Chapter 1
Introduction
Fundamental issues in intelligent systems 

Overview

The course follows the Association of Computing Machinery's Curriculum 2001 guide for university computer science programs. The guide provides the following description:

The field of artificial intelligence (AI) is concerned with the design and analysis of autonomous agents. These are software systems and/or physical machines, with sensors and actuators, embodied for example within a robot or an autonomous spacecraft. An intelligent system has to perceive its environment, to act rationally towards its assigned tasks, to interact with other agents and with human beings. These capabilities are covered by topics such as computer vision, planning and acting, robotics, multiagents systems, speech recognition, and natural language understanding. They rely on a broad set of general and specialized knowledge representations and reasoning mechanisms, on problem solving and search algorithms, and on machine learning techniques.

Furthermore, artificial intelligence provides a set of tools for solving problems that are difficult or impractical to solve with other methods. These include heuristic search and planning algorithms, formalisms for knowledge representation and reasoning, machine learning techniques, and methods applicable to sensing and action problems such as speech and language understanding, computer vision, and robotics, among others. The student needs to be able to determine when an AI approach is appropriate for a given problem, and to be able to select and implement a suitable AI method.

Definition of AI?

• One pragmatic definition of AI is that if the problem is hard, not well-understood how to solve and novel it is probably AI.

• Another is anything that is generally assumed to require human intelligence.

• Accounting applications and pay-roll systems were once exclusively performed by humans; while difficult solutions and methods are well-understood; the problem has been solved by hand for hundreds of years and recently on computer but is not AI.

• Translating Russian to English is something humans can do but is has proven difficult on computers; natural language processing is still part of AI.

• Occasionally, AI problems are solved well enough for commercial applications (e.g. speech recognition (of individual words), expert systems for factory automation (beer brewing), loan application vetting).

1.1 What is AI? (History of artificial intelligence)

Four approaches:

1. Acting humanly: The Turing Test approach

   Alan Turing proposed that an entity (computer) displayed intelligence if its actions were indistinguishable from a human's.

   • Human interrogates computer via text only, no other physical interaction
   • Computer requires:
     • natural language processing to communicate in human language
     • knowledge representation to store what it knows and hears
     • automated reasoning

   • Modified Extract from Turing paper

     Interrogator: In the first line of your sonnet which reads `Shall I compare thee to a summer's day', would not `a spring day' do as well or better?

     Computer: It wouldn't scan.

     Interrogator: How about  'A winter's day'? That would scan all right?

     Computer: Yes, but nobody wants to be compared to a winter's day.

     Interrogator: Would you say Mr. Pickwick reminded you of Christmas?

     Computer: In a way.

     Interrogator: Yet Christmas is a winter's day, and I do not think Mr. Pickwick would mind the comparison.

     Computer: I don't think you're serious. By a winter's day one means a typical winter's day, rather than a special one like Christmas.

   • Eliza Session

     The earliest attempt was a system called Eliza (1966).

     Eliza: What brought you here to see me today?

     Visitor: Oh, nothing much. I'm feeling a bit tired, that's all.

     Eliza: Why do you think you're feeling a bit tired?

     Visitor: Well, I've been traveling a lot, and away from home.

     Eliza: Tell me more about your family.

     Visitor: The trouble is, my mother's ill.

     Eliza: How long has she been ill?

     Visitor: The trouble is, my mother's Irish.

     Eliza: How long has she been Irish?

     Visitor: The trouble is, my mother's poodle.

     Eliza: How long has she been poodle?

      

2. Thinking humanly: The cognitive modeling approach

   • Cognitive modeling: Determine how we think by introspection or experimentation

     Self-awareness is important: "I think therefore I am"

     Humans feel emotions and apparently don't always think (or act) rationally

      

   • One weakness in modeling human thought; if aeronautical engineers modeled flight based on birds, airplanes would have feathers.

      

3. Thinking rationally: The "laws of thought" approach

   • The laws of thought:

     eg "Socrates is a man. All men are mortal. Therefore Socrates is mortal"

     Codifying rational thinking started with Aristotle (at least in the West)

     The study of logic has greatly influenced AI

      

4. Acting rationally: The rational agent approach

   • The rational agent: perform actions which will (most likely) achieve one's goals

   Knowledge may not be perfect | we need to go beyond strict rational thought in general

   The rational agent view is the basis of Artificial Intelligence: A Modern Approach

1.2 Foundations of AI

• The big question: How does the mind arise from the brain?

• Philosophy,  logic, methods of reasoning

• Mathematics formal representation and proof algorithms, computation, (un)decidability, (in)tractability, probability

• Psychology adaptation, perception

• Linguistics knowledge representation, grammar

• Neuroscience physical substrate for mental activity

• Control theory simple optimal agent designs

1.3 Potted history of AI

• 1943     Boolean circuit model of brain

• 1950     Turing test

• 1950s    Early AI programs: checkers program, automated mathematical proofs

• 1956     Dartmouth meeting: "Artificial Intelligence" adopted

• 1962     Early neural networks

• 1966-74 AI discovers computational complexity. Neural network research almost disappears

• 1969-79 Early development of knowledge-based systems

• 1980-88 Expert systems industry booms

• 1988-93 Expert systems industry busts: "AI Winter"

• 1985-95 Neural networks return to popularity

• 1988      Resurgence of probabilistic methods. Rapid increase in technical depth of mainstream AI. Data mining, Bayesian networks

• 1995    Intelligent agents

1.4 State of the Art

Which of the following can be done at present?

• Play a decent game of table tennis

• Drive along a curving mountain road

• Drive across an open desert

• Play a decent game of bridge

• Discover and prove a new mathematical theorem

• Write an intentionally funny story

• Give competent legal advice in a specialized area of law

• Translate spoken English into spoken Swedish in real time