Aesthetic archives is a collaborative project between a human curator and an AI programme investigating the importance of order and categorisation regarding networked items in digital archives. Digital archives are predominantly organised by provenance, similar to how they are catalogued in physical archives. Current digital cataloguing methods restrict creative research and limit results. This project uses supporting visual data from the Science Museum Group, the Tate, and the Baltic to recognise the gaps in these landscapes. Following a unique methodology to source and group existing archival items, new networks have been catalogued under adjective definitions.
Condensed methodology = Online generated random adjectives + numbers > Numbers put into open-access archival finding aid > Human selects work and downloads high res image > ChatGPT generates an item description from item name > Original title + new description from ChatGPT > Human assigns adjective category and creates cluster to creatively network images = aesthetic archives.
Aesthetic archives: How do networked items in online archives need to be categorised, catalogued and curated for creative research purposes?.
*Visual data from the digital finding aids of the Science Museum Group, the Tate, and the Baltic.