Project FRACTAL MIND: Recursive Mathematical Imagery and Altered Perception in AI

Abstract

Project FRACTAL MIND is an independent research study investigating how artificial intelligence image-generation systems interpret recursive mathematical imagery under different cognitive and perceptual framing conditions. The study will compare three AI systems—ChatGPT, Gemini, and Grok—using two mathematical fractals, the Mandelbrot set and Buddhabrot. Each system will generate images under three standardized conditions: neutral mathematical representation, altered-perception framing, and perceptual reintegration. Eight independent generations per combination will produce 144 images for comparative analysis. The research will examine whether changes in prompting influence the emergence of recognizable forms, anthropomorphic features, symbolic structures, visual complexity, and perceived depth. Images will be evaluated using predefined criteria and independent human assessment. Inspired by questions surrounding psychedelic perception, pattern recognition, and artificial cognition, Project FRACTAL MIND seeks to investigate how generative AI represents visual patterns associated with altered states of perception. The study does not presume that AI systems possess consciousness or subjective experiences. Instead, it aims to establish a reproducible experimental framework for examining the relationship between mathematical recursion, generative imagery, and perception-related language.