Nothing wrong with running their emails or servers but helping it identify targets with AI brings it very close to everyone's nightmare scenario of fully autono

Nothing wrong with running their emails or servers but helping it identify targets with AI brings it very close to everyone's nightmare scenario of fully autonomous killer drones. I don't know the details, probably everything will be verified by a human (for now) and we're maybe talking only structures and patterns instead of persons. But things may evolve quickly. AI can already drive a car and diagnose some illnesses better than humans can (there was a recent article about AI detecting melanoma). The step to "decide who to kill better than a human" is awfully close. Personally I'm not against it all if better really means better. It's just crazy that cutesy Google with their maps and emails are now doing this. It's shocking enough that Intercept would pounce on it. If it was Raytheon nobody would care. I wonder if they'll update their terms & conditions. "Any selfies may be used to kill you". Al Qaeda better (not) be using Google Photos.