The Machine Does Not Know What It Is Doing. Neither Does the Artist: A Case Study on Coinsnobiety.
Generative AI is often dismissed because the machine does not know what it is doing. This paper accepts the premise and rejects the conclusion: artists do not know either. From Plato’s Ion and Kant’s genius to Duchamp’s “art coefficient”, not-knowing has belonged to the description of the artist, never to the case against authorship. What separates artist from machine is not foresight but the handling of the unforeseen. We call this Productive Non-Knowing: the process by which a divergence is recognised, used to reorient the work and taken up as a decision someone answers for. Recognition is read as abduction in Peirce’s sense and tested by three criteria, fertility, revision and exposure, that distinguish it from retrospective projection. Within the framework of Augmented Imagination, imagination is attributed to the human, generativity to the human–machine system, and authorship to whoever answers for the work. The case is Coinsnobiety (2024–2026), ten AI-generated coins whose misspelled luxury names became a script, then a foil-printed edition.
Authors
- Gabriele Moschin (ORCID: https://orcid.org/0009-0000-4691-4666)
- Andrea Filippi (ORCID: https://orcid.org/0009-0007-7963-723X)
Institutions
- Domus Academy (IT)
Publication Details
- Journal
- Zenodo (CERN European Organization for Nuclear Research)
- Published
- 2026-11-20
- DOI
- https://doi.org/10.5281/zenodo.24382283
- Primary Topic
- Art, Technology, and Culture
- Type
- preprint