The project makes reference to 'The Next Documenta Should Be Curated by an Artist' (e-flux, 2003) - which questioned the structures of the art world and the privileged position of curators within it – and extends this questioning to machines. Under this overarching concept, two parallel experiments have been realised to date, applying various machine learning techniques (a subset of AI) to ‘curate’ datasets derived from various biennial exhibitions.

Experiment B³(NSCAM) is a collaboration with artists Ubermorgen, commissioned by The Whitney Museum of American Art for its online platform artport and Liverpool Biennial. It uses datasets from the two commissioning institutions, amongst others, archival curatorial statements, artists names and biographies from past editions. Processing datasets linguistically and semiotically, the AI algorithms ‘learn’ their style and content, calculating a future probability for words to appear, cutting and mixing them together, to generate endless combinations of possible instances of Biennials in flux.

A parallel experiment, AI-TNB is a collaboration with Eva Cetinić and MetaObjects with Sui, exploring human-machine co-authorship. It uses data from Liverpool Biennial 2021 – the photographic documentation of artworks, their titles and descriptions – and applies Open AI’s ‘deep learning’ model CLIP to generate new interpretations of, and connections between, works and texts. As visitors navigate the project, they create their own paths through the material, each journey becoming a co-curated human-machine online iteration of the Biennial saved to the project’s public repository.