Machine Learning and the Philosophy of Photography

This chapter develops a model of vision that distributes information processing, storage, and recall between ‘hardware and wetware’, demonstrating how individual perceptions can be aggregated into a larger, collective perceiving machine. The design and potential reorganisation of such cognitive assemblages are explored through a practical experiment: repurposing a typical computer vision classifier using focal length data. The resulting system is re-engineered to disregard
what
images depict, becoming sensitive only to
how
they depict. This discussion of the computation of space concludes the chapter and section, hinting at the complementary aspect of photographic imaging: the mechanical calculation of time.