with a “crappy AI app” but couldn’t cheat at the actual job, then your interview doesn’t test what the job requires.
Test communication. Test judgment. Test what it’s like to actually work with someone.
And maybe—just maybe—look beyond the resume pile. The best juniors aren’t applying cold. They’re out there being useful, building trust, waiting to be noticed.
Tomorrow: What Anthropic Learned About AI and Skill Development
If you’re a junior struggling right now, DM me on LinkedIn. I can’t hire everyone, but I can point you in the right direction.
