I've been skeptical about new camera tech in the past, like the ill-fated, consumer-oriented Lytro, which was hamstrung by a prohibitive proprietary image type, but I think the idea of the computational camera is going to be one that becomes more and more important going forward. If you notice the latest trends in smartphone cameras—an undeniable force in the photography world at the moment, like it or not—the focus has clearly shifted slightly from improving hardware to making the cameras "smarter." The iPhone camera, for instance, hasn't drastically changed in terms of pure imaging hardware in some time, but there are reportedly 800 people working on the iPhone camera team. My guess is that the bulk of that work has to do with using software and computing power to do things with image data outside the bounds of simply making it look better. If smartphone cameras can do impressive things with one tiny imaging device and lots of processing power, it's only logical that 16 will make things even more interesting.