We all know what Captcha’s are (ok, if you don’t, they’re the annoying things that you have to figure out what letters are in an image then type them in to get page X to do Y), some of us find them annoying as bloody hell as well. I personally don’t like them because I’ve seen how badly they interfere with disabled users.
Some companies have even gone so far as to allow people to listen to Captchas so disabled users can solve them.
Anyway, someone has come up with an uterly brilliant idea that is starting to change my mind about Captchas and the like. While it still has issues, it takes on human knowledge and problem solving and turns it into a global cause. For instance, when you solve a Captcha, that knowledge is already there - the computer knows what you need to type to get access to whatever. ReCaptcha takes that to the next level and throws a second word into the mix.
Yes - I said second word… The second word is a word from a scanned book or other source. When you type in both words, the computer figures out what the second word means. Basically, when the OCR scans the book, the printing and font confuse the machine and you help solve it. Check out http://recaptcha.net/learnmore.html for a full explanation on this process and graphics.
One of the developers explains this process in his blog as well. Oddly, from what I see, the idea basically came from the same guy that created the Google Picture Labeler.
