As LLMs automate coding tasks, developers feel a tension between creative cleverness and maintainable code. The article argues that clean code for humans and clean code for LLMs aren't the same; LLMs need explicit contracts and structural regularity to refactor safely, while humans rely on context and intuition.

The solution isn't to abandon elegance but to relocate it. Adopt a principle of "clever inside, boring outside."" Keep public module APIs and function signatures predictable and explicit for both humans and models. Reserve algorithmic elegance, domain modeling, and strong abstractions for private implementations where the joy of engineering art belongs.

This approach reconciles automation with creativity. Examples in Ruby, Go, and TypeScript show how to create stable, LLM-friendly boundaries—like explicit Result types or HTTP handlers—while building beautiful, invariant-enforcing domain logic inside. The result is code that survives refactors, whether human or AI-driven, without killing the pleasure of building clever solutions.