Description
The Athena Dataset (2022) by Jenn Karson is an interdisciplinary project that merges artistic authorship with machine learning processes. At its core is a dataset of hand-created glyphs, designed by the artist and used to train generative AI models. Through this system, the work explores how images are produced, selected, and interpreted by both humans and machines.
The project treats the dataset not as neutral input, but as a site of authorship and bias—revealing how decisions embedded in data shape the outputs of artificial intelligence. By translating abstract computational processes into visual, material, and time-based forms, the work maps the otherwise invisible terrain of “latent space,” where AI systems encode and transform information.
Spanning prints, animations, sculptures, and installations, The Athena Dataset emphasizes the entanglement of analog and digital practices, questioning boundaries between original and generated imagery. It foregrounds the human, technical, and cultural infrastructures behind AI, positioning machine-generated images as collaborative artifacts rather than autonomous creations.
