As advances in generative artiicial intelligence (Al)continue at an unprecedented pace,large languagemodels (LLMs)are emerging as transformative toolswith the potential to redefine the job landscape.Therecent advancements in these tools,ike GitHub’sCopilot,Micjourney and ChatGPT,are expectedto cause significant shifts in global economies andlabour markets.These particular technological
advancements coincide with a period of considerablelabour market upheaval from economic,geopolitical,green transition and technological forces.TheWorld Economic Forum’s Future of Jobs Report2023 predicts that 23%ofglobal jpbs wil changein the next five years due to industry transformation,induding through artificial intelligence and other text,image and voice processing technologies.
This white paper provides a structured analysis ofthe potential direct,near-term impacts of LLMs onjobs.With 62%of total work time invohing language-based tasks,’the widespread adoption of LLMs,such as ChatGPT,could significantly impact a broadspectrum of job roles.
To assess the impact of LLMs on jobs,this paperprovides an analysis of over 19,000 individualtasks across 867 oocupations,assessing thepotential exposure of each task to LLM adoption,classifying them as tasks that have high potentialfor automation,high potential for augmentation,lowpotential for either or are unaffected (non-languagetasks).The paper also provides an overview of newroles that are emerging due to the adoption of LLMs.
The longer-term impacts of these technolbgiesin reshaping industries and business models arebeyond the scope of this paper,but the structuredapproach proposed here can be applied to otherareas of technological change and their impact ontasks and jobs.
The analysis reveals that tasks with the highest
potential for automation by LLMs tend to be routineand repetifive,while those with the highest potentialfor augmentation require abstract reasoning andproblem-sohing skills.Tasks with lower potentialfor exposure require a high degree of personalinteraction and collaboration.
-The jpbs ranking highest for potential automationare Credit Authorizers,Checkers and Clerks (81%of work time could be automated),ManagementAnalysts (70%),Telemarketers (68%),StatisticalAssistants (61%),and Tellers (60%).Jobs with the highest potential for taskaugmentation emphasize mathematicaland scientific analysis,such as hsuranceUnderwriters (100%of work time potentiallyaugmented),Bioengireers and BiomedicalEngineers (84%),Mathematicians (80%),andEditors (72%).Jobs with lower potential for automation or
augmentation are jobs that are expected to
remain largely unchanged,such as Educational,Guidance,and Career Counsellors and Advisers(84%of time spent on bw exposure tasks),Clergy (84%),Paralegals and Legal Assistants(83%),and Home Health Aides (75%).
In addiion to reshaping existing jobs,
the adoption of LLMs is likely to create newroles within the categories of Al Developers,Interface and Interaction Designers,Al ContentCreators,Data Curators,and Al Ethics andGovernance Specialists.
An industry analysis is done by aggregatingpotential exposure levels of jobs to the industrylevel,noting that jpbs may exist in more thanone industry.Resuts reveal that the industrieswith the highest estimates of total potentialexposure (automation plus augmentationmeasures)are both segments of financial
services:financial services and capital marketsand insurance and pension management.Thisis followed by information technology and digitalcommunications,and then media,entertainmentand sports.Additional lists of jpbs ranked byhighest exposure potential for each majorindustry category are compiled in the appendix.
Similarhy,a function group analysis revealsthat the two thematic areas with the greatesttotal potential exposure to LLMs are informationtechnology,with 73%of working hours
exposed,and finance,with 70%of workinghours exposed.As with the industry groups,additional lists of jobs ranked by highestexposure potential for each function group
are compiled in the Appendices.
These new findings connect directly to earlier
work done by the Centre for the New Economy
and Society in the Future of Jobs Report 2023.Many of the jpbs found to have high potentialfor automation by LLMs were also expectedby business leaders to undergo employmentdecline within the next five years,such as banktellers and related clerks,data entry derks,and administrative and executive secretaries.Meanwhile,jpbs with high potential for
augmentation are expected to grow,such asAl and Machine Learning Specialists,DataAnalysts and Sdientists,and Database andNetwork Professionals.Together,these twopublications identify and reaffirm salient themesin the connection between technological changeand labour market transformation.
本文来自知之小站
PDF报告已分享至知识星球，微信扫码加入立享3万+精选资料，年更新1万+精选报告
（星球内含更多专属精选报告.其它事宜可联系zzxz_88@163.com）
