5 points by jcd
From the point of view of rubber ducky-ing “with” an LLM, it is probably a good idea to skew it towards disproving your assumptions.