Wk 1 Ses 1-3 Intro to AI
description
Transcript of Wk 1 Ses 1-3 Intro to AI
Artificial Intelligence
Introduction to Artificial Intelligence * Property of STI Page 1 of 8
TOPIC TITLE: INTRODUCTION TO ARTIFICIAL INTELLIGENCE Specific Objectives: At the end of the topic session, the students are expected to: Cognitive:
1. Define what AI is. 2. Classify the categories of AI. 3. Describe each category of AI. 4. Recognize the foundations of AI. 5. Identify the history of AI.
Affective:
1. Listen to others with respect. 2. Actively participate in class discussions.
MATERIALS/EQUIPMENT:
o Topic slides o OHP
TOPIC PREPARATION:
o Read about Artificial Intelligence. o Have the students research on the following:
Definition of AI Foundation and history of AI
o It is imperative for the instructor to incorporate various kinds of teaching strategies while discussing the suggested topics. The instructor may use the suggested learning activities below to facilitate a thorough and creative discussion of the topic.
o Prepare the slides to be presented in class.
TOPIC PRESENTATION: The topic will revolve around the basic concepts of Artificial Intelligence. Discussions include definitions of AI, foundations of AI, and history of AI. This will be the suggested flow of discussion for the course topic:
1. Ask a student to briefly discuss what AI is based on his/her research. Then discuss the definitions of AI and its concept as presented in the slide.
2. Illustrate the process in the Turing Test approach. 3. Discuss the foundations of AI. 4. Explain the history of AI. 5. Ask the students to write an essay on what they have learned on
the topic presented.
Artificial Intelligence
Introduction to Artificial Intelligence * Property of STI Page 2 of 8
What is Artificial Intelligence? Page 1 of 15
Introduction to Artificial Intelligence
Artificial Intelligence
* Property of STI
Page 1 of 15
What is Artificial
Intelligence?
branch of computer science that is concerned with making computers act like humans
was coined in 1956 by John McCarthy
includes: Game playing Expert system Natural language Neural networks Robotics
What is Artificial Intelligence? AI is a branch of computer science that is concerned with making computers act like humans. Its term was coined in 1956 by John McCarthy at the Massachusetts Institute of Technology. AI includes the following:
Game playing – this is programming computers to play games such as chess, scrabble, etc.
Expert system – this is programming computers to make decisions in real life situations. An example would be some expert systems that help doctors diagnose disease based on symptoms.
Natural language – this is programming computers to understand natural human languages.
Neural networks – these are systems that stimulate intelligence by attempting to reproduce the types of physical connections that occur in animal brains.
Robotics – these are programming computers to see, hear, and react to other sensory stimuli.
As of now, there are no computers exhibit full artificial intelligence (able to stimulate human behavior). In May 1997, IBM developed a computer chess program called Deep Blue, which defeated the world chess champion Gary Kasparov in a chess match. In the field of robotics, computers are used in assembling plants. However, their capabilities are very limited in performing tasks. Robots have difficulty in identifying objects based on appearance or feel and they awkwardly move and handle objects. Natural language processing allows people to interact with computers without the need of any specialized knowledge. You could simply walk up to a computer and talk to it. Unfortunately, it is difficult to program computers to understand natural languages. It would require some rudimentary translation systems to translate from one human language to another. However, these systems are not nearly as good as human translators. Also, there is voice recognition system that can convert spoken sounds into written words, which only take dictations. These systems are also limited because you should speak slowly and distinctly. In the early 1980s, expert systems were believed to represent the future of AI and of computers in general. But they have not lived up to expectations. Many expert systems help human experts in the fields of medicine and engineering, but they are costly to produce and are only helpful in special cases. Today, the main focus of AI is on neural networks, which proves to be successful in voice recognition and natural language processing. Various programming languages are known as AI languages since they are used almost exclusively for AI applications. The two most commonly known are LISP and Prolog. According to several textbooks, the definitions of AI vary into two main dimensions. As illustrated in the table below, the ones on top are concerned with thought processes and reasoning, while the ones below address behavior.
System that thinks like humans System that thinks rationally
“The exciting new effort to make “The study of mental faculties
Artificial Intelligence
Introduction to Artificial Intelligence * Property of STI Page 3 of 8
Page 2 of 15
Introduction to Artificial Intelligence
Artificial Intelligence
* Property of STI
Page 2 of 15
What is Artificial
Intelligence?
Dimensions of AI:
System that thinks like humans
System that thinks rationally
“The exciting new effort to make computers think…machines with minds, in the full and literal sense.” (Haugeland, 1985)
“*The automation of+ activities that we associate with human thinking, activities such as decision-making, problem solving, learning…” (Bellman, 1978)
“The study of mental faculties through the use of computational models.” (Charniak and McDermott, 1985)
“The study of the computations that make it possible to perceive, reason, and act.” (Winston, 1992)
System that acts likehumans
System that acts rationally
“The art of creating machines that performs functions that require intelligence when performed by people.” (Kurzweil, 1990)
“The study of how to make computers do things at which, at the moment, people are better.” (Rich and Knight, 1991)
“Computational Intelligence is the study of the design of intelligent agents.” (Poole et al., 1998)
“AI… is concerned with intelligent behavior in artifacts.” (Nilsson, 1998)
Page 3 of 15
Introduction to Artificial Intelligence
Artificial Intelligence
* Property of STI
Page 3 of 15
What is Artificial
Intelligence?
Four categories of AI:
Thinking humanly
Thinking rationally
Acting humanly
Acting rationally
Thought process and reasoning
Or
Behavior
Measure success: Human or Rationality
Page 4 of 15
Introduction to Artificial Intelligence
Artificial Intelligence
* Property of STI
Page 4 of 15
What is Artificial
Intelligence?
Acting Humanly: The Turing Test Approach
devised by Allan M. Turing
determines whether or not a computer can be said to think like a human brain
devised a subjective test to answer the question, “Can machines think?”
a human interrogator is isolated and given the task of distinguishing between a human and a computer based on their replies to questions that the interrogator poses
AI System
Human
Interrogator
?Turing Test
computers think…machines with minds, in the full and literal sense.” (Haugeland, 1985) “[The automation of] activities that we associate with human thinking, activities such as decision-making, problem solving, learning…” (Bellman, 1978)
through the use of computational models.” (Charniak and McDermott, 1985) “The study of the computations that make it possible to perceive, reason, and act.” (Winston, 1992)
System that acts like humans System that acts rationally
“The art of creating machines that performs functions that require intelligence when performed by people.” (Kurzweil, 1990) “The study of how to make computers do things at which, at the moment, people are better.” (Rich and Knight, 1991)
“Computational Intelligence is the study of the design of intelligent agents.” (Poole et al., 1998) “AI… is concerned with intelligent behavior in artifacts.” (Nilsson, 1998)
The definitions on the left measures success in terms of fidelity to human performance, while on the right measure against an ideal concept of intelligence, which is called rationality. A system is rational if it does the “right thing”, given what it knows. Thus, AI can be viewed into four categories:
Thinking humanly Thinking rationally
Acting humanly Acting rationally
Acting Humanly: The Turing Test Approach Allan M. Turing devised a test called the Turing Test that determines whether or not a computer can be said to think like a human brain. In an attempt to cut through the philosophical debate on how to define “thinking,” Turing devised a subjective test to answer the question, “Can machines think?” and reasoned that if a computer acts, reacts, and interacts like a sentient being, then call it sentient. The test is simple: a human interrogator is isolated and given the task of distinguishing between a human and a computer based on their replies to questions that the interrogator poses. After a series of tests are performed, the interrogator attempts to determine which subject is human and which is an AI. The computer’s success at thinking can be quantified by its probability of being misidentified as the human subject. Thinking Humanly: Cognitive Approach You must have some way to determine how humans think if you want to say that a given program thinks like a human. You need to get inside the actual workings of human minds. These can be done in two ways:
1. Through introspection – this is trying to catch your own thoughts as they go by
2. Through psychological experiments Once you have a sufficient precise theory of the mind, this theory can be expressed as a computer program. If the program’s I/O and timing behavior matches the human’s behavior, that is evidence that some of program’s mechanisms may also be operating in humans.
Artificial Intelligence
Introduction to Artificial Intelligence * Property of STI Page 4 of 8
Page 5 of 15
Introduction to Artificial Intelligence
Artificial Intelligence
* Property of STI
Page 5 of 15
What is Artificial
Intelligence?
Thinking Humanly: Cognitive Approach
finding a way to determine how humans think
need to get inside the actual workings of human minds
• Through introspection
• Through psychological experiments
once you have a sufficient precise theory of the mind, this theory can be expressed as a computer program
if the program’s I/O and timing behavior matches the human’s behavior, that is evidence that some of program’s mechanisms may also be operating in humans
Cognitive science• Interdisciplinary
– Computer models from AI
– Experimental techniques from psychology
• Try to construct precise and testable theories of working with human mind
Page 6 of 15
Introduction to Artificial Intelligence
Artificial Intelligence
* Property of STI
Page 6 of 15
What is Artificial
Intelligence?
Thinking Rationally: Logic Approach
Aristotle attempts to codify “right thinking” –indisputable reasoning processes
Syllogism• Provided patterns for argument structures that
always gave correct conclusions given correct premises
• Example: “Juan is a man; all men are mortal; therefore, Juan is mortal”
Logic• Provides a precise notation for statements on all
kinds of things in the world and the relations between them
The so-called logicist tradition within AI hopes to build on such programs to create intelligent systems
Two obstacles:• It is not easy to take informal knowledge and
state it in the formal terms required by logical notation, particularly when the knowledge is less than 100% certain.
• Being able to solve a problem “in principle” and doing so “in practice” is different.
Page 7 of 15
Introduction to Artificial Intelligence
Artificial Intelligence
* Property of STI
Page 7 of 15
What is Artificial
Intelligence?
Acting Rationally: Rational Agent
An agent is something that perceives and acts.
Computer agents are expected to have other attributes that distinguish them from mere “programs,” such as:
• operating under autonomous control
• perceiving their environment
• persisting over a prolonged time period
• adapting to change
• being capable of taking on another’s goals
A rational agent is someone that acts so as to achieve the best outcome or, when there is uncertainty, the best expected outcome.
Making correct inferences is part of being a rational agent (however, correct inference is not all of rationality).
The interdisciplinary of cognitive science brings together computer models from AI and experimental techniques from psychology to try to construct precise and testable theories of the workings of the human mind. AI and cognitive science continue to fertilize each other, especially in the areas of vision, natural language, and learning. Vision in particular has recently made advances via an integrated approach that considers neurophysiologic evidence and computational models. Thinking Rationally: Logic Approach Aristotle was one of the first to attempt to codify “right thinking,” that is, indisputable reasoning processes. His syllogisms provided patterns for argument structures that always gave correct conclusions given correct premises. An example would be “Juan is a man; all men are mortal; therefore, Juan is mortal.” These laws of thoughts were supposed to rule the operation of the mind, and initiated the field of logic. The development of logic way back centuries ago provided an exact notation for statements on all kinds of things in the world and the relations between them. By 1965, programs existed that could, given enough time and memory, take a description of a program in logical notation and find the solution to the problem, if it exists. The supposed logicist tradition within AI hopes to develop on such programs to create intelligent systems. The logic approach has two obstacles:
It is not easy to take informal knowledge and state it in the formal terms required by logical notation, particularly when the knowledge is less than 100% certain.
Being able to solve a problem “in principle” and doing so “in practice” is different.
Even though these obstacles apply to any attempt to create computational reasoning systems, they appeared first in the logicist tradition since the power of the representation and reasoning systems are well-defined and well understood. Acting Rationally: Rational Agent An agent is something that perceives and acts. It comes from the Latin word “agree”, which means to do. However, computer agents are expected to have other attributes that distinguish them from mere “programs,” such as:
operating under autonomous control perceiving their environment persisting over a prolonged time period adapting to change being capable of taking on another’s goals
A rational agent is someone that acts so as to achieve the best outcome or, when there is uncertainty, the best expected outcome. The emphasis of the logic approach to AI was on correct inferences. Part of being a rational agent is making correct inferences since one way to act rationally is to reason logically to the conclusion that a given action will achieve one’s goal, and then to act on that conclusion. In contrast, correct inference is not all of rationality. There are ways of acting rationally that does not involve inference (e.g., reflex action).
Artificial Intelligence
Introduction to Artificial Intelligence * Property of STI Page 5 of 8
Page 8 of 15
Introduction to Artificial Intelligence
Artificial Intelligence
* Property of STI
Page 8 of 15
What is Artificial
Intelligence?
Acting Rationally: Rational Agent
All the cognitive skills required for the Turing Test are there to allow rational actions.
• Represent knowledge and reason with it
• Generate comprehensible sentences in natural language
• Learning
• Visual perception
Two advantages of studying AI as rational agent:
• It is more general than the logic approach.
• It is more amenable to scientific development than approaches based on human behavior or human thought.
All the cognitive skills required for the Turing Test are there to allow rational actions. Therefore, the abilities needed are the following:
Represent knowledge and reason with it – this enables you to reach good decisions in a wide variety of situations
Generate comprehensible sentences in natural language – the sentences help you get by in a complex society
Learning – this enables you to generate more effective strategies for dealing with it
Visual perception – this is to have a better idea of what an action might achieve
Studying AI as rational agent encompasses two advantages and these are as follows:
1. It is more general than the logic approach. Correct inference can be useful only for achieving rationality and it is not necessarily required.
2. It is more amenable to scientific development than approaches based on human behavior or human thought. The standard of rationality is clearly defined and completely general.
[What is Artificial Intelligence?, Pages 1-8 of 15]
Foundations of AI Page 9 of 15
Introduction to Artificial Intelligence
Artificial Intelligence
* Property of STI
Page 9 of 15
Foundations of AI
To clearly understand how artificial intelligence was conceptualized, we need first to understand the foundations of AI.
Its foundation can be understood through a series of questions.
Can formal rules be used to draw valid conclusions?
How does the mental mind arise from a physical brain?
Where does knowledge come from?
How does knowledge lead to action?
The objective of research in the foundations of AI is to address some of the basic questions.
What is a theory in Al?
What are the most abstract assumptions underlying the competing visions of intelligence?
What are the basic arguments for and against each assumption?
Foundations of AI You need first to understand the foundations of AI to be able to clearly understand how AI was conceptualized. This can be done through a series of questions as follows:
Can formal rules be used to draw valid conclusions? How does the mental mind arise from a physical brain? Where does knowledge come from? How does knowledge lead to action?
The objective of research in the foundations of AI is to address some of the basic questions such as:
What is a theory in Al? What are the most abstract assumptions underlying the
competing visions of intelligence? What are the basic arguments for and against each
assumption? The pursuit of AI does not occur in isolation. The fields of philosophy, linguistics, psychophysics, and theoretical computer science have exercised a historical influence over the field. Today, there is much dialogue as ever, specifically with the new field of cognitive science. One consequence of dialogue is that criticisms of positions held in one discipline frequently apply to positions held in other disciplines. The following are issues that have become focal points of debate and serve as dividing lines of positions.
1. Pre-eminence of knowledge and conceptualization. Intelligence that goes beyond insect-level intelligence needs declarative knowledge and some form of reasoning similar to computation. Core AI is the study of the conceptualizations of the world assumed and used by intelligent systems during cognition.
Artificial Intelligence
Introduction to Artificial Intelligence * Property of STI Page 6 of 8
Page 10 of 15
Introduction to Artificial Intelligence
Artificial Intelligence
* Property of STI
Page 10 of 15
Foundations of AI
Issues that have become focal points of debate and serve as dividing lines of positions
Pre-eminence of knowledge and conceptualization
Disembodiment
Kinematics of cognition are language-like
Learning can be added later
Uniform architecture
Other fields that were taken into the foundations of AI
Economics
Neuroscience
Psychology
Computer engineering
Linguistics
2. Disembodiment. Cognition and the knowledge it assumes can
be studied largely in abstraction from the details of perception and motor control.
3. Kinematics of cognition are language-like. It is possible to
describe the trajectory of knowledge states or informational states created during cognition using a vocabulary much like English or some regimented logico-mathematical version of English.
4. Learning can be added later. The kinematics of cognition and the domain knowledge required for cognition can be studied separately from the study of concept learning, psychological development, and evolutionary change.
5. Uniform architecture. There is a single architecture that underlies virtually all cognition.
There were other fields that were taken into the foundations of AI and these are as follows:
Economics – Is AI capable of making the right decisions to boost the economy?
Neuroscience – Can AI really imitate how the brain process information?
Psychology – How do humans and animals think and act? Computer Engineering – Can AI help build an efficient
computer? Linguistics – Can AI communicate efficiently and accurately?
[Foundations of AI, Pages 9-10 of 15]
History of AI Page 11 of 15
Introduction to Artificial Intelligence
Artificial Intelligence
* Property of STI
Page 11 of 15
History of AI
Human race has labeled itself as Homo sapiens – man the wise.
Tasks that are taken for granted require intelligence.
For over 2000 years, philosophers have been trying to understand how human intelligence works and how it should be d
One of the early theories revolves around the mind and body dualism.
the mind does not necessarily depend on the body for it to function properly
History of AI Human race has labeled itself as Homo sapiens – man the wise. This is because of the huge role that intelligence plays in our everyday life. Tasks that are taken for granted require intelligence. For example, the task of looking at an object and identifying it. This requires that a very complex set of operations take place so that a visual image of the object forms in the brain. After which, another set of operations is performed to map a description of that object so that its identify is known. As illustrated in the example, the task of identifying an object is very complex and difficult to model in a computational environment. But what exactly is intelligence and how does it works? How is it possible for a tiny brain to perceive, understand, and manipulate a world much larger than it? How do you go about making smart decisions? This is one of the ultimate puzzles of humanity. For over 2000 years, philosophers have been trying to understand how human intelligence works and how it should be done. Many have proposed theories and ideas and have changed and modified through the test of time. One of the early theories revolves around the mind and body dualism. This theory stated that the mental events making up conscious thought and reasoning are not necessarily dependent on certain physical events for both their existence and content. In simpler terms, this means that the mind does not necessarily depend on the body for it to function properly. This theory would be hard to prove that the creation of an artificially intelligent agent is possible.
Artificial Intelligence
Introduction to Artificial Intelligence * Property of STI Page 7 of 8
Page 12 of 15
Introduction to Artificial Intelligence
Artificial Intelligence
* Property of STI
Page 12 of 15
History of AI
Thomas Hobbes proposed the possibility of the creation of artificial life by man.
Nature, the art whereby God has made and governs the world, is by the art of man, as in many other things, so in this also imitated-that it can make an artificial animal. For seeing life is but a motion of limbs, the beginning whereof is in some principal part within, why may we not say that all automata (engines that move themselves by springs and wheels as does a watch) have an artificial life? For what is the heart but a spring, and the nerves but so many strings, and the joints but so many wheels giving motion to the whole body such as was intended by the artificer?
– The Leviathan
Page 13 of 15
Introduction to Artificial Intelligence
Artificial Intelligence
* Property of STI
Page 13 of 15
History of AI
Ratiocination is the main element of natural reason that involves the gaining of experience either through cause of effects or effects of cause.
By ratiocination, I mean computation. Now to compute is either to collect the sum of many things that are added together, or to know what remains when one thing is taken out of another. Ratiocination, therefore, is the same with addition and subtraction.
Page 14 of 15
Introduction to Artificial Intelligence
Artificial Intelligence
* Property of STI
Page 14 of 15
History of AI
1956 John McCarthy coined the term “Artificial Intelligence” at a Dartmouth computer conference.Demonstration of the first running AI program called the Logic Theorist at Carnegie Mellon University.
1958 John McCarthy invented the LISP language –an AI programming language – at Massachusetts Institute of Technology.
1964 Danny Bobrow showed that computers can understand natural language enough to solve algebra word problems correctly.
1965 Joseph Weizenbaum built ELIZA, which is an interactive program that carries on a dialogue in English on any topic.
1969 SRI robot, Shakey, demonstrated combining locomotion, perception and problem solving.
1979 The first computer-controlled autonomous vehicle named the Stanford Cart is built.
1983 Danny Hillis co-founds Thinking Machines, the first company to produce massively parallel computers.
However, on the other hand, Thomas Hobbes was one of the first philosophers to propose the possibility of the creation of artificial life by man. In his book entitled “The Leviathan”, he states that: Nature, the art whereby God has made and governs the world, is by the art of man, as in many other things, so in this also imitated-that it can make an artificial animal. For seeing life is but a motion of limbs, the beginning whereof is in some principal part within, why may we not say that all automata (engines that move themselves by springs and wheels as does a watch) have an artificial life? For what is the heart but a spring, and the nerves but so many strings, and the joints but so many wheels giving motion to the whole body such as was intended by the artificer? Here, Hobbes implies that since God created natural life, why is it impossible for man to create artificial life? In addition, he took this one step further by stating: By ratiocination, I mean computation. Now to compute is either to collect the sum of many things that are added together, or to know what remains when one thing is taken out of another. Ratiocination, therefore, is the same with addition and subtraction. Ratiocination is the main element of natural reason that involves the gaining of experience either through cause of effects or effects of cause. This statement is one of the foundations of modern AI because it breaks down intelligence into a computational process, which is what AI is trying to model in computational machines. The table below summarizes the major events of modern history of AI.
1956 John McCarthy coined the term “Artificial Intelligence” at a Dartmouth computer conference. Demonstration of the first running AI program called the Logic Theorist at Carnegie Mellon University.
1958 John McCarthy invented the LISP language – an AI programming language – at Massachusetts Institute of Technology.
1964 Danny Bobrow showed that computers can understand natural language enough to solve algebra word problems correctly.
1965 Joseph Weizenbaum built ELIZA, which is an interactive program that carries on a dialogue in English on any topic.
1969 SRI robot, Shakey, demonstrated combining locomotion, perception and problem solving.
1979 The first computer-controlled autonomous vehicle named the Stanford Cart is built.
1983 Danny Hillis co-founds Thinking Machines, the first company to produce massively parallel computers.
1985 The drawing program named Aaron, which was created by Harold Cohen, is demonstrated at AI conference.
1990s Major advances in all areas of AI, with significant demonstrations in machine learning, intelligent tutoring, case-based reasoning, multi-agent planning, scheduling, uncertain reasoning, data mining, natural landscape understanding and translation, vision, virtual reality, and games.
1997 The chess program built by IBM named Deep Blue beats
Artificial Intelligence
Introduction to Artificial Intelligence * Property of STI Page 8 of 8
Page 15 of 15
Introduction to Artificial Intelligence
Artificial Intelligence
* Property of STI
Page 15 of 15
History of AI
1985 The drawing program named Aaron, which was created by Harold Cohen, is demonstrated at AI conference.
1990s Major advances in all areas of AI, with significant demonstrations in machine learning, intelligent tutoring, case-based reasoning, multi-agent planning, scheduling, uncertain reasoning, data mining, natural landscape understanding and translation, vision, virtual reality, and games.
1997 The chess program built by IBM named Deep Blue beats the world chess champion Garry Kasparov in a chess match.
Late1990s
Web crawlers and other AI-based information-extraction programs become essential in widespread of the World Wide Web.
2000 Interactive robot pets (a.k.a. "smart toys") become commercially available, realizing the vision of the 18th cen. novelty toy makers.
the world chess champion Garry Kasparov in a chess match.
Late 1990s Web crawlers and other AI-based information-extraction programs become essential in widespread of the World Wide Web.
2000 Interactive robot pets (a.k.a. "smart toys") become commercially available, realizing the vision of the 18th cen. novelty toy makers.
[History of AI, Pages of ]
EVALUATION/GENERALIZATION:
o AI is a branch of computer science that is concerned with making computers act like humans.
o AI includes game playing, expert system, natural language, neural networks, and robotics.
o The four categories of AI are thinking humanly, thinking rationally, acting human, and acting rationally.
o To clearly understand how artificial intelligence was conceptualized, we need first to understand the foundations of AI.
o Human race has labeled itself as Homo sapiens – man the wise. o Ask the students to write an essay on the topic presented.
REFERENCES:
Russell and Norvig, (2003), Artificial intelligence: a modern
approach (2nd
ed.), Prentice Hall Luger, George F., Wesley, Addison, Artificial intelligence:
structures and strategies for complex problem solving (4th ed.)