Computational Creativity in Media Production: At the Crossroad of Progress and Peril

Abstract

This study focuses on an approach to generate suggested video stories from raw source footage by using a multilayered hierarchical classification structure and narrative generation through the application of common patterns found within specific story genres. It frames the potential for synthetically facilitating these complex editorial decisions by analyzing current theories that define creativity and artistic quality and explores the demand for this type of engine within the creative community. The study defines the possible impact of Artificial Intelligence (AI) on artists and storytellers by examining the parallels between the effect of previous technological advancements and AI's bearing on contemporary creatives who are searching for effective modes of implementation. The potential of machine learning and artificial intelligence is rapidly advancing; therefore, this thesis traces the technology's development as a means to project future applications, specifically within media production. The Storytelling Model proposed in this report combines a multitude of algorithms that are either currently in the marketplace or exist as proof-of-concept studies. What is unique about the proposed model is the sequential and layered application of these algorithms for synthesizing a suggested video story. The recommended approach aims to provide users with the flexibility to easily modify the generated story through a prompt-feedback loop. The study's theoretical model is presented here in the hopes that it can be used as a starting point for future development.