1Slide 1Tomorrow’s Technologyand You 8th Edition© 2008 Prentice-Hall, Inc.Slide 2Tomorrow’s Technologyand You 8/eChapter 15Is Artificial Intelligence Real?© 2008 Prentice-Hall, Inc.Slide 3Tomorrow’s Technology and You 8/eChapter 15ObjectivesExplain the two basic approaches of artificial intelligenceresearch.Describe several hard problems that artificial intelligenceresearch has not been able to solve yet.Describe several practical applications of artificial intelligence.Explain what robots are and give several examples illustratingwhat they can—and can’t—do.© 2008 Prentice-Hall, Inc.2Slide 4Tomorrow’s Technology and You 8/eChapter 15 Alan Turing, Military Intelligence, andIntelligent Machines Alan M. Turing was the British mathematician who designedthe world’s first operational electronic digital computerduring the 1940s: Turing effectively launched the field of artificial intelligence (AI) witha 1950 paper called “Computing Machinery and Intelligence.” In 1952 he was professionally and socially devastated when he wasarrested and injected with hormones for violation of British anti-homosexuality laws. The 41-year-old genius apparently committed suicide in 1954, yearsbefore the government made his wartime heroics public. Four decades after his death, Turing’s work still has relevance tocomputer scientists, mathematicians, and philosophers.© 2008 Prentice-Hall, Inc.Slide 5Tomorrow’s Technology and You 8/eChapter 15Thinking about Thinking MachinesCan Machines Think? The Turing test: The test involves two people and a computer. One person, the interrogator, sits at a terminal and types questions. The questions can be about anything—math, science, politics, sports,entertainment, art, human relationships, emotions, etc. As answers to the questions appear on the screen, the interrogatorattempts to guess whether those answers were typed by the otherperson or generated by the computer. According to Turing, by repeatedly fooling interrogators into thinking it isa person, a computer can demonstrate intelligent behavior. If it actsintelligently, it is intelligent.© 2008 Prentice-Hall, Inc.Slide 6Tomorrow’s Technology and You 8/eChapter 15Thinking about Thinking Machines Turing did not intend this test to be the only way to demonstrate machineintelligence; he pointed out that a machine could fail and still beintelligent. Even so, Turing believed that machines would be able to pass his test bythe turn of the century. So far no computer has come close, in spite of 40 years of AI research. While some people still cling to the Turing test to define artificialintelligence, most AI researchers favor less stringent definitions.© 2008 Prentice-Hall, Inc.3Slide 7Tomorrow’s Technology and You 8/eChapter 15Thinking about Thinking MachinesWhat Is Artificial Intelligence? Many computer scientists believe that if a task is easy toperform with a computer, it can’t be an example ofartificial intelligence. A more recent textbook definition reflects this point of view: Artificial intelligence is the study of how to make computers do thingsat which, at the moment, people are better. —Elaine Rich, in ArtificialIntelligence Artificial intelligence is the study of the computations that make itpossible to perceive, reason, and act. —Patrick Henry Winston, inArtificial Intelligence© 2008 Prentice-Hall, Inc.Slide 8Tomorrow’s Technology and You 8/eChapter 15 Thinking about Thinking Machines Two common approaches to AI Use computers to simulate human mental processes Three inherent problems: Most people have trouble knowing and describing how they do things. There are vast differences between the capabilities of the human brainand those of a computer. Even the most powerful supercomputers can’t approach the brain’sability to perform parallel processing. The best way to do something with a machine is often very differentfrom the way people would do it.© 2008 Prentice-Hall, Inc.Slide 9Tomorrow’s Technology and You 8/eChapter 15 Thinking about Thinking Machines A second approach to AI involved designing intelligentmachines independent of the way people think. This is a more common approach. Human intelligence is just one possible kind of intelligence. A machine’s method for solving a problem might be different fromthe human method, but no less intelligent. Many problems are far too complex to solve all at once. Break these problems into smaller problems that are easier to solve. Create programs that can function intelligently when confined tolimited domains.© 2008 Prentice-Hall, Inc.4Slide 10Tomorrow’s Technology and You 8/eChapter 15 Thinking about Thinking MachinesOpening Games One of the first popular domains for AI research was thecheckerboard. Some AI techniques are still used today in a variety ofapplications: Searching: Looking ahead to the possibilities generated by eachpotential move The staggering number of decision points makes brute-force searchingimpractical. Searching is generally guided by a planned strategy and by rules knownas heuristics.© 2008 Prentice-Hall, Inc.Slide 11Tomorrow’s Technology and You 8/eChapter 15 Thinking about Thinking Machines Heuristics: Rule of thumb Heuristics guide us toward judgments that experience tells us are likelyto be true. For example, in checkers, “Keep checkers in the king’s row as long aspossible.” Pattern recognition: identifying recurring patterns in input data The goal of pattern recognition is understanding or categorizingthat input. The best human chess and checkers players remember thousands ofcritical board patterns and know the best strategies for playing whenthose patterns or similar patterns appear. Game-playing programs recognize recurring patterns, too, but notnearly as well as people do.© 2008 Prentice-Hall, Inc.Slide 12Tomorrow’s Technology and You 8/eChapter 15 Thinking about Thinking Machines Computer game programs often have trouble identifying situations that aresimilar but not identical. Pattern recognition is probably the single biggest advantage a human gameplayer has over a computer opponent. Machine learning: learn from experience If a move pays off, a learning program is more likely to use that move (orsimilar moves) in future games. Most AI researchers have moved on to more interesting and practicalapplications.©