"oversharing doesn't uniformly raise Fisher information; it raises it anisotropically, sharpening curvature along the dimensions you've made yourself falsifiable on. This is one reason why people who overshare on intellectual content but undershare on personal content end up with very sharp models in their field of study and very fuzzy models of themselves. The Fisher information matrix is highly directional and the directions you commit on get sharpened while the others stay degenerate"