A social norm against unjustified opinions? — AI Alignment Forum