Despite the many advances of artificial intelligence (Al) technology, most notably in the area of expert systems, efforts to build intelligent systems that would approach the commonsense reasoning and sensory abilities of even a small child have not been rewarding. A small, but growing number of researches believe that the existing AI toolboxes of symbolic representation and heuristic search may not hold the answer, and that massively parallel networks of simple neuron-like processing elements may hold the key. In this overview article, we examine the use of neurally inspired concepts in the construction of intelligent machines, and address their practical applications, advantages, and limitations.
Dadashzadeh, Mohammad and Bahr, Behnam
"Neural networks: The natural road to artificial iiitelligence,"
Journal of International Information Management: Vol. 2
, Article 7.
Available at: http://scholarworks.lib.csusb.edu/jiim/vol2/iss1/7