Implementation of the efficient truly two-dimensional artificial life simulation

Abstract

This work aims to develop an efficient implementation of the two-dimensional artificial life simulation Fungera (Yavorskyi, 2021, GitHub). Initial Python-based implementation of the Fungera allowed to verify that the two-dimensional approach is viable to study evolutionary processes. Nevertheless, this prototype implementation is too slow to study emerging evolutionary innovations and emerging ecological relations efficiently. One-dimensional simulator Tierra, the first successful artificial life evolution simulator, required several million iterations for the emergence of parasites, hyperparasites, “social behavior”, and so on. Because of its twodimensional nature, Fungera would require orders of magnitude more time for such events leading to many weeks of simulations using Python-based implementation. So, this work aims to develop efficient implementation based on C++. It should support easy modification of the instruction sets used by the simulation. The other way this project can be considered as an experiment on two-dimensional computer architectures, inspired by the esoteric programming language Befunge.