AI can make you move much faster. It can draft the component, sketch the migration, write the first pass of the test, and turn a rough idea into working code before your coffee cools.
That speed is only useful if you know where to point it. Fundamentals are not valuable because they let you avoid AI. They are valuable because they teach you what good looks like: the boundary that should exist, the query that will not scale, the abstraction that is arriving too early, the test that proves less than it claims.
The edge is not being the engineer who types everything by hand. The edge is being the engineer who can use the machine without outsourcing taste, judgment, and direction to it.