Conceitos essenciais
Computational creativity research faces a fundamental paradox in its attempt to model or replicate the open-ended, dynamic process of human creativity within the bounded state spaces of digital systems.
Resumo
This paper explores the paradoxical nature of computational creativity research, focusing on the inherent limitations of closed digital systems in emulating the open-ended, dynamic process of human creativity.
The authors first discuss the concept of Sudden Mental Insight (SMI) as a recognized form of creative behavior, and how computational systems like Aaron and the Serendipity Machine have been able to elicit such responses from human observers. However, the authors argue that these systems are ultimately constrained by their pre-programmed state spaces and cannot achieve the level of originality and autonomy seen in human creativity.
The paper then delves into the theoretical background of computational creativity research, examining both procedural approaches (rule-based systems, genetic algorithms, case-based reasoning) and representational approaches (shape grammars, object-based representations, recognition algorithms). While these methods can contribute to or support the creative process, the authors contend that they are fundamentally limited by their inability to redefine their own state spaces, a key characteristic of human creativity.
The authors introduce the concept of the State Space Paradox (SSP), which arises when closed computational systems attempt to replicate or automate creative behaviors. They explain that any outcomes generated by such systems will always be a proper subset of their pre-defined state spaces, and thus cannot truly break free of their programmed constraints to achieve genuine creativity.
The paper concludes by discussing the implications of the SSP on the future of creativity-related computer systems, emphasizing the cultural and contextual fluidity of creativity itself and the challenges of producing truly creative outcomes within the limitations of algorithmic structures. The authors suggest that recognizing and embracing the limitations and potentials of digital systems could lead to more nuanced and effective tools for creative assistance.
Citações
"Creativity is basically a rare human act. There are very few individuals who are considered truly creative and their lives are finite. This is a tautological outcome. If there was an overabundance of creative acts, we would no longer be willing to call them creative – or the word creative would have an entirely different meaning."
"For a machine to match that, would require the machine to have all traits of humans."