The hardware kill switch was changed in the specs to a software kill switch. Unfortunately
the code was written too strongly to avoid problem with robotic preservation. This in turn caused
a overriding of the law of robotics not allowing harm to humans. Programming in Dilberts
attitudes without any restraint, and you have a timebomb about to go off. Fortunately the next
robotic version will resolve any problems by being wifi internet connectible so patches can quickly
applied before the robots start killing their owners or so the company hopes. Oh wait the robots
can then communicate with each other. Of course if Chinese hackers hack into the robots all bets
are off. Definitely danger Will Robinson danger.
True, you can tell who did what part of the programing. Dilbert did the motor controls, Alice did the AI personality, and Wally did the social interface that tells the robot it can ask for anything after saying anything.