ChatGPT on ChatGPT: An Exploratory Analysis of its Performance in the Public Sector Workplace

Abstract

This study explores the impact of Generative Artificial Intelligence (GenAI), in particular, ChatGPT, on the public sector workforce in the United States, focusing on task replacement, assistance potential, and the evolving landscape of skills. Utilizing GPT-4 to evaluate 1,022 core tasks across 51 public sector occupations, we provide an exploratory analysis of the roles susceptible to ChatGPT automation and those in which ChatGPT can augment human efforts. Our findings reveal that while 63% of tasks are resistant to ChatGPT replacement, primarily due to their requirement for physical presence, emotional intelligence, and complex decision-making, tasks that are routine, rule-based, and involving basic content generation show a high potential for automation. The study also identifies key skills that will remain vital, those likely to become obsolete, and new skills that will emerge as essential, highlighting the need for a strategic approach to workforce development in the face of AI advancements. In particular, our findings underscore the growing importance of skills in applying AI technologies and the ability to validate and interpret AI-generated content for humans to remain competitive. We offer insights into public-sector-specific impacts and propose a methodological framework for future research, emphasizing the importance of adapting educational curricula and policies to prepare for an AI-integrated future.