Everyone by now presumably knows about the danger of premature optimization. I think we should be just as worried about premature design - designing too early what a program should do.

Graham is extending a well-known software maxim—avoid “premature optimization”—to the earlier stage of deciding what software should be. The warning is that locking in requirements, abstractions, and architecture too soon can be as damaging as micro-optimizing code: it commits a team to assumptions that haven’t been validated by real use. “Premature design” can lead to overbuilt systems, brittle interfaces, and wasted effort solving the wrong problem. Implicitly, the quote argues for iterative development: build something small, learn from feedback, and let the program’s purpose and structure emerge from actual needs rather than speculative planning.