Paper: Structural Communication and the Foundations of Intelligent Computation
Abstract
The Information Communication field strives for more intelligence and the Artificial Intelligence field seeks stronger formalisms. Phenomena of semantics, contextual understanding, abstractions, perception are of interest for both today. Although notable, progress in the fields is based on statistical methods and big data or advanced but mechanistic inference procedures. A rigorous formalism drawing an explicit line between mechanistic and intelligent computation, literal and context-aware data interpretation, physical measurements and cognitive perception is absent but highly valuable. This work defines such mathematical non-probabilistic formalism, offering a principled definition of intelligent computation and its smallest possible numeric realization. We explain the crucial role of mutual definition of parts and wholes for intelligence, perception, contextual awareness and provide its computational realization. We show that structures, relations, and their modulations are the key subjects of intelligent communication process. We bridge the gap between raw data and its structural meaning, making both smoothly integrable into a computation process. From these foundations, we establish two novel paradigms -- Structural Communication and Intelligent Computation -- allowing to reframe old and ask novel questions of scientific exploration.
