You’ve all seen those ubiquitous wavy, squashy words that have to be typed into boxes before you can sign onto websites. Those oddly shaped words, called ‘captcha’ (Completely Automated Public Turing test to tell Computers and Humans Apart), were developed as a way to thwart computer robots. Although very difficult for computers to decipher, the words are not usually a problem for people. They’ve been used as security measures since the late 90’s.
Turing tests, named for Alan Turing, a WWII mathematician who broke the Nazi Enigma code machine, are designed to distinguish humans from computers. A computer that can fool a human being in a natural language conversation into thinking it is another human is said to have passed the test. To date, no computer has done such a thing.
On the other hand, computers are able to spam email, send automated posts to blogs or forums, and otherwise wreak havoc. Captcha was designed to prevent this sort of mayhem by locking out non-human responders.
Unfortunately, as spammers learn new ways to get around these roadblocks, security measures must be constantly updated.
Danny Cohen-Or of Tel Aviv University found a novel way of impeding computer hackers. He and his colleagues created ‘emergent images’, images that could only be interpreted by watching them move. For example, an image might blend into the background until it moved. At that point, humans can recognize it, but computers still can’t.
Rather than deciphering a wavy word and typing it into a box, captcha in the future might consist of a brief video and then a multiple choice question about what was seen.