Information Technology Reference
In-Depth Information
to a specific culture, social group, style or individual. After all, what is taught at art
schools? Students learn the basic craft of their medium, they are exposed to many
exemplars, they try and fail, try again, receive critique and feedback with a hope of
improving with experience. But as has been pointed out by Harold, rule following
isn't enough, art is an ongoing dialogue.
A lot of generative art software encodes specific forms of aesthetic judgement.
The artist/programmer carefully chooses specific rules so as to create a system that
generates pleasing aesthetics for
them
(which in turn may change after being ex-
posed to computer aesthetics or even the aesthetics of the artwork-in-progress).
Therefore, in a sense, this software is “evaluating” what it is doing (as it is doing
it), but not in the way that a human does. It is an evaluation done for aesthetic pur-
poses. However, the judgement originates with the programmer, not the program, so
it becomes a continuous scale of how much is imbued to each.
A program that can adapt can learn, and hence change its judgement. That we
know to be possible (using evolutionary algorithms or machine learning for exam-
ple), but as Frieder points out, the baby may never get out of its tiny pink shoes.
Perhaps we need to wait until machines have their own social evolution.
Frieder also raises the point that aesthetics is tied to the phenomenology of sen-
sual perception—how else could we appreciate work like that of the artist James
Turrell for example? It is difficult to imagine a machine experiencing such a work
and coming to a similar aesthetic understanding, unless that machine had very simi-
lar phenomenological perception to us, or had sufficient knowledge about us (our
perception, cognition, experience) and physics, to infer what our understanding
would be. The same provisos apply to a machine originating such a work.
But while there may be many areas of human aesthetics, cognition and perception
that are currently “off limits” to machines, it does not necessarily preclude machines
that may be able to originate something that humans find aesthetically valuable.
Indeed, a lot of “computer art” has given us very new aesthetics to contemplate.
Paul Brown (PB):
I am very aware that writing too briefly opens up the oppor-
tunity for misunderstanding (I suspect Darwin said this?). But, to try:
One of the major themes in human development has been the revealing of struc-
ture (logic) through the observation and analysis of phenomena. Let me suggest
that this point of view, in it's extreme perhaps, believes that all phenomena can be
explained in some rational manner. In the history of art this complements the “classi-
cal” roots of art and leads directly to the work of Peirce, Saussure, Cezanne, Seurat,
etc., and then into the 20th century experiments in constructivism, rational aesthet-
ics, analytical philosophy, cybernetics, conceptualism, systems art, and so on. . . We
could call this approach Modernist but this term is fraught with misunderstanding,
especially as it is so (ab)used within the art world.
Another major theme suggests that understanding comes via entering into a rela-
tionship with the phenomena that enables the spontaneous emergence of meaning.
We use terms like “intuition” and “inspiration”. The extreme of this point of view
suggests that critical analysis is unnecessary and may actually be counter-productive
(and in theological “controlling” manifestations that it should be suppressed).
I know of several artists who, after pursuing PhD “practice-based” research, are now
