that probably has the same effect as "think harder" which is a legitimate prompt. The models are fairly could at interpreting intent.