The research field of Cognitive Science asks, then attempts to answer, many
basic questions that relate to how machines can function (apparently) in ways similar to intelligent entities:
p.ix"What does it mean to have intelligence in a computing device?
Can computers have intentions or exhibit purpose? Can computers learn new things or understand the meaning of sentences? Just
what is the nature of intelligence? Can intelligence exist outside the human person? Can it be described by a set of abstract
laws? What is consciousness and what is its relationship to intelligence? What is perception and how can abstract concepts
arise from perceptual experience? What is the mind and how does it relate to the body? How can immaterial and abstract ideas
affect material reality? These and similar questions make up the subject matter of cognitive science."
If you think about it, nothing has meaning by itself, in isolation from all other things. And even
when we consider things in their proper environment, in interaction, meaning is really only meaning to us.
p.3"Meaning is not a thing; it involves what is meaningful to us. Nothing is meaningful in itself.
Meaningfulness derives from the experience of a being of a certain sort in an environment of a certain sort... George Lakoff
(1987)"
Intelligence and intelligent behavior might be reducible to a common set of principles. We
could establish a science to study these principles.
p.4"Cognitive science begins with the assumption that there is a common set of principles underlying
all instances of intelligence."
Intelligent behavior requires knowledge, but it is really knowledge about how to acquire knowledge
that forms the basic behavioral trait called intelligence. Ken Jennings prepared for his successful winning
run on the American TV game show Jeopardy by competing on Brigham Young University's
Quiz Bowl Team. Jennings also studied the book How to Get on Jeopardy! and Win by 1996 Jeopardy! Tournament of Champions
winner, Michael Dupee [source: Wikipedia]. He likely did not spend much time studying how to rebuild
an automatic transmission or memorizing the value of pi to 1000 places. He knew that those questions would not be asked. Jennings
knew exactly how to acquire knowledge that would be useful on a quiz show. Jennings, amazingly, is a model for an intelligent
system.
p.8"Intelligence is the ability to acquire knowledge, and not the knowledge itself. Although it
may be relevant to distinguish between intelligence and knowledge, it is also the case that it requires knowledge to acquire
knowledge."
An intelligent system needs to find a way to become sensitive to changes in its environment in order
to 1) react to and 2) identify cues which indicate that a change might happen in the future.
p.9"Reactivity, as defined here, relates to intelligence by the fact that any system's ability to
adapt to the environment is limited by its sensitivity to changes within that environment."
Intelligent systems might be classified as strong method or weak method systems, depending on the amount of knowledge
used in the process of determining behavior.
p.10"Strong method problem solvers depend on a large amount of knowledge for their intelligent behavior."
Here we have Luger's definition of Artificial Intelligence:
p.29"Artificial intelligence is that branch of computer science concerned with the creation
of a kind of software that might be considered to exhibit 'intelligent' behavior."
The focus of AI must be, ultimately, on creating software that is determined to be 'successful' by a pre-determined set
of goals or tests. Successful software might or might not represent the human approach to solving a problem.
p.30"The focus of AI is on building computational structures and designing strategies to search
these structures for a solution. From an AI perspective, problem solving is construed as a series of operations on the models
of a domain... The primary goal of the AI practitioner is to build successful software, and not necessarily to implement
human intelligence!"
Intelligent beings need to be able to evaluate non-identical things in their world and form equivalencies:
p.75"Since no organism can cope with infinite diversity, one of the most basic functions of all
organisms is the cutting up of the environment into classifications by which non-identical stimuli can be treated as equivalent...
Eleanor Rosch (1978)"
Sometimes things that are simple become complex according to 1) the environment they are placed
in and 2) the interaction they have with the things located in that environment:
p.255"An ant, viewed as a behaving system, is quite simple. The apparent complexity of its behavior
is largely a reflection of the complexity of the environment in which it finds itself. - Herbert Simon (1981)"
Luger now discusses strong method or knowledge based problem solving. Here he sketches
the architecture of such a system, and we can see the similarity to the design for the evaluation function presented in A
Proposed Heuristic:
p.319"There are three components of the blackboard system: 1) the central global data base
called the blackboard, 2) one or more independent knowledge sources, KSs, each operating independently on
the global data base, using the data base to receive its data for processing as well as for posting the results of its work,
and 3) a control regime for handling the interaction of the knowledge sources with the blackboard... each knowledge
source (KS) gets its data from the blackboard, processes it, and returns its results to the blackboard to be used by the other
knowledge sources. Each KS is independent in that it is a separate process operating according to its own specifications..."
