Generation and Use of Knowledge in a Collective of Agents

Abstract

A computer model of generation and use of information in a collective of agents is constructed and analyzed in the current article. The model is based on observations of persistence hunting on antelope in the hot Kalahari Desert (South Africa), as described in L. Liebenberg's book “The Origin of Science”. The model examines the behavior of a group of agents-hunters pursuing an antelope agent. The hunt is conducted unarmed until either the antelope or the hunters are exhausted. The hunters exchange information about their own location and that of the antelope. This information exchange increases the hunting success of the hunting collective. Experienced hunters can form and use simple hypotheses about the location of antelope and their goal-directed actions. The use of such hypotheses leads to increased hunting success. The formation and use of these hypotheses can be viewed as the origin of certain precursors to scientific cognitive processes. The model demonstrates the effectiveness of information exchange and hypothesis formation in a team of hunting agents.