CLUSTER CONCEPT DYNAMICS LEADING TO CREATIVE IDEAS WITHOUT CRITICAL SLOWING DOWN

Abstract

We present algorithmic procedures for generating systematically ideas and solutions to problems which are perceived as creative. Our method consists of identifying and characterizing the most creative ideas among a vast pool. We show that they fall within a few large classes (archetypes) which share the same conceptual structure (Macros). We prescribe well defined abstract algorithms which can act deterministically on arbitrary given objects. Each algorithm generates ideas with the same conceptual structure characteristic to one of the Macros. The resulting new ideas turn out to be perceived as highly creative. We support our claims by experiments in which senior advertising professionals graded advertisement ideas produced by our method according to their creativity. The marks (grade 4.6±0.2 on a 1–7 scale) obtained by laymen applying our algorithms (after being instructed for only two hours) were significantly better than the marks obtained by advertising professionals using standard methods (grade 3.6±0.2)). The method, which is currently taught in USA, Europe, and Israel and used by advertising agencies in Britain and Israel has received formal international recognition.