Information Technology Reference
In-Depth Information
Chapter 7
IDyOT: A Computational Theory of Creativity
as Everyday Reasoning from Learned
Information
Geraint A. Wiggins and Jamie Forth
Abstract
We present progress towards a computational cognitive architecture,
IDyOT (Information Dynamics of Thinking) that is intended to account for cer-
tain aspects of human creativity and other forms of cognitive processing in terms of
a pre-conscious predictive loop. The theory is motivated in terms of the evolutionary
pressure to be efficient. It makes several predictions that may be tested by building
computational implementations and studying their behaviour.
7.1 Introduction
The aim of this chapter is to explain more of the theoretical detail of a cognitive
architecture that was outlined byWiggins [
55
], with a view to developing aminimalist
theory of cognition that encompasses many human behaviours within as simple a
functional framework as possible. The architecture is called IDyOT, for “Information
Dynamics of Thinking”.
The overarching proposal is that information-theoretically regulated probabilistic
prediction, which serves as a general mechanism for managing information and
events in the world, is the mechanism underlying non-conscious creativity—that
creativity is, essentially, a by-product of mechanisms which have evolved to serve
prosaic, not poetic, forms of existence.
It is important to emphasise from the start that taking such a reductionist position
about the
source
of creativity does not in any sense diminish the value of its applica-
tion, as exhibited in magnificently various forms and magnitude throughout human
society, and, less consistently, by other species.
In the current chapter, then, we will present the abstract structure of the pro-
posed architecture, which is based on function only, and not on any kind of phys-
iological model of a brain, human or otherwise. Nevertheless, we will refer to
Search WWH ::
Custom Search