Print Price: $43.95

Format:
Hardback
400 pp.
104 illustrations, 183 mm x 257 mm

ISBN-13:
9780199357789

Publication date:
April 2017

Imprint: OUP US

After Digital

Computation as Done by Brains and Machines

James A. Anderson

Current computer technology doubles in in power roughly every two years, an increase called "Moore's Law." This constant increase is predicted to come to an end soon. Digital technology will change. Although digital computers dominate today's world, there are alternative ways to "compute" which might be better and more efficient than digital computation. After Digital looks at where the field of computation began and where it might be headed, and offers predictions about a collaborative future relationship between human cognition and mechanical computation.

James A. Anderson, a pioneer of biologically inspired neural nets, presents two different kinds of computation - digital and analog - and gives examples of their history, function, and limitations. A third, the brain, falls somewhere in between these two forms, and is suggested as a computer architecture that is more capable of performing some specific important cognitive tasks-perception, reasoning, and intuition, for example- than a digital computer, even though the digital computer is constructed from far faster and more reliable basic elements. Anderson discusses the essentials of brain hardware, in particular, the cerebral cortex, and how cortical structure can influence the form taken by the computational operations underlying cognition. Topics include association, understanding complex systems through analogy, formation of abstractions, the biology of number and its use in arithmetic and mathematics, and computing across scales of organization. These applications, of great human interest, also form the goals of genuine artificial intelligence. After Digital will appeal to a broad cognitive science community, including computer scientists, philosophers, psychologists, and neuroscientists, as well as the curious science layreader, and will help to understand and shape future developments in computation.

Readership : Well-educated public that is interested in computers and where they are going. This book was developed for a freshman seminar at Brown University, though it was not designed as a text. It could be a useful text in a computer topics course that was not concerned with programming but with the larger cultural significance of computers now and in the future. (I.e. computer science for poets).

Reviews

• "After Digital is an incisive and eloquent milepost in our accelerating journey to human-like artificial intelligence. Anderson, a seminal figure in the field of neural networks and one of its most prescient practitioners, gives us a first hand look at how we got here and where we are going. An accessible and essential book that will appeal to both scientists and non-scientists alike."
  
  --Michael J. Tarr, PhD, Professor and Head, Department of Psychology, Carnegie Mellon University
  
  
• "If you want to understand the arrival of the computer age in the 20th century, you will value this book. To chart the course we are on, Anderson combines a physicist's knowledge of the technology with a psychologist's appreciation of human aspirations and cognitive capabilities. Informative, thoughtful, accessible."
  
  --Pentti Kanerva, PhD, Redwood Center for Theoretical Neuroscience
  
  
• "Professor James A. Anderson presents historical and evolutionary arguments emphasizing current attempts to understand biological computation within a digital computer context [that] may lead to computationally easy or sociologically popular solutions which unfortunately are dangerously misleading or even wrong. Anderson's writing style is best described as a fun, fascinating, and entertaining dinner conversation where seemingly innocent conversation threads consisting of interesting anecdote sequences incrementally reveal novel insights about the strengths, limits, and future of both biological and digital computation. Required reading for anyone interested in how biological systems compute as well as anyone interested in exploiting biological constraints for developing smart machines!"
  
  --Richard M. Golden, PhD, Professor of Cognitive Science, University of Texas at Dallas