|
_Can Machines Think _
By: Bob Fraiman
Artificial Intelligence is a broad topic, consisting of different fields, from
machine vision to expert systems. Nowadays with the use of high level
techniques, computers have been programmed to solve many difficult problems,
but the products that are available today are only a sample of what is coming
in the future. AI has an interesting history and has always been on the
pioneering end of computer science. In order to classify machines as
"thinking", it is necessary to explain intelligence. What is intelligence
exactly? How do you get intelligent? All these questions are still unanswered
and all of them have helped to create both problems and solutions that found
the heart of modern AI. Instead of building a theory of intelligence, there
are a great variety of theories, such as the psychological, the cognitive
theories and other theories. All these theories are reaching to the
conclusion, that intelligence is composed from various mental power and
abilities. Psychologists assume that intelligence is based on adaptation,
adjustment of self and behavior. Therefore it is important to understand the
determinant factors of a complicated problem or a complex situation. The
cognitive theory is focused on mental representations of information, such as
images, ideas or symbols and on processes to operate with these signs. It
considers an intelligent being as a kind of information-processing system. If
we assume that a machine has the same components which mirror a human body, it
could theoretically be possible that this machine thinks. From this point of
view a digital computer could be described as an instantiation of a computer
program and all humans are nothing more than instantiations of computer
programs and they are able to think! Even if we are unable to imitate mental
processes, it is possible for AI to reproduce the information processing, like
a human brain does. Computers could be programmed in the same way as the brain
processes information. To simulate human intelligence the computer must first
simulate and understand the human behavior. But even after that, the computer
must be able to understand emotions and how those emotions make us feel. A
computer can appear to think but in reality it is just following instructions.
The goal of AI is to stimulate human performance at a task, not only do they
want the program to do the best as it can but to also fail where a person
would fail. There are some techniques which have been developed by
psychologists that can be used to do this analysis. A digital computer is
purely a machine to manipulate ones and zeros, given a specific program. But
humans are able to understand what the "meaning" of the phrase was, however a
computer would not. The separation between a human and a machine is easy
because one has emotions and the other one doesn't. Machine based decisions
are more logical and rational and are without any emotional baggage. But on
the other hand emotions are useful to form decisions. There are only two
possible ways of dealing with this problem. Emotions can be provided for
intelligent machines by duplicating a sub-system of a human brain or on the
other hand we could have machines without any emotions. If that's the case,
then machines will not take account of any moral issues of our society. AI is,
as shown in the article, a powerful scientific invention with far-reaching
possibilities that offers great changes. But when will we accept the changes
and allow this to be a possibility? If we ever will then we have to be
prepared for the worst, that could happen in AI, in regard to ultraintelligent
machines. Hopefully this will be just a futuristic fiction. I agree with
Charles W. Bailey, Jr., when he said computers will never simulate human
intelligence in that way human brain does. So the answer to the question: "can
machines think?", must be denied. It will never be possible to attribute human
qualities to a machine, because of lacking emotions or feelings. Computer
systems could never interact in same environment as a human does, maybe they
perfectly simulate this so these systems appear to be intelligent. The complex
human brain and its understanding will never be duplicated. But the AI
technology will be a useful tool to utilize for our society. AI based software
will have more influence in various areas like Simulation, Language and
Speech, Automatic Programming, Sensor Systems, Robots, Expert Systems and
Parallelism - Neural Networks. These examples can be seen as partially
artificially intelligent, but have not yet reached the full capacity.
Artificial intelligence will most definitely exist at some point in the future
and technology will influence further conception. The advances that are
impossible today will be similar to reality tomorrow. To reach this stage
there are many technical problems to solve. Till we have a system, which will
be able to perform a multitude of so called humanoid tasks, scientists will
have to do a lot more research before this goal will eventually be achieved.
Word Count: 833
|
