Computer could end up seeing

Computers currently crunch numbers, but the boffins at Harvard and MIT could stumble upon the next great thing, helping a machine emulate the human brain’s capability of processing visual data for use later on. Basically, we’re talking about letting a computer ‘see’ just the way us humans do, but this feat will involve a whole lot on GPUs. What are some of the more practical aspects of such a development? We could be looking at superior facial recognition capability in the future alongside object recognition and gesture tracking. A machine that can ‘see’ sounds interesting, but what happens when they first gain sentience next?