Carnegie Mellon is doing an experiment that will use CAPTCHA as a sort of Human OCR. (CAPTCHA refers to those characters that you have to type when creating an account like on Yahoo to prove that you are human. en.wikipedia.org/wiki/Captcha And OCR refers to Optical Character Recognition or a way of digitizing printed materials en.wikipedia.org/wiki/Optical_character_recognition)
Here is how the system would work. The Captcha would consist of two words, one that was known to Carnegie and the other that was from a document that was not known. It would be assumed that if two people typed in the known word correctly than the identity of the unkown one would also be know.
So the next time you are signing up for an account and have to type in those funny letters, you might be helping to digitize some old manuscripts.
You can read more at it.slashdot.org/it/07/10/02/1224204.shtml