Allen Newell

from Wikipedia, the free encyclopedia

Allen Newell (born March 19, 1927 in San Francisco , † July 19, 1992 in Pittsburgh ) was an American computer scientist and cognitive psychologist . Newell is considered to be one of the fathers of artificial intelligence and cognitive science .

Newell studied physics at Stanford University and mathematics at Princeton . From 1950 to 1961 he worked for the think-tank RAND Corporation . However, inspired by the development of computer technology and the formulation of cybernetics by Norbert Wiener , he soon began to be interested in machine problem solving . Together with Herbert A. Simon , he developed some of the earliest programs in artificial intelligence.

Newell was a professor at Carnegie Mellon University in Pittsburgh. In 1972 he was elected to the American Academy of Arts and Sciences and the National Academy of Sciences . In 1975 he was awarded the Turing Prize together with Herbert A. Simon . In 1992 he died of cancer .

The ACM-AAAI Allen Newell Award is named after him.

Early work with Herbert Simon

In 1956 Newell developed the Logic Theorist with Herbert A. Simon . For the first time, this program was able to prove a set of logical theorems . Specifically, the logic theorist proved 38 theorems from the Principia Mathematica by Bertrand Russell and Alfred North Whitehead . This result was a milestone in artificial intelligence, as it was shown that programs are capable of actions for which a human needs intelligence.

The General Problem Solver (GPS) was the next development by Newell and Simon. He could solve a lot more problems than the logic theorist. The GPS is able to prove theorems and play games like chess or the Towers of Hanoi . The GPS approaches a problem in which it formulates a main goal and then determines a series of intermediate goals, which must be reached in order to ultimately reach the main goal. Despite these achievements, GPS remained limited to applications in a small area. It was also an application area where many of the problems of everyday intelligence had no applications. For example, there was no ambiguous information or unpredictable events in the tasks that the GPS could solve.

Soar

With Soar , Newell wanted to describe the beginning of a unified theory of human cognition . Soar is a cognitive architecture , i.e. a computer program that brings together and realizes models of human cognitive abilities. At Soar, the results of modern cognitive psychology are integrated into the program as far as formally representable. The model grows with the increase in cognitive psychological knowledge and can therefore predict human behavior better and better. Soar also has some commercial applications. In addition to Soar, there are many other well-known cognitive architectures (e.g. ACT-R and EPIC ).

Publications (selection)

  • An example of human chess play in the light of chess playing programs. Carnegie Mellon University, Pittsburgh 1964
  • On the analysis of human problem solving protocols. Carnegie Mellon University, Pittsburgh 1966
  • with George W. Ernst : GPS: a case study in generality and problem solving. Academic Press, New York 1969
  • with Herbert A. Simon: Human Problem Solving. Prentice-Hall, Englewood Cliffs 1972
  • Productions systems: models of control structures. Carnegie Mellon University, Pittsburgh 1973
  • Reasoning, problem solving and decision processes: the problem space as a fundamental category. Carnegie Mellon University, Pittsburgh 1979
  • The knowledge level. Carnegie Mellon University, Pittsburgh 1981
  • with Stuart K. Card and Thomas P. Moran : The Psychology of Human-Computer Interaction , 1983
  • Two Soar studies: toward chunking as a general learning mechanism. Carnegie Mellon University / Department of Computer Science, 1985
  • Unified theories of cognition. Harvard University Press, Cambridge 1990

literature

Web links