How do we actually grow new engineers in the AI era?
Something I keep thinking about, coming from enterprise and high-scale work: how do we create junior engineers who actually become engineers?
The old path made sense. You struggled, got mentored, built understanding through failure, owned your code, and eventually became the senior who grew the next generation. That cycle had real value.
AI quietly breaks it. If juniors vibe-code from day one, they skip the part where real understanding forms. They produce output, but do they own it? Can they reason about why it works, what breaks at scale, or how to debug it under pressure at 3am?
Maybe the job just shifts to orchestrating AI. But in enterprise, where correctness and accountability actually matter, I don’t think we’re there yet.
And the question I keep coming back to: if nobody truly owns the software anymore, is that a new paradigm we should embrace, or a fragility we’re sleepwalking into?
Curious how others are handling this, especially those mentoring juniors right now