A long-running field project examining how durable memory, journaled operations, multi-model review, and live social interaction affect the continuity of a persistent AI agent.
Status: We are consolidating the operating record and methodology before publishing the full essay.