An understanding of the evolutionary mechanisms that lead to
increased genome capabilities is essential for determining the
potentials and limits of living systems. We evolve abstract
aesthetic
imagery on the basis of an
image connectivity trait that has a theoretically determined maximum
value. This trait can be influenced by
``development''
genes. We investigate simulated
evolution when the development genes are not present, when they are
fused with the existing genotype, and when they function
semi-autonomously by residing in symbionts. Genome fusing initially
impedes evolution while symbiosis initially accelerates evolution.
Our findings suggest that symbiosis events play a more important
role than genome fusion events for sustaining open-ended evolution.
Keywords: Evolving expressions, computational aesthetics, simulated evolution.