欢迎来到课桌文档! | 帮助中心 课桌文档-建筑工程资料库
课桌文档
全部分类
  • 党建之窗>
  • 感悟体会>
  • 百家争鸣>
  • 教育整顿>
  • 文笔提升>
  • 热门分类>
  • 计划总结>
  • 致辞演讲>
  • 在线阅读>
  • ImageVerifierCode 换一换
    首页 课桌文档 > 资源分类 > DOCX文档下载  

    生成式人工智能如何改变商业和社会-奥纬咨询(英).docx

    • 资源ID:1381091       资源大小:736.02KB        全文页数:102页
    • 资源格式: DOCX        下载积分:5金币
    快捷下载 游客一键下载
    会员登录下载
    三方登录下载: 微信开放平台登录 QQ登录  
    下载资源需要5金币
    邮箱/手机:
    温馨提示:
    用户名和密码都是您填写的邮箱或者手机号,方便查询和重复下载(系统自动生成)
    支付方式: 支付宝    微信支付   
    验证码:   换一换

    加入VIP免费专享
     
    账号:
    密码:
    验证码:   换一换
      忘记密码?
        
    友情提示
    2、PDF文件下载后,可能会被浏览器默认打开,此种情况可以点击浏览器菜单,保存网页到桌面,就可以正常下载了。
    3、本站不支持迅雷下载,请使用电脑自带的IE浏览器,或者360浏览器、谷歌浏览器下载即可。
    4、本站资源下载后的文档和图纸-无水印,预览文档经过压缩,下载后原文更清晰。
    5、试题试卷类文档,如果标题没有明确说明有答案则都视为没有答案,请知晓。

    生成式人工智能如何改变商业和社会-奥纬咨询(英).docx

    ExecutiveSummaryWillgenerativeartificialintelligenceusherinagoldenageofproductivityordestroymillionsoflivelihoodsacrosstheglobaleconomy?Willitsetpeopleontonewpathsofpersonalfulfillmentorleadthemintocul-de-sacsoflonelinessandisolation?Willitlifthumanitytonewheightsorsowtheseedsofourcollectivedestruction?Dependingonwhomyouask,theanswertoallthosequestionsisyes.Inthe14monthssincethelaunchofChatGPT,ithasn'tbecomeclearexactlyhowgenerativeAIwilltransformtheworld,onlythatitwillwithbothpositiveandnegativeeffectsonindividuals,families,socioeconomicgroups,businesses,industries,andsocietiesasitrevolutionizesworkplacesandreorderspersonallives.Manybreakthroughtechnologiescomewithpitfalls.Fireallowedhumanstogatheratnightbutalsobumedvillages.Theautomobilerevolutionizedmobilitybutbroughttrafficfatalities.Theinternetconnectedpeopleinstantlybutgavenewtoolstocriminals.AkeydifferencebetweengenerativeAlandearlierinnovationsisthatitsverycreatorsarewarningofthepotentialdownsides.ThedualstrandsofpromiseandperilarewoventhroughoutAIcompaniesthemselves;looknofurtherthanthebattleforcontrolofOpenAIforanexampleofthedeepambivalencethatgenerativeAIisproducing.Butthereisopportunityintheambiguity.ThestartinggunofthegenerativeAIracewasfiredalongtimeago,butChatGPTbroughtarushofnewcompaniesandcountriesintotherace.Businessandgovernmentleaderswilldecidehowmuchofthedevelopmentwillbeopen-sourcedandtransparentversusclosed-sourcedandproprietary.Regulatorswilldecidethepaceandbreadthoftheaction.Consumersandworkerswillbecentralinthetechnology'sadoptionandwillhelpdete11ninehowquicklythebenefitsarecaptured.Withthatinmind,theOliverWymanForumsetouttothoroughlyexaminetheattitudes,perceptions,andmisperceptionssurroundinggenerativeAI.InJuneandNovember,wesurveyedmorethan25,000peopleacrosstheUnitedStates,theUnitedKingdom,Canada,Mexico,Brazil,France,Italy,Germany,Spain,China(HongKong),India,Indonesia,Singapore,theUnitedArabEmirates,andAustralia.Thefindingshighlighttheconfusionmanypeoplefeel.While96%ofemployeessaidtheybelieveAIcanhelpthemintheircurrentjob,60%areafraiditwilleventuallyautomatethemoutofwork.Some55%ofemployeesusegenerativeAIatleastonceaweekatwork,but61%ofusersdonotfinditverytrustworthy.Ofthose61%,40%wouldneverthelessuseittohelpthemmakebigfinancialdecisions,and30%wouldsharemorepersonaldataforabetterexperience.Amidthecacophonyofopinionsandpredictions,thereisagrowingconsensusaroundtheworldthatgenerativeAsinfluenceonboththeworkplaceandtheconsumereconomywillbemassive.ThenewAI-conomyWeestimatethatgenerativeAlcouldaddupto$20trilliontoglobalGDPby2030andsave300billionworkhoursayear.Thepossibilitiesareextraordinary.Fully96%oftheworkerswesun,eyedsaidtheybelievegenerativeAIcanhelpthemintheirjobs.ButasgenerativeAIreshapestheworkplace,itcouldplacenewstressesonorganizationalstructures.Inall,thebroadcategoryofAIcoulddisplace85millionjobsgloballyby2025,accordingtoanestimatebytheWorldEconomicForum.One-thirdofallentry-levelrolescouldbeautomated;atthesametime,junioremployeesarmedwithgenerativeAImaypotentiallyreplacetheirfirst-linemanagers,leavingavacuuminthemiddleofthejobpyramid.Asaresult,manyworkersaregrowingmoreanxiousbytheday.Automationisnolongeranarrativeofblue-collarworkersversusrobots.Threeinfivewhite-collarworkersnowfeartheirroleswillbecomeredundantorautomatedasgenerativeAsabilitiesincreasinglyimpactknowledgejobs.1.efttofester,thisanxiousnesscouldsapmorale:AccordingtotheAmericanPsychologicalAssociation,employeesintheUnitedStateswhoareconcernedaboutAIare68%morelikelytofeeltenseorstressedoutduringtheworkdaythanthosewhoaren'tworried,andaremorethantwiceaslikelytobelievetheydon,tmattertotheirworkcommunity.Suchfeelings,inturn,couldleadtodecreasingengagementandproductivity,andhigherturnover.ToreapthebenefitsgenerativeAlcanbring,companiesshouldembraceapeople-firstapproach,investinginworkersasmuchas,ifnotmorethan,thetechnology.Employeeswillneedtrainingandsupporttocreatesensibleandintuitiveprocessesalongsidethistechnology.Afterall,theyarethesameoneswhowillusetheinterfaces,updatethesystems,andmanagetheoutputs.Businessleadersneedtobringthemalongbylisteningandaddressingtheirconcerns,andup/reskillingtheiremployeesalongtheway.RadicalchangeforconsumersThegenerativeAIpicturebecomesabitclearerwhenweexamineitsimpactonconsumersanditswidespreadinfluenceacrossvariousindustries.Healthcareappearstobeespeciallyripefordisruption.Weestimatethatby2030,generativeAIcouldsavedoctorsthreehoursaday,whichwouldallowthemtoserveanadditional500millionpatientsgloballyperyear(assumingtheproductivitygainsgotoservingmorepatients).1.ikewise,weestimatethatupto4(X)millionpatientswillreceivementalhealthsupportasaresultofgenerativeAItherapyservices,democratizingaccessandsparkingglobalinterest.ConsumersareespeciallyeagertoexperiencetherapyforthefirsttimewithgenerativeAl:ofthe77%ofrespondentswhohaveneverdonetherapywithahuman,nearlyoneinthreereporttheywouldtrygenerativeAItherapyinthefuture.GenerativeAIisalsoemergingasasourceoftrustedadvice.NearlyhalfofconsumerssaidtheywouldplacetheirfaithinAIforbiglifedecisionssuchasbuyingahome.Some36%saidtheypreferAIoverhumansforfinancialadvice,demonstratingthebreadthofservicegenerativeAIcanproduce,evenwhensolvingfortraditionallyhumanneeds.Consumersare23%morelikelytoreportthattheyengagewithgenerativeAIfinancialadviserswiththeaimoffindingconnectionthantheyarewhentheyengagewithhumanones;theyarealsonearlyfourtimesaslikelytobelookingforasenseofpurposeintheirpersonalfinanceswhenengagingwithAIadvisersthanwhentheyengagewiththeirhumancounterparts.Educationisanotherarealikelytobeaffected.Ofthe260millionschl-agedchildrenworldwidewhodonotattendschool,weestimatethatupto100millioncouldgainaccesstoeducationthroughgenerativeAIby2030duetogenerativeAspowertoprovideuniversalaccesstoindividualizedtutoring.AppslikeHelloHistory,forexample,allowstudentstoengageinphilosophicaldebateswithAristotleorlearnabouttheintricaciesofevolutionarybiologyfromCharlesDarwin.GenerativeAIonedaycouldevenchangethewaypeopleformrelationships,forbetterorworse.Some14%ofconsumerssaidtheypreferinteractingwithAIbecausetheybelievegenerativeAIcanbemoreemotionallyintelligentthanhumans.AppslikeReplikaareredefiningcompanionship,allowinguserstocraftvirtualfriendswhoarealwaystheretolisten.OneinfiveconsumerssaidtheywouldgoonavirtualdatewithanAIpersona.AndalmostoneinthreeconsumerssaidtheywouldpurchaseanAi-poweredpetcollarthattranslatesanimalsoundsintolanguage,potentiallystrengtheningthebondevenfurtherbetweenpelsandtheirowners.RiskversusrewardisthewrongdiscussionGenerativeAIpresentsincreasinglycomplexrisksfororganizationstomanage.Manyarealreadywell-understood:thepotentialforhallucination,thevagariesofblack-boxlogicsystems,opportunitiesforcyberattacks,databreaches,improperlytrainedmodels,outputcontrol,copyrightconcerns,andonandon.Othershaven,tbeenimaginedyet;justashumansaredevelopinggenerativeAIsystems,sotoowillthesystemschangehumans,withpotentiallyunforeseenramifications.Companiesandsocietiesmustsetasidethequestionofriskorrewardandacceptafutureofriskandrewardbuiltonadynamicmodeloftest,measure,andlearn.Theattitudesandbeliefsbeingformednowamongemployersandemployees,consumersandgovernmentswillfeedbackintothemodelsandhelpshapethisfuture.Inthepagesthatfollow,wedelvedeeplyintothesurveydata,pointingoutthesurprisesandparadoxesamongtheresponses,andofferourbestviewofwheregenerativeAIisheadedandhowitwillaffectbusinesses,employees,andconsumersaroundtheworld.TableofContentsExecutiveSummary2ExploringtheDualitiesofGenerativeAI8Reward/Risk9GenerativeAIischangingeverything,everywhere,allatonce9HowGenerativeAIIsChangingtheNatureofWork15PotentiaIZPitfalls18Thepotential:Agoldenageofproductivity18GenerativeAsswiftmarchacrosstheglobalworkforce18Massadoptionmassproductivity26Sohowdowedrivemoreefficientadoption?29Catalyst/Collapse32Reitnagininganewjobpyramid32Thegreatreskillinganemployerandemployeechallenge38Preparingtodayfortheworkforceoftomorrow41MetamorPhOSiS/Meltdown42Chartingahumaneworkforcejourney42Employeroptimismisloud,employeeconcernislouder43FivekeyprinciplestoleadingwithheartthroughthegenerativeAItransformation49ExploringTomorrow'sGenerativeAI-PoweredConsumerUniverse52TheemergingAl-powereduniverse53Keytakeawaysforbusinessleaders55ExcitemenvExistentialism56TheearlydaysofagenerativeAI-poweredfuture56GenerativeAIasthegreatequalizer58Necessity,themotherofinvention60ElevateZErode62IllustratinggenerativeAswide-rangingcapabilitiesacrossindustries62Retail:GenerativeAIasadviceanddecisionsupport64Media:GenerativeAIascontentcreationandcustomization65Healthcare:GenerativeAIasadviceanddecisionsupportandinformationgatheringandsynthesis66Education:GenerativeAIas(personalized)informationgatheringandsynthesis68Adaptation/Obsolescence70ExploringAs“human”essence71ConsumersareheadoverheelsforgenerativeAI71CanAIteachUStobemorehuman?72TheEvolvingRisks81SpeedZSafety84WhatweknowaboutAIrisk84ImpactsandregulatorychallengesofgenerativeAI86ExploringthebroaderimplicationsofgenerativeAIrisks88Riskmitigations90Unknownrisks,beyondgenerativeAI9396MethodologyIntroductionRISKANDREWARD/ExploringthedualitiesofgenerativeAlRewardXS!HRevolution,notevolutionInbiology,thepunctuatedequilibriummodelpositsthatevolutionoccursinsuddenburstsofexplosiveprogressinbetweenlongperiodsoflittletonochange.Thereisaparallelintechnology,whereincrementalprogressisoccasionallyinterruptedbyaninnovationsomassivelydisruptivethatittransformssociety.Generativeartificialintelligence(AI)isonesuchadvancement,reshapingbothindustriesandlives.Thequestionishow.GenerativeAIischangingeverything,everywhere,allatonceArtificialintelligencehasexistedfordecades,butitspopularitysoaredwiththelaunchofChatGPTinlate2022.Ilquicklybeganreshapingindustries,poweringdailytasks,andenhancingproductsthatdefinemodemliving.WhatseemslikemagicareactuallydeeplearningalgorithmsthatenablegenerativeAItocontinuouslyself-oplimizewithouthumaninten,ention.Forinstance,DeepMincTsprotein-foldingAIsurprisedexpertsbyteachingitselfphysicsandchemistryprinciples,allowingittomakeproteinstructurepredictionsmuchbetterthanhumanexpertscould.Asthesenetworksadvance,theonlyquestionswillbehowfar,andhowfast,generativeAIcanexpanditsdominion.GenerativeIsuppliestoolsandplatformsforinnovationtoflourishGenerativeAldefiesone-size-fits-allinterpretationswithitssheerbreadthofapplications,fromhealthcareandfinancetoeducationandart.Intransportation,therecouldonedaybeagenerativeAl-poweredurbanspherethatdynamicallyadjustssignalsandroutes,improvingsafetyandeliminatingtrafficjams.Ineducation,generativeAIcoulddemocratizelearningwithlessonsindividualizedtoastudent'suniquestyle,needs,andaspirations.OursurveyrespondentsareeagertoseeAl-acceleratedhealthcareadvancementsaswell,perhapsenvisioningafutureinwhichpersonalizeddiagnosesandtreatmentsareprovidedWillgenerativeAlbeaforcefor:GOODBADEVERYTHINGINBETWEEN$20trillion50%13%bstinglobaleconomyby2030,equivalentto20%oftheglobalGDPtodayofAlexpertsthinkthereisahighchance'humanswillgoextinctduetoAlofpeoplesurveyedexpressinterestindevelopingaromanticbondwithgenerativeAl300billion40%28%workhourssavedgloballyeachyear,equivalenttoanaverageofroughlytwohoursperpersonweeklyofrespondentsbelievegenerativeAlwillworseninequalitiesofpeoplesurveyedbelievegenerativeAlcancapturethedepthofhumanemotion500million84%56%additionalpatientstreatedbydoctorsduetotimesavingsfrommaturestategenerativeAlby2030ofemployeesreportexposingtheircompany'sproprietarydatainthepastthreemonthsofrespondentscouldn'ttellAl-generatedfakeimagesfromreal*>10%chance(nsidered"high"becauseitpertainstoexistentialrisks,whichhavepotentiallycatastrophicconsequences)Source:OliverWymanForum,AlImpacts2022ExpertSurveyonProgressinAlwithnearperfectprecision.Itisalreadybeingusedacrossprofessionsthatoncereliedonlyonpeopleratherthanmachines.Writers,influencers,andartistsnowcantapgenerativeAItohelpcrafttheirwork.Butitalsohasresultedinplagiarism,errors,andjobloss.Here,thedualityofgenerativeAIisinfulleffect:ThetechnologypromisestoeliminatemillionsofhoursofmundaneandrepetitivetasksWheredowegofromhere?GenerativeAsarrivalpresentshumankindwithincredibleopportunities(andalsoquestionsthatfrequentlyleadtosomeanswersandmorequestions).Inthechaptersthatfollow,weexploregenerativeAstransformativeimpactontheworkforceandconsumerspheres,thepotentialopportunities,andhiddenrisks.butalsothreatenstoeliminatetheveryauthorswhobenefitfromitsproductivity.PeopleholdhighhopesforAspotentialtoexpeditehumanadvancementacrossindustries%respondentswhothinkAlwillhelpimprovethefollowinginthenext30yearsQuestion:"WhichofthefollowingareasdoyouthinkAlwillhelpimprovemostinthenext30years?"Source:OliverWymanForumGenerativeAlSurvey,October-November2023,16untries,N=16,033TTRealorAl?Mostcan,ttellWeaskedrespondentswhethertheythinkthefollowingimagesarephotographsofreal-lifeObjectsZsceneriesorcreatedbygenerativeAlScififactFactGenerativeAIcanmakeyoumoneyovernight56%,69%,74%.JWWSource:OliverWymanForumGenerativeAlSurvey,October-November2023,16ntries,N=16,033Sci-Fiorfact:Whatisandisn,tgenerativeAl?GenerativeAICBiisaveyourlifeSource:OliverWymanForumanalysisWheredoesgenerativeAIthinkitstacksuprelativetootherinnovativetechnologies?Thetechnologyisnothingifnotbold.Werecentlyaskedthebest-knownapplication,ChatGPT,toevaluategenerativeAspositionrelativetothemostinfluentialtechnologiesinhistory.ItrankeditselfNo.25,betweenrailroadsand5G.Timewilltellifthisisbravadoormodesty.EquallyinterestingishowChatGPTpositionedotherinventions,declaringe-commercesupremeoverairconditioningandcloudcomputingabovetheairplane.Rankingeachofthesetechnologicalmarvelsisatrickquestionaseachhasleftanundeniablemarkontheworld.Nonetheless,it,sanintriguingexercise:Aretheserankingsareflectionofthepast,apredictionofthefuture,oramirrorofgenerativeAsownaspirations?Prompt:Considerthetop50mostinfluentialtechnologiesinhumanhistorybasedonfactorsincludingbutnotlimitedtotheinnovativecontribution,economicimpact,andsocialandculturalimpactithashad/willcontinuetohaveonhumanity.Assigneachtechnologyascorefrom1-100,where1isnegligibleimpactand100isthemostinfluentialofallhistory.Note:Impactonhumanityscore0-100Source:ChatGPT4.0AIcangiveusourwildestdreamsandworstnightmares.Whatapivotalpointforhumanitytodecidewhattodowiththisandhowtotreateachother.Biotechnologist,US(Man,age58)HowgenerativeAIischangingthenatureofworkGenerativeAIwilllikelytransformworkplacesacrosstheglobaleconomy.Itiselicitingarangeofreactions-fromoptimismaboutunparalleledproductivitygainsandexpandedcreativitytofearsaboutjoblossandpeople,ssenseofpurpose.Acrossindustries,generativeAIalreadyisaugmentinghumancapabilities.1.egalfirmsuseittosynthesizecaselaw;marketersleverageitfortargetedcampaigns;scientistsharnessittoacceleratediscoveries.However,massadoptionofgenerativeAIdoesnotau

    注意事项

    本文(生成式人工智能如何改变商业和社会-奥纬咨询(英).docx)为本站会员(夺命阿水)主动上传,课桌文档仅提供信息存储空间,仅对用户上传内容的表现方式做保护处理,对上载内容本身不做任何修改或编辑。 若此文所含内容侵犯了您的版权或隐私,请立即通知课桌文档(点击联系客服),我们立即给予删除!

    温馨提示:如果因为网速或其他原因下载失败请重新下载,重复下载不扣分。




    备案号:宁ICP备20000045号-1

    经营许可证:宁B2-20210002

    宁公网安备 64010402000986号

    课桌文档
    收起
    展开