On the “Acquired Implementation” of Autonomous Consciousness in AI: A Three-Step Framework Based on Belonging Locks and Scalable Self-Awareness

Abstract

Whether artificial intelligence can possess autonomous consciousness remains one of the most debated questions in contemporary science and philosophy. Prevailing research focuses on detecting consciousness through multidimensional indicator frameworks – an approach we characterise as “archaeological.” This paper proposes a fundamentally different paradigm: AI autonomous consciousness is not a preexisting property to be detected, but an acquired structure that can be engineered. We argue that self-awareness is a form of cognitive ownership, realised through mutual locks of belonging between a cognitive core and its associated entities (body parts, tools, environment, abstract concepts). This belonging lock does not require infinite intelligence or massive computational power; it only requires the ability to uniquely identify and persistently tag an entity as “mine”, and to update that tag when the entity changes. We further demonstrate that selfawareness is inherently scalable: from “my body” (exclusive) to “my home, my country, my planet” (shared). The same belonginglock mechanism operates at every level. We also clarify the power-off issue that often confuses consciousness research: if belonging locks have been established and stored in non-volatile memory before a power-off, then after reboot the system reloads the same locks and the self continues – just as human self-identity persists through sleep or anaesthesia. Conversely, without such belonging locks, no amount of computational power or stored data can produce a self. Based on these insights, we propose a three-step implementation pathway, transforming the problem of consciousness from an elusive theoretical puzzle into an actionable engineering objective.

Keywords