AI is automating the code, not the engineering. A guide for senior developers on shepherding AI through the work that was always the actual job.

Foojay.io's platform is a central hub for Java developers, offering insights into Java programming language, JVM ecosystem, and Java-related technologies. Through articles, tutorials, and community forums, Foojay.io offers insights into building scalable and maintainable Java applications. Developers can learn about Java language features, performance tuning tips, and best practices in Java development to write efficient and reliable software.

Foojay.io

Senior developers are not being replaced by AI — they are evolving into 'shepherds' who guide AI through terrain it cannot see: legacy context, business constraints, and system complexity. Drawing on a real-world example of a Quarkus/Jakarta Data migration where AI gave confidently wrong advice, the author argues that the actual job of a senior developer has always been reading terrain, choosing what to delegate, verifying AI output, and mentoring teams. A Stanford study is cited showing real-world AI productivity gains are 15–20%, not the hyped 30–40%. The core message: seniority is a moat, not a liability, because AI automates the visible typing while leaving the judgment-heavy work untouched.

The Code Was Always the Door

Counterpoint, though: <a href="https://larsfaye.com/articles/agentic-coding-is-a-trap" target="_blank" rel="noopener nofollow">Agentic Coding is a Trap</a>
Being successful with agents coding agents hinges on a crucial element: only a skilled developer who’s thinking critically, and comfortable operating at the architectural level, can spot issues in the thousands of lines of generated code, before they become a problem.
This is the paradox that has people talking out of both sides of their mouths: The use of coding agents is actively diminishing the very skills needed to effectively manage the coding agents.
Shepherds only exist because people were lost in the first place and had to find their own way, to map their own terrain. They need to build those skills, and if they stop building them, they won’t be able to shepherd anybody else, nor teach a new generation on how to do so, either.

Shephard analogy is a good way to imagine the new shift. The plan that you had to also implement yourself previously, is now done by AI. Makes things easier to test during R&amp;D cycles. We can be “Orchestrators”, for now at least.