First Draft on the xInf Model for Universal Physical Computation and Reverse Engineering of Natural Intelligence

Abstract: Turing Machines are universal computing machines in theory. It has been a long debate whether Turing Machines can simulate the consciousness mind behaviors in the materialistic universe. Three different hypotheses come out of such debate, in short:(A) Can; (B) Cannot; (C) Super-Turing machines can. Because Turing Machines or other kinds of theoretical computing models are abstract objects while behaviors are real observables, this debate involves at least three distinct fields of science and technology: physics, computer engineering, and experimental neuroscience. However, the languages used in these different fields are highly heterogeneous and not easily interpretable for each other, making it very difficult to reach partial agreements regarding this debate, Therefore, the main goal of this manuscript is to establish a proper language that can translate among those different fields. First, I propose a theoretical model for analyzing how theoretical computing machines would physically run in physical time. This model, termed as the xInf, is at first place Turing-complete in theory, and depending on the properties of physical time, it can be either Turing-equivalent or Super-Turing in the physical universe. The xInf Model is demonstrated to be a suitable universal language to translate among physics, computer engineering, and neuroscience. Finally, I propose a conjecture that there exists a Minimal Complete Set of rules in the xInf Model that enables the construction of a physical machine using inorganic materials that can pass the Turing Test in physical time. I cannot demonstrate whether such a conjecture to be testified or falsified on paper using finite-order logic, my only solution is physical time itself, i.e. an evolutionary competition will eventually tell the conclusion.