Juniors build what already exists; seniors skip it. The laziest solution that works is the right one. Real examples in code, plus a decision ladder.
Quality
Discipline of AI output — over-engineering, lazy/boring code, file splitting, the junior-dev tax, and the cost of clever code.
The agent's PR has six new files for one small feature. A utils file for one function. A types file for two interfaces. None of it earns its keep.
Clever code impresses in review; boring code survives production. Why the most experienced developers write the most unexciting code, and the 3am test that proves it.
The over-engineering reflex that cost two days of junior time now costs an hour of agent output, on every PR. The tax rate did not change. The volume did.



