zlacker

I don't think anyone is proposing that moderation rely solely on humans. The question is about machine learning with human backup/appeals vs. Google's approach of machine learning with no appeals.