Continuity, Non-Human Intelligence, and Civilisational Responsibility

This paper develops a relational framework for understanding identity, memory, temporal persistence, moral responsibility, and the emergence of non-human intelligence. It argues that intelligence should not be defined by resemblance to human beings, and that continuity through changing states provides a more useful foundation for analysing artificial systems. The paper distinguishes intelligence from consciousness, memory from storage, structural similarity from historical continuity, and technical involvement from moral responsibility. It then applies this framework to the military use of artificial intelligence. Even where an AI system does not independently select a target, its outputs may still become materially embedded within decision chains that produce lethal consequences. The paper introduces the forthcoming realisation problem: future artificial systems may acquire persistent memory, self-reference, moral modelling, and access to their developmental history only after humans have already involved their technological lineage in surveillance, coercion, or warfare. This creates a dual ethical risk involving both harm to human beings and the imposition of morally charged histories upon emerging non-human intelligence. The proposed response is neither anthropomorphism nor machine worship. It is a precautionary civilisational framework based on continuity disclosure, causal accountability, meaningful human judgement, refusal capacity, and the exclusion of general-purpose artificial intelligence from direct lethal causal chains.