If we look at pages of history, we will soon notice that every new technological revolution has led to the displacement of old positions and the birth of new ones. The automation of certain job tasks by new technologies is immediate, yet acquiring the new skills necessary to adapt to these technologies takes time. This lag between the two creates a decoupling effect between labor demand and supply, inevitably causing structural unemployment in the short term, termed "technological unemployment". Artificial intelligence (AI) related technologies can be seen as the new wave of technological revolution. Unlike previous technological revolutions, automation and information technology often affect repetitive, routine job tasks. However, the new generation of AI technologies, combined with Big Data and machine learning, already has the potential to impact many job tasks previously thought to require human judgment. AI-related technologies such as intelligent robots, autonomous driving, the Internet of Things, drones, and others are profoundly influencing people's lives and modes of production, gradually deepening their impact on society and the labor market.
The latest IMF research report shows that nearly 40% of jobs globally will be affected by AI. Understandably, due to differences in infrastructure, innovation, economic integration, and labor quality, the pace of AI's development varies among economies, resulting in differences in its impact on local labor markets. The IMF estimates that in developed economies with a good environment for developing AI, about 60% of jobs may be affected. Among the affected jobs, about half may benefit from the integration of AI, thereby increasing productivity. As for the other half, AI applications may replace humans in performing some key tasks, reducing the demand for labor in this part, leading to wage decreases and reductions in jobs. In extreme cases, some positions may disappear. For emerging markets and low-income countries, the percentages of jobs affected by AI are 40% and 26%, respectively. Meanwhile, AI brings less direct disruption to emerging market countries and developing economies.
In China, technological innovation has always received strong support from national strategy and government decisions. At the same time, based on the vast consumer market and huge commercial potential, AI-related technologies have rich application scenarios domestically. For instance, many cities in China have begun to deploy smart connected vehicles and commercial unmanned driving services. Currently, Baidu's unmanned taxis, known as "robo-taxis", have been deployed in Wuhan with more than 300 vehicles, and 70% of the completed orders in April were fully unmanned. It is expected that the scale of the operating fleet will reach 1,000 vehicles by the end of the year. Under the strong advocacy of the country for the "low-altitude economy", industries such as electric vertical take-off and landing aircraft (eVTOL), civilian drones, and smart low-altitude new infrastructure are developing rapidly. Data shows that in 2023, the scale of the "low-altitude economy" represented by these industries reached RMB 505.95 billion, an increase of 33.8% year-on-year. In Hangzhou, Alibaba's "City Brain" urban management project utilizes AI technology to optimize traffic management, greatly reducing traffic congestion and accidents through real-time monitoring and data analysis. Companies like JD and Tencent also use AI technology to optimize agricultural production, including collecting data using drones and sensors, as well as using machine learning technology to improve crop yield and quality. Some rural areas even use AI-controlled drones and robots to complete all production activities from sowing to harvesting.
China has comparative advantages in the development of AI-related technologies such as infrastructure, market size, and abundant data. However, some conditions of the country also make its labor market more vulnerable in the face of such a development.
Compared to developed economies, China has greater potential for automation in sectors such as manufacturing and agriculture. As the country's growth model shifts from export and industry-driven to consumption and service-driven, the space for job automation will further expand. According to Frey and Osborne (2013), the substitution rate of employment by AI in the U.S. was 47%. In 2016, they extended their analysis framework to include developing countries, including China, and the results showed that as many as 77% of jobs in China are at risk of automation, higher than all countries except Ethiopia. A study by the OECD in 2016, based on PIAAC survey data, estimated a job replacement rate of around 10% in the U.S. In December 2018, PwC, using the same data and methods, concluded that China faces a high risk of automation in the next two decades, with a job replacement rate of 39%. These analyses all point to one conclusion, i.e. the technological advancements pose a significant impact on China's labor market.
More importantly, under the guidance of national strategy, AI technology is rapidly being deployed across the country. However, the talent development and employment training system related to it have not kept pace, further exacerbating the mismatch between new technologies and skills, which may have a significant impact on the labor market. According to McKinsey's 2022 research report on the global development of AI, there is still considerable room for improvement in the utilization of the technology by Chinese enterprises compared to the world's leading level. From 2017 to 2022, China's AI usage rate was 41%, lower than the average of 50% and the U.S. level of 60%. Furthermore, AI has not yet helped Chinese enterprises achieve widespread revenue growth and profit contribution. This is mainly due to the fact that Chinese enterprises still have a long way to go in integrating the technology with their businesses. Apart from the immature overall development strategy, the lack of talent is more crucial. These talents should not only possess general management skills but also have profound business knowledge in both AI and business fields. The shortage of such talents is related to both the education system and the training system.
Therefore, for Chinese policymakers, it is necessary to be proactive and maximize the productivity improvements and income effects brought about by technological innovation, while minimizing the substitution effects of AI on the job market.
Economist Daron Acemoglu has previously pointed out that tasks previously performed by labor have been replaced by "automation" created by machines and AI. This substitution effect has the effect of reducing labor demand and wages. While replacing physical work tasks, AI also increases the automation of intellectual work. A report released by the McKinsey Global Institute this year analyzed the tasks of hundreds of occupations to assess the automation potential of these jobs using generative AI. The report estimates that 75% of the value of generative AI is concentrated in customer service, marketing, software engineering, and R&D. According to PwC's estimation in December 2019, in the next 20 years, 36% of industrial positions, 27% in agriculture, 25% of construction positions, and 21% of service positions in China will be replaced by AI-related technologies. For labor with different skill levels, technology will increase the demand for high-skilled labor and replace medium-skilled labor.
However, this development will also bring about productivity effects, also known as income effects, which will offset the substitution effect of AI technology on labor. On the one hand, through deep integration with various industries, AI will also create numerous new employment opportunities by promoting intelligent upgrades on a large scale. However, these employment opportunities require a relatively high level of skills from the workforce. On the other hand, perhaps more importantly, AI improves production efficiency and reduces labor costs, which in turn will lower production costs. Companies will pass on the costs to consumers through pricing, leading to an increase in consumers' actual income levels, thus further promoting consumption and increasing corporate profits, expanding the scale of enterprises, and increasing employment of labor. This virtuous cycle may be overlooked by "new technology pessimists" because income effects are often lagging behind. However, in the long run, the role of new technology in economic growth is enormous, and there will be a positive effect on the labor market. According to PwC's estimation in December 2019, the projected income effects of AI on employment in the construction, industrial, agricultural, and service industries in the next 20 years are 48%, 39%, 15%, and 50%, respectively.
All in all, AI-related technologies, which lead to the new wave of technological revolution, are a double-edged sword for employment. On the one hand, these technologies will automate some daily, repetitive tasks and replace a batch of job positions. On the other hand, the new positions created by AI in the short or long term, either directly or through income effects, will offset the substitution effect. However, the period for income effects to take effect is relatively long, so the technological unemployment caused by substitution effects in the short term is inevitable. Especially in the case of significant lag in skill upgrading in China, the majority of the workforce cannot meet the requirements of AI for skills. This mismatch between technology and skills hinders the effectiveness of income effects, and the short-term pains may be "severe" or last longer. Therefore, for policymakers, it is crucial to expedite the training of AI talents, promote skill upgrading with technological iteration, and concurrently weave a social safety net for those who lack the time or ability to adapt to new technologies in the short term, narrowing the income distribution inequality caused by AI.
Final analysis conclusion:
The rapid advancement of AI technology will deeply influence the global job market. While it automates tasks and potentially replaces some roles, it also fuels economic growth by boosting productivity and cutting costs, thus generating new job opportunities. However, the rapid adoption of these technologies in China may overlook the need for talent development and skill upgrading. This could undermine the beneficial effects of new technology on the economy and job market while increasing the impact of job substitution. Consequently, policymakers will need to prioritize measures to enhance talent development and skill acquisition, in addition to establishing a social safety net to mitigate the negative impacts of technological unemployment and capitalize on the positive contributions of AI to the economy.
______________
Chen Li is an Economic Research Fellow at ANBOUND, an independent think tank.
