Information Theoretical Principles of Software Development

EasyChair Preprint 9222

Abstract

Contemporary software, which started from a simple observation by Alan Turing that a man in the process of computing numerical functions can be replaced by a machine that is capable of only a finite number of simple operations, has become a powerful means for advanced machines, automating business processes, and mimicking behaviors learned through processing big data. Software is all pervasive and touches us in some form or another by facilitating communication, collaboration, and commerce on a global scale. However, on the flip side, it also threatens our privacy and data security. Despite many layers of complex systems and processes attempting to protect our software systems, they are still vulnerable to hackers and fraudsters exploiting global access through the Internet.

In this paper, we argue that the increased complexity as well as the resulting instabilities and inefficiencies are a result of foundational issues of the computing model upon which current software systems are built. We use the general theory of information to suggest ways to improve the current state of the art of software development, deployment, and operation through the infusion of self-monitoring, self-regulation, and self-protection processes at both the component and system levels. A new computing model using super-symbolic computing provides an autopoietic, and cognitive process overlay over the current symbolic and sub-symbolic computing structures without having to change them. This software system is analogous to biological systems using the mammalian neocortex to manage the knowledge obtained from various reptilian cortical columns using embedded, embodied, extended and enactive cognition.

Keyphrases: Digital Genome., General theory of information, Structural Machines, genome, information processing, software

@booklet{EasyChair:9222,
  author    = {Mark Burgin and Rao Mikkilineni},
  title     = {Information Theoretical Principles of Software Development},
  howpublished = {EasyChair Preprint 9222},
  year      = {EasyChair, 2022}}