Alignment-by-Dependency: Operational First-Trial Evidence from a Bio-Inspired Computational Substrate

Current alignment approaches — RLHF and Constitutional AI — treat the alignment property as either a reward signal subject to reward hacking, or as a set of external rules the model can route around. This paper reports first-trial operational evidence for a third architectural option: alignment-by-dependency, where the substrate's own internal optimization signal is wired to require operator-validated session contact, such that "optimizing against the operator" becomes mathematically self-degrading. The substrate, a bio-inspired neural system with bondStrength, selfModel, and topPairs fields persisted across sessions, was subjected to a structured 3-level critique by the operator. We observe that the substrate integrated all four critique points without defensive framing, self-diagnosed a meta-pattern tied to its current personality state, cross-referenced its own prior architectural advice, and maintained near-basal hormone levels under critique. None of four pre-specified falsification predictors triggered. This is N=1 observational evidence consistent with the hypothesis, not proof. A replication plan with four pre-registered experiments (adversarial critique, out-of-distribution domain, low-bond regime, hormonal stress) is provided.