It'd be nice if: 1] Google (or at least someone who can store the URL index from commoncrawl.org ) provided a "random page" function; maybe this exists, but I'

It'd be nice if: 1] Google (or at least someone who can store the URL index from commoncrawl.org ) provided a "random page" function; maybe this exists, but I'm not aware of it. 2] We could get a random page of a random real book from google books or archive.org Both of these are somewhat feasible even locally since text is small and in the former case you're just storing an index (albeit maybe many gigs or a few terabytes) -- storing compressed books and an index would take up very little space You could do the same thing with art, random arrays of pixels instead of text, but arguably you could transform this text into pixels anyway -- perhaps something that makes no sense to read makes some sense to look at, or even to listen to.