On the Role of Computers in Creativity-Support Systems - Computational Creativity Research: Towards Creative Machines - page 117

that her kitchen was a manufacturing facility for her business. The judges, while
finding the argument 'ingenious and appealing,' ruled it 'insufficient' nonetheless.
3. John Casti ([ 5 , pp. 7-10]) comments on the fate of Immanuel Velikovsky's theory
as outlined in Worlds in Collision, which hypothesized Earth's encounters with
a large comet expelled from Jupiter and provided explanations for many biblical
events. Velikovsky's theory proposes a novel understanding of the Solar system,
but the scientific community has not accepted it.
Considering such examples, we suggest that this usefulness aspect of creativity
remains essentially non-algorithmic, but not because humans are special, and cog-
nitive processes cannot be computational, but because nature is not bounded by the
limits of our cognitive models. To elaborate this further, whether an object flies or
not (which is the usefulness of the idea) does not depend on how beautiful or elegant
the theory is, or how much effort and emotional energy the creator has invested in
the object, and so on. In other words, usefulness cannot be addressed from within
the cognitive model, but it must be applied and tested in the real world.
Thus, creativity, in our view, represents the open-endedness of our interaction
with the environment, and cannot be captured in a cognitive or computational model.
Nonetheless, we can have computational models of creativity in limited domains, and
computational systems can be designed to stimulate and enhance general creativity
in people.
We can contrast our position with some other approaches to model creativity,
notably among them being the FACE and IDEA models proposed Colton and his
colleagues [ 10 , 11 , 49 , 50 ]. The FACE model formalizes novelty by explicitly iden-
tifying eight dimensions along which an object can be considered novel. To comple-
ment this, the IDEA model formalizes the impact of the artifact (so the usefulness )by
assuming how it affects an ideal audience. Two dimensions are identified to measure
the impact on the audience: one refers to how the well being of the audience has
changed in response to the artifact or the work, and the other refers to the cognitive
effort required to appreciate the artifact.
Our position is consistent with the FACE model, except that the FACE model is
more detailed in its explicit identification of various ways in which novelty can be
generated. However, the position we have articulated here with respect to usefulness
essentially implies that the goal of the IDEA model is not attainable. First of all,
it is very difficult to characterize an ideal audience. Moreover, as the character of
the audience—and here we include both the nature of the individual members of the
audience and the membership of the audience—changes as a result of interacting
with the artifact, and changes in quite unpredictable ways, it is nearly impossible to
measure the two parameters posited in the IDEA model. One only has to consider the
history of art genre like atonal music, minimalist music, abstract visual art, conceptual
art, and how they gradually became accepted by the audience to appreciate this point.
Finally, for scientific creativity and real-world problem solving, the audience is the
nature or the real world, which ultimately accepts or rejects the artifact, and this
response cannot be modeled, as we have argued above. Nevertheless, in restricted