Feedback Needs Somewhere to Slot In

A follow-up to last week's post on the quantity trap. It's true that good advice for building anything is to seek fast feedback. Build, learn, iterate, repeat. The faster you can run that loop, the faster you learn: What is worth building, and how?

But if you just throw stuff at the wall to see what sticks, you're doing "Build, iterate, repeat". You do something, see what happens, then do something else. For true learning to occur, the observation needs to slot into an existing hypothesis: What, exactly, did you expect to happen, and why? How does what actually happened confirm or invalidate that hypothesis, and what in your worldview needs to change because of that?

In short, you need a theory, expressed as some self-consistent set of beliefs, and that theory needs to generate testable hypotheses. And those, in turn, will tell you exactly what experiments to run. Sounds like science, because it is.