the process of image building as a machine fascinates me.
these generated images are not maps. they do not, nor will ever, function as maps in the traditional sense. they are instead indicators of a network and system that both
flattens and expands.
isn't flattening and expanding what maps do?
reducing everything down, but also taking a single thing and expanding it
i notice how much of machine generated images, videos, and realities are in service of self-driving cars. content is created as data for these cars, and the machines that run them.
(it is turtles all the way down)
to generate videos, like dash-cam videos, you predict typical scenes and patterns. you can then multiply the possibilities and train with them. to learn to generate a map, you learn how to make new ones by learning typicality, and respond to new landscapes. (that is the hope at least).
within a map, the world is delineated into three colors
{urban: beige, forest: green, water:blue}
roads, the access points for cars, the place you
(the machine) can be, are white.
cars do not need to run through the forest, or the ocean. in these world(s) that the cars live in, the forests are contained.
or they are replaced with something that serves the cars.
or they are simply gone.
where do we fit in to the world of machines?
where do i fit, except, economically, as physician or trainer, realistically as fodder, and simply as a breeding ground for soft technologies of industry and control?
where do the forests fit in?