And human-refined training sets are also selected for the accuracy of their knowledge. Book authors mostly default to assuming they know. There's just comparat

And human-refined training sets are also selected for the accuracy of their knowledge. Book authors mostly default to assuming they know. There's just comparatively not much reference material for someone correctly self-assessing that they don't know.