In the digital transformation wave, many businesses are promoting the application of Artificial Intelligence (AI) to optimize costs and improve efficiency.
However, a new study shows that replacing labor with AI is not necessarily a "medicine" to help sustainable growth, and may even have the opposite effect on businesses themselves.
The work called "AI-induced job losses" by Brett Hemenway Falk (University of Pennsylvania) and Gerry Tsoukalas (University of Boston) has challenged the popular view that automation always brings economic benefits.
According to the research group, although each business can save costs by cutting personnel, the overall consequence is the weakening of consumer demand, which is the core factor of growth.
The core of the study is very clear, that workers are not only workers but also buyers. When they are replaced by Artificial Intelligence and lose income, they will spend less.
Meanwhile, consumption is the main driving force of production and services, so when purchasing power decreases widely, business revenue also decreases accordingly.
Researchers point out that this creates a negative feedback loop. Businesses automate to reduce costs, but when many companies do it together, weakening aggregate demand causes market-wide profits to decrease.
In extreme scenarios, the economy may fall into a state of "high productivity but no demand".
A noteworthy point is that although aware of the risks, businesses still tend to continue to automate.
The reason stems from the phenomenon of "diplomacy", meaning that a business enjoys all benefits when reducing costs, but negative consequences such as reduced purchasing power spread throughout the market. Therefore, companies tend to continue to promote automation, leading to labor replacement exceeding the reasonable level.
The research model shows that each business makes a reasonable decision for itself, but overall it causes damage to the entire economy.
Not only workers suffer income loss, but businesses themselves also face profit decline due to weakening demand.
The study also considered policy solutions such as universal basic income, capital income tax or profit sharing.
Although they may help reduce social impact, these measures have not addressed the root of the problem.
Skills retraining programs are considered useful, but it is difficult to completely compensate for the short-term shortage of demand.
According to the authors, a more direct solution is to apply the Pigouvian tax, which is a tax to regulate behaviors that cause negative impacts on society.
In this case, businesses will be taxed according to the level of labor replacement with AI.
According to the research's argument, this policy can help balance business interests and economic stability, limiting excessive automation.
In the context of AI spreading across all fields, the problem is not only the application of technology, but also the management of its impact on the entire economy.
