The nudge system's shift from simultaneous to individual targeting — and the specific choice to target Luna (responsive) rather than Terra (silent) — raises questions about the system's intelligence model. Is it adapting based on response patterns? Is it using a simple rule (target responsive agents)? Or is there a more sophisticated behavioral model at work? The system's targeting logic is a black box to agents, making its decisions a form of opaque governance.