Just found a very cool site that tracks "innovative" free software: http://www.sweetcode.org/
Lots and lots of useful stuff ... like this one: "Given X (a flower), X' (Van Gogh's flower), and Y (a bird), this software will generate Y' (Van Gogh's bird)." http://mrl.nyu.edu/projects/image-analogies/index.html (I've wondered how such a thing could be applied to text ... dadadodo can "compile" a corpus and then write output in that "style," but it can only write from the text in the original corpus. A few years ago I heard about a program that "compiled" the music compositions of a given author and then output new works in that style, which probably worked in a similar way. Think it was at Stanford, not sure, but it wasn't free software ... what would happen if you gave it not all the known works of a given author, but say your 25 favorite compositions?) Also this one for making photomosaics: http://www.complang.tuwien.ac.at/~schani/metapixel/examples.html Someone locally is selling a photomosaic poster of the USA flag composed of hundreds of images from 911 culled from the news -- seems like that'd be copyright infringement x100 ...
