We propose a framework for evaluating the potential impacts of large-language models (LLMs) and associated technologies on work by considering their relevance to the tasks workers perform in their jobs. By applying this framework (with both humans and using an LLM), we estimate that roughly 1.8% of jobs could have over half their tasks affected by LLMs with simple interfaces and general training. When accounting for current and likely future software developments that complement LLM capabilities, this share jumps to just over 46% of jobs. The collective attributes of LLMs such as generative pretrained transformers (GPTs) strongly suggest that they possess key characteristics of other “GPTs,” general-purpose technologies (1, 2). Our research highlights the need for robust societal evaluations and policy measures to address potential effects of LLMs and complementary technologies on labor markets.
