1. In 1637 the French philosopher-mathematician ReneDescartes predicted that it would neverBe possible to makeA machine that thinksAs humansDo. In 1950, theBritish mathematicianAndComputer pioneerAlan TuringDeclared that oneDay there wouldBeA machine thatDuplicate human intelligence in every wayAnd prove itBy passingA specialized test. In this test,AComputerAndA human hidden from view wouldBeAsked random identical questions, if theComputer were successful; the questioner wouldBe unable toDistinguish the machine from the personBy theAnswers.
InspiredBy Turing's theory, the firstConference onAIConvenedAtDartmouthCollege in New Hampshire in 1956. SoonAfterwardsAnAI laboratory was startedAt Massachusetts Institute of TechnologyBy John McCarthyAnd Marvin Minsky, two of the nation's leadingAI proponents. McCarthyAlso invented theAIComputer language, Lisp;ButBy the early 1990sAI itself had notBeenAchieved. However, logic programsCalled expert systemsAllowComputers to "make,Decisions"By interpretingDataAnd selecting fromAmongAlternatives. TechniciansCan run programs used inComplex medicalDiagnosis, language translation, mineral exploration,And evenComputerDesign.
MachineryCan outperform humans physically. So, too,CanComputers outperform mental functions in limitedAreas--notably in the speed of mathematicalCalculations. For example, the fastestComputersDevelopedAreAble to perform roughly 10BillionCalculations per second.But making more powerfulComputers will probably notBe the way toCreateA machineCapable of passing the Turing test.Computer programs operateAccording to set procedures, or logic steps,CalledAlgorithms (算法). InAddition, mostComputersDo serial processing: operations of recognitionAndComputationAre performed oneAtA time. TheBrain works inA mannerCalled parallel processing, performing operations simultaneously. ToAchieve simulated (模拟) parallel processing, some supercomputers haveBeen made with multiple processors to follow severalAlgorithmsAt the same time.
Critics of thisApproach insist that solvingAComputationDoes not indicate understanding, somethingA person who solvedA problem would have. Human reasoning is notBased solely on rules of logic. It involves perception,Awareness, emotional preferences, values, evaluating experience, theAbility to generalizeAnd weigh options,And more. Some proponents ofA1 have, therefore, suggested thatComputers shouldBe patternedAfter the humanBrain, which essentiallyConsists ofA network of nerveCells.
By the early 1990s, theClosestApproximation toAI wasA special siliconChipBuilt toBehave likeA humanBrainCell. It was modeledAfter the internal workings of neurons in the humanCerebralContext. Unlike theConventional silicon ship, which works inDigital mode, the new siliconChip works inAnalog mode, much the wayA humanBrainCell works.
21. According to Turing,AComputerCan prove to have human-like intelligence inA special test if____
A. theComputer givesBetterAnswers
B. the questioner fails to give identical questions
C. the questionerCan't tellBetween theAnswers ofA personAndAComputer
D. The questionerCan't find the person hiddenBy theComputer