Facebook disputes that its AI can’t detect hate speech or violence consistently | Hacker News Reader