Do you honestly think that a real soldier would be better at telling apart a child with a toy gun from a child soldier in a conflict zone?
A real soldier would be making this decision while being afraid for his life and life of his comrades. A drone operator or AI would be making this decision while being afraid for a cheap (especially by military standards) machine.
Guess which decision process produces more false positives?
A real soldier can't watch 100 drones at the same time.
The problem with false positives is if you increase accuracy, but dramatically increase testing you end up with more false positives. So, we can trade 1 snap second judgment by a person, with 100 rapid judgments from an overworked person. A 10% chance of killing one person is better than a 1% chance of killing 100 people.
> A 10% chance of killing one person is better than a 1% chance of killing 100 people.
It's only true if you forget about the original objective: killing bad guys. This objective is in place not because of some sports, but because bad guys actually kill other innocent people.
So if you calculate how many innocent lives you save in the second option, it still looks quite better.
You don't understand the point. I am saying a drone operator that's 10x times as good is not enough if you see 100x as many people. And, the suggestion was sending 100 drones instead of a single soldier so 100x really is the right ballpark.
A drone operator may be better, but are they really going to be 100x as good? Of course not.
Don't forget drones can't detain people, the options are observe, ignore, or kill. Drones have already been responsible for a massive increase in civilian deaths, because they make attacks seem less risky.
