What I learned from intentionally building alongside AI instead of just using it for quick answers — and why the biggest shift wasn't speed, it was clarity.
I've been intentionally practicing what people call "vibe coding" — building alongside AI instead of just using it for quick answers.
Two things pushed me there.
A billiards league teammate — a seasoned software engineer — challenged himself to build a full site using only AI. The result was a working web app comparing pro stats and payouts. Not a gimmick. Functional. Around the same time, I started a ZTM course focused on this approach.
So I tested it myself.
Not copy/paste coding. More like pressure-testing design decisions, generating scaffolding to focus on structure, and surfacing edge cases earlier.
The biggest shift wasn't speed. It was clarity.
If I can't clearly define what I'm building, AI exposes that quickly. Vague prompts produce vague results. The discipline required to work effectively with AI is the same discipline that makes you a better engineer without it — you have to know what you want before you can ask for it.
I'm not talking about accepting whatever the model produces and shipping it. That's not vibe coding — that's outsourcing your judgment.
What I mean is using AI as a thinking partner:
My portfolio projects all have an AI-NOTES.md file that documents exactly this — what AI suggested, what I changed, and why. Not to prove I'm clever, but because the reasoning behind a decision matters as much as the decision itself.
I've been reading about a concept called "Comprehension Debt" — the idea that as more AI-generated code gets committed, overall understanding of that code declines. It might be technically correct. It might even pass review. But that doesn't mean it aligns with your architecture, your team's mental model, or long-term maintainability.
Having worked in healthcare and aerospace, that gap concerns me deeply. In regulated environments, "it works" isn't enough. You need to know why it works, under what conditions, and what breaks it.
AI is a powerful tool. It doesn't replace responsibility.
Used intentionally, AI doesn't replace judgment — it sharpens it.
The engineers who will thrive in an AI-assisted world aren't the ones who use it most. They're the ones who use it most deliberately — who know when to accept a suggestion, when to push back, and when to throw it out entirely.
That requires domain knowledge. Context. Experience. Things AI don't have and won't develop.
Ask it questions. Use its help. But own your work.
Tools don't replace ownership. People do.