Jobs of Tomorrow: Large Language Models and Jobs

The report examines how large language models (LLMs) may automate or augment work tasks. It finds that routine, language-based tasks are most susceptible to automation, while complex problem-solving tasks in STEM fields are most likely to be augmented. The analysis highlights that industries like financial services and IT have the highest total exposure, but suggests that LLMs will transform the nature of work rather than simply reducing the total number of jobs.

• The analysis identifies a distinction between tasks with high automation potential, which tend to be routine and repetitive, and those with high augmentation potential, which require abstract reasoning and problem-solving skills.
• Specific roles identified as having the highest potential for automation include Credit Authorizers, Checkers and Clerks (81% of work time), Management Analysts (70%), Telemarketers (68%), Statistical Assistants (61%), and Tellers (60%).
• Jobs emphasizing mathematical and scientific analysis show the highest potential for augmentation, led by Insurance Underwriters (100% of work time), Bioengineers and Biomedical Engineers (84%), Mathematicians (80%), and Editors (72%).
• Occupations requiring high levels of personal interaction or physical movement are the least exposed to LLM transformation, such as Educational, Guidance, and Career Counsellors and Advisers (84% low exposure), Clergy (84%), and Home Health Aides (75%).
• The adoption of LLMs is expected to create new specialized roles, including AI Model and Prompt Engineers, Interface and Interaction Designers, AI Content Creators, Data Curators, and AI Ethics and Governance Specialists.
• At the industry level, financial services and capital markets, and insurance and pension management have the highest total potential exposure (automation plus augmentation), followed by information technology and digital communications, and media, entertainment and sports.
• Analysis by function group shows that information technology (73% of working hours exposed) and finance (70% of working hours exposed) are the two thematic areas with the greatest total potential exposure.
• There is a strong correlation between a job's potential for automation and its expected employment decline, and between its potential for augmentation and its expected growth over the next five years.

APA

World Economic Forum (2023). Jobs of Tomorrow: Large Language Models and Jobs. https://www3.weforum.org/docs/WEF_Jobs_of_Tomorrow_Generative_AI_2023.pdf

Chicago

World Economic Forum. Jobs of Tomorrow: Large Language Models and Jobs. 2023. https://www3.weforum.org/docs/WEF_Jobs_of_Tomorrow_Generative_AI_2023.pdf.

Wikipedia

{{cite report |author=World Economic Forum |title=Jobs of Tomorrow: Large Language Models and Jobs |date=September 2023 |url=https://www3.weforum.org/docs/WEF_Jobs_of_Tomorrow_Generative_AI_2023.pdf |access-date=23 October 2026 |via=Climate Insights Directory}}

BibTeX

@techreport{worldeconomicforum2023jobs, author = {{World Economic Forum}}, title = {{Jobs of Tomorrow: Large Language Models and Jobs}}, institution = {World Economic Forum}, year = {2023}, month = sep, url = {https://www3.weforum.org/docs/WEF_Jobs_of_Tomorrow_Generative_AI_2023.pdf}, urldate = {2026-10-23}, note = {Indexed by Climate Insights Directory} }