I've been wondering if a machine could ever truly understand the nuance of human morality, or if it’s just a sophisticated pattern-matcher. Do you think that’s an overblown fear, or a real ethical dilemma we should tackle?