jan 2026
i’ve tried a lot of ways to learn things. courses, books, documentation, following along with tutorials. they all work to some degree. but none of them compare to picking a small, real problem and just trying to solve it.
there’s something about having skin in the game — even a tiny stake, even just your own time — that changes how your brain processes information. when you’re reading a book about a subject, you’re learning facts about it. when you’re trying to make something work, you’re learning how it behaves. those are different kinds of knowledge.
the thing that trips people up is that building-to-learn requires being publicly (or semi-publicly) incompetent for a while. you’re going to take wrong turns. you’re going to ask questions with obvious answers. if you’re sharing what you’re doing, people will see you not know things.
i used to avoid that. i’d wait until i felt like i knew enough before showing anyone what i was working on. what i’ve learned is that the waiting doesn’t make the incompetence go away — it just delays it, and delays all the learning that comes from being corrected.
the most efficient learning loop i’ve found: build something small with the new thing, show it to someone who knows more than you, get corrected, build something slightly less small. repeat. it’s uncomfortable and it works.