Guideless Artificial Life Model for Reproduction, Development, and Interactions.

Area of Science:

• Artificial Life (ALife)
• Computational Biology
• Evolutionary Computation

Background:

• Existing ALife models like Tierra and cellular automata have limited morphological/behavioral expression.
• Karl Sims' virtual creatures offer high expressiveness but lack complex reproduction and development.
• Genetic algorithms in current models can reduce biological diversity due to predefined fitness functions.

Purpose of the Study:

• To propose a new ALife model integrating Tierra/cellular automata mechanisms with free 3D cell movement.
• To enable simulation of reproduction, development, and interactions with high expressive power.
• To demonstrate a model that evolves via mutation and natural selection, avoiding predefined fitness functions.

Main Methods:

• Developed a novel ALife model combining Tierra and cellular automata principles within a 3D freely moving cell.
• Implemented a simulation environment without predefined fitness functions or creature forms.
• Utilized mutation and natural selection as the sole evolutionary drivers, akin to Conway's Game of Life.

Main Results:

• Proof-of-concept demonstration of the model's capabilities.
• Emergence of two distinct creature types: dumbbell-shaped and reticulated.
• Observed unique reproductive strategies in virtual creatures, closely resembling natural organisms.

Conclusions:

• The proposed model successfully simulates reproduction, development, and interactions with high morphological and behavioral expressiveness.
• The model's evolution through mutation and natural selection demonstrates potential for generating complex virtual ecologies.
• This approach offers a promising avenue for modeling intricate biological processes and creating novel virtual life forms.