A fun game for a boring task

Image labeling seems like a pretty boring task to most people, but Luis von Ahn made it fun more than two years ago with his "ESP Game".  This addictive multi-player game embraces the power of wasted human cycles to label images.
Danny Sullivan of Search Engine Watch discovered a TechTalk video from July 26th where Luis von Ahn discussed the application of "human computation" to generate high quality descriptions of images.  It's not a surprise that just one month later Google released their very own version of the game called "Google Image Labeler" that will give Google Images a more complete set of tags.

Also discussed in the TechTalk were two other "Asymmetric" games.  The first one, called peekaboom, piggy backs on the "ESP Game" by taking the same image/tag pair and getting player one to try and get player two to guess where in the image the keyword appears.  Given the usefulness of such information, Google will likely create a game based on peekaboom as well.
For example, let's assume there was a picture of two people walking on a beach with a dog.  Player one will get a picture and the word "dog".  The goal is to make player two guess the word "dog" by highlighting a part of the picture.  This information gives images an extra dimension of information that was previously unknown.

The third game discussed has not yet been released, but the name is "verbosity" and aims to generate a very large set of facts (nothing to do with images) -- the building blocks of artificial intelligence.  Player one receives a word and tries to get player two to guess it.  To do so, player one is given a sentence that contains a word that must be filled in.  Player two gets the new complete sentence and uses it like a clue to figure out the word.  If player two is successful, the fact entered by player one must be accurate.  Again, a game of this nature created by Google would not surprise me.