I just found an article online that refutes the claims Timo Arnall makes about his “Robot Readable World” video that I provided a link for in a previous post.
According to Jay Owens, “Arnall’s video is actually a depiction of the debug output of machine vision, processed and formatted to be human-readable. It looks the way it does because programmers threw together a visualization to help them understand why the machines weren’t seeing what they were supposed to be seeing, or to confirm that they were seeing what they were supposed to be seeing when everything seemed to work. It’s an attempt to peer into the mind of an algorithm. Its aesthetic core comes from the same place as scrolling lines of program output in a VT-100 terminal or the bright orange of safety vests.”
In other words, Owens argues that the images in Arnall’s video do not depict robot perception and it is perhaps impossible for us to see how a robot sees.
Owens also discusses what he perceives as “an unavoidable mistake” at the core of the New Aesthetic that is “the multiplication of entities and agents” with our unintentional ( or intentional?) personification of robots.