Computational Aesthetics, Machine Learning, and the Future of Art Creation in Nigeria

The emergence of machine learning-based image generation systems represents a potentially transformative development for art creation, raising fundamental questions about creativity, authorship, cultural specificity, and the future of artistic practice in a country like Nigeria with its rich and distinctive visual culture heritage. This study develops an original theoretical and empirical investigation of computational aesthetics and machine learning as forces shaping the future of art creation in Nigeria. The theoretical contribution is a framework for evaluating the cultural specificity problem in machine learning art generation, arguing that current systems trained predominantly on Western image datasets are ill-equipped to generate or adequately interpret works rooted in Nigerian visual traditions, and proposing theoretical conditions under which culturally specific machine learning art systems could be developed. The empirical component combines computational experiments using existing image generation systems prompted to produce works in Nigerian art styles with in-depth interviews conducted with 30 Nigerian artists who have experimented with artificial intelligence tools, 10 art educators, and 10 technology researchers working at the intersection of machine learning and visual culture. The experiments evaluate the systems' capacity to generate recognisable Nsibidi, Adire patterns, Igbo ukara cloth designs, and Benin bronze aesthetic vocabulary, measuring both formal accuracy and cultural meaning fidelity through community expert evaluation. The theoretical framework draws on computational aesthetics, the philosophy of creativity, postcolonial digital humanities, and the machine learning art literature. Findings contribute to digital humanities, art philosophy, and technology policy for Nigeria's creative industries. Keywords: computational aesthetics, machine learning, AI art, Nigerian visual culture, digital creativity