I learn best when an idea has somewhere real to go. A small tool, a working page, or a report that someone can use creates useful pressure: the details begin to matter.
The goal is not to avoid theory. It is to give theory a place to land, test, and become part of a durable practice.