A 48-second generated loop stays out of Git, and ~21 MiB of tracked source, manifests and receipts identify ~2.2 GiB of untracked media and verify any retained or reconstructed copy. That buys identity, not availability or reproducibility — the gates taught the rest.
A small isolated-completion pilot asked whether a model can report influences on its own output. External-reconstruction controls qualitatively reproduced the effects, so the useful residue was operational, not introspective.
AI generation creates more disposable realizations, but an artifact becomes derived only when assurance can move from its exact implementation to a durable generation and verification contract.
A model can be replaced only when the application retains authority, canonical state, evidence, verification, and recovery outside the model's control.
AI weakens the old correlation between producing a plausible patch and understanding a project, so OSS contribution shifts toward evidence, verification, standing, and durable ownership.