This short film by Tino Arnall was linked by Warren Ellis back in February. I find myself returning to or thinking about it again every few weeks. Warren likens it to a nascent machine intelligence learning to see, which is fascinating. But I keep considering all of these pieces of footage as visualizations of domain-specific heuristics of attention. Most of these are not novel ways of seeing, but rudimentary versions of ways that humans see already. We have special psychology for the recognition of faces, and a heightened capacity to recognize moving figures over stationary ones. This video makes me aware of the different qualities of my own perception, how the character of “paying attention” changes with the subject to which my attention is paid.