Good is ethical and AI machines are not ethical. They create the most probable knowledge, and yet that might be ethically very bad. Like the prediction of a war here or there and the number of people who will be killed, deported, maimed, and so many other horrible things. And AI machine could have predicted how many thousand children would be separated from their families and what came and comes next if Trump had asked an AI machine to do the calculations, but Trump did not ask because he does not give a damn. And the dispatching of kids was done in a blind way so that an AI machine cannot even access the data. An AI machine collects the data and gives the plain results of its calculations: statistical mapping of the data most of the time. And the AI machine might be wrong because it does not always take the proper parameters into account, because it is programmed and it self-learns what it can self-learn according to its program.