UP

Notes by RBJ on

The Philosophy of Artificial Intelligence

Edited by Margaret A. Boden

These notes relate only to the Introduction, notes on individual papers when available will be linked to from here.

ChAuthorsTitle
  Introduction
1Warren S. McCulloch
Walter H.Pitts
A Logical Calculus of the Ideas Immanent in Nervous Activity
2Alan M. TuringComputing Machinery and Intelligence
3John R. SearleMinds, Brains and Programs
4Margaret A. BodenEscaping from the Chinese Room
5Allen Newell
Herbert A Simon
Computer Science as Empirical Enquiry
6David C. MarrArtificial Intelligence: A Personal View
7Daniel C. DennettCognitive Wheels: The Frame Problem of AI
8Patrick J. HayesThe Naive Physics Manifesto
9Drew McDermottA Critique of Pure Reason
10Aaron SlomanMotives, Mechanisms and Emotions
11Geoffrey E. Hinton
James L.McClelland
David E. Rumelhart
Distributed Representations
12Andy ClarkConnectionism, Competence and Explanation
13Hubert L. Dreyfus
Stuart E. Dreyfus
Making a Mind Versus Modellng the Brain: Artificial Intelligence at a Branch Point
14Paul M.ChurchlandSome Reductive strategies in Cognitive Neurobiology
15Adrian CussinsThe Connectionist Construction of Concepts

Introduction

Definitions of AI are considered:
  1. the study of how to build and/or program computers to enable them to do the sort of things that minds can do
  2. making computers do things that would require intelligence if done by people
  3. the development of computers whose observable performance has features which in humans we would attribute to mental processes
  4. the science of intelligence in general
  5. the intellectual core of cognitive science
Boden's leaning seems to be toward the latter two.

She considers the main division in AI to be between "GOFAI" (Good Old Fashioned AI) and Connectionism, claiming that both have a common ancestry in the paper by McCulloch and Pitts in Chapter 1. This reflects Boden's perception of AI as a part of cognitive science concerned with the modelling of mental phenomena, rather than (say) as an engineering enterprise attaching no special significance to the structure of the brain.

Having discussed the significance of the McCulloch and Pitt paper Margaret moves on (or rather back) to Turing's paper (chapter 2), briefly describing the "Turing Test".

Three kinds of objection to Turings position are mentioned:

  1. That it does not provide a good definition of intelligence
  2. That computers could not be intelligent (however well they might perform in a Turing Test). A 1979 paper by Dreyfus is included under this heading, along with the paper by Searle in Chapter 3 of this volume. Margaret Boden's own response to this attack is also included as Chapter 4 of the volume.
  3. That computers never will pass the Turing Test.

Moving on from the criticisms of Turing, we come to the paper by Newell and Simon, in Chapter 4, where the concept of a Physical-Symbol System is introduced. In this paper computer science is characterised as empirical enquiry.


UP HOME © RBJ created 1996/4/10 modified 1997/9/4