"Amazon won’t say much about how the system works, other than to say it involves sophisticated computer vision and machine learning software." Why do I have the feeling this probably actually translates to humans staring at video feeds a lot of the time?