You’ve all heard of CAPTCHAs, haven’t you? Those annoying, squiggly letters and numbers you’ve had to make sense of and then type out so as to verify you’re a real person and not a malicious string of code? Well, did you know that besides serving the purpose of proving you’re a human, you’re also creating a digital archive of several books, word by word, along with millions of other CAPTCHA users across the globe?
Luis von Ahn, professor at Carnegie Mellon University, is one of the brains behind this novel and useful idea. CAPTCHA was initially started to serve just the purpose of web security. However, after realizing that all CAPTCHA users taken together were collectively spending thousands of hours every day to type out CAPTCHAs, von Ahn and his team came up with the brilliant idea of utilizing CAPTCHAs for digitizing books without compromising internet security. This project, called reCAPTCHA, works like this: Now, instead of typing out just one word for the CAPTCHA, you are made to type out two. One of them is already known by the system, while the other is unknown. The user is unaware as to which word is known by the system and which word isn’t. The system verifies that you have typed one of the words correctly, and thus assumes you’re human, and that you’re capable of getting the other word right. In this manner, through ten seconds of your time, another word is added to a digital archive. The overall rate at which this digitization takes place is quite impressive; about 2.5 million books are digitized in a year! The reCAPTCHA project has also resulted in the creation of a meme called captcha art, where the two words of the CAPTCHA are taken and a funny picture associated with the words is attached to it.
After the success of the CAPTCHA project, von Ahn realized the true potential of collective human efforts via the internet. He decided to take it one step further by posing this question to his graduate student, Severin Hacker: How do we get a hundred million people to translate the web into every major language for free? This seems to be a very far-reaching vision, but they came up with an answer.
Translating the web would require the contribution of a large number of bilinguals, and there are not a large number of them. Also, why on earth would anybody want to do this seemingly tedious task for free? These two serious hurdles were found to have one elegant solution: language education. People wishing to learn a new language can now do so for free, using Duolingo, the idea that was born from von Ahn’s vision. You first start off by translating simple sentences given the meaning of each word. As your level progresses, you are given more and more complex sentences and are asked to translate them. Your translation is then compared with other users’ and a reasonably accurate final translation is obtained. This strategy has been tested and found to work quite well, with people translating web pages almost as accurately as a professional translator, but free of cost. Users are also motivated to learn as they are given real content to translate such as Wikipedia pages and news articles. Also, a new language can be learned without even having to pay a single dollar! C’est formidable, n’est-ce pas?
Leave a Reply