How Should Companies Hire Programmers in the AI Era?

If most business code is eventually generated by AI, the question arises: how should companies recruit programmers?

The author believes interview focus will move from pure code‑implementation ability to assessing system‑design capability and AI‑collaboration skills.

System design here means more than drawing diagrams; it involves taking a requirement and organizing the whole system—defining module boundaries, data ingress and flow, API contracts, code placement, isolation, and anticipating future extension points.

These design tasks are considered more important than writing code itself.

While some may argue AI could eventually handle such design, the author contends that AI can propose designs but cannot replace human decision‑making because software architecture lacks a single correct answer; the optimal solution depends on the company’s specific situation.

For example, a coupon system could be split into three services or five; decisions about table redundancy, shared services, or pre‑reserved extension points each have trade‑offs. Factors such as team size, growth rate, system lifespan, overseas expansion plans, and maintenance capability influence the final architecture.

AI can generate five or ten design options, but a human must make the final call. Choosing the wrong direction can be extremely costly: a coding mistake is fixed by changing a method or class, whereas an architectural mistake may require adjusting dozens of services and is often discovered only after months of operation.

Consequently, no company will readily hand over such decisions entirely to AI.

The truly valuable programmer of the future will be able to evaluate AI‑generated proposals, know when to adopt, reject, or redesign them.

Interview formats will evolve accordingly: rather than whiteboard algorithm questions, candidates will be given realistic business scenarios—such as designing a coupon or order system—and assessed on how they decompose modules, design data models, and weigh different architectural trade‑offs.

Another increasingly critical skill is proficiency with AI tools; effective programmers will leverage AI to quickly produce designs, generate code, detect issues, and then apply their own judgment to refine solutions.

Ultimately, a programmer’s core value remains understanding the business, making sound technical decisions, and steering AI assistance.