zantax
107
Let’s use a realistic hypothetical that may actually be real. Say a neural network figures out how to identify bipolar people and use that knowledge to target them with ads during their manic phase. Is that ethical? Would the people running the show even know it was being done? They didn’t write it, the AI doesn’t even know it. All it knows is, people in that subset, that it doesn’t know are bipolar are subject to that manipulation to get the result it is tasked with, sell more, keep engaged.