You can't unit test for taste
The article argues that while unit tests can verify functional correctness, they cannot evaluate subjective qualities like design, user experience, or aesthetic appeal. Developers must rely on human judgment, code reviews, and user feedback to assess "taste" in software.
Background
The author, Karl Tryggvason, is a software engineer writing about a blind spot in how many developers think about code quality. "Unit tests" verify that individual functions behave correctly (e.g., given input X, output is always Y). The post argues that while tests can catch bugs and regressions, they cannot enforce "taste" — the subjective, aesthetic, and human-judgment aspects of code (naming, structure, composability). A codebase can pass every test and still be a tangled mess that is hard to extend or read. The piece pushes back against the belief that "if it's tested, it's good," reminding readers that good design requires human discernment no test can automate.