|
|
|
|
|
|
|
(2) Each agent in the site executes uptake of resources produced at the site. |
|
|
|
|
|
|
|
|
(3) Each agent in the site is charged a "maintenance" cost, which must be payed by the "subtraction" of specified resources from its reservoir. If the cost cannot be met, the agent is deleted (some of its resources may be returned to the site, depending on the particular model). Each agent also has a small random chance of being deleted ''without cause." |
|
|
|
|
|
|
|
|
(4) Each agent in the site tests to see if it has accumulated enough resources in its reservoir, via steps (1) and (2), to make a copy of its chromosomes. If so, it replicates itself, with infrequent mutations. |
|
|
|
|
|
|
|
|
(5) Each agent in the site, other than the replicates produced in step (3), tests to see if it has acquired at least one of the resources it currently needs for reproduction. If not, the agent migrates to one the neighboring sites. |
|
|
|
|
|
|
|
|
(6) The production function associated with the site adds a specified number of units of each resource to the site, for later uptake in step (2). |
|
|
|
|
|
|
|
|
The details of this basic cycle can be filled out in a variety of ways, depending upon the particular range of gedanken experiments of interest. Even the simplest models show surprisingly sophisticated evolutions. One of the earliest models produced evolving sequences of agents with ever longer, more complicated chromosomes, accompanied by a corresponding increase in the complexity of their interactions. The result was a "biological arms race" (Dawkins 1986), wherein defense tags became ever longer and offense tags developed ever more sophisticated matches to overcome the increasing defensive capabilities. More recent models, by a modification of the basic cycle, provide for the evolution of "metazoans"connected communities of agents that have internal boundaries and reproduce as a unit. With this provision, agents belonging to a connected community can specialize to the advantage of the whole community. For example, one kind of agent belonging to the community can specialize for offense, while a second kind specializes in resource acquisition (somewhat reminiscent of the stinging cells and cavity cells of the hydra). It is easy to show that intracommunity trading between these specialists yields a net increase in the reproduction rate of both. As a consequence the metazoans come to occupy a significant place in the overall ecology of agents. Many of the mechanisms investigated by Buss (1987) can be imitated by this model, including the evolution of cooperation between cell lines (cf. Axelrod and Hamilton 1981) and the origin of such developmental mechanisms as induction and competence (see Figure 21). |
|
|
|
|
|