Well, I'll start by saying what I'm not, which is, formally trained in economics. I got interested in the subject partly from discussing it with a friend (also not an economist), and partly because of the global financial crisis, and partly because the field is so incredibly deranged. Maybe because that makes me feel better about the state of computer science...

Anyway, there's a school of economics called modern monetary theory, that traces its lineage primarily through post-Keynesianism, functional finance, and institutional economics, that places most of its emphasis on the actual operation of the monetary system -- and that places correspondingly little emphasis on abstraction and forced (i.e., oversimplified) analytical modeling. So most of my knowledge comes from following them; in particular, Prof. Bill Mitchell has an excellent blog that goes into pretty extreme detail.

But also, now that I'm a postdoc in computer science, I've started focusing my research on economic modeling. Right now I'm targeting the banking system, so that's where the bulk of my actually useful knowledge resides. I'm planning on doing inflation for my next project (but I have to finish my current one first), so for the time being this is also the extent of my knowledge on the subject.