A new study released by OpenAI and the University of Pennsylvania titled "An Early Look at the Labor Market Impact Potential of Large Language Models" has explored the potential effects of GPT-4 on various professions. The study indicates that while some jobs might be safe from automation for now, the future is uncertain as technology continues to evolve rapidly. In this article, we will discuss the impact of GPT-4 on different professions, how education and income levels correlate with exposure to GPT-4, and the potential consequences for the future of work.

GPTs are GPTs: An Early Look at the Labor Market Impact Potential of Large Language Models
We investigate the potential implications of large language models (LLMs),such as Generative Pre-trained Transformers (GPTs), on the U.S. labor market,focusing on the increased capabilities arising from LLM-powered softwarecompared to LLMs on their own. Using a new rubric, we assess occupations b…

Understanding the Concept of Exposure to Large Language Models

The concept of "exposure" is critical when discussing the impact of GPT-4 on jobs. Exposure refers to how much a task can be impacted by large language models like GPT-4. The more exposed a task is, the quicker and easier it becomes to complete using AI tools. Exposure does not necessarily mean a job will be entirely automated, but rather that certain tasks within a job could be made more efficient with the help of AI.

The findings reveal that 80% of the workforce and 10% of work tasks are affected by the introduction of LLMs. Furthermore, 19% of workers have 50% of their tasks impacted. To assess the potential impact of LLMs on different occupations, researchers developed a rubric that measures the exposure of tasks to LLM capabilities. OpenAI employed human annotators and GPT-4 classifiers to apply this rubric to occupational data sourced primarily from the O*NET database.

The exposure measure reflects an estimate of the technical capacity for human labour to be replaced efficiently by LLMs. However, various social, economic, and regulatory factors imply that technical feasibility does not guarantee labour productivity or automation outcomes.

Major Exposure Based on Education and Income Levels

The study revealed that individuals with higher education levels, such as bachelor's, master's, and professional degrees, are more exposed to GPT-4 than those without formal education. This is because roles heavily reliant on science and critical thinking skills show a negative correlation with exposure. On the other hand, programming and writing skills are positively associated with GPT-4 exposure. This means that as education levels increase, so does the average exposure to GPT-4, which correlates directly with median income.

The Relationship Between Wage Levels and Exposure to Large Language Models

The projected effects of LLMs span across wage levels, with higher-income jobs facing greater exposure to LLM capabilities and LLM-powered software. The impacts are not restricted to industries with higher productivity growth. The analysis suggests that with access to LLMs, 15% of worker tasks could be completed faster and at the same level of quality. When incorporating software tooling built on LLMs, this increases to between 47% and 56% of tasks.

This finding implies that LLM-powered software has the potential to bring substantial scaling to the economic impacts of the underlying models.

Exposure to LLMs Increases with Wage Levels

The research has shown that higher-wage jobs tend to have greater exposure to LLM capabilities and LLM-powered software. This finding is contrary to previous evaluations of exposure to machine learning and automation, which generally suggested that lower-wage jobs were more vulnerable. There are several reasons why higher-wage jobs might experience greater exposure to LLMs:

Higher-wage jobs often require specialized skills: Many higher-wage jobs require specialized skills and knowledge, such as programming, data analysis, or content creation. LLMs excel in these areas, particularly in code generation and natural language processing, making them suitable for tasks that require these specialized skills.

Greater reliance on language and communication: Higher-wage jobs often involve more complex language processing and communication tasks. LLMs are designed to understand and generate human-like text, making them particularly relevant to jobs that require advanced language and communication skills.

Increased demand for efficiency and productivity: In higher-wage jobs, there is often a greater emphasis on efficiency and productivity. LLMs can automate or augment various tasks, streamlining workflow and reducing the time and effort required to complete them. This can make higher-wage jobs more attractive targets for LLM adoption.

Examples of Higher-Wage Jobs with Increased Exposure

Lawyers and legal professionals: LLMs can assist in drafting legal documents, conducting research, and analyzing case data, which can significantly impact the work of lawyers and legal professionals. These tasks typically require specialized knowledge and strong language skills, making them vulnerable to LLM capabilities.

Financial analysts and advisors: Financial analysts and advisors often rely on data analysis, forecasting, and communication with clients. LLMs can help in processing large amounts of data, generating financial reports, and providing insights, which can impact the job market for these professionals.

Marketing and advertising specialists: These professionals often work with content creation, data analysis, and communication tasks. LLMs can generate creative content, analyze consumer data, and help with campaign planning, potentially automating or augmenting portions of their work.

The Relationship Between Education Levels and Exposure to Large Language Models

The paper also suggests that exposure to LLMs increases with higher education levels.

Why Exposure Increases with Education Levels

Greater reliance on language and communication skills: Higher education levels typically correlate with jobs that demand more advanced language and communication skills. Since LLMs excel in natural language processing and generation, they have a greater potential to impact occupations that involve complex language and communication tasks, which are often associated with higher education levels.

Higher focus on programming and writing skills: Jobs requiring programming or writing skills are more exposed to LLMs, as these models have demonstrated impressive capabilities in code generation and natural language processing. Occupations that demand these skills are often found in industries that require higher education levels, such as technology, finance, and research.

More complex and specialized tasks: Jobs that require higher education levels often involve more complex and specialized tasks. While these tasks may not be routine, they can still be influenced by LLMs' advanced capabilities, especially when it comes to tasks related to data analysis, decision-making, and problem-solving. LLMs can provide valuable insights and aid in these tasks, increasing exposure for jobs with higher education requirements.

Implications of Increased Exposure to Education Levels

Shift in the nature of work: As exposure to LLMs increases with education levels, the nature of work in highly skilled jobs might change. Professionals in these roles may need to focus more on tasks that require creativity, critical thinking, and emotional intelligence, which are less susceptible to automation by LLMs.

Need for continuous learning and upskilling: In a labour market where exposure to LLMs increases with education levels, the importance of continuous learning and upskilling becomes paramount. Workers will need to adapt to the changing landscape by acquiring new skills and staying updated on technological advancements.

Collaboration between humans and AI: The increased exposure of highly educated workers to LLMs highlights the need for collaboration between humans and AI. Rather than replacing jobs, LLMs can augment human capabilities, enabling professionals to work more efficiently and effectively.

Impact of GPT-4 on Various Professions

The study found that approximately 80% of the US workforce could have at least 10% of their work affected by GPT-4, with 19% of workers potentially seeing at least 50% of their tasks impacted. Some professions that face the most disruption include interpreters, translators, creative writers, public relations specialists, authors, mathematicians, tax preparers, and blockchain engineers. While these professions have a high exposure to GPT-4, it is important to note that this does not necessarily mean their jobs will be fully automated. Instead, GPT-4 could help save workers in these areas a significant amount of time on their tasks.

Key Traits of Jobs Affected by Large Language Models

High reliance on language and communication: Jobs that involve substantial language processing, interpretation, and communication are more susceptible to the effects of LLMs. These models excel at understanding and generating human-like text, which can automate or augment various language-related tasks.

Routine and repetitive tasks: LLMs can automate tasks that follow a predictable pattern or require minimal variation. Jobs that involve routine and repetitive tasks are more likely to be affected by LLMs.

Programming and writing skills: Occupations that require programming or writing skills are more exposed to LLMs. These models have demonstrated impressive capabilities in code generation and natural language processing, making them suitable for tasks that involve writing or coding.

Examples of Jobs That Might Be Affected

Customer service representatives: LLMs can handle various customer inquiries and provide relevant information, potentially automating tasks that customer service representatives perform. This can streamline customer support processes and reduce the need for human intervention.

Content creators and copywriters: LLMs can generate high-quality written content, such as articles, blog posts, and social media updates. This can impact the demand for content creators and copywriters, as LLMs can produce a vast amount of content quickly and efficiently.

Translators and interpreters: LLMs have shown impressive capabilities in understanding and translating languages. This can affect the job market for translators and interpreters, as LLMs can potentially provide faster and more accurate translations.

Data entry clerks: Data entry tasks often involve routine and repetitive work, making them susceptible to automation by LLMs. These models can quickly process and input data, reducing the need for human data entry clerks.

Software developers: LLMs can generate code and help developers with programming tasks. These models can potentially automate parts of the software development process, which may impact the job market for developers.

Adapting to the Rise of AI in the Workplace

As AI continues to advance, it is crucial for individuals to adapt and embrace these new technologies. This means learning how to use AI tools to one's advantage and staying informed about the latest advancements. In addition, society as a whole needs to address ethical and safety concerns surrounding the use of AI in various industries. By doing so, we can ensure a smooth transition into the future of work.

As AI and large language models (LLMs) continue to permeate various industries, workers need to adapt and learn how to work alongside these advanced technologies. By developing new skills and focusing on areas where AI and LLMs can complement human expertise, workers can mitigate potential risks and maximize their value in the evolving labour market.

Embrace Lifelong Learning

To keep up with rapid technological advancements, it's crucial to invest in continuous learning and skill development. Stay updated with the latest trends in AI and LLMs and identify the skills most relevant to your profession. Online courses, webinars, and workshops can be excellent resources for learning about AI, LLMs, and their applications in your industry. By staying informed and proactive, you'll be better equipped to adapt to the changing landscape of your profession.

As AI and LLMs become integral to various tasks and processes, it's essential to develop relevant skills that can help you work effectively with these technologies. Some key skills to consider include data analysis, programming, and AI ethics. Additionally, being adept at Prompt Engineering can be invaluable, as it involves crafting effective prompts to communicate with AI models and generate desired outputs. Mastering this skill can help you harness the full potential of AI and LLMs in your day-to-day tasks.

Focus on Human-centric Skills

While AI and LLMs are incredibly powerful, there are certain areas where human expertise remains irreplaceable. Emphasize developing and honing human-centric skills such as empathy, creativity, critical thinking, and emotional intelligence. These skills can help you navigate complex interpersonal situations and solve problems that require a nuanced understanding of human emotions and behaviour, setting you apart from AI-driven solutions.

Learn to Collaborate with AI

AI and LLMs have immense potential to augment human capabilities when used collaboratively. Embrace the idea of working with AI and LLMs as partners rather than competitors. Learn how to leverage their strengths to enhance your productivity and efficiency, while also understanding their limitations and providing the necessary human touch when required. By fostering a collaborative mindset, you can create a symbiotic relationship that benefits both you and the technology.

Advocate for Responsible AI Use

As AI and LLMs become more prevalent, it's vital to ensure their responsible and ethical use in the workplace. Stay informed about the ethical implications of AI and LLMs and advocate for transparency, fairness, and accountability in their deployment. By promoting responsible AI use, you can help create a more equitable and inclusive work environment that harnesses the benefits of AI without compromising on human values and ethics.

Subscribe to PromptEngineering.org to keep up to date on AI

Challenges for Policymakers

The eventual trajectory of LLM development and application poses significant challenges for policymakers to predict and regulate. As general-purpose technologies, LLMs have the potential to emerge in a broad range of economically valuable use cases, including the creation of new types of work. Policymakers must be prepared to adapt to the rapidly evolving impact potential of LLMs in the labour market.


This study "An Early Look at the Labor Market Impact Potential of Large Language Models" highlights the pervasive and transformative effects GPT-4 and other LLMs may have on various professions. As the labour market continues to evolve, individuals need to adapt to the changing landscape by embracing lifelong learning, developing AI-related skills, focusing on human-centric skills, and learning to collaborate effectively with AI systems. Simultaneously, policymakers must grapple with the challenges presented by the rapid advancements in AI and LLM technologies, ensuring that they create appropriate regulatory frameworks to protect workers and foster responsible AI use.

The future of work will undoubtedly be shaped by AI and LLMs, but by staying informed, proactive, and adaptable, both workers and policymakers can navigate the uncertain terrain ahead and create a more equitable, inclusive, and innovative labour market. The key lies in harnessing the power of AI while preserving and enhancing the unique qualities and skills that make us human.

Share this post