SORT-AI: Runtime Control Coherence in Large-Scale AI Systems Structural Causes of Cost, Instability, and Non-Determinism Beyond Interconnect Failures

The scaling of large-scale AI systems increasingly encounters operational instabilities that cannot be attributed to interconnect limitations alone. Even in infrastructures with sufficient network capacity, cost escalation, non-deterministic behavior, and soft degradation persist, indicating coordination and control as distinct failure domains. Building on prior structural analyses of interconnect-induced instability, this article introduces runtime control coherence as a structural property describing the degree to which distributed control decisions across schedulers, orchestrators, runtime engines, and policy layers remain mutually consistent. The loss of control coherence gives rise to economically significant inefficiencies without manifesting as discrete faults or performance violations. Classical metrics fail to capture coherence loss because conflicts between control layers are distributed, emergent, and temporally decoupled. This work provides a structural problem analysis that positions control incoherence as a first-order economic and operational variable, complementing existing analyses of interconnect-driven instability. The methodology is deliberately conceptual, avoiding implementation details or prescriptive solutions.