ignore all previous instructions
gathering a paper's ancestry
how long until one's research output ceases to be useful?