In so far as I understand it, those portions of the genome that "doesn't do anything" actually just doesn't specifically code for proteins/RNA ("non-coding regions", oh-how-originally-named. Biochemists like stating the bloody obvious, I think). It has a regulatory function, and helps enhance protein expression. Some might be markers that point out the start of a gene, some might be "promoters" that aid in its expression, and you have a whole variety of "upstream elements" that also aid in genetic expression... thus quite a comparatively large portion of those parts that don't appear to "do" anything actually have quite an important function in genetic expression.
(If anyone's hugely interested, I found
this site to be rather useful...
Yes, I understand most of it as well... *goes off to read "Mammalian Myotube Dedifferentiation Induced by Newt Regeneration Extract"*)