xx This book [published in 2002] is about robotic brains as they should
be implemented in the very near future. The biologists and psychologists will find the engineering views on cognitive processes
in [the] human brain.
p.1 The phenomenon of intelligence is demonstrated as a computational
phenomenon. It emerges as a result of the joint functioning of several operators: grouping, focusing attention, searching,
and formation of combinations. When information is processed by these operators, multiresolutional systems of knowledge develop,
and nested loops of knowledge processing emerge. This conceptual structure permits the explanation of most of the processes
characteristic of intelligent systems.
p.2 What is lacking is a general theoretical model of intelligence that ties all these separate
areas of knowledge into a unified framework... it is emphasized that an intelligent system always has an architecture
that develops as a result of the joint functioning of certain computational operators. The architecture that emerges as a
result has multiple closed loops, each loop at a particular level of resolution.
Our model is to be used for the analysis of all kinds of intelligent systems, including
animals, humans, automated machines, robots, autonomous vehicles, and integrated manufacturing systems.
p.3 Intelligence is the ability of a system to act appropriately in an uncertain environment,
where an appropriate action is that which increases the probability of success, and success is the achievement of behavioral
subgoals that support the system's ultimate goal.
p.11 In this book, we approach the phenomenon of intelligence from a systems engineering viewpoint.
Our goal is to develop engineering guidelines that enable the design and construction of intelligent systems that rival natural
intelligence in performance of significant tasks in the natural world. In the following pages we set forth an outline
for a theory of intelligence and propose a reference model architecture that can serve as a guideline for engineering intelligent
systems.
p.17 The ability to transform "reality" into a "representation of reality" can be considered to be one of
the most important phenomena linked with intelligence.
p.44 Grouping includes various procedures of clustering, class formation, unifying of sets, construction
of strings, n-tuples, and so on.
p.44 The scope of representation is satisfied by focusing attention upon a subset of the world sufficient
for problem solving. After this, the rest of the information about the world can be neglected... We will never use
information more precise (and detailed) than that available at the resolution of the level of consideration... Condition of
windowing. This condition is satisfied by proper selection of the size and the law of motion within the scene for the sliding
window applied for the consecutive clustering and generalization procedures... All three conditions are prerequisites for
the operation of search
p.80 The development of human intelligence would be impossible without attention focusing
p.83 No matter what the package is that one acquires for an intelligent system, no matter
from which company it is bought, this package, if it is complete, will perform attention focusing, grouping, and combinatorial
search.
p.350-351 in reactive behavior, robots usually react to the present situation. In a system with
planning they react to the anticipated future.
Thus planning can be considered an anticipatory reactive behavior. The difference
is in the fact that anticipation requires richer representation than the simple reactive behavior requires.
