Artificial intelligence (AI) can offer solutions to more fields and further enhance productivity. However, some worry that AI may lead to job losses and widen income inequality. Like the internal combustion engine and the computer, AI is considered a general-purpose technology, one that has the potential to impact economic conditions and cause labor displacement. Every historical technological revolution has brought concerns about “technological unemployment.” While transitions often involve growing pains, we have always managed to create more jobs than were lost.
Automation typically increases production efficiency, and the resulting output can be used to develop new products and services, generating entirely new productivity demands. As these demands expand, more jobs that cannot be replaced by machines will emerge. MIT’s David Autor found that 60% of the jobs in the U.S. in 2018 didn’t exist in 1940. The iterations of technology and labor redistribution demand new skills from workers, who must continuously pursue training and education. If the economic transformation is not smooth, macroeconomic and financial stability can also affect labor structures. The decades-long polarization of the labor market shows that income distribution can be influenced by the automation process.
Will the AI era follow historical trends, or will it open a new chapter? Like many technologies, AI has the potential to free up productivity in many automated jobs. With its ability to continuously learn from databases and integrate implicit knowledge, AI holds the potential to perform non-routine tasks. Some might believe that AI will one day outperform humans in most jobs — that day may come, but it’s not here yet. At this stage, AI can independently complete a few cognitive tasks and assist humans in more complex ones.
A study suggests that knowledge-intensive industries like finance, advertising, consulting, information, and communication are more likely to be affected by AI. In contrast, industries with lower knowledge intensity, such as mining, construction, administrative support, and transportation, have less connection to AI. This raises concerns about technological disruption once again — automation has already caused many blue-collar workers to lose their jobs. Could highly skilled workers be next? Is AI a substitute or a catalyst? Both are important, depending on the industry. Jobs that don’t require judgment on ambiguous events can be automated. In knowledge-intensive industries, using AI to augment workers’ abilities is the better choice. AI may replace some jobs, but it can also enhance productivity for others. While AI can provide valuable tools for work, it may introduce bias or misleading judgments in high-stakes decisions, meaning human judgment cannot be entirely replaced.
The cost, as well as the ability to allocate resources, adjust production processes, and adapt business models, will affect how efficiently AI is adopted and its productivity impact. The costs of developing, applying, and operating AI models are high, and their energy-intensive nature may threaten sustainability goals in various countries. Additionally, bottlenecks in data and skills may hinder the implementation of AI projects. As tech giants leverage AI’s advantages to strengthen their market positions, the gap between them and other businesses will widen. Therefore, regulating the AI race will be a crucial step in encouraging widespread innovation and productivity growth.
Many economists remain optimistic, believing that breakthroughs in AI could permanently increase the growth rate of productivity. However, looking back at history, the path of technological innovation and application has always been uneven, often influenced by external factors. For now, both the fear of mass unemployment and the anticipation of significant productivity growth may be premature. Nonetheless, AI can revolutionize work methods, influence business models, and reshape distribution patterns. Studies suggest that AI could provide more support to low-skilled workers, thereby helping to reduce inequality.
Ultimately, will AI serve as an automation worker bee that leads to unemployment and wage cuts, or will it assist in creating new products, services, business models, and jobs? Governments should formulate policies that increase investments in education, promote skills training, provide career counseling and job assistance, ensure fair competition, optimize labor and capital taxation policies, and prevent AI misuse to ensure a smooth and positive AI transition.
Leave a comment