Problem is, it takes a lot longer to find work as a post-doc than as a programmer. The money may run out on predictable, yearly intervals, but what do you do once it's gone? Get a faculty position? (Because we all know that's easy.)