Tuesday, November 01, 2011

TEDx Tidbits - Those annoying squiggly words actually do some good!


This weekend, I had the good fortune to attend TEDx Midatlantic, a super-inspiring day of "leading thinkers, activists, artists, philosophers and entertainers from the MidAtlantic region." Each of these amazing speakers - scientists, artists, doctors, educators, chefs, and more - shares their most influential ideas and innovations in 18 minutes or less. It's an intense day, but it's such a rush to hear and share so many great ideas! Over the next few blog posts, I'll be sharing some of my favorite TEDx Tidbits, ideas that can change your life or change the world...

On of my favorite talks was given by Luis von Ahn, who was described as a "crowdsourcing innovator." He started by showing this image at the top of this page, which makes most of us cringe. How many times have you tried buying tickets or making a purchase online, only to be faced by those annoying squiggly words. Half the time, I have no idea what the letters are! Ugh.

Well.... Luis von Ahn invented that.

It was originally invented to keep robots (is that the right word? I'm so old-school) from buying, say 10,000 tickets to a concert that you and I want to go to. Which also would be annoying. The squiggly words ensure that a human, and not a machine, is buying a ticket.

Luis said that at first he felt great about his invention, but then he started to feel guilty. It takes the average person 10 seconds to fill out that form, called a Captcha. He realized that millions of hours were being lost every year (150,000 hours of work each day) through this process, by people all over the globe, and he thought to himself: How can I use this program for the greater good? That's when reCAPTCHA was born.

There is a huge, worldwide effort to digitize books, which will eventually make books more accessible to a far greater number of people. Scanners are used to digitize the books, but the scanners can't read all the words. Especially with old books, there are words that are smudged, there are words where the ink had faded, etc. Around 30% of the words cannot be read by the scanners. That's where reCAPTCHA comes in.

When you get a CAPTCHA, there are two words. One of those words is a straight-up CAPTCHA word, which is randomly selected and helps figure out if you are human. The other word is a word from a digitized book that a scanner could not read. When lots of people fill in the letters, they are essentially voting on what they think that word is. When reCAPTCHA gets enough agreed-upon answers, that word in the book can be officially digitized.

Now hang on to your hat: 900 MILLION people worldwide have helped digitize books through this process. That's 13% of the world population!

What an incredible idea! Definitely makes me fill better about the squiggles...

No comments: