A lot of people are blaming AI for why junior engineers feel lost.
That's lazy thinking.
The real issue is that:
- juniors stopped getting real ownership
- work got fragmented into tickets
- learning-by-doing quietly disappeared
AI just exposed it.
The apprenticeship that vanished
Ten years ago, junior engineers learned by:
- sitting next to seniors
- watching decisions get made
- making mistakes in safe contexts
- owning small but real things
Then the industry "scaled."
Teams got bigger. Processes got heavier. Work got sliced into tickets so small that no one person understood the whole.
Juniors stopped learning how to think.
They learned how to complete tasks.
The ticket trap
Here's what modern engineering onboarding often looks like:
- Get assigned a ticket
- Follow the spec exactly
- Get code review feedback
- Ship
- Repeat
At no point does the junior:
- understand why this feature exists
- see how it connects to user problems
- make a real decision
- experience the consequence of that decision
They're executing. Not learning.
Why this matters now
AI is incredibly good at executing.
If a junior's entire value was "translate spec to code" — that value just got commoditized.
But here's the thing:
AI didn't cause this problem. It revealed it.
Juniors who learned through real apprenticeship — who understand systems, tradeoffs, and user context — are doing fine.
Juniors who learned through ticket factories are struggling.
The difference isn't AI. It's how they were taught.
What real apprenticeship looks like
It's not formal training programs. It's not onboarding docs.
Real apprenticeship means:
- Ownership of small, real things — not tickets, but outcomes
- Proximity to seniors — seeing how they think, not just what they ship
- Safe failure — making mistakes early, when stakes are low
- Context, not just instructions — knowing why, not just what
This takes time. It takes intention. And most modern teams don't prioritize it.
AI could actually help — if used right
Here's an unpopular take:
AI could restore apprenticeship.
If juniors use AI as a thinking partner — not a shortcut — they can:
- explore decisions faster
- understand tradeoffs earlier
- prototype ideas without waiting for senior availability
- learn by doing, at scale
But this only works if:
- they're taught to question AI outputs
- they're given ownership, not just tasks
- someone is still guiding their judgment
AI can accelerate learning. But it can't replace mentorship.
The industry must redesign how juniors learn
We can't keep blaming AI for a problem we created.
If juniors feel lost, it's because we:
- fragmented work into meaningless slices
- removed ownership from early careers
- stopped investing in real mentorship
The fix isn't "protect juniors from AI."
The fix is give them back the apprenticeship we took away.
At Sodiarc, we don't just teach coding — we teach how to think like a product engineer. Join our free working sessions to see if it's right for you.