So it seems that letting LLMs to write sloppy posts for us can be useful after all. May be c/privacy should implement an automatic AI reformating XD