A Personal Perspective into the Future for Computational Creativity - Computational Creativity Research: Towards Creative Machines - page 391

in the past to foretell this type of information have generally failed. What I thought
I might do is to see how the road travelled so far, and the experience accumulated in
the process, might give me some hints as to what the path ahead may bring. I have
done this in three different ways One is to try and explain which of the processes of
the conceptual schema presented above are still outstanding, and how likely I think
they are to be addressed in the near future. Another is to consider what were the
factors that played a significant role in the success (or lack thereof at given moments
of time) of disciplines that have faced similar challenges. Finally, a different way is
to consider the visions expressed by others, and try to value where they are likely to
lead in terms of the initially stated aims.
19.3.1 Processes of the Conceptual Schema to be Addressed
in the Future
Of the basic proceseses for computational generation of novel artefacts described
in Sects. 19.2.1 and 19.2.2 , I have outlined in Sect. 19.2.3 which ones have been
addressed so far in existing efforts on Computational Creativity. With respect to
the future, the important consideration is which of those processes are likely to
undergo further evolution, which ones may be good candidates for the development
of practical applications, and which of them have more potential in terms of long
term impact on society.
It is clear that those processes that have received less attention in the past—1, 2,
3, 4, 7 and 8, covering selection of constructive procedure or ingredients, valid-
ity and novelty criteria, and, most importantly reflection over results leading to
improvement—are likely to experience significant growth in the future. For a start,
these processes constitute virgin territory, where low hanging fruit may be waiting
for any pioneers brave enough to consider them. But remember that, as argued in
Sect. 19.2 , these are the subtasks that human creators would least consider delegating.
They correspond to the role of the engineer designing a robot, rather than the factory
manager supervising its mass production. Or to the role of the sculptor conceiving a
piece, as opposed to the foundry manager that supervises the production of the final
cast in bronze. In a way, they embody the core of the creative task, and it is very
possible that a large percentage of the mistique of creativity arises from decisions
taken in these processes. The remaining processes 5 and 6 correspond to mass pro-
duction and quality control, and are much less likely to involve essential ingredients
of creativity. There are some aspects of process 8 for establishing novelty of particu-
lar artefacts that have particular potential. In contrast to other processes that rely on
human abilities that are hard to model in a computer [ 2 , 19 ], the simple establishment
of novelty may be significantly helped by massive storage capabilities and high rates
of comparison, abilities that computers can achieve with significantly more ease than
humans. This has already been demonstrated by the success of computer solutions
for detecting plagiarism, which can be seen as a related problem. In contrast, those