I honestly have a difficult time understanding the AC paradoxes, too. I recognize that it's totally necessary for many things (i.e. multidimensional integrals) and when it's consistent with a logical conclusion I have little trouble "seeing" it. Otherwise, though, not so much.
I think it's interesting to see the connection between the AC and the Law of Excluded Middle which also causes paradoxical issues from time to time.
http://en.wikipedia.org/wiki/Axiom_of_choice#Law_of_the_excl...
Also, I have trouble with "incrementally" understanding this. I can understand that the limit of 1/n (n->INF) is 0, because even with large finite numbers we're getting close to 0, but no matter how many times I cut up an orange I detect no Banach-Tarski strangeness.