Tag AI-Sandbox

This paper develops a Tag AI-Sandbox through competition between different subjects and the survival aspect of a game of tag.Using genomes, we create sophisticated entities that coordinate themselves for the benefit of their survival.Through a number of tryouts using reinforced learning with genomes as the base, we look for the fittest survival genes and self-development strategies.We later introduce their clever reasoning development abilities through the use of more complex environments at random which leads to more human-like skills.This investigation has been done before on other games like Open-AI hide & seek, but for this research we added more human-like behaviors and used genomes to see their ingenuity from a biology perspective.This research supports the development of the curriculum by introducing students from the Department of Computer Science (CS) to the development of artificial intelligence, and how other areas like biology can bring further development into the technology that they develop.We compare how this relates to the biological structure of organisms towards their decision-making based on tasks, random initialization, and constraints.Based on their behavior, we will propose fine-tuning ways to improve the research further towards a replication of human behavior through the use of genomes.