60humansupervisionandbeabletomakeatleastafewdecisionsforthemselvesgoalsthatposearealchallenge."Whileweknowhowtotellarobottohandleaspecificerror,"saysDaveLavery,managerofaroboticsprogramatNASA."wecan'tyetgivearobotenough'commonsense'toreliablyinteractwithadynamicworld."Indeedthequestfortrueartificialintelligencehasproducedverymixedresults.Despiteaspellofinitialoptimisminthe1960sand1970swhenitappearedthattransistorcircuitsandmicroprocessorsmightbeabletocopytheactionofthehumanbrainbytheyear2010,researcherslatelyhavebeguntoextendthatforecastbydecadesifnotcenturies.Whattheyfound,inattemptingtomodelthought,isthatthehumanbrain'sroughlyonehundredbillionnervecellsaremuchmoretalentedandhumanperceptionfarmorecomplicatedthanpreviouslyimagined.Theyhavebuiltrobotsthatcanrecognizetheerrorofamachinepanelbyafractionofamillimeterinacontrolledfactoryenvironment.Butthehumanmindcanglimpsearapidlychangingsceneandimmediatelydisregardthe98per-centthatisirrelevant,instantaneouslyfocusingonthemonkeyatthesideofawindingfor-estroadorthesinglesuspiciousfaceinabigcrowd.ThemostadvancedcomputersystemsonEarthcan'tapproachthatkindofability,andneuroscientistsstilldon'tknowquitehowwedoit.46.Humaningenuitywasinitiallydemonstratedin[A]theuseofmachinestoproducesciencefiction.[B]thewideuseofmachinesinmanufacturingindustry.[C]theinventionoftoolsfordifficultanddangerouswork.[D]theelite'scunningtacklingofdangerousandboringwork.47.Theword"gizmos"(line1,paragraph2)mostprobablymeans[A]programs.[B]experts.[C]devices.[D]creatures.48.Accordingtothetext,whatisbeyondman'sabilitynowistodesignarobotthatcan[A]fulfildelicatetaskslikeperformingbrainsurgery.[B]interactwithhumanbeingsverbally.[C]havealittlecommonsense.[D]respondindependentlytoachangingworld.49.Besidesreducinghumanlabor,robotscanalso[A]makeafewdecisionsforthemselves.www.kaoshidian.com考试点更多精彩视频http://www.kaoshidian.com61[B]dealwithsomeerrorswithhumanintervention[C]improvefactoryenvironments.[D]cultivatehumancreativity.50.Theauthorusestheexampleofa...
