The potential impact of artificial intelligence on wages may mimic the traditional boom-and-bust cycle, according to a recent paper from the Brookings Institution. While the cycle usually pertains to market dynamics, this analysis focuses on wage fluctuations in response to AI advancements.
Initially, the integration of automation technologies can lead to increased productivity and, consequently, higher wages. However, as AI technologies become more proficient in performing tasks traditionally done by humans, the demand for human labor in those areas might decrease. This shift could force workers into lower-value or slower-growing job sectors, negating the initial wage growth.
Through simulations, Konrad Kording, a professor at the University of Pennsylvania Integrates Knowledge (PIK), and Ioana Marinescu, an associate professor at Penn’s School of Social Policy & Practice, observed that “automation in the intelligence sector first increases and then decreases wages.” They noted that after an initial boost in productivity, the “negative effects dominate as most workers get iced out of intelligence tasks.”
A Wage Boom Followed by a Decline
The researchers developed an interactive model to demonstrate the transition from human-led to machine-led intelligence. The model illustrates a sharp increase in wages due to heightened productivity from AI, followed by a stagnation and subsequent decline as automation becomes more widespread.
Despite the continued rise in output, wages decline, suggesting that the benefits increasingly favor capital over labor. As cognitive work becomes more automated, workers tend to shift towards slower-growing physical jobs, such as construction and caregiving, leading to reduced wages. This trend results in a hump-shaped curve, depicting an initial wage boom that is corrected as the digital economy surpasses physical industries.
The authors remarked, “Even as automation first increases wages, it can eventually lead to strong wage declines.”
The Case for a Balanced AI Rollout
The paper critiques both the utopian vision of limitless abundance through AI and the dystopian fear of total job loss. Instead, Kording and Marinescu advocate for a balanced approach termed “intelligence saturation.” This concept suggests that while AI can enhance economic productivity, its benefits will eventually slow, as it still depends on human labor and physical tools.
To avoid adverse effects on workers, the authors recommend moderating the pace of automation and investing in physical capital like machines and equipment. This strategy could help maintain productivity levels even as digital tasks diminish. Additionally, they propose imposing taxes on virtual substitutes for in-person services, echoing Sen. Bernie Sanders’ proposal for a “robot tax” on businesses that use AI to replace human jobs.
