This paper argues that the standard definition of the robot — a physically embodied, programmable machine that acts in the physical world — is a historical artifact of technological constraint rather than a principled definition of what a robot is. Building on the Law of Intelligence Propagation (LIP) and the tool inversion analysis of Paper Four, this paper introduces the concept of Distributed Agency Computation (DAC): the capacity to generate meaningful change in a system through computational signal rather than physical actuation. The argument proceeds through three independent lines of evidence. First, the ISO 8373:2021 definition of robotics is analyzed to show that its physical embodiment requirements encode a technological constraint of the 1950s–2000s rather than a structural property of agency. Second, Latour's (1992) actor-network theory and Gibson's (1979) ecological theory of affordances together establish that agency is a relational property distributed across agent-environment configurations, not a property of physical form. Third, plant electrical signaling (Fromm Hedrich et al., 2016) provides the biological extreme case: functionally significant agency does not require a nervous system, a body plan, or locomotion. The convergence of these lines establishes the Independence Illusion: the belief that AI requires physical embodiment to act in the world. This illusion misidentifies the body as the agent when it was always the interface.
Kyungae Ahn (Wed,) studied this question.