文档介绍:该【世界经济论坛-未来的工作:大型语言模型和工作——一个商业工具包(英) 】是由【youyicheng】上传分享,文档一共【17】页,该文档可以免费在线阅读,需要了解更多关于【世界经济论坛-未来的工作:大型语言模型和工作——一个商业工具包(英) 】的内容,可以使用淘豆网的站内搜索功能,选择自己适合的文档,以下文字是截取该文章内的部分文字,如需要获得完整电子版,请下载此文档到您的设备,方便您编辑和打印。IncollaborationentureJobsofTomorrowLargeLanguageModelsandJobs–ABusinessToolkitWHITEPAPERDECEMBER2023Images:GettyImagesContentsForeword34ExecutivesummaryIntroduction51Consideration1:::,?ndings,interpretationsandconclusionsexpressedhereinarearesultofacollaborativeprocessfacilitatedandendorsedbytheWorldEconomicForumbutwhoseresultsdonotnecessarilyrepresenttheviewsoftheWorldEconomicForum,northeentiretyofitsMembers,Partnersorotherstakeholders.?,includingphotocopyingandrecording,–ABusinessToolkit2December2023JobsofTomorrowLargeLanguageModelsandJobs–ABusinessToolkitForewordMaryKateMorleyRyanElselotHasselaaranization/HumanPotentialManagingDirector,AccentureHeadofMission,Work,WagesandJobCreation,WorldEconomicForumGenerativearti?cialintelligence(AI)and,inparticular,largelanguagemodels(LLMs),underpinnedbyadvancementsinmachinelearningandnaturallanguageprocessing,representaparadigmshiftinhowweinteractwithinformationand,byextension,,generateinsightsfromlargeamountsofdata,translatelanguageswithnear-?ciencyofthesetechnologies,includingnewLLM-poweredhuman-machineinterfacessuchasintelligentagents,,-uponthepreviouseditionoftheJobsofTomorrowseries,JobsofTomorrow:LargeLanguageModelsandJobs,whichtookastructuredapproachtounderstandingthedirectimpactofLLMsonspeci?,enablingstakeholders–businessleaders,policy-makers,workersandthebroaderpublic–,?cantproductivitygainsandthecreationofnewtypesofjobs,thereisalsoariskthattheycoulddisplaceexistingroles,,,aswellasfortheentureteam,,aglobalcoalitionofministersandchiefexecutiveof?cersthatpromotesabetterfutureofwork,throughboostinglabourmarketforesight,drivingjobcreationandimprovingjobqualitywhileenablingjobtransitionsandwillserveasakeytoolfortheGoodWorkAlliance,aglobal,cross-,–ABusinessToolkit3ExecutivesummaryBusinesseswillneedtostrategicallynavigateconcernsoverjobdisplacementrisk,?cialintelligence(AI)technologies,suchasGitHub’sCopilot,MidjourneyandChatGPT,haveproducedAI’s?rsttruein?exionpointinpublicadoption,demonstratingthetechnology’,includingimage,video,reation,largelanguagemodels(LLMs)-to-,addressingconcernslikejobchangeandjobdisplacementrisk,,,providingpracticalstrategiestonavigatethechangingjoblandscapeinthreeprimaryareasofconsideration:1)jobchangeandjobdisplacementrisk,2)jobquality,and3):1)promotingworkerawareness,2)anizationalchange,and3)-%ofglobaljobsareexpectedtotransformwithinthenext?veyearsduetorapidtechnologicaladvances,,LLM’:LargeLanguageModelsandJobswhitepaper,publishedbytheWorldEconomicForumentureearlierin2023,foundthat40%ofworkinghourscouldbetransformed,affectingrolesfromspecialized?–ABusinessToolkit4IntroductionUpto40%ofworkinghourscouldbetransformedbyLLMs–%ofcurrentpoint,demonstratingthetechnology’,ChatGPT,OpenAI’strainedlanguagemodel,,ChatGPThadreachedonehundredmillionmonthlyactiveusers,makingitthefastest-growingconsumerjobstotransformwithinthenext?,,technologyadoptionisdrivingchangeinlabourmarkets:82%ofbusinessleadersexpectincreasedadoptionofnewtechnologiestodrivebusinesstransformation,while37%ofbusinessleadersanticipatethatnewjobcreator,and21%,possessingthecapabilitytogeneratenovelimages,videos,music,,largelanguagemodels(LLMs)andtheiruniquelanguage-,suchasthosepoweringChatGPT,,toincreaseunderstandingofhowLLMscouldimpactjobs,theWorldEconomicForum,incollaborationwithAccenture,publishedtheJobsofTomorrow:,speci?cally,generativearti?cialintelligence(AI)hasdevelopedatarapidpaceintermsofcapabilitiesandadoption,,threepaniesacrosstheglobeareexpectedtoadopttechnologiesthatincludegenerativeAIinthenext3to5years,and98%ofglobalexecutivesagreeAIfoundationmodelsanizations’’sCopilot,MidjourneyandChatGPTareacaseinJobsofTomorrowLargeLanguageModelsandJobs–ABusinessToolkit5Over40%ofworkingimpactofLLMsonjobsand?ndsthatitwillbothautomateandaugmentjobtasks,:LargeLanguageModelsandJobswhitepaper,itiscriticalthatbusinessesadoptaproactive,human-%oftotalworktimeacrossoccupationsinvolveslanguage-basedtasks–,000individualtasksacross867occupationswereassessedtounderstandthepotentialexposureofeachtasktoLLMadoption,classifyingthemastaskswithhighpotentialforautomation(thetaskcouldbeperformedbyLLMs,withouthumans),highpotentialforaugmentation(thetaskwillcontinuetobeperformedbyhumans,withLLMsincreasinghumanproductivity),lowpotentialforeither(humanswillcontinuetoperformthetaskwithnosigni?cantimpactfromLLMs)orunaffected(non-languagetasks).Thiswhitepaperisafollow-uptotheJobsofTomorrow:?rstisjobchangeandjobdisplacementrisk:asLLMsarecapableofperformingmanyofthelanguagetasksusedonthejob,andupto60%ofworktimeuseslanguagetasks,:thedeploymentofthesetechnologicaltoolscouldaffectnotonlythenumberofjobsbutalsothenatureoftheworkitself,potentiallyaffectingthemeaningfulnessofwork,thewell-:toadjusttoandembracetechnologicalchangerequirescontinuouslearning;%,-solvingskills,especiallythoseinscience,technology,engineeringandmathematics(STEM)?,puterprogrammers,plusrolesthatrequirehumanvalidation,suchasassessors,-to-municationandinterpersonalinteractionsareexpectedtobelessexposedtotheimpactsofLLMs,andthoseemphasizingnon-languagetaskswillbeevenlessexposedornotexposedInthefollowingsections,eachconsiderationis?rstde?:1)promotingworkerawareness,whichinvolvesinformingandeducatingemployeesaboutthepotentialimpactsofLLMsontheircurrentrolesandfuturejobprospects,empoweringthemtoproactivelyengagewiththeevolvingtechnologicallandscape,2)anizationalchange,inpaniesadapttheirstructures,processesandstrategiestoharnessthebene?tsofnewtechnologieswhilemitigatingtherisksassociatedwithsuchtransformations,and3)shiftingworkplacenormsandculture,toreshapetheunderlyingattitudes,,thepaperalsofoundthatnewjobs4couldemerge,forexample,AImodelandpromptengineers,interfaceandinteractiondesigners,AIcontentcreators,datacuratorsandtrainers,,pairedwithitsTABLE1SummaryofbusinessstrategiesineachoftheprimarythreeconcernsJobchangeandjobdisplacementriskJobqualitySkillingandlearningWorkerawarenessmunicateexpectations4EnsureinclusivedesignofAIandLLMapplications7RolloutbroadLLMliteracyprogrammesalongsideAI-speci?cjobsOrganizationalchange2Setupandmakeuseofinternaljobmarkets5ImplementtransparentgovernancestructuresforLLMdeployment8Applyaskill-anizationwideShiftingnormsandculture3Createincreasedworkforceagility6IncreaseawarenessofLLMbene?tsintheworkplace9OfferworkplacelearningopportunitiesJobsofTomorrowLargeLanguageModelsandJobs–,meaningthesegenerativeAItechnologies,speci?callyLLMs,couldleadtotheautomatio