Before Consciousness: Persistence, Identity, and Human Agency in Artificial Agents

Contemporary debate on artificial intelligence asks whether artificial systems can become conscious or self-conscious and what moral status they might have. This editorial argues that two prior questions must be addressed before judging an artificial system conscious, autonomous, responsible, or an agent: what makes it the particular entity it is, and whether its identity persists through relevant changes. Drawing on personal-identity debates, especially Parfit, and work on artificial individuation, agency without consciousness, and responsibility, it argues that artificial agents raise a serious problem of identity and persistence even without settling questions of consciousness. Updating, memory transfer, replication, Claude Opus ۳’s retirement, and an AI agent’s unexpected action in a sportsclub booking system make the problem concrete. The editorial proposes three exploratory dimensions for examining artificial-agent identity: agentic memory, relations with the environment and user, and decision-making structure, and argues that none is sufficient on its own in borderline cases. Before asking “Is the machine conscious?” we should ask “Which machine?” and “Is it still the same machine?”

کلیدواژه ها:

artificial intelligence ، personal identity ، Persistence ، Individuation ، artificial agency ، moral responsibility ، consciousness ، human agency