This study explores the intersection of creativity and computation through a multi-criteria approach to evolutionary art. By leveraging evolutionary algorithms, particularly genetic algorithms, the research examines how images can be iteratively refined through computational techniques. Aesthetic evaluation remains a core challenge, and this work introduces a multi-metric assessment system incorporating simplicity, contrast, symmetry, and self-similarity to bridge human perception with algorithmic outputs. Inspired by G. Seurat’s pointillism, the study evaluates how these factors influence artistic judgment. Experimental results indicate that while fitness functions optimize pixel similarity, they do not fully capture human aesthetics. The findings emphasize the necessity of multi-criteria evaluation in evolutionary art and suggest future work on adaptive weighting of aesthetic metrics to improve automated assessments.