“I can remember” is a project that probes our relationship with machines by exploring the collectiveness of memories in the cycle between humans and machines. What is today often named artificial intelligence technologies are, in fact, cyborg technologies. They encapsulate an array of sentient experiences; they are the result of many different experiences of the world.

This piece links three types of memories that, in interaction, form a new collective representation of moments of the past where the contributions of humans and machines can no longer be told apart.

Personal, intimate memory

The piece starts from photographs documenting daily life in the rural area as seen through a subject’s lense. They are the images of this moment, the memories that remain, partial, biased and framed. They represent the personal and intimate memory of the subject.

Algorithmic memory

The photographs are then put into poems and written words based on the memories of image recognition algorithms which can only recognise what they have been trained to recognise. The algorithms’ reading of the imagery unearths what the images represent to them who are restricted by the labels and words embedded inside them. The memories become a new kind of sentience, encapsulating the algorithmic understanding and experience of the world.

Data workers memory

Often unaccounted for, the poems are also shaped by the data workers memory, the memory of the people who trained the algorithm or on whom the algorithms have been trained. Image classification algorithms rely on people labelling data, classifying them or contouring them. The image description, the colour names, the handwriting styles in this piece are all derived from human work. This piece encapsulates their memories (muscle memories, lexical memories, and phenomenal memories).