At scale, hostile comments and coordinated harassment are near-universal creator experiences, not personal failures - Response protocol: don't engage (it rewards the algorithm), block aggressively, document patterns, separate from real feedback, tell someone you trust