Past Issues

Studies in Informatics and Control
Vol. 3, No. 4, 1994

BOOK REVIEW - Natural and Artificial Intelligence Misconceptions about Brains and Neural Networks

Angela lonita
Abstract

A first version of the book described an improbable brain model but did not show why any other paths were impracticable. These explanations might have helped researchers not to repeat errors of the past. This is the second version of the book. The book topic is "large scale architecture" as defined in Minsky (1990). Modelling brains is like designing a processing plant. An engineer must connect many devices in an efficient way. The brain designer has a large choice of components in the machinery of regulation systems, instrumentation, neural networks, and artificial intelligence. The design task is to reproduce natural intelligence in a robot by integrating all these devices. Discovering how brains work is a detective job. A motto of this book could be: "When you have eliminated the impossible, whatever remains, although improbable, must be the truth" (Sherlock Holmes, Conan Doyle, 1884).

Keywords

by Armand M. de Callatay

View full article