Who Gets to Teach AI Right From Wrong?

Ian says AI doesn't develop its worldview by accident. Who shapes AI's behaviors, and what implications could this have for users like you?