The creative conundrum of artificial intelligence: are machines truly capable of creativity?

• Main Author: Lam, Ronald Kei Yin
• Other Authors: Melvin Chen
• Format: Final Year Project
• Language: English
• Published: Nanyang Technological University 2023
• Subjects: Humanities::Philosophy
• Online Access: https://hdl.handle.net/10356/171593
• Institution: Nanyang Technological University

In the age of burgeoning technological advancements, a compelling inquiry surfaces: Can artificial intelligence (AI) genuinely be creative? This exploration delves into key AI concepts such as the Turing Test and the Lovelace Objection, before addressing the attributional and definitional challenges of creativity. Upon examining eliminativist, experientialist, and functionalist accounts of creativity, it becomes clear that no single viewpoint is superior due to inherent limitations. The complexities embedded in these accounts emphasize the necessity of first defining our understanding of creativity itself. Following, I explore two historically accepted criteria of creativity: novelty and value. While I attempt to strengthen value by introducing the criteria of epistemic value, I acknowledge it falls short due to exceptions. As our analytical lens shifts to agency, I argue AI fails to satisfy two indispensable conditions intrinsic to genuine creativity: volition and responsibility, which confines it to the realm of instrumental assistance rather than that of independent creation. I then conclude that because of AI’s inherent inability to manifest agentic responsibility and volition, it cannot therefore truly be creative.