It'll never happen (due to a lack of any formal on-paper education), but my ideal career would be me researching new things that are interesting to me in regards to math and programming. I'd love to research news things in the Join and Pi process calculi (and other interesting things with distributed computing) and write papers in LaTeX and whatnot; honestly, the more theoretical, the better.

Sadly, I think those jobs are mostly reserved for people who went through academic programs and have a bunch of letters after their name, so I stick around in the land of pragmatism.