Understanding One Another: Making Out AI Meanings with Boolean Equations

Autonomous AI agents raise the issue of semantic interoperability between independently architectured and differently embodied intelligences. This paper offers an apporach to the issue that is close in spirit to the way humans make out meanings. Using a mathematical model of cognition, it is shown how autonomously developed conceptualizations can bootstrap and unravel each other’s meanings ad hoc. The domain general methodology is based on own Boolean capabilities, and any shared outside environment. No prior provisions are required. The formalized cognitive process consists of constructing, and solving, Boolean equations that are grounded in the shared environment. The process yields a testable conjectured conceptualization of the other, along with a testable conjectured translation that maps from that conceptualization to one’s own.