Curb Your Self-Modifying Code

Abstract

Self-modifying code has many intriguing applications in a broad range of fields including software security, artificial general intelligence, and open-ended evolution. Having control over self-modifying code, however, is still an open challenge since it is a balancing act between providing as much freedom as possible to not limit possible solutions and restriction to avoid security issues and invalid code or solutions. In the present study, I provide a prototype implementation of how one might curb self-modifying code by introducing a control mechanisms for code modifications within specific regions and for specific transitions between code and data. I show that this is possible to achieve with the so-called allagmatic method -an approach to formalise, model, implement, and interpret complex systems inspired by Gilbert Simondon's philosophy of individuation and Alfred North Whitehead's philosophy of organism. Thereby, the allagmatic method serves as guidance for self-modification based on concepts defined in a metaphysical framework.

Programming Data Structures for Large-Scale Desktop Simulations of Complex Systems

“…These model building blocks can be put together in an automatic way by concretising them into more and more concrete models of complex systems. In self-modifying code, on the other hand, it allows controlled modification of specific code regions [12]- [14].…”

“…The allagmatic method was first developed in C++ using classes and generic programming (template metaprogramming) [15] to describe the system metamodel [6]. In later studies on self-modifying code, the system metamodel was reimplemented in C# [12]- [14] because it allows compiling and running code as an object, provides reflection capabilities, and generally comes with a vast API. Compiling and running code as an object in C# is faster than writing code into a file, call the system to compile and then run the executable in C++.…”