Artificial Life as a Petri Dish
Artificial life simulations reinforce hypotheses that possibly a species’ total skill set necessary for optimum survival may first be assembled through symbiosis, and only later encapsulated as heritable traits in single individuals.
If new species can be created through genetic integration of traits first sampled in a symbiotic relationship, the Baldwin Effect would overcome the problem that most random variations are detrimental. The Baldwin effect, proposed by James Mark Baldwin, suggested that an organism’s ability to learn new behaviors (e.g. to acclimatize to a new stressor) would affect its reproductive success and so affect the genetic makeup of its species through natural selection. Baldwin proposed his hypothesis in the 1890s at about the time that August Weismann was attacking Lamarck, and Baldwin was dismissed in his day as aligned with Lamarck. But today the Baldwin effect is generally recognized as part of the Modern Evolutionary Synthesis.
Although natural selection, rather than self-organization into a cooperating unit, is typically seen as the engine of evolution, some artificial life models suggest that the process of ordering, rather than solely the selection of pre-ordered objects, may drive evolutionary innovation.
Artificial life explores evolution by recreating it in synthetic systems. In multi-dimensional agent gaming systems, such as the Prisoner’s Dilemma, and simulations of artificial life, game rules build hypothetical worlds. Scenario-building sheds light, not only on the origin and evolution of life, but also on our own problem-solving methods. Artificial life, and robotic research, raise a question: Is intelligence composed of principles that modify behaviors housed in various physical embodiments? If so, can artificial life programs evolve plausible designs for life that does not yet physically exist? Although intelligence was traditionally seen as an attribute of complex life, some wonder if it might in the future be an attribute of behaviors of objects not biologically defined as alive, such as robots or artificial life.