WORKS – ARTIFICIAL AGENTS – SOCIAL EVOLUTION
R.C. Hoetzlein (c) 2006-2008
Version Beta. Geneve, Switzerland. Oct 31-Dec 15th, 2008.
Social Evolution is an experiment in simulated societies. Character agents are introduced that engage in a range of activities with internal parameter optimization carried on to offspring. Actions include standing, sitting, eating, sleeping, killing, mating, walking, raking (harvesting) and seeding (planting).
Suprising outcomes of the project – not pre-programmed – include the group ability of the characters to cluster into towns to conserve energy, local sharing of resources, patterns in which individuals chase one another in circles (3 and 4-way cycles), and social hierarchies similar to caste systems. The last phenomenon is particularly surprising considering that all characters begin as equals. It was found that simulations in which land ownership is included (30% of the people owned 95% of the land), there are discrete bands of wealth based on second and third order modes of behaviors such as stealing from those who steal. In addition, the poorer individuals were found to cluster in groups around the wealthy despite these disadvantages.
Social Evolution was exhibited at Version Beta for the final exhibition of the Center for the Contemporary Image in Geneve (2008), Switzerland, for the Digital Days Festival at the Arlington Theatre in Santa Barbara (2005), at the 2nd Beijing International Arts & Science Exhibition (2006), Tsinghua Univ, China, and at the 25th Dorkbot So. Cal (2007).