Xiong Bingqi: In the AI era, employment-friendly development must move beyond replacement thinking

Main text

Recently, the State Council issued the "Opinions on Adapting to the Development of Artificial Intelligence Technology and Helping to Build an Employment-Friendly Development Model," which explicitly proposes tapping the employment potential of AI in traditional fields and mitigating the employment substitution impact brought about by AI applications. The document proposes giving play to the role of AI in helping traditional industries transform and upgrade, implementing human-machine collaboration in industry and agriculture, and tapping digital and intelligent employment opportunities; expanding new scenarios for service consumption and enlarging the employment capacity of the service sector. It also proposes giving priority to promoting AI in positions facing labor shortages and in high-risk posts, steadily grasping the pace of application in labor-intensive industries, and encouraging enterprises to establish unemployment buffer periods, notify workers in advance, and provide support for job transfers and transitions.

As AI technology is rapidly implemented and its promotion and application accelerate across industries, many traditional positions have been affected. Some employers use AI applications to cut recruitment quotas and eliminate traditional positions, giving rise to the worry that "AI is competing with job seekers for jobs." How to use AI to raise enterprise labor productivity while avoiding a simple job-displacement effect, mitigating the employment substitution impact brought by AI, and truly achieving AI-empowered employment is a real problem that must be confronted directly.

It is the general trend of technological change for enterprises to embrace AI and use AI technology to take over mechanical and repetitive work. But if the goal of application is merely to reduce staffing and lower labor costs, it will intensify the pressure on new youth employment and also leave incumbent workers facing the risk of unemployment. If both public opinion and enterprises one-sidedly magnify AI's substitution effect, emphasizing only the elimination of old positions without exploring position upgrading and the creation of new opportunities, the value of AI development will be greatly diminished. This State Council document makes arrangements for mitigating the impact of AI substitution and sets out the direction for governance.

The document explicitly requires giving priority to promoting AI in positions facing labor shortages and in high-risk environments. This is the most commendable value of AI technology: replacing high-risk, high-intensity, and high-danger human labor with machines greatly improves workers' working environments and reduces the risk of safety accidents, and is an important manifestation of technological progress benefiting society.

But it must also be viewed dialectically: after high-risk posts and heavy manual labor posts are replaced by AI, the job risks disappear, but workers' survival problems still exist. Most frontline workers who are eliminated or replaced have single skills and narrow reemployment channels. If only technological substitution is discussed and personnel placement is not, technological progress will become a livelihood crisis for some groups.

In reality, many scenic areas retain the positions of mountain porters, which is a highly instructive governance sample. Today, drones, robot dogs, and intelligent transport equipment can completely replace human labor in carrying loads up mountains, with higher efficiency and lower costs. But the scenic areas have not fully replaced people with machines; instead, they appropriately retain traditional human positions. The reason is very simple: although the work of mountain porters is hard, it is the most core, or even the only, source of income for local middle-aged and elderly workers. Full machine substitution would amount to directly depriving them of their livelihood. This prudent stance that takes both technological efficiency and people's livelihood and employment into account is precisely the balancing thinking most needed in current AI applications, and has therefore won broad social recognition.

On this basis, building an employment-friendly society in the AI era should establish a higher-level development principle: when an enterprise replaces an old position, it should simultaneously upgrade and create a new position; when an industry eliminates traditional business forms, it should simultaneously spawn new employment forms. It cannot only do "subtraction layoffs"; it must simultaneously do "addition creation."

AI greatly improves labor productivity, and this should be a dividend for the whole society. It should not merely be converted into enterprise profits, but should also be converted into workers' well-being and the stability of social employment. For example, against the background of technology improving efficiency, enterprises can fully explore a new employment model of "improving efficiency without reducing staff, improving quality while shortening working hours": relying on AI to replace redundant labor, optimize production processes, and increase overall production capacity, but not simply reduce hiring and lay off workers; converting the efficiency dividend brought by technological progress into shorter working hours per person, optimized labor intensity, and balanced job loads, using working-hour optimization in exchange for stable employment capacity.

At a U.S. university graduation ceremony this year, an entrepreneur talked freely about AI's replacement of traditional positions, and the college students in the audience booed. This does not mean that young people resist artificial intelligence; rather, when the technology narrative contains only content about elimination and squeezing employment, it is difficult to win everyone's recognition. Only by continuously cultivating new employment opportunities different from traditional positions and grounding technological change in human development will AI be accepted by workers.

Of course, positions being replaced by AI is difficult to avoid in the process of technological iteration, but enterprises cannot only do simple layoffs and compensation. First, enterprises should bear the responsibility for employee transition training. Enterprises cannot promise employees lifelong employment, but they have the responsibility to help employees develop lifelong employability adapted to the AI era, and cannot wait until positions are replaced and then merely pay compensation to terminate labor contracts. In reality, there have already been many relevant labor arbitration cases. Even if arbitration rulings protect workers' rights and interests, many outcomes are still that enterprises complete layoffs after paying compensation, and workers lack room for transition buffers and find it difficult to cope smoothly with job changes.

Second, the endpoint of AI application should be industrial upgrading, not merely replacing human labor. Enterprises should rely on AI to transform original businesses, spawn new business tracks and new job demands, and open channels for original employees to move to new AI-related positions. That is to say, they cannot do the subtraction of "AI replacing people" within the original business volume, but should rely on technology to expand development boundaries, create new employment through industrial expansion, and achieve mutual progress in both industrial upgrading and human development.

In short, the spread of artificial intelligence is an irreversible trend of the times. The relationship between AI and workers should not be one of either-or replacement, but rather one of collaborative upgrading and mutual empowerment. When enterprises advance intelligent transformation, they must not focus only on cutting headcount; they should also use new technologies to empower incumbent employees, expand industrial tracks, and create new jobs. Only by moving beyond the simple logic of elimination and replacement, building an employment-friendly development model, and moving toward an empowerment model of human-machine collaboration can technological progress truly benefit enterprises, workers, and society as a whole. This is precisely the question of the era that the AI age must answer. □Xiong Bingqi

(Source: Southern Metropolis Daily) Mid-article embedded advertisement