(Originally posted in my blog on GovLoop)
This weekend I had the opportunity to attend the TedX Mid-Atlantic conference in DC. For those of you who are unfamiliar, TED is a nonprofit "devoted to ideas worth spreading". TED has annual conferences in California, and independently organized TEDx conferences throughout the year in cities all around the world.
Let me tell you, this year's TEDx was full of inspiring people. It was an all day event, and at 11:30am I had heard so many great speakers, I couldn't believe that the conference wasn't even half over. Each speaker was incredibly passionate about their ideas, and was clearly eager to share them with the room. Their topics of expertise ranged from cooking and feeding people around the world (Jose Andres), to creating the newest telescope to replace the Hubble (Jane Rigby).
Although I was blown away several times by the energy and innovation of the speakers at this conference, the one that resonated the most with me was the presentation by Luis von Ahn. This associate professor in the Computer Science Department at Carnegie Mellon University was the founder of reCAPTCHA, the system that tests to see if you are human or not upon entering/making purchases on certain websites:
Inevitably we have all seen these images before, and have completed these puzzles countless times. Until this past Saturday I found them completely irritating and irrelevant. But now, I am tempted to complete them just for fun.
Luis shared with us that originally the displayed words were completely random. But then he realized that over 100 million people were completing these puzzles every day, and with each puzzle taking 10 seconds on average to complete, about 550,000 years of time was being spent on these puzzles each day. So he thought to himself, "what if we can use this to achieve a common goal?". Now, the reCAPTCHA puzzles serve a new purpose - one word is a "known" word (to test that you are actually human), and the second word is a "new" word from a scanned book that is in the digitization process. After 10 people come up with the same solution, the "new" word becomes a "known" word.
As humans, one thing that we can do that computers cannot do (yet, at least) is recognize scanned words that are slightly distorted. By completing these puzzles, we are simultaneously digitizing books that would have taken years to convert. So far with this new system, millions of books have been digitized for free.
Luis also shared his newest idea, called Duolingo, which is a website where you can learn a new language for free, and simultaneously translate internet webpages to and from English. Luis said if 5 million English speakers used Dualingo to learn Spanish, the entirety of Wikipedia pages in English would be translated to Spanish in 80 hours.
At the end of his presentation, I was nothing short of awestruck. What a brilliant way to engage the public, and complete tasks that would benefit the public, for free. Especially with the economy in the state that it is in now, there is no reason why the government shouldn't be coming up with ideas such as this to complete some of their projects.
No comments:
Post a Comment