As if the prospect of being watched by security cameras wasn’t unsettling enough, they may not only be watching you. Computerised surveillance software currently under development will also let them predict what you’re about to do.
The project, funded by the U.S. government and being worked on by researchers at Carnegie Mellon University, ultimately aims to make humans obsolete when it comes to parsing security camera footage. By using technology similar to that used in Microsoft’s Kinect and Google’s self-driving cars, the cameras are getting better and better at spotting and identifying certain objects. The step the researchers are on now, is creating a “cognitive engine” so that a computer can decide whether or not those objects are misbehaving and whether or not they are about to.
According to the researchers involved, this technology could be applied to both military and civilian situations, like watching for someone in an airport who is suspiciously leaving a package somewhere. Obviously, a computers advantage over humans here is that a computer can watch everything all at once. The creepy side-effect, however, is that a computer can watch everything all at once, including you.
The project isn’t complete yet, but it seems to be coming along swimmingly though some of the conferences regarding the tech are reserved for those with top secret clearance. It’s only a matter of time, everyone. Skynet is incoming.