The hiring question changed
For years, engineering teams hired capacity. More backlog meant more engineers. AI complicates that equation. If one good engineer with the right tools can do the work of two average engineers, the case for bulk hiring weakens.
That does not mean engineering hiring stops. It means the department becomes more selective about leverage.
What employers should screen for
Ask candidates to review generated code, write a test plan, diagnose a production failure, and explain when they would not use an LLM. These screens reveal judgment faster than trivia.
The best teams will make AI usage explicit in interviews. Hiding it creates theater. Good builders use the tools and still know what good looks like.
What candidates should target
Backend systems, platform engineering, AI integrations, QA automation, observability, and security-adjacent engineering are better lanes than generic web implementation.
The message to candidates is simple: show ownership of a working system. Not just a repo. A system.