How Will Artificial Intelligence Affect Our Work?

Sep 23-2025   

*This article is based on the keynote speech by Zhang Dandan, Vice Dean of the National School of Development and Professor of Economics at Peking University.

On the one hand, our traditional job positions are undergoing changes, becoming increasingly informal, fragmented, and task-oriented. On the other hand, these phenomena reflect rapid technological advancements. So, in this era, what kind of situation will we face? Will our jobs still be needed? Do we still need to work?

How Will Large Language Models Affect the Chinese Labor Market?

Next, let's examine the impact of AI language models on the Chinese labor market. Below are the top 20 professions with the highest and lowest exposure to AI language models. A high level of exposure means that the work of this profession can be largely completed by AI language models, while a low level of exposure indicates that current technology cannot do the work of this profession.

Professions with high exposure to AI language models include finance-related positions such as accounting and auditing, as well as editors, sales engineers, and computer programmers—all of which rank at the top. Professions with low exposure include cleaners, porters, restaurant chefs, and dishwashers. Thus, our conclusion is that large AI language models have a greater impact on mental labor.

Our recent research has also focused on changes in skill demands. In the era of AI language models, the demand for some skills is decreasing, but this does not mean that these skills are no longer needed overall. These skills include communication skills, learning abilities, emotional intelligence, adaptability, and critical thinking abilities. The fastest decline in demand is seen in "communication skills" and "learning abilities." Large AI language models have strong learning capabilities and can assist users in achieving leaps in learning. In terms of emotional intelligence (also known as EQ), large AI language models also perform well.

On the other hand, with the emergence of AI language models, employers’ requirements for certain skills will increase. These skills include professional expertise, management skills, self-motivation, problem-solving abilities, and collaboration skills.

What Will Future Work Look Like?

From the perspective of technological progress, economists believe that technological progress has a "double-edged sword" effect on individual professions. The x-axis in the figure represents process substitution—a negative value that indicates a destructive effect, meaning technological progress will completely replace this type of job. The y-axis represents factor enhancement, which means technological progress will improve work efficiency and make one’s work more valuable to society.

▲ Autor D, Chin C, Salomons A M, et al. “New Frontiers: The Origins and Content of New Work, 1940–2018.” Quarterly Journal of Economics, 139(3), 2024.

This provides us with room to mitigate the impact of technological progress and exert our subjective initiative. If we adapt well, we can gain an advantage; if not, we may be at a disadvantage. Of course, this is also related to the speed of technological progress.

In addition, there may be two trends regarding the impact of future technological advancements on work: one is work polarization, and the other is work taskization. With technological advancements, the demand for highly skilled and low-skilled occupations will increase, while the demand for middle-skill occupations will decrease. On the other hand, the number of jobs that require no specialized skills will also increase. These are task-based jobs, falling on the low-skill end, but with significant demand.

▲ Daron Acemoglu, David Autor. Chapter 12 - Skills, Tasks and Technologies: Implications for Employment and Earnings. In: Handbook of Labor Economics, Elsevier, Volume 4B, 2011: 1043-1171.

However, all our previous discussions are based on two assumptions: "work is necessary" and "each of us needs to work to support ourselves." Suppose AI continues to develop to the point where it can create large quantities of scarce resources—then we will no longer need so much work. If AI can provide for our material needs, then work may no longer be necessary.