This contribution documents a series of artworks that explore how human and machine come together within ‘AI’ smartphone image optimisation. In smartphone camera systems, machine learning infrastructure creates a computational equivalent to various aspects of human experience, such as recognition, recollection, and anticipation. As humans interact with machine learning infrastructure through smartphone camera systems, they are mediated by these computational processes.

The impossibility of perfect translation is a widely acknowledged trope, yet translation remains a powerful act of meaning-making. This research-creation project investigates not what is lost, but what is gained through translation, by presenting and reflecting on our artistic re-interpretation of Dichterliebe, Robert Schumann’s nineteenth-century song cycle on texts by Heinrich Heine.