#ethics

기계 지원 번역 초안 (Korean) for "AI Alignment": The research field focused on ensuring that AI systems pursue goals that match human values and intentions. A misaligned AI might optimize for a metric that appears correct but produces harmful or unintended outcomes at scale.

“예문 초안: AI alignment researchers worry that optimizing for user engagement could misalign with genuine user wellbeing.”