Drafting a Covid Survey — LessWrong