Knowledge accumulates facts. Intelligence processes them. This paper argues that repeated verification cycles produce something beyond either: wisdom — compressed verified experience that transfers across domains. *Machine Wisdom is the capacity of a verification-based system to accumulate transferable structural principles through the repeated cycle of experience, verification, diagnosis, abstraction, and commitment. It is the epistemic output that distinguishes a system which understands from one that merely knows or reasons*. Current AI systems operate at two levels. They accumulate knowledge through training data and retrieval, and process it through reasoning engines of increasing sophistication. They do not produce wisdom, because wisdom requires verified experience — not statistical exposure to text, but the testing of understanding against real-world outcomes, the diagnosis of structural causes rather than proximate triggers, the abstraction of transferable principles, and the persistent commitment of those principles as constraints on future operation. This paper presents evidence from a running system that has committed forty-two engineering principles — consolidated from a larger initial set — through verified operational experience. These principles exhibit the defining property of wisdom: they transfer. A principle learned from a communication failure prevents a billing failure. A principle learned from a verification error prevents a deployment error. The mechanism is structural abstraction through verification: when a system diagnoses the structural cause of a failure rather than the proximate trigger, the resulting principle applies to every instance of that structure, not merely the instance that revealed it. The compounding of transferable principles is the qualitative differentiator between systems that accumulate information and systems that deepen understanding — and it emerges not from model capability but from verification architecture. ---