A New Guide to Artificial Intelligence

In this textbook the author endeavors to cover the large and growing field of artificial intelligence (AI) in some detail. While there are books that examine and discuss global perspectives on AI, they make no attempt to cover the diversity of theories and programs. A global perspective on the subject is provided by this volume, but in conjunction with an exhaustive survey of the field. It covers all recognized AI work in sufficient detail to allow a critiquie from general concerns to be anchored, whenever possible, in the structure of specific AI programs. It can be used as a supplement to other AI texts, providing broader perspectives on the wealth of details that such texts contain. It can also be considered as a companion to the current AI literature for it is only in conference proceedings and journals that these up-to-date details are usually found.

Table of Contents:
Preface               xviii 1 AI: what is it?               1  definitions: what would it look like if I saw one?               1  True AI Story: 1.1-ELIZA meets PARRY: the syntax is willing but the semantics is weak               7  a history of scaling down               8  categorizations of AI work               14  the goals of AI research               15  the heuristic programming approach               18  the Samuel phenomenon               22  2 AI and the Science of Computer Usage: The Forging               23 of a Methodology  how to use the essential tool?               23  first specify, then verify               26  the nature of AI problems               27  a methodology of incremental exploration               30  rapid prototypes to the rescue?               32  supportive environments               33  forging a new methodology               34  is AI so different?               37  3 The Major Paradigms               39  symbolic search spaces               40  planning intelligent solutions               45  SSSP infrastructure               54  the pivotal role of searching strategies               56  heuristic pruning               59  connectionism: a possible alternative?               62  connectionism: the second coming               64  on not losing their inhibitions               66  the need for decay               67  subsymbolic connectionism: the good news               68                 when is an AI system like a piece of fine china?               70  subsymbolic connectionism: the real news               71  reasoning with amorphous complexity               72  the myth of empirical guidance               73  what's the stopping rule?               77  single-minded models               78  philosophical objections               79  potential solutions to the dilemma               81  formal analysis               81  software support systems               83  approximate translation-the truth about mendacity               84  the SSSP and the CP: integration, bifurcation, or annilation?               86  simulated evolution: guess and try it out               89  'bad' paradigms               90  4 The Babel of AI Languages               98  it's all done by manipulating symbols               98  LISP               100  flexibility               100  the magic of recursion               101  code-data equivalence               103  the special assignment               105  lists of properties               106  PROLOG               108  the independence of declaration               109 loss of control: better or worse?               111  extralogical pollutants               116  negation as failure               120  verify or compute               120  bidirectionality               121  pattern matching               121  the promises of PROLOG               122 parallelism               122  a specification language               123  heuristic controls               124  object-oriented programming               125  programming environments               131  LISP environments               131 LOOPS               132  POPLOG               134  True AI -Story: 4.1. DIMWIT (Do I Mean What I Tell): A PA (Programmer's Assailant) System               136  5 Current Expert Systems Technology (CEST)               139 experts with tunnel vision               140 the basic assumptions and the criticisms               141 what can be CESTed?               145 explanations and context sensitivity               146 updating knowledge bases and machine learning               150 let's dig deeper               155 logical decision making               159 human and computer decision making               161 classes of human decision making               163 connectionism: a possible answer?               165 knowledge elicitation               167 knowledge engineers and the third degree               167 automatic learning from examples               168 empirical techniques               168 CEST: where is it and where is it going?               169 6 Knowledge Representation: A Problem of Both               171 Structure and Function                why networks?               171 why neurons?               176 pandering to evolution: beware of classical reconditioning               178 neural architectures: in the beginning               179 knowledge representation: structure and function               183 the SSSP and the CP: representational issues               188 knowledge representation in the CP               189 functionally distributed representations               189 symbolic connectionist representations               189 winner-takes-all subnets               194 hybrid connectionism               195 totally distributed representations               196 path-like architectures in the CP               202 bath-like architectures in the CP               204 knowledge representation in the SSSP               209 logic-based representations               210 procedural representations of knowledge               213 semantic networks               217 elements of structured knowledge: frames, scripts, and schemata               226 7 Vision: Seeing is Perceiving               228 bottoming in: operators canny, uncanny, and cannyless               233 pixel processing               233  edges and lines               237  vertices or junctions               238  texture: a truly superficial feature               239  illumination, reflectance, and other sources of nuisance               241  the intrinsic image               241  model-based vision systems               244  True AI Story: 7.1               247  beer cans, broomsticks, etc.               248 seeing as perceiving               250  oversight and hallucination               252  the modularity of human vision               255  eyeballs and nervous optics               256  biological feature detectors               256 human perceptual behavior               261  breaking up context               262 structuring top-down information               262  a cognitive model of word recognition               264  the eye of the robot               267  general theories of visual perception               273  the vision of connectionists               278 8 Language Processing: What You Hear is What You Are               283 natural language               286 what mode of natural language?               287 the goals of AI-NLP               289 natural language: the essential ingredients               289 phonetics and phonology               290 the lexical level and above               291  generation and analysis               291 natural language generation (NLG)               293 text generation systems               297 empirical guidance for NLG               298 natural language understanding (NLU)               299 syntax, grammars, and parsing                grammars               300 furious transformational grammarians               sleep curiously               303  transition networks: augmented and otherwise               305 unification and the new grammatism               308 semantic definite clause grammars (SDCG)               309 NLP and a formal complaint               314 semantics               318  the meaning of semantics               319  the atomic struture of meaning               320 the case of the missing-blocks world               322  True AI Story: 8.1 SHRDLU and a "SORRY" story               324  revolting computational linguists               324  scripted NLU and its dependencies               326  True AI Story: 8.2 Try it again SAM               328  the conceptual dependency notation               329  a Swale of a tale               331  True AI Story: 8.3 Another SWALE of a tale               332  giving semantics preferential treatment               333 bidirectional NLP               335 pragmatics?               336  machine translation (MT)               339 natural language interfaces (NLI)               341 networks for NLP               344  9 Learning To Do it Right                351 can we have intelligence without learning?               354 can we have AI without learning?               355 learning paradigms in AI               356 learning as the accretion of symbolic structures               358 learning as the adjustment of link weights               361 external tutoring: learning by being told               363 learning on the path               370 learning in a bath - taking the plunge               381 climbing hills because the 're there               384 rote learning: if it might be useful, store it               390 learning generalities               391 induction               393 overgeneralization and refinement               395  a first guess and generalization               397  True Al story: 9.1. Underneath the arches: an everyday story of concept learning               399 competitive learning               404  learning particularities: removal of unwanted generalization               406  EBG, or is it EBL?               409  the EBL viewpoint               416  mechanized creativity               421 learning by introspection               423 rediscovering things               427 learning by analogy               432  learning at the knowledge level               433  soaring through search spaces               437  the more you know the slower you go               443 on finding needles in haystacks               448 when to learn and what to learn               448 giving credit where it is due               450 unlearning               453 10 Foundations of AI: Can we find any?               458 foundations: why dig for them?               459 formal foundations               460 a disinterested user's guide to the FOPC               462 the curse of nonmonotonicity               467 logical odds and ends               471  True AI Story: 10 .1 It is not a closed world after all               471 the epilogic               474 methodological foundations               477  the roles of programs in AI               478  programs as theories               479 programs as experiments               483 rational reconstructions in AI               483  sorting out AI methodologies               486 philosophical foundations               488 there's nothing special about you, or me               488 building the foundations on the CP               490 undermining the foundations of the CP               491 total disbelief: let's not be Searle-ish               492 11 Prognostications, or W(h)ither AI?               496  abstract AI and concrete AI               496  is the mind an appropriate object for scientific study?               498                 True AI Story: 11.1. Sand in the works               499 AI as a magnifying glass               501  AI: can it be practically useful?               503 AI: just wait till we get into parallel hardware               503 last words               506 References               507  Author Index               531  Subject Index