Human beings have excellent cognitive abilities. They can create new business opportunities by making the use of their excellent abilities. However, it is unfortunate that human beings cannot fully utilize their potential abilities as humans are clumsy in 'incentive' design. Recently, several principles of success have been discovered.
Human beings tend to entrust computers with boring, troublesome, dangerous and complicated matters. A robot that fights and cleans up instead of humans is a good example. Such specialized computer exercises and excellent abilities in a specific field, but it is not in the fields that demand a high-level of intelligence. Accordingly, human beings' abilities should be utilized in the fields, where computers fall behind man's own ability.
One of the fields where computers cannot catch up with humans is pattern recognition. Human beings can easily interpret situations that are contained in a photo, however, there is little that a computer can do in such a case. As a result, technology that uses pattern recognition capabilities is able to differentiate between man and computer. A notable example of this is CAPTCHA. When we join an online site, we sometimes come across strangely twisted numbers, letters, or character patterns that we need to type into a textbox. If we input the visible characters correctly, it will accept us as a member. In case of a failure, it displays other things. Therefore, the more mistakes we make, it lowers our chance of being treated as a human. The name, CAPTCHA, was devised by Luis von Ahn of Carnegie-Mellon in 2000.
Let's briefly look into Luis von Ahn. He was born in Guatemala in 1979 and graduated from Duke University in 2000 and obtained a doctorate at Carnegie-Mellon in 2005. He won the title of MacArthur Fellowship, which is given to top geniuses in the United States.
Luis von Ahn became famous in 2005 by introducing the game called ESP, a sort of GWAP (Games with a Purpose), in his doctoral thesis. Although everyone agrees that attaching a tag to a photo is important, although few people do so. Accordingly, a study to attach a tag by using a computer has been carried out for a long time, but the progress is slow. Nevertheless, he was able to make a very simple game. The game shows a photo to two people who do not know each other, and makes them input a word that they think it is, and if the words reach an accord, it gives a mark. People, who do not attach a tag to their own photos, attach a tag while playing this game and do it with pleasure. It became clear that if a proper 'incentive' is given, people do what they dislike. The key point is 'incentive'.
In 2007, Luis von Ahn developed a new technology, called reCAPTCHA. Google makes many documents that we cannot scan search-possible files by using character recognition technology. But, the problem lies in the fact that it cannot recognize all the characters. Accordingly, it shows twisted characters by using CAPTCHA in advance, and then lets a person type in characters as it shows. If succeeded, it is almost certain that the person is a human being. As a follow-up step, the recognition shows difficult words. By doing so, 40 million words are being corrected a day. It is called reCAPTCHA because it uses CAPTCHA technology twice.
Luis von Ahn explored a new academic field, called the human-based computation, and opened the possibility of crowd computing. We continuously have made attempts to solve the computer problem, completely excluding human beings. Namely, we have strived to make smart devices.
Now, the attempt of tiding over the technical limit by using human beings' ability in smart devices offers new possibilities. To bind smart devices with smart people, proper 'incentives' should be provided and voluntary participation should be made.
63336, a pay short message service, which seeks an answer to a question, or GTIP, is an example of human-based computation. Citizen science projects such as Stardust@home, Galaxy Zoo, fold.it are also included in this category. As human beings' intuition and cognitive ability still takes precedence over computers, but human-based computation is still attractive.
Today, research is going on for attaching a keyword to the old theses papers stored in libraries that do not have a keyword. The researchers are trying to apply machine learning technology to the keyword project. Computers cannot scan all the theses and if they did recognize the theses, the computers still will not understand the theses, then what is the point in finding the keywords. However, there is a simple way to solve this problem. The original writer of the thesis will insert the keywords themselves into the thesis. Is there a machine or a person that knows the paper better than the author In case of failing to find out the former contributor, we can request the author quoting the paper. It is the human-based computation.
