Behind the scenes of Google’s neatest new tricks is an artificial mind that’s getting smarter… We are getting closer to Skynet and Terminator…
And you are building the face recognition database via Facebook and iPhotos that Skynet will use to identify you and your loved ones… oh silly people…
This year’s Google I/O was kind of a quiet one. There were no fancy new watches, no self-driving cars, and definitely no sky-diving from blimps. But behind smaller (and still awesome!) announcements like a photo service that will store and organize all your photos for you, and a version of Google Now that can intelligently suggest reminders, is a titanic achievement Google only hinted at onstage: A robotic brain that can learn.
Taking in the world
To be fair, this isn’t the first time Google’s toyed with a robot brain. In 2012, an army of 16,000 computers hidden away in the deep, dark recesses of the experimental Google X labs accomplished a complex and impressive feat, at least by the standards of a computer brain working with silicon neurons.
It taught itself to recognize a cat.
By pouring over some 10 million images from YouTube videos and thinking really, really hard (for a computer), this network actually taught itself to conceive of the idea of cat, and put a furry face to that concept. It’s easy to gloss over the magnitude of that; these are cats after all. Remember the one with the keyboard? Lol. The cheeky headlines basically write themselves.
But now, three years later, we’re starting to really see the fruits of that labor pay off. Google’s new Photo’s app seems to be much the same experiment but put towards a practical goal. Instead of a million YouTube images, this brain is looking through thousands of personal photos. Instead of a cat, it’s learning to recognize your friends and family. Also, coincidentally, your cat.