雅思阅读科学类3Word格式文档下载.docx
《雅思阅读科学类3Word格式文档下载.docx》由会员分享,可在线阅读,更多相关《雅思阅读科学类3Word格式文档下载.docx(18页珍藏版)》请在冰点文库上搜索。
【B】Thefieldwaslaunched,andtheterm‘artificialintelligence’coined,ataconferencein1956,byagroupofresearchersthatincludedMarvinMinsky,JohnMcCarthy,HerbertSimonandAlanNewell,allofwhomwentontobecomeleadingfiguresinthefield.Theexpressionprovidedanattractivebutinformativenameforaresearchprogrammethatencompassedsuchpreviouslydisparatefieldsasoperationsresearch,cybernetics,logicandcomputerscience.Thegoaltheysharedwasanattempttocaptureormimichumanabilitiesusingmachines.Thatsaid,differentgroupsofresearchersattackeddifferentproblems,fromspeechrecognitiontochessplaying,indifferentways;
AIunifiedthefieldinnameonly.Butitwasatermthatcapturedthepublicimagination.
【C】MostresearchersagreethatAIpeakedaround1985.Apublicrearedonscience-fictionmoviesandexcitedbythegrowingpowerofcomputershadhighexpectations.Foryears,AIresearchershadimpliedthatabreakthroughwasjustaroundthecorner.MarvinMinskysaidin1967thatwithinagenerationtheproblemofcreating‘artificialintelligence’wouldbesubstantiallysolved.Prototypesofmedical-diagnosisprogramsandspeechrecognitionsoftwareappearedtobemakingprogress.Itprovedtobeafalsedawn.Thinkingcomputersandhouseholdrobotsfailedtomaterialise,andabacklashensued.‘Therewasundueoptimismintheearly1980s,’saysDavidLeake,aresearcheratIndianaUniversity.‘Thenwhenpeoplerealisedthesewerehardproblems,therewasretrenchment.Bythelate1980s,thetermAIwasbeingavoidedbymanyresearchers,whooptedinsteadtoalignthemselveswithspecificsub-disciplinessuchasneuralnetworks,agenttechnology,case-basedreasoning,andsoon."
【D】Ironically,insomewaysAIwasavictimofitsownsuccess.Wheneveranapparentlymundaneproblemwassolved,suchasbuildingasystemthatcouldlandanaircraftunattended,theproblemwasdeemednottohavebeenAIinthefirstplace.‘Ifitworks,itcan’tbeAI,’asDrLeakecharacterisesit.Theeffectofrepeatedlymovingthegoal-postsinthiswaywasthatAIcametoreferto‘blue-sky’researchthatwasstillyearsawayfromcommercialisation.ResearchersjokedthatAIstoodfor‘almostimplemented’.Meanwhile,thetechnologiesthatmadeitontothemarket,suchasspeechrecognition,languagetranslationanddecision-supportsoftware,werenolongerregardedasAI.YetallthreeoncefellwellwithintheumbrellaofAIresearch.
【E】Butthetidemaynowbeturning,accordingtoDrLeake.HNCSoftwareofSanDiego,backedbyagovernmentagency,reckonthattheirnewapproachtoartificialintelligenceisthemostpowerfulandpromisingapproacheverdiscovered.HNCclaimthattheirsystem,basedonaclusterof30processors,couldbeusedtospotcamouflagedvehiclesonabattlefieldorextractavoicesignalfromanoisybackground—taskshumanscandowell,butcomputerscannot.‘Whetherornottheirtechnologylivesuptotheclaimsmadeforit,thefactthatHNCareemphasisingtheuseofAIisitselfaninterestingdevelopment,’saysDrLeake.
【F】AnotherfactorthatmayboosttheprospectsforAIinthenearfutureisthatinvestorsarenowlookingforfirmsusingclevertechnology,ratherthanjustacleverbusinessmodel,todifferentiatethemselves.Inparticular,theproblemofinformationoverload,exacerbatedbythegrowthofe-mailandtheexplosioninthenumberofwebpages,meansthereareplentyofopportunitiesfornewtechnologiestohelpfilterandcategoriseinformation—classicAIproblems.Thatmaymeanthatmoreartificialintelligencecompanieswillstarttoemergetomeetthischallenge.
【G】The1969film,2001:
ASpaceOdyssey,featuredanintelligentcomputercalledHAL9000.AswellasunderstandingandspeakingEnglish,HALcouldplaychessandevenlearnedtolipread.HALthusencapsulatedtheoptimismofthe1960sthatintelligentcomputerswouldbewidespreadby2001.But2001hasbeenandgone,andthereisstillnosignofaHAL-likecomputer.Individualsystemscanplaychessortranscribespeech,butageneraltheoryofmachineintelligencestillremainselusive.Itmaybe,however,thatthecomparisonwithHALnolongerseemsquitesoimportant,andAIcannowbejudgedbywhatitcando,ratherthanbyhowwellitmatchesuptoa30-year-oldscience-fictionfilm.‘Peoplearebeginningtorealisethatthereareimpressivethingsthatthesesystemscando.’saysDrLeakehopefully.
27howAImighthaveamilitaryimpact
28thefactthatAIbringstogetherarangeofseparateresearchareas
29thereasonwhyAIhasbecomeacommontopicofconversationagain
30howAIcouldhelpdealwithdifficultiesrelatedtotheamountofinformationavailableelectronically
31wheretheexpressionAIwasfirstused
12.What’ssofunny?
JohnMcCronereviewsrecentresearchonhumor
Inboxes14-20onyouranswersheet,write
TRUE
ifthestatementagreeswiththeinformation
FALSE
ifthestatementcontradictstheinformation
NOTGIVEN
ifthereisnoinformationonthis
Thejokecomesovertheheadphones:
‘Whichsideofadoghasthemosthair?
Theleft.’No,notfunny.Tryagain.‘Whichsideofadoghasthemosthair?
Theoutside.’Hah!
Thepunchlineissillyyetfitting,temptingasmile,evenalaugh.Laughterhasalwaysstruckpeopleasdeeplymysterious,perhapspointless.ThewriterArthurKoestlerdubbedittheluxuryreflex:
‘uniqueinthatitservesnoapparentbiologicalpurpose.
Theoriesabouthumourhaveanancientpedigree.Platoexpressedtheideathathumorissimplyadelightedfeelingofsuperiorityoverothers.KantandFreudfeltthatjoke-tellingreliesonbuildingupapsychictensionwhichissafelypuncturedbytheludicrousnessofthepunchline.ButmostmodernhumortheoristshavesettledonsomeversionofAristotle’sbeliefthatjokesarebasedonareactiontoorresolutionofincongruity,whenthepunchlineiseitheranonsenseor,thoughappearingsilly,hasacleversecondmeaning.
GraemeRitchie,acomputationallinguistinEdinburgh,studiesthelinguisticstructureofjokesinordertounderstandnotonlyhumorbutlanguageunderstandingandreasoninginmachines.Hesaysthatwhilethereisnosingleformatforjokes,manyrevolvearoundasuddenandsurprisingconceptualshift.Acomedianwillpresentasituationfollowedbyanunexpectedinterpretationthatisalsoapt.
Soevenifapunchlinesoundssilly,thelistenercanseethereisacleversemanticfitandthatsuddenmental‘Aha!
’isthebuzzthatmakesuslaugh.Viewedfromthisangle,humorisjustaformofcreativeinsight,asuddenleaptoanewperspective.
However,thereisanothertypeoflaughter,thelaughterofsocialappeasementanditisimportanttounderstandthistoo.Playisacrucialpartofdevelopmentinmostyoungmammals.Ratsproduceultrasonicsqueakstopreventtheirscufflesturningnasty.Chimpanzeeshavea‘play-face’—agapingexpressionaccompaniedbyapanting‘ahah’noise.Inhumans,thesesignalshavemutatedintosmilesandlaughs.Researchersbelievesocialsituations,ratherthancognitiveeventssuchasjokes,triggertheseinstinctualmarkersofplayorappeasement.Peoplelaughonfairgroundridesorwhentickledtoflagaplaysituation,whethertheyfeelamusedornot.
Bothsocialandcognitivetypesoflaughtertapintothesameexpressivemachineryinourbrains,theemotionandmotorcircuitsthatproducesmilesandexcitedvocalisations.However,ifcognitivelaughteristheproductofmoregeneralthoughtprocesses,itshouldresultfrommoreexpansivebrainactivity.
PsychologistVinodGoelinvestigatedhumourusingthenewtechniqueof‘singleevent’functionalmagneticresonanceimaging(fMRI).AnMRIscannerusesmagneticfieldsandradiowavestotrackthechangesinoxygenatedbloodthataccompanymentalactivity.Untilrecently,MRIscannersneededseveralminutesofactivityandsocouldnotbeusedtotrackrapidthoughtprocessessuchascomprehendingajoke.Newdevelopmentsnowallowhalf-second‘snapshots’ofallsortsofreasoningandproblem-solvingactivities.
AlthoughGoelfeltbeinginsideabrainscannerwashardlytheidealplaceforappreciatingajoke,hefoundevidencethatunderstandingajokeinvolvesawidespreadmentalshift.Hisscansshowedthatatthebeginningofajokethelistener’sprefrontalcortexlitup,particularlytherightprefrontalbelievedtobecriticalforproblemsolving.Buttherewasalsoactivityinthetemporallobesatthesideofthehead(consistentwithattemptstorousestoredknowledge)andinmanyotherbrainareas.Thenwhenthepunchlinearrived,anewareasprangtolife—theorbitalprefrontalcortex.Thispatchofbraintuckedbehindt