In his talk, Professor von Ahn pointed out that in 2003, the world spent 6 billion hours playing solitaire. This was contrasted with the 7 million hours necessary to construct the Empire State Building or the 20 million man hours required to complete the Panama Canal. Professor von Ahn proposes that: Using human skills and abilities, it is possible to solve large-scale computational problems and teach computers many of the basic human talents.

The first example provided is reCaptcha. Professor von Ahn developed reCaptcha, a new form of Captcha that helps digitize books. In reCaptcha, the images of words displayed to the user come directly from old books that are being digitized; they are words that optical character recognition could not identify. It is estimated that 200 million Captcha’s are solved dail by humans around the world. reCaptcha was designed with the purpose of using crowdsourcing, to identify distorted text. Once a piece of text has been agreed upon by the necessary number of humans, it is re-inserted to help in the digitization of the book.

Professor von Ahn next turned his attention to the use of computer games to achieve his goal of solving large-scale computational problems that computers aren’t capable of solving. Google images provided a perfect case study. In this situation, a user may search for an image by keyword, however, the implemented method of tagging these images was woefully inadequate. Far too often, images had no relation to the keywords associated with an image. To solve this problem von Ahn invented the ESP Game. ESP takes two players that don’t know one another with an objective to get them to type the same word without communication. Both players share an image and words are provided by the players that describe the shared image. As users agree on keywords they become taboo or unusable. This helps spur the further identification of keywords to be associated with an image.

Professor von Ahn stated that already 50 million tags had been agreed upon by 200,000 players. A surprising number of which play for more than 20 hours a week and sessions of 15 hours are not unknown. These numbers lead to belief that if 5,000 players played simultaneously, all of the images in Google would have been agreed upon by humans in a few weeks. While the ESP game is rather addictive and has proven a hit (licensed to Google), the obsessive nature it fostered, bothered Professor von Ahn and a maximum of 15 consecutive hours (10 for .edu) has been imposed, after which, the the system will lock you out.

An interesting side note mentioned regarding the ESP game project, was the acquisition of user data. Students constructed prgrams that could accurately determine several demographics of ESP users. After assessing a couple of responses from players, in 98% of the cases, gender could be accurately identified, while age could be fairly accurately determined in 85% of the cases. This could potentially provide very powerful tools for marketers, as well as training material for computers.

Professor von Ahn mentioned the strong bond created between partners that play the ESP game. Partners that perform well together not surprisingly, find a warmth and connection. These have led to some users wanting to date their anonymous partners and describe the game as “anonymous intimacy”, “like crack”, “a great way to learn English”

In what manner can this be applied to today’s classroom activities? Where can crowdsourcing provide learning opportunities? Small inefficiencies in procedures or schedules can grow over the school year to an enormous figure. For example: Two minutes, five periods per day of wasted time, turns into 30 hours over a 180 day school year. How can we recycle these wasted human cycles?

L:uis von Ahn has several current projects including the ones mentioned in this article. See the URLs below for more information:

• http://gwap.com
• Luis von Ahn’s blog