I've thought in the past, once the compute is available in a low-enough-power form, you can continually increase the resolution (frame rate) that autonomous veh

I've thought in the past, once the compute is available in a low-enough-power form, you can continually increase the resolution (frame rate) that autonomous vehicles view the world. Essentially giving them 'The Flash' like vision. You could then have all kinds of predictive models based on various patterns.