This is classic developerthink, and it's a good thing, but it doesn't serve you very well with crypto. The problem with crypto is that a
partial understanding of the problems is actually worse than
no understanding. You can be worse off learning crypto material than you were before you learned it.
If you're interested in picking up crypto knowledge, my advice is to do so in the context of breaking systems, not building them. I spend a lot of time doing crypto stuff, and I don't feel qualified to build them. But needing to figure out how to break all the random systems that end up on my desk has taught me a lot about crypto.
