How a robot could know when it harms a human?

Artificial intelligence and robotics are claimed to be regulated. Experts say they could turn against us.

My point is: I assume that robots are not aware of nothing, they're just able to measure what they're programmed to. In many years, if we had enough developed humanoid robots for elderly, for instance, how could the robot distinguish when it helps and when hurts us?

• please be clear in what you mean to say ..... robots are not aware of nothing does not mean that robots are unaware ..... it means that robots are not aware of nothingness, not aware of void – jsotola Mar 11 '19 at 1:45