Goto
Part IX: The Conclusion The concept of artificial intelligence began with Charles Babbage; it was the start of the computer's evolution. Babbage's main attempts were in the construction of a working brain. Since then, there have been a number of scientists who have continued with and improved upon his theoretical ideas, most notably, Alan Turing. "To Turing, the comparison between the brain and the computer was irresistible. Each was, in its own way, a binary computer. It would rather gather data in an on/off manner, either receiving and storing the information or rejecting it and shipping it to another destination. The two were composed of hardware and circuitry and could be programmed, he believed, in practically any manner" (Ladd, 1986, p.26). Studies of the brain have led to the existence of artificial neural networks. The brain can be seen as a central base which is able to gather and interpret sensory data as neural impulses. It is known to be comprised of approximately 100 billion neurons which are responsible for such operations as "sexual activity, passion, emotion, pain, behaviour patterns, body temperature, dreams, and a range of unconscious activity" along with motor control and coordination (Ladd, 1986, p.4). At one time, computer scientists believed that a reconstruction of the neurons in the brain would produce an artificial brain. This has been proven to be untrue-- the brain is much more complex than any of the early theorists had ever imagined. Nonetheless, research involving the workings of the brain has become extremely beneficial in the area of artificial intelligence.
Warren McCulloch was instrumental in the development of artificial neural networks. McCulloch believed that the brain consisted of a definitive set of rules and regulations and that the cells of the brain logically connected into networks (Ladd, 1986, p.28). McCulloch introduced the theory of neural networks. Neurons were thought of in states of excitement, either on or off, and that these signals were the key to the transference of knowledge. These states of excitement are the core of the computer; they are represented as binary states which are transferred into digital signals which can be operated on.
The concept of parallel processing, originally explored with Psychology, has proved beneficial in the area of Computer Science. Psychological research in the 1960's supported findings of a shift from the tradition theory of serial processing in the mind to a more complex and efficient parallel processing. It was proposed that the mind consisted of a wide range of individual neurons which were able to act on multiple commands currently. Although computer scientists have been able to approximately equate the number of switching circuits in the computer to the number of neurons in the mind, the progress towards parallel processing has been relatively slow. The mind is able to accomplish many activities at one time but, until recently, the computer could process information only serially with its single central processing unit (CPU). Single processor computer can take advantage of parallel techniques through the advent of computer networks; distributed operating systems allow the individual systems to function as one. After years of research and work, "neural network computers [now] usually involve several 'processing units'. They work together, in parallel, and transfer information among themselves to accomplish complex tasks rapidly" (Reed, 1992, p.34).
Goto
Part IX: The Conclusion
Goto the Effects of Psychology on Computer Science
