dc.description.abstracten |
Artificial life is the field of study where researchers use simulations to understand
natural life. One of the notable simulations of artificial life is called Tierra. In Tierra,
self-replicating programs acting as organisms compete for CPU time and RAM un-
der the pressure of natural selection. In time, various types of organisms develop,
including parasites, hyperparasites, and even some form of social relationships. The
problem with this simulation is that informational complexity stalls after some initial
growth; evolution stops producing new types of organisms and is not open-ended.
Within the Tierra simulator, the memory address space is one-dimensional; no mat-
ter how far away the resource is located, the effort is the same. This thesis is focused
on replacing one-dimensional or pseudo-two-dimensional memory space of Tierra
and Tierra-like simulators, like Avida or Amoeba, with real two-dimensional space
and refining the concepts of the location and local access to achieve a more open-
ended evolution. |
uk |