How to do this? One idea is to write an appropriate prompt for GPT-4. Something along the lines of “if you were HN moderator or HN community, would you censor this story? Please give numerical score.” Then post a much smaller list with top scores. That would be useful I think.