Firstly we are not talking about sane programmers. We are talking about humans. Secondly, much like the emotions, one need not install a program specifically for the purpose of removing humans. Thirdly, what “sane programmer” would program a drone in Iraq to fire a hellfire missile at an Iraqi tank with people in it? If an android could not kill, there would be little use for it in the military. Yet the military is the very first place such things are developed and used.
A computer cannot be programmed to simply not kill if the computer is very intelligent, because the computer could deduce that almost anything it did would shift events and thus if not directly kill, indirectly cause death some time in the future. By doing work for you, it takes exercise away from you. By doing work for you, it takes employment away from you. By making complex computations for you, it takes mental practices away from you. By making life easier for you, it takes self-control away from you. Anything that it does for you, it takes something away from you. And if it isn’t taking it away from you, it is taking it away from someone else who could have been doing it for you instead.
So the reality of the situation is that an android would have to attempt to cause the least death, but at what cost? If by not directly killing a certain few people, perhaps the unemployment rate will go up and cause the death of a great many. Who gets to make that decision? If a baby is on the train tacks, is it to derail the train and thus probably kill many people on board? Or is it to go ahead and kill the baby? Who gets to decide which live and which die?
Because of those kinds of issues, the programming has been merely, “kill when we tell you to kill”. And even that is more precisely “do what we say regardless of who’s in the way”. Police androids have absolute authority in all matters simply because distant authority sends them into the region to accomplish a task and thus, in effect, “the king has commanded his solders” and no one is to get in their way. Any and all resistance is “terrorism” and thus a death sentence. The king’s solders are far more valuable than people unless those people are of particular significance to the king.
All the androids are doing is trying to accomplish the task given to them by a higher authority. What are they to do when people get in their way or try to stop them? What does every king do when that happens? He simply removes the people involved. But of course as a clever king, he will not be seen doing it, nor any of his androids. Clever schemes are required to be rid of the “bad people”, such as false flag attacks on the androids, justifying counter attacks by the androids who were merely protecting themselves (just as was done with the police in Canada not long ago along with hundreds of other times less noted).
The truth is, they cannot program an android to NOT kill and there would be little to no use for one so programmed. The androids are for the king, not for the consumers who paid for them. They are there to watch out for, protect, and serve… the king. And the more the king uses androids, the less he needs people.