The project studies the evolution of some digital organisms in a simple world whose rules they learn by trial and genetic evolution. The creatures and the world are modeled to resemble a simple natural ecosystem. The world is a cellular automaton responsible of food supply and some other factors. The creatures are digital organisms capable of reproduction, eating, movement on the 'world' grid and decision making in response to the information they receive from the world. The world challenges the creatures by adding complexity to the rules of interaction. In different scenarios I experimented with, the food supply decreased to a very low level, a different type of terrain - a trap - was added or some portions of terrain contained poisonous food supplies. The creatures are forced to learn the new rules (or they die). The evolutionary changes in the genotype show that a benefic evolution occurs very fast - at a scale of 100 generations or less. The computer simulation generates many types of information regarding the evolution that can be analyzed with a statistical kit. Beside that, theprogram can be run in user mode - the user is free to explore the behavior of the individual creatures.