Muutke küpsiste eelistusi

AI and Human Thought and Emotion [Kõva köide]

2.50/5 (2 hinnangut Goodreads-ist)
  • Formaat: Hardback, 252 pages, kõrgus x laius: 234x156 mm, kaal: 560 g, 1 Tables, black and white
  • Ilmumisaeg: 20-Aug-2019
  • Kirjastus: CRC Press
  • ISBN-10: 0367029294
  • ISBN-13: 9780367029296
Teised raamatud teemal:
AI and Human Thought and EmotionSuurem pilt
  • Formaat: Hardback, 252 pages, kõrgus x laius: 234x156 mm, kaal: 560 g, 1 Tables, black and white
  • Ilmumisaeg: 20-Aug-2019
  • Kirjastus: CRC Press
  • ISBN-10: 0367029294
  • ISBN-13: 9780367029296
Teised raamatud teemal:
Püsilink: https://www.kriso.ee/db/9780367029296.html
Märksõnad:
The field of artificial intelligence (AI) has grown dramatically in recent decades from niche expert systems to the current myriad of deep machine learning applications that include personal assistants, natural-language interfaces, and medical, financial, and traffic management systems. This boom in AI engineering masks the fact that all current AI systems are based on two fundamental ideas: mathematics (logic and statistics, from the 19th century), and a grossly simplified understanding of biology (mainly neurons, as understood in 1943). This book explores other fundamental ideas that have the potential to make AI more anthropomorphic.

Most books on AI are technical and do not consider the humanities. Most books in the humanities treat technology in a similar manner. AI and Human Thought and Emotion, however is about AI, how academics, researchers, scientists, and practitioners came to think about AI the way they do, and how they can think about it afresh with a humanities-based perspective. The book walks a middle line to share insights between the humanities and technology. It starts with philosophy and the history of ideas and goes all the way to usable algorithms.

Central to this work are the concepts of introspection, which is how consciousness is viewed, and consciousness, which is accessible to humans as they reflect on their own experience. The main argument of this book is that AI based on introspection and emotion can produce more human-like AI. To discover the connections among emotion, introspection, and AI, the book travels far from technology into the humanities and then returns with concrete examples of new algorithms. At times philosophical, historical, and technical, this exploration of human emotion and thinking poses questions and provides answers about the future of AI.
Author xiii
0 Introduction 1(10)
0.1 Frustrations and Opportunities in AI Research
1(3)
0.2 Central Questions
4(1)
0.3 Structure of This Volume
5(1)
0.4 How to Read This Book
6(5)
Part I Intelligence In Computers, Humans And Societies
1 Artificial Intelligence as It Stands
11(16)
1.1 About AI
11(4)
1.1.1 AI's Relation to Psychology, Cognitive Science, etc
11(2)
1.1.2 What Are Intelligence, Consciousness, and Introspection
13(1)
1.1.3 Defining and Viewing AI
14(1)
1.2 First Approach: Logic and Mathematics
15(1)
1.3 Second Approach: Biological Inspiration
16(1)
1.4 A Half-Approach, and a Point or Two
17(1)
1.5 Watson
18(4)
1.5.1 Explicit Motivations
19(1)
1.5.2 Arguments against Introspection
19(1)
1.5.3 Interesting Points
20(2)
1.5.4 Watson - Summary
22(1)
1.6 Simon
22(3)
1.6.1 Economics
23(1)
1.6.2 Hostile to Subjectivity - Rationalistic
23(1)
1.6.3 Artificial Intelligence
24(1)
1.6.4 Against His Critics
25(1)
1.6.5 Flirting with Subjectivity
25(1)
1.7 AI as It Stands - Summary
25(2)
2 Current Critiques of Artificial Intelligence
27(22)
2.1 Background: Phenomenology and Heidegger
28(2)
2.1.1 Phenomenology
28(1)
2.1.2 Heidegger
28(2)
2.2 The Cognition vs Phenomenology Debate
30(4)
2.3 Dreyfus
34(5)
2.3.1 Part I - Ten Years of Research in Artificial Intelligence (1957-1967)
34(1)
2.3.2 Part II - Assumptions Underlying Persistent Optimism
35(1)
2.3.3 Part III - Alternatives to the Traditional Assumptions
36(2)
2.3.4 Dreyfus's Updated Position
38(1)
2.4 Winograd and Flores
39(3)
2.4.1 Cognition as a Biological Phenomenon
40(1)
2.4.2 Understanding and Being
40(2)
2.4.3 Language as Listening and Commitment
42(1)
2.5 Hermeneutics and Gadamer
42(3)
2.5.1 Hermeneutics
42(1)
2.5.2 The Hermeneutics of Heidegger and Gadamer
43(2)
2.6 AI's Inadequate Response to Dreyfus and Other Critiques
45(1)
2.7 Locating This Project amongst Existing Thinkers
46(1)
2.8 Current Critiques of AI: Summary
46(3)
3 Human Thinking: Anxiety and Pretence
49(16)
3.1 Individual Thinking
50(4)
3.1.1 Our Thinking Processes Are Embarrassing
50(1)
3.1.2 Anxiety, Pretence, Stories, and Comfort
51(1)
3.1.3 Can We Even Tell the Truth?
52(1)
3.1.4 Motivations
53(1)
3.2 Society's Thinking
54(4)
3.2.1 Politics
54(1)
3.2.2 Social Perceptions of Science
55(1)
3.2.3 Interrelation of Politics and Science
56(1)
3.2.4 Distinct Disciplines and Education
56(1)
3.2.5 Education as Indoctrination
57(1)
3.3 Adapting to Social Norms
58(2)
3.3.1 Social Pressure - the Game of Life
58(1)
3.3.2 Conforming
59(1)
3.3.3 Escape to a Role, Arrogance
59(1)
3.3.4 Needs Must
60(1)
3.4 Relevance to AI
60(2)
3.4.1 Anxiety and Pretence Are Immediately Relevant to Thinking
60(1)
3.4.2 Implications for AI, a Rudimentary Human-Like Mind
61(1)
3.4.3 Meaning-for-Me vs Big Data
61(1)
3.4.4 Relevance to AI - the Future
62(1)
3.5 Human Thinking: Anxiety and Pretence: Summary
62(3)
4 Prevailing Prejudices Pertaining to Artificial Intelligence
65(20)
4.1 A History of an Idea: Positivism
66(2)
4.2 Knowledge
68(3)
4.2.1 Truth Exists, Is Knowable, and Can Be Expressed in Language
69(1)
4.2.2 There Is Only One Truth System
69(1)
4.2.3 Kinds of Illumination
70(1)
4.2.4 Polarisation of Knowledge and Doubt
71(1)
4.3 Science
71(3)
4.3.1 The Scientific Clean Sweep
71(1)
4.3.2 Science Is Distinct from Magic or Religion
72(2)
4.3.3 The World Is Modular, Logical Atomism, Determinism
74(1)
4.4 "Wooly" vs "Rigorous" Thinking
74(3)
4.4.1 Secularisation
74(1)
4.4.2 Philosophy Is Seen as Bad
75(1)
4.4.3 Especially, Continental Philosophy Is Seen Negatively
75(2)
4.5 Humans and Minds
77(3)
4.5.1 The Human Mind Is a Natural Kind
77(1)
4.5.2 Humans Are Like Computers
77(1)
4.5.3 Lower and Higher Human Functions
78(1)
4.5.4 Humans Are Rational
78(2)
4.6 Other Worries about Religions
80(2)
4.6.1 Genesis
80(1)
4.6.2 Heresies
81(1)
4.7 Prejudices Pertaining to AI: Summary
82(3)
Part II An Alternative: AI, Subjectivity, And Introspection
5 Central Argument Outline
85(14)
5.1 Context for Central Argument
86(4)
5.1.1 Science vs Technology and Human-Like vs Rational
86(1)
5.1.2 Philosophy of AI
87(1)
5.1.3 Philosophy of Technology
88(2)
5.2 Notions of Truth
90(4)
5.2.1 The Idea of a Single Truth
90(1)
5.2.2 Perspectivism
90(2)
5.2.3 Perspectives, Realities, Agendas, Occam
92(1)
5.2.4 In What Sense Is This Book True?
93(1)
5.2.5 Notions of Truth: Summary
93(1)
5.3 Outline of "Is Recommended for Developing"
94(3)
5.3.1 "Recommended"
94(1)
5.3.2 "For"
95(1)
5.3.3 "Developing"
95(2)
5.4 Central Argument Outline - Summary
97(2)
6 Main Term: "Anthropic AI"
99(18)
6.1 Human vs Ideal/Rational
99(1)
6.2 Motivations for Human-Like AI
100(3)
6.2.1 Rational AI's Interaction Is "Clunky"
100(1)
6.2.2 The Versatility of Human Intelligence
101(1)
6.2.3 Getting along with People
102(1)
6.3 Characteristics of Human-Like AI
103(1)
6.4 Human-Like vs Anthropic
104(1)
6.5 Perspectives and Levels in Human Modelling
105(5)
6.5.1 Are There Really Levels or Layers in the Mind/Brain?
105(1)
6.5.2 Multiple Levels of Discussion
106(3)
6.5.3 The Cognitive Level Is Problematic
109(1)
6.5.4 Simultaneous Multiple Levels in Computers
110(1)
6.6 Anthropic AI so far
110(2)
6.7 Knowing That vs Knowing How, and a Hint on Data Structure
112(2)
6.8 Metaphysical Non-problems
114(1)
6.9 Ethics
115(1)
6.10 Anthropic AI: Summary
116(1)
7 Main Term: "Introspection"
117(16)
7.1 Studying Subjectivity
117(5)
7.1.1 Why Subjectivity?
118(1)
7.1.2 Locating Subjectivity
118(1)
7.1.3 What Is Subjectivity
119(1)
7.1.4 Subjectivity Can Be Studied
120(1)
7.1.5 Phenomenology, Hetero-Phenomenology
121(1)
7.2 Defining Introspection
122(1)
7.3 A Boundary between Introspection and Science Collapses
123(7)
7.3.1 "Thinking Aloud" (TA) Can Be Seen as Introspective
124(1)
7.3.2 Two Distinctions between TA and Introspection
125(2)
7.3.3 Inferences and Confusion
127(1)
7.3.4 Non-inferential Observation Is Impossible
128(1)
7.3.5 A Boundary between Introspection and Science Collapses: Conclusion
129(1)
7.4 What Kind of Introspection Is Recommended
130(1)
7.5 Main Term: "Introspection": Summary
131(2)
8 Introspection Is Legitimate
133(24)
8.1 Introspection as "Impossible"
134(1)
8.2 Introspection as "Forbidden"
135(7)
8.2.1 Watson
135(1)
8.2.2 Cognitive Psychology's Attitude to Introspection
136(2)
8.2.3 Other Objections
138(1)
8.2.4 Contexts of Discovery and Justification
138(1)
8.2.5 Truth in Science vs Technology
139(2)
8.2.6 Example and Summary of "Introspection Is Forbidden"
141(1)
8.3 Introspection as "Commonplace"
142(6)
8.3.1 Sweeping Testimony
142(2)
8.3.2 Specific Apparent Cases
144(2)
8.3.3 Mainstream Cognitive Science Uses Introspection
146(1)
8.3.4 Introspection Is "Commonplace": Summary
147(1)
8.4 Introspection as "Desirable"
148(2)
8.4.1 Introspection and Phenomenology
148(1)
8.4.2 The Neisser-Dreyfus Debate
149(1)
8.4.3 Introspection vs Phenomenology
149(1)
8.5 Introspection as "Unavoidable"
150(1)
8.6 A Hybrid Position
150(2)
8.7 Types of Truth in Introspection
152(3)
8.8 Introspection Is Legitimate: Summary
155(2)
9 Introspection Is Likely to Be Profitable
157(16)
9.1 Conceptual Arguments
158(1)
9.2 An Argument from Education
158(8)
9.2.1 Skill Questions
159(1)
9.2.2 Teaching Skills
160(1)
9.2.3 Self-Observations
161(1)
9.2.4 Mental Self-Observation Is Introspection
162(1)
9.2.5 Examples of Mental Skills Being Transmitted by Introspection
163(2)
9.2.6 Skills Only Part-Acquired by Explicit Instruction
165(1)
9.2.7 An Argument from Education: Summary
165(1)
9.3 Programming Impossible without Introspection
166(3)
9.3.1 Role-Playing
166(2)
9.3.2 Programming Is Introspective
168(1)
9.3.3 If So, What Is the Point of This Book?
168(1)
9.4 Introspection Is Likely to Be Profitable: Summary
169(4)
Part III Getting Practical
10 Details and How to Use Introspection for Artificial Intelligence
173(18)
10.1 Definitions and Delineations
174(5)
10.1.1 Definition for "AI Based on Introspection"
174(1)
10.1.2 Non-human-Like Inspirations
175(1)
10.1.2.1 Genetic Algorithms (Twice)
175(1)
10.1.2.2 Neural Nets
176(1)
10.1.3 Human-Like Inspirations (Non-introspective)
176(1)
10.1.4 Types of Introspection for AI
177(2)
10.2 The Process of Introspection for AI
179(2)
10.3 Comments on the Process of Introspection for AI
181(7)
10.3.1 Introspection Is a Witness Account
181(1)
10.3.2 Looking/Listening For
182(1)
10.3.3 Pollution
183(2)
10.3.4 Introspection: Is It Above or Below the Culture Line?
185(1)
10.3.5 Interpolation and Approximation
185(2)
10.3.5.1 The Holes in Introspection
185(1)
10.3.5.2 Opportunistic Approximation
186(1)
10.3.5.3 Analogue Cannot Arise Out of Digital
186(1)
10.3.5.4 Being Analogue Does Not Mean It Is Not Digital
186(1)
10.3.6 Multiple Iterations, Multiple Mechanisms
187(1)
10.3.7 Personnel
187(1)
10.4 Project Expectations
188(1)
10.5 Testing and Evaluation
189(2)
11 Examples
191(14)
11.1 Fuzzy Logic
192(2)
11.2 Case-Based Reasoning
194(1)
11.3 AIFO
195(7)
11.3.1 Introspection
196(1)
11.3.2 Implementation
196(1)
11.3.3 Example Run, Statistics
197(3)
11.3.4 Discussion
200(7)
11.3.4.1 Details and Parameters
201(1)
11.3.4.2 Why This Is More Anthropic
201(1)
11.3.4.3 Similarity
202(1)
11.4 AIF1
202(3)
12 A More Sophisticated Example
205(14)
12.1 Introspection
205(1)
12.2 Introspective Model
206(1)
12.3 Software Design
207(7)
12.3.1 Preliminary: Sequences in Software
207(1)
12.3.2 A Novel Data Type
208(2)
12.3.3 Decision Process
210(1)
12.3.4 More Details of AIF2's Implementation
211(1)
12.3.5 Dynamics of the Sequences Table
212(1)
12.3.6 Initial Conditions and Decisions
213(1)
12.3.7 Further Parameters
213(1)
12.4 AIF2 Example Runs
214(2)
12.4.1 Learn 1
214(1)
12.4.2 Learn 2
215(1)
12.4.3 Learn 3
215(1)
12.5 Discussion of AIF2
216(1)
12.6 Consequences of the Examples
216(3)
12.6.1 AIF Is More Like CBR Than Like Reinforcement Learning
216(1)
12.6.2 The "Sequence" Data Type
217(1)
12.6.3 Dynamic Symbols
217(1)
12.6.4 How AIF Is Gadamerian
218(1)
13 Summary, Consequences, Conclusion
219(12)
13.1 Summary
219(2)
13.2 Future Technical Work
221(2)
13.3 Possible Consequences for Cognitive Science
223(2)
13.3.1 Models for Scientific Psychology
223(1)
13.3.2 A Response to Dreyfus's Critique of AI
224(1)
13.3.3 Natural Language Processing
225(1)
13.3.4 Cognitive Models
225(1)
13.4 "Underpinning" Models in Philosophy
225(3)
13.4.1 Wittgenstein's "Seeing As"
226(1)
13.4.2 Gadamer
226(1)
13.4.3 Dreyfus's Demands from AI
226(1)
13.4.4 Wheeler's Action-Oriented Representations
227(1)
13.4.5 Adhyasa/Superimposition
228(1)
13.5 Open Questions
228(2)
13.5.1 Dilthey vs Gadamer
228(1)
13.5.2 Further Unexplored Terrain
229(1)
13.6 Conclusion
230(1)
Bibliography 231(12)
Index 243
Sam Freed is a researcher in the COGS (Centre for Cognitive Science) at the University of Sussex. This centre spans departments as diverse as neuroscience and philosophy, psychology, and computer science. Freeds background includes a career as a technologist starting as a research assistant in computer science at the age of 15 (at the Hebrew University, Jerusalem), work on Lotus 1-2-3 during Irelands 1990s boom, and managing international projects in internet security during the .com bubble. His education includes a BA in philosophy and comparative religion, an MA in cognitive science (both from the Hebrew University), and a PhD in informatics (from Sussex). His work centers on the relevance of the humanities to technology and on the historical analysis of the mindset behind current technology.