GPTs are GPTs: An Early Look at the Labor Market Impact Potential of Large Language Models
Tyna Eloundou, Sam Manning, Pamela Mishkin, Daniel Rock · arXiv / OpenAI · 2023
Read the original paperPlain-English Summary
An analysis of which jobs and tasks are most exposed to automation by large language models. The study found that roughly 80% of the US workforce could have at least 10% of their tasks affected by LLMs, with higher-wage occupations facing greater exposure.
Why This Paper Matters
As AI systems become integrated into workplaces, understanding which jobs and tasks face the most disruption is essential for workers, policymakers, and educators. This paper provides one of the first systematic assessments.
Key Concepts
-
Task-level exposure: Rather than asking "will this job be automated?", the paper examines which specific tasks within each job can be augmented or replaced by LLMs.
-
Wage and exposure: Counter-intuitively, higher-wage knowledge workers face more task exposure than lower-wage physical laborers, inverting previous assumptions about automation primarily affecting blue-collar work.
-
Complementary tools: When including software tools built on top of LLMs, the exposure increases significantly, suggesting that the indirect effects of AI may be even larger than direct model use.
Discussion Questions
- How should governments prepare workers for AI-driven changes in the labor market?
- Does this research change your view on who benefits from and who is harmed by AI advancement?