How Training Datasets Shape AI Aesthetics 🎨🤖

This presentation delves into the transformative role of datasets as aesthetic infrastructures, examining their influence on AI outputs and artistic expression. Topics include creative labelling, the political dimensions of datasets, and case studies such as Anna Ridler's curated artistry and insights from Kate Crawford's 'Atlas of AI'. We explore challenges like bias, dataset neutrality, and ethical implications, highlighting how artists contribute to transparency and innovation. By...