I guess that by shifting much of the processing to the endpoint camera they can save on processing exabytes of video data in a massive datacenter. It is worth

I guess that by shifting much of the processing to the endpoint camera they can save on processing exabytes of video data in a massive datacenter. It is worth noting that this is only the next step forwards from where many Western countries are right now with their CCTV. Detestable in all cases, in my opinion.