Captcha, an example of human ingenuity.

When thinking about society, one visualizes many individual parts making up the whole. Each of these parts work separately to perform tasks. With each person representing a discrete packet of energy, then only certain limited tasks can be accomplished. But what if all the discrete packets of energy came together to perform a task, then perhaps many of society’s problems could be tackled. These are the ideas that keep going through the head of a genius who last winter, was awarded a $500,000 MacArthur genius grant, and in April, was awarded another $200,000 as one of Microsoft’s New Faculty fellows. The way this man’s brain works is very unlike how ours works. And, what makes him so different is what may contribute to solving large scale societal problems. You may be surprised how his thinking leads to a form of computerized collaborative programing.

From an article in the Jul., ‘07, Wired magazine, Mike McGregor introduces us to the genius, Luis von Ahn, who wants “to harvest every idle moment in our lives and turn it to productive use”. In other words, think of it this way, if the world’s computer solitaire players could be coaxed into enjoying a game that contributed to solving some sort of computer problem, then billions of man-hours of labor could be harnessed each year. [That’s a lot of man-hours].

What makes von Ahn think this way is different than the way you or I think. He is probably better known for thinking about ways people break security systems and especially the problem of how to verify that someone is a person and not some automated spam program. His way of thinking paid off when Yahoo, desperate to foil spam bots that were running wild on the company’s site, looked for help to solve this huge problem.

So, they turned to Manuel Blum for help. He is a computer science professor at Carnegie Mellon and also just happens to be von Ahn’s PhD adviser. So, von Ahn went about writing a program that generates four random letters and numbers, distorts them, and places them on a fuzzy background. Since people can recognize these patterns as letters and numbers, they pass the test. However, automated spamming software could not. Thus, the “Captcha” was created, or what von Ahn dubbed as, “Completely Automated Public Turing Test To Tell Computers and Humans Apart”. [Now that’s a name]. Yahoo began using it along with AOL, Microsoft and Ticket-master. And, the rest is history.

So, not resting on his laurels, von Ahn kept thinking that if people could recognize pictures of letters and numbers, then they could be coaxed to use this ability to identify and label the massive numbers of images on the Web. Thinking like a cryptographer, he came upon an idea in labeling images in which he would get strangers online to verify each others’ output in order to eliminate the odds of one person screwing it up. His thinking was that if two people agreed on the same word to describe an image, each would be holding the other accountable.




However, he needed massive numbers of people to identify all the images on the Web. He figured the only way to get people to do this would be to make it a game. So, in order to test his idea, he threw together some code and came up with The ESP Game and emailed the URL to his friends. When his program was “Slashdotted”, within days his server nearly crashed from the traffic of new players. What he watched happen was that 13,000 players produced 1.3 million labels for some 300,000 images. He even reported that a few hardcore fans clocked more than 50 hours of play.

Other computerized image-search technologies struggle with labeling images. They do it by associating words with a picture, such as the name given to the image, words in the page around it, or links pointing to it which could be highly inaccurate. With von Ahn’s people program, he found the image labeling to be highly accurate. When Google heard about it, they approached von Ahn and commercialized his idea into a licensed game, and in August 2006, they launched it as the Google Image Labeler. Here’s the funny part — Google is quietly using the results of the game to make the company’s database of images better and smarter. And, it doesn’t cost anybody anything.

This is the beginning for von Ahn. He and his team of 10 students are working on other computer applications that will harness human potential to do important work for which computers are ill suited. The trick, of course, is to get people to have fun so that they will do it for free. Unlike von Ahn’s programs, there are others who employ the same power of human potential but they pay people to do it. There is Amazon’s Mechanical Turk who farms out informational piecework which are quick-hit recognition jobs that computers can’t perform. PriceGrabber.com uses thousands of Web surfers to update its catalog by hunting down and inputting descriptions on products. There is also a mapping firm called Geospatial Vision that hires people to pore over satellite pictures of cities to identify tiny features like lampposts and road signs.




But this is where von Ahn is unique. He does the same thing but gets people to do it for free. The challenge is daunting at times and impossible at other times but he keeps thinking of new ways to make things fun. Von Ahn has recently even gotten into the dating game by offering a game called “Matching” that offers players a side benefit. If they’re intrigued by a partner’s answers, they can click a button to introduce themselves.

There’s no end to von Ahn’s imagination. With over 50 million Captchas being solved every day, he is even thinking of ways of capturing that 20 seconds of manpower to constitute the world’s fastest and most accurate character-recognition computer. This could tackle digitizing millions of public-domain books and putting them online for free. He is also planning to launch a company called “Games With a Purpose”, to digitize a major newspaper’s 150-year back catalog archive. Then there’s law firms who have massive amounts of information that could be digitized, and banks who use people to verify the number on a check against the amount written in long hand.

Von Ahn says, “its not yet clear what the limits of such an enormous human computer would be”. He speculates that if we have millions of people doing some little part, then something “insanely huge could be accomplished for humanity”. He says that we just aren’t thinking big enough. It’s nice to know that we have people like von Ahn in the world who aren’t afraid to think big. Who knows what the next von Ahn global solution will be. He has only just begun.