Brain Evolution as an Information Flow Designer: The Ground Architecture for Biological and Artificial General Intelligence

Abstract

For centuries, neuroscientists have identified a number of neural systems involved in sensory, motor, state control, and cognitive functions. Modern comparative studies have proposed their diversity, origins, and basic functionality across animal phyla. Despite a number of attempts, however, a common functional plan of the complex brain remains controversial. For example, there is currently no prominent theory of how neural networks are structurally comparable between phylogenetically distant animals such as vertebrates, octopuses, worms, and insects, in which there are distinguishably different brain architectures. This chapter attempts to identify the types of information flow patterns that were specialized during brain evolution, when these patterns appeared as a prototype, and how the flow systems have been shaped based on the common morphological architecture. In a notable case, a number of sensory associative centers show comparable patterns in mammalian, insect, and octopus brains, representing a common input and output flow of information. One can speculate that a common underlying structure is shared between various animals because of common functionalities that produce highly effective learning, memory, and autonomous cognitive tasks. Such an underlying structure could help establish a large-scale framework for comparison between phylogenetically distant animal brains and perhaps even form the groundwork for artificial general intelligence.