Increasingly, artists are using AI technologies such as machine learning in their creative practice. However, these technologies are often seen as a ‘black box’, lacking the visibility and transparency of more traditional techniques and media.

Some artists ‘open’ this black box by making the processes at work within machine learning systems visible to audiences. The Creative AI Lab highlights the importance of such work and conducts interviews and studio visits to explore and support it. We have found that the artists who engage in these techniques are contributing to Explainable AI and that their work invites a general public to engage with technical aspects of AI systems and to acquire urgently needed AI literacy. Examples of artists using these techniques are as follows.

Adam Harvey traces the origin of data sets that train facial recognition.

Allison Parrish allows visitors to manipulate language systems by playing with different sliders or weights.

Refik Anadol visualises the range of data within a cultural archive.

dmstfctn explores the training of machine learning systems used in cashierless Amazon supermarkets.

The Creative AI Lab is a collaboration between the Department of Digital Humanities, King’s College London, and the Research & Development Platform of the Serpentine Galleries, and was founded by Mercedes Bunz and Eva Jäger in 2019.

Project Team

Eva Jäger
Serpentine Galleries

Mercedes Bunz
Daniel Chavez Heras
Alaisdair Milne
Joanna Zylinska
Department of Digital Humanities, Faculty of Arts & Humanities, King’s College London

Funder

King’s College London, Serpentine Galleries, AHRC and Google’s Artists + Machine Intelligence Grant.

Acknowledgements

The Lab has also collaborated with Digital Theory Lab, NY and Rhizome, New York.