Last update: January 28, 2010CMSC 421, Artificial IntelligenceChapter 1Chapter 1 1What is AI?Try to get computers to be intelligent. But what does that mean?Chapter 1 2What is AI?Try to get computers to be intelligent. But what does that mean?♦ Systems that think intelligently, or systems that act intelligently?♦ Do it like humans do it, or some other way?Chapter 1 3What is AI?Try to get computers to be intelligent. But what does that mean?♦ Systems that think intelligently, or systems that act intelligently?♦ Do it like humans do it, or some other way?Computers thatthink like humans think rationallyact like humans act rationallyChapter 1 4Acting like humansthink like humans think rationallyact like humans act rationallyHow many of you have heard of the Turing Test?Chapter 1 5Acting like humans: The Turing testTuring (1950) “Computing machinery and intelligence”:♦ the Imitation GameAI SYSTEMHUMAN? HUMANINTERROGATORChapter 1 6Acting like humans: The Turing testTuring (1950) “Computing machinery and intelligence”:♦ the Imitation GameAI SYSTEMHUMAN? HUMANINTERROGATOR♦ Predicted that by 2000, a machine might have a 30% chance offooling an ordinary person for 5 minutes♦ Anticipated all major arguments against AI in following 50 years♦ Suggested major components of AI:knowledge, reasoning, language, understanding, learningProblem: Turing test is not reproducible, constructive, oramenable to mathematical analysisChapter 1 7Thinking like humans: Cognitive Sciencethink like humans think rationallyact like humans act rationallyInformation-processing psychologyScientific theories of internal activities of the brain– What level of abstraction? “knowledge” or “circuits”?– How to validate? Requires1) Predicting and testing behavior of human subjects (top-down)or 2) Direct identification from neurological data (bottom-up)Both approaches (roughly, Cognitive Science and Cognitive Neuroscience)are now distinct from AIBut both share with AI the following characteristic:We don’t yet have theories that explain or produceanything resembling human-level general intelligenceChapter 1 8Thinking rationally: Laws of Thoughtthink like humans think rationallyact like humans act rationallyNormative (or prescriptive) rather than descriptiveAristotle: what are correct arguments/thought processes?The ancient Greeks developed various forms of logic:notation and rules of derivation for thoughtsDirect line through mathematics and philosophy to modern AIProblems:1) Not all intelligent behavior is mediated by logical deliberation2) What is the purpose of thinking? What thoughts should I haveout of all the thoughts (logical or otherwise) that I could have?Chapter 1 9Acting rationallythink like humans think rationallyact like humans act rationallyRational behavior: doing the right thingThe right thing: that which is expected to maximize goal achievement,given the available informationDoesn’t necessarily involve thinkinge.g., blinking reflexbut means that thinking should be in the service of rational actionAristotle (Nicomachean Ethics):Every art and every inquiry, and similarly everyaction and pursuit, is thought to aim at some goodChapter 1 10Rational agentsAn agent is an entity that perceives and actsRussell & Norvig’s book focuses on designing rational agentsAbstractly, an agent is a function from percept histories to actions:f : P∗→ AFor any given class of environments and tasks, we seek theagent (or class of agents) with the best performanceCaveats:♦ Computational limitations make perfect rationality unachievable→ design best program for the available machine resources♦ Not an absolute standard of rationalityInstead, rationality relative to the performance measureWhen (and why) should we believe a performance measure is adequate?Chapter 1 11Ideas Adapted from Other FieldsPhilosophy logic, methods of reasoningmind as physical systemfoundations of learning, language, rationalityMathematics formal representation and proofalgorithms, computation, (un)decidability, (in)tractability,probabilityPsychology adaptationphenomena of perception and motor controlexperimental techniques (psychophysics, etc.)Economics formal theory of rational decisionsLinguistics knowledge representation, grammarNeuroscience physical substrate for mental activityControl theory homeostatic systems, stabilitysimple optimal agent designsChapter 1 12Brief History of AI1943 McCulloch & Pitts: Boolean circuit model of brain1950 Turing’s “Computing Machinery and Intelligence”1952–69 Look, Ma, no hands!1950s Early AI programs, including Samuel’s checkers program,Newell & Simon’s Logic Theorist, Gelernter’s Geometry Engine1956 Dartmouth meeting: “Artificial Intelligence” adopted1965 Robinson’s complete algorithm for logical reasoning1966–74 AI discovers computational complexityNeural network research almost disappears1969–79 Early development of knowledge-based systems1980–88 Expert systems industry booms1988–93 Expert systems industry busts: “AI Winter”1985–95 Neural networks return to popularity1988– Resurgence of probability; general increase in technical depthSoft computing (fuzzy logic, genetic algorithms, etc.)1995– Popularity of the notion of “agents”2003– Human-level AI back on the agendaChapter 1 13State of the artWhich of the following can be done at present?♦ Drive safely along a curving mountain roadChapter 1 14State of the artWhich of the following can be done at present?√Drive safely along a curving mountain road♦ Drive safely along US 1Chapter 1 15State of the artWhich of the following can be done at present?√Drive safely along a curving mountain road×Drive safely along US 1♦ Buy a week’s worth of groceries on the webChapter 1 16State of the artWhich of the following can be done at present?√Drive safely along a curving mountain road×Drive safely along US 1√Buy a week’s worth of groceries on the web♦ Buy a week’s worth of groceries at the local GiantChapter 1 17State of the artWhich of the following can be done at present?√Drive safely along a curving mountain road×Drive safely along US 1√Buy a week’s worth of groceries on the web×Buy a week’s worth of groceries at your local Giant♦ Play a decent game of bridgeChapter 1 18State of the artWhich of the following can be done at present?√Drive safely along a curving mountain road×Drive safely along US 1√Buy a week’s worth of groceries on