Yeah, it's funny, in a way, that I'm raising this objection, sitting here typing to you on this computer, which has such unfathomable layers of complexity that it's practically magic. (And we thought pencils were complicated!) And even though I understand the stack, at least in theory, down to the physical layer, my working schema is much simpler: I'm sending you a message.

If you handed a computer to someone from 1500 (after representational art and the printing press, among other things) and said to him: "Behold, I am magically sending a message to my friend through the ether!" they would get like 90% of the idea, don't you think? It might even be a little easier in 1500 than, say, 1900 -- you'd convince him you were some kind of magician, which might well be plausible according to his worldview, and he'd be sorted. The communicative act is fundamentally the same, even though the medium is not, the manner is not.

So:

Knowing what I know of the math behind it, there's just so much spooky weirdness that defies intuition. And not just PKC, but all of these abstractions. The idea that the mere knowing of a thing makes it yours, for instance; or that your entire history of transactions can be revealed if you can guess a number, but that the number of guessable numbers is so large that guessing any of them is impossible --

I dunno. The more we talk about it the less convincing I find my own argument. But I'm also acutely aware of how long I've been thinking about this topic and how foreign it still seems.