November 9, 2014

Why you are helping Google each time you sign in

I've read the book Big Data - A Revolution That Will Transform How We Live, Work, and Think. The book promises to give the reader a brief introduction to the world of big data. While some say big data will revolutionize the world and transform it in ways we've never thought of before, others say that big data is just another bubble. I guess only time will tell us the answer.

The basic idea behind big data is that if we analyze gigantic amounts of data we will discover what we otherwise couldn't have discovered. The book is filled with examples from the world of big data, ranging from how to discover flue trends with the help of what we search for in Google to new alarm systems that analyze the seat-position of a car driver and sends a signal to the police if the car doesn't recognize the driver's position. It also discusses what might be possible in the future. Will we be able to jail criminals before they committed a crime if the data said they would commit the crime in a near future?

The company the book talks the most about is Google - a company that has access to gigantic amounts of data. Each time we search for something in their search engine, they save the data and are improving their systems with the help of that data. But we are also helping Google in other ways each time we sign in. As you probably know, Google has been driving around in their cars and photographed heaven and earth. The result is Street View.  But they have a problem. The images they took are not accurate enough so they can determine the street number of all buildings. That's where you and me are expected to help them each time we sign in somewhere using their system. 

At the height of the tech bubble, in 2000, a guy called Luis von Ahn was tired of automatic computers who wrote "spam" all over the Internet. He came up with the idea to force you and me to write numbers and letters that are difficult for a computer to read automatically. He called the system Captcha. But what if he could use you and me to also do some useful work and not just write random numbers and text each time we sign in. Google had the same idea and acquired the technology from Luis von Ahn's company in 2009. 

So in 2009, we were all hired by Google to translate text. Google has scanned almost all books in the world and they are now searchable at Google Books. But to make them searchable by you and me, Google had to translate the scanned images to text. Their computers could translate some text, but not all, and it would be too expensive to hire translators, so Google began to send out images to you and me each time we signed in. If maybe 5 of us translated the image of text the same way, Google took the translation and added it to the book they scanned.

I believe Google has run out of books to translate, so they now need help to translate street numbers. That's why you and me now have to write a number each time we sign in:

No comments:

Post a Comment