Information Technology Reference
In-Depth Information
the processes at work when software produces pieces of art. However, our position
is that how the work is produced should form only one part of the framing of gener-
ative artworks, and they would be culturally more important if the pieces themselves
offered a reason for audiences to think about certain issues, or if they invoked certain
feelings or moods.
Another context within which our project can be seen is that of the graphics sub-
field of
Non-Photorealistic Rendering
(NPR). Here, the emphasis is on producing
software which simulates natural media such as paints, pencils, canvases, pastels,
and their usage in paint strokes, filling regions of colour, etc. Much of the pioneer-
ing work in this area has ended up in software such as Adobe Illustrator, which
give artists new digital tools and media to work with. As a good example, James
Faure-Walker (
2006
) mixes simulated paint with real paint in his art practice. NPR
software is designed along solid software engineering and Human-Computer Inter-
action lines to be useful and reliable tools for artists and designers. Moreover, given
that the consumers of such software are largely within the creative industries (and
hence possibly perceived to be worried about creative software taking over some
of their responsibilities), there have occasionally been mistakes of judgement from
NPR experts keen to downplay claims of creativity in their software. In particular,
in a standard NPR textbook, Strothotte and Schlechtweg state that:
Simulating artistic techniques means also simulating human thinking and reasoning, espe-
cially creative thinking. This is impossible to do using algorithms or information processing
systems. (Strothotte and Schlechtweg,
2002
, p. 113)
It is difficult to tell whether this statement is denying the subfield of Computational
Creativity research, or the entire field of Artificial Intelligence. In any case, the state-
ment attempts to reinforce the myth that creativity is beyond scientific study, which
is one of the main issues addressed within creativity studies and Computational Cre-
ativity research in particular, as addressed most vocally by Boden (
2003
).
Other NPR researchers are more enlightened, however, and supply their tech-
niques with more intelligent abilities. For instance, with their saliency-adaptive
painting research, Collomosse and Hall (
2006
) enabled their NPR system to de-
termine the most important regions in an image using an evolutionary search. This
enabled the production of painterly renditions of digital images with special atten-
tion paid to the most salient regions, which is more in line with the way in which
painters understand the content of the pictures they are painting.
To summarise our placing of The Painting Fool project into various contexts, we
observe that it is generative art software which has evolutionary search and non-
photorealistic rendering abilities, in addition to the ability to construct scenes in a
similar fashion to AARON. It is being engineered and further trained to transcend
most generative art projects by addressing higher level artistic behaviours such as
critical ability and cultural awareness. As such, it is designed not as a tool for artists
to employ, but rather as a creative collaborator, or even an independent artist.
