CSC 480 Artificial Intelligence Dr Franz J Kurfess

CSC 480: Artificial Intelligence Dr. Franz J. Kurfess Computer Science Department Cal Poly © 2000 -2012 Franz Kurfess Introduction

Course Overview u Introduction u Knowledge u Intelligent u Agents u Search u u problem solving through search informed search u Games u games as search problems © 2000 -2012 Franz Kurfess u u u and Reasoning reasoning agents propositional logic predicate logic knowledge-based systems u Learning u u learning from observation neural networks u Conclusions Introduction

Chapter Overview Introduction u Logistics u Foundations Intelligence u Motivation u Objectives u What is Artificial Intelligence? u u u definitions Turing test cognitive modeling rational thinking acting rationally © 2000 -2012 Franz Kurfess of Artificial u u u philosophy mathematics psychology computer science linguistics u History of Artificial Intelligence u Important Concepts and Terms u Chapter Summary Introduction

Instructor u Dr. Franz J. Kurfess u Professor, CSC Dept. u Areas of Interest u u u Artificial Intelligence Knowledge Management, Intelligent Agents Neural Networks & Structured Knowledge Human-Computer Interaction User-Centered Design u Contact u u preferably via email: fkurfess@calpoly. edu Web page http: //www. csc. calpoly. edu/~fkurfess phone (805) 756 7179 office 14 -218 © 2000 -2012 Franz Kurfess Introduction

Logistics u Introductions u Course Materials u textbook u handouts u Web page u Term Project u Lab and Homework Assignments u Exams u Grading © 2000 -2012 Franz Kurfess Introduction

Course Material u on the Web (http: //www. csc. calpoly. edu/~fkurfess) u syllabus u schedule u project information u homework and lab assignment descriptions u most lab assignment submissions u Poly. Learn u grades u assignment and some lab submissions u student presentation schedule u TRAC Wiki © 2000 -2012 Franz Kurfess u project documentation by students Introduction

Term Project u development u of a practical application in a team prototype, emphasis on conceptual and design issues, not so much performance u implementation u must be accessible to others e. g. Web/Java u milestones/deliverables u mid-quarter and final presentation/display u peer evaluation u each team evaluates the system of another team u information u u exchange on the Web course Web site TRAC Wiki for documentation of individual teams v team accounts © 2000 -2012 Franz Kurfess Introduction

Homework and Lab Assignments u individual assignments u some lab exercises in small teams u documentation, hand-ins usually person u may consist of questions, exercises, outlines, programs, experiments © 2000 -2012 Franz Kurfess Introduction

Exams u experiment with weekly quizzes instead of the midterm/final as described below u coordination with online Stanford AI course? u one midterm exam u one final exam u typical exam format 5 -10 multiple choice questions u 2 -4 short explanations/discussions u v v u explanation of an important concept comparison of different approaches one problem to solve may involve the application of methods discussed in class to a specific problem © 2000 -2012 Franz Kurfess Introduction v usually consists of several subtasks v

Motivation u scientific u try curiosity to understand entities that exhibit intelligence u engineering u building challenges systems that exhibit intelligence u some tasks that seem to require intelligence can be solved by computers u progress in computer performance and computational methods enables the solution of complex problems by computers u humans may be relieved from tedious tasks © 2000 -2012 Franz Kurfess Introduction

Objectives u become familiar with criteria that distinguish human from artificial intelligence u know about different approaches to analyze intelligent behavior u understand the influence of other fields on artificial intelligence u be familiar with the important historical phases the field of artificial intelligence went through © 2000 -2012 Franz Kurfess Introduction

Exercise: Intelligent Systems u select a task that you believe requires intelligence u examples: playing chess, solving puzzles, translating from English to German, finding a proof for a theorem u for that task, sketch a computer-based system that tries to solve the task u architecture, components, behavior u what are the computational methods your system relies on u e. g. data bases, matrix multiplication, graph traversal u what are the main challenges u how do humans tackle the task © 2000 -2012 Franz Kurfess Introduction

Trying to define AI u so far, there is no generally accepted definition of Artificial Intelligence u textbooks either skirt the issue, or emphasize particular aspects © 2000 -2012 Franz Kurfess Introduction

Examples of Definitions u cognitive u u approaches emphasis on the way systems work or “think” requires insight into the internal representations and processes of the system u behavioral u approaches only activities observed from the outside are taken into account u human-like u try to emulate human intelligence u rational u u systems that do the “right thing” idealized concept of intelligence © 2000 -2012 Franz Kurfess Introduction

The Turing Test u proposed by Alan Turing in 1950 to provide an operational definition of intelligent behavior u the ability to achieve human-level performance in all cognitive tasks, sufficient to fool an interrogator u the computer is interrogated by a human via a teletype u it passes the test if the interrogator cannot identify the answerer as computer or human © 2000 -2012 Franz Kurfess Introduction

Basic Capabilities for passing the Turing test u natural language processing u communicate u knowledge u store with the interrogator representation information u automated u answer u machine reasoning questions, draw conclusions learning u adapt behavior u detect patterns © 2000 -2012 Franz Kurfess Introduction

Relevance of the Turing Test u not much concentrated effort has been spent on building computers that pass the test u Loebner Prize u there is a competition and a prize for a somewhat revised challenge u see details at http: //www. loebner. net/Prizef/loebner-prize. html u “Total Turing Test” u includes video interface and a “hatch” for physical objects u requires computer vision and robotics as additional capabilities © 2000 -2012 Franz Kurfess Introduction

Cognitive Modeling u tries to construct theories of how the human mind works u uses computer models from AI and experimental techniques from psychology u most AI approaches are not directly based on cognitive models u often difficult to translate into computer programs u performance problems © 2000 -2012 Franz Kurfess Introduction

Rational Thinking u based on abstract “laws of thought” u usually with mathematical logic as tool u problems and knowledge must be translated into formal descriptions u the system uses an abstract reasoning mechanism to derive a solution u serious real-world problems may be substantially different from their abstract counterparts u difference © 2000 -2012 Franz Kurfess between “in principle” and “in practice” Introduction

Rational Agents u an agent that does “the right thing” u it achieves its goals according to what it knows u perceives information from the environment u may utilize knowledge and reasoning to select actions u performs actions that may change the environment © 2000 -2012 Franz Kurfess Introduction

Behavioral Agents u an agent that exhibits some behavior required to perform a certain task u the internal processes are largely irrelevant u may simply map inputs (“percepts”) onto actions u simple behaviors may be assembled into more complex ones © 2000 -2012 Franz Kurfess Introduction

Foundations of Artificial Intelligence u philosophy u mathematics u psychology u computer science u linguistics © 2000 -2012 Franz Kurfess Introduction

Philosophy u related questions have been asked by Greek philosophers like Plato, Socrates, Aristotle u theories of language, reasoning, learning, the mind u dualism (Descartes) ua part of the mind is outside of the material world u materialism u all (Leibniz) the world operates according to the laws of physics © 2000 -2012 Franz Kurfess Introduction

Mathematics u formalization of tasks and problems u logic u propositional logic u predicate logic u computation u Church-Turing thesis u intractability: NP-complete problems u probability u degree of certainty/belief © 2000 -2012 Franz Kurfess Introduction

Psychology u behaviorism u only observable and measurable percepts and responses are considered u mental constructs are considered as unscientific v knowledge, beliefs, goals, reasoning steps u cognitive psychology u the brain stores and processes information u cognitive processes describe internal activities of the brain © 2000 -2012 Franz Kurfess Introduction

Computer Science u provides tools for testing theories u programmability u speed u storage u actions © 2000 -2012 Franz Kurfess Introduction

Linguistics u understanding u sentence and analysis of language structure, subject matter, context u knowledge representation u computational linguistics, natural language processing u hybrid field combining AI and linguistics © 2000 -2012 Franz Kurfess Introduction

AI through the ages © 2000 -2012 Franz Kurfess Introduction

Conception (late 40 s, early 50 s) u artificial neurons (Mc. Culloch and Pitts, 1943) u learning in neurons (Hebb, 1949) u chess programs (Shannon, 1950; Turing, 1953) u neural computer (Minsky and Edmonds, 1951) © 2000 -2012 Franz Kurfess Introduction

Birth: Summer 1956 u gathering of a group of scientists with an interest in computers and intelligence during a two-month workshop in Dartmouth, NH u “naming” of the field by John Mc. Carthy u many of the participants became influential people in the field of AI © 2000 -2012 Franz Kurfess Introduction

Baby steps (late 1950 s) u demonstration of programs solving simple problems that require some intelligence u Logic Theorist (Newell and Simon, 1957) u checkers programs (Samuel, starting 1952) u development of some basic concepts and methods u Lisp (Mc. Carthy, 1958) u formal methods for knowledge representation and reasoning u mainly of interest to the small circle of relatives © 2000 -2012 Franz Kurfess Introduction

Kindergarten (early 1960 s) u child prodigies astound the world with their skills u General Problem Solver (Newell and Simon, 1961) u Shakey the robot (SRI) u geometric analogies (Evans, 1968) u algebraic problems (Bobrow, 1967) u blocks world (Winston, 1970; Huffman, 1971; Fahlman, 1974; Waltz, 1975) u neural networks (Widrow and Hoff, 1960; Rosenblatt, 1962; Winograd and Cowan, 1963) u machine evolution/genetic algorithms (Friedberg, 1958) © 2000 -2012 Franz Kurfess Introduction

Teenage years (late 60 s, early 70 s) u sometimes also referred to as “AI winter” u microworlds aren’t the real thing: scalability and intractability problems u neural networks can learn, but not very much (Minsky and Papert, 1969) u expert systems are used in some real-life domains u knowledge representation schemes become useful © 2000 -2012 Franz Kurfess Introduction

AI gets a job (early 80 s) u commercial applications of AI systems u R 1 expert system for configuration of DEC computer systems (1981) u expert system shells u AI machines and tools © 2000 -2012 Franz Kurfess Introduction

Some skills get a boost (late 80 s) u after all, neural networks can learn more -in multiple layers (Rumelhart and Mc. Clelland, 1986) u hidden Markov models help with speech problems u planning becomes more systematic (Chapman, 1987) u belief networks probably take some uncertainty out of reasoning (Pearl, 1988) © 2000 -2012 Franz Kurfess Introduction

AI matures (90 s) u handwriting and speech recognition work -- more or less u AI is in the driver’s seat (Pomerleau, 1993) u wizards and assistants make easy tasks more difficult u intelligent agents do not proliferate as successfully as viruses and spam © 2000 -2012 Franz Kurfess Introduction

Intelligent Agents appear (mid-90 s) u distinction between hardware emphasis (robots) and software emphasis (softbots) u agent architectures u SOAR u situated u agents embedded in real environments with continuous inputs u Web-based agents u the agent-oriented perspective helps tie together various subfields of AI u but: “agents” has become a buzzword u widely (ab)used, often indiscriminately © 2000 -2012 Franz Kurfess Introduction

AI Disappears (~2000) u more and more AI approaches are incorporated into generic computing approaches u planning, scheduling u machine learning u natural language processing u reasoning u autonomy © 2000 -2012 Franz Kurfess Introduction

A Lack of Meaning (~ 2005) u most AI methods are based on symbol manipulation and statistics u e. g. search engines u the interpretation of generated statements is problematic u often left to humans u the Semantic Web suggests to augment documents with metadata that describe their contents u computers still don’t “understand”, but they can perform tasks more competently © 2000 -2012 Franz Kurfess Introduction

Outlook u concepts u u and methods many are sound, and usable in practice some gaps still exist: “neat” vs. “scruffy” debate u computational u u aspects most methods need improvement for wide-spread usage vastly improved computational resources (speed, storage space) u applications u u u reasonable number of applications in the real world many are “behind the scene” expansion to new domains u education u u established practitioners may not know about new ways newcomers may repeat fruitless efforts from the past © 2000 -2012 Franz Kurfess Introduction

Important Concepts and Terms u u u agent automated reasoning cognitive science computer science intelligent agent knowledge representation linguistics Lisp logic machine learning microworlds © 2000 -2012 Franz Kurfess u u u u natural language processing neural network predicate logic propositional logic rational agent rationality Turing test Introduction

Chapter Summary u introduction to important concepts and terms u relevance of Artificial Intelligence u influence from other fields u historical development of the field of Artificial Intelligence © 2000 -2012 Franz Kurfess Introduction

© 2000 -2012 Franz Kurfess Introduction