Why Jobs Won't Decrease in the Era of AI Efficiency: Future Strategies Revealed by a 19th-Century Paradox
The rapid advancement and widespread adoption of AI (Artificial Intelligence) have sparked widespread concern that it might "take away human jobs." On the other hand, through new demand and changes in industrial structure created by efficiency and automation, it is also possible that employment could expand in ways previously unimagined. In fact, we can derive hints to explain these phenomena from a theory proposed in the 19th century called Jevons Paradox.
In this article, we will first outline Jevons Paradox, and then present a perspective on applying it to the labor market in the AI era. Finally, based on current trends and future prospects, we will propose perspectives and preparations that readers should adopt.
1. What is Jevons Paradox?
1-1. Definition and Historical Background
British economist William Stanley Jevons pointed out in his 1865 book, "The Coal Question," that improvements in steam engine and fuel efficiency paradoxically led to an increase in coal consumption.
At the core of this idea is the rebound effect, which is counterintuitive at first glance: when the cost of using a resource decreases due to efficiency, people end up using more of that resource.
This theory is also referenced in environmental policy and energy efficiency discussions, serving as a warning that "efficiency" does not necessarily mean "reduced consumption."
1-2. The Mechanism of Efficiency and Demand Expansion
For Jevons Paradox to hold, efficiency is said to be able to trigger demand expansion when the following conditions overlap:
Efficiency improvement: Costs or inputs per unit decrease
Price reduction: Consequently, usage and consumption costs decrease
Elasticity of demand: Cost reduction can induce consumption expansion
Creation of secondary demand: New uses or needs beyond basic applications emerge
For example, if power generation and factory operations become cheaper due to coal efficiency, more factories will operate, and as a result, coal demand will increase.
This paradox provides a perspective that efficiency can have the potential to unlock latent demand rather than just being a form of savings.
2. Why the paradox can be applied to "AI × Labor Market"
2-1. The Case of Radiologists: A Real-World Example Where AI Adoption Expanded Demand
As mentioned in the text provided, even though Geoffrey Hinton predicted that "AI would surpass radiologists within five years," that did not actually happen. Conversely, as medical image analysis AI has become widespread, the demand for radiologists is actually at a historic high.
The key to explaining this phenomenon is the cycle: "Scans become cheaper and can be done in large quantities → The number of examinations increases → Cases requiring more advanced diagnosis and treatment increase → The involvement of radiologists becomes indispensable." This is a classic example of the paradox where efficiency (the introduction of AI) lowers costs, increases usage, and expands the workload.
Although there are industry-specific constraints such as medical, insurance, and liability issues, the principles of technological complementarity and the uncovering of latent demand are widely applicable.
2-2. Complementarity and Substitutability of Labor in the AI Era
The debate over whether AI will act as a substitute for humans or as a complement is a critical proposition in labor economics. Recent research suggests that the effect of complementarity may actually be greater than that of substitution.
Specifically, as AI technology advances, humans are expected to shift their roles toward tasks AI struggles with—such as contextual judgment, ethical decision-making, and interpersonal coordination—and work in collaboration with AI.
In this scenario, a paradoxical expansion of demand can occur: efficiency gains lower operational costs → this enables more work or more complex work → human judgment and value-added tasks increase.
3. Current Trends and Why the Paradox Is Not Always Correct
3-1. Empirical Trends in AI Substituting and Changing Jobs
Goldman Sachs analyzes that while AI adoption will improve productivity, it could push up the unemployment rate by about 0.3 percentage points in the short term. However, they expect most of this to recover within two years.
Brookings' analysis suggests that "more than 30% of all workers could see over half of their job duties changed by generative AI."
Additionally, the U.S. Bureau of Labor Statistics (BLS) has pointed out the possibility that the duties of credit analysts could be partially replaced by AI, projecting a change of approximately –3.9% by 2033.
These trends support a moderate view that differs from both the optimism that "all jobs will be protected" and the pessimism that "everything will be replaced."
3-2. Cases and Limitations Where the Paradox Is Hard to Apply
Jevons' paradox is not universal. Under conditions such as the following, efficiency gains may struggle to expand demand:
Cases where demand is inelastic
For example, in fields like food or medicine where consumption is constrained beyond a certain point, even if costs decrease, the expansion of consumption may be limited.Structural and institutional constraints
In fields where regulations, laws, insurance systems, or ethical constraints intervene and human involvement is essential, efficiency gains are less likely to lead directly to replacement (e.g., medicine, law, education).Areas with low complementarity and high substitutability
Tasks that can be completely standardized, have high error tolerance, and require no human intervention may see efficiency gains lead directly to replacement.Situations where the rebound effect is small
This applies to structures where the benefits of cost reduction from efficiency gains are not passed on to prices, and demand does not fluctuate.
4. Practical Perspective: What Should We Consider as Readers?
4-1. A Perspective of Changing Job Forms, Not Assuming "Jobs Will Disappear"
Efficiency and automation not only simplify and accelerate tasks but also provide an opportunity to redesign the scope and roles of work.
For example, areas requiring human judgment, creativity, and ethics—such as "operating and supervising AI," "collaborating with AI to make complex decisions," and "correcting AI-generated proposals from a human perspective"—will become even more important in the future.
4-2. Strategies for Skill Shifts and Reskilling
Research indicates that skills complementary to AI (digital literacy, critical thinking, interpersonal communication, etc.) are seeing increased demand and are more likely to command wage premiums.
Therefore, reinforcing specialized knowledge while honing general-purpose skills and interdisciplinary integration abilities will serve as long-term preparation.
4-3. Perspectives on Innovation and Entrepreneurship
Efficiency can lead to the creation of business models that are only viable because costs have decreased. Following Jevons' paradox, the reduction in cost itself becomes the starting point for fostering demand.
As AI becomes cheaper, faster, and more accurate, identifying areas where products and services that were previously unfeasible can now be offered will provide hints for startups and new business ventures.
Conclusion
Jevons' paradox provides a powerful framework for discussing the impact of the AI era on employment. However, whether it holds true depends on many factors, including industry, systems, skill structures, and social institutions.
As presented in this article, AI efficiency does not necessarily suppress labor; rather, it has the potential to unlock latent demand and generate employment. Furthermore, as agents of the future, readers should focus on maintaining an attitude of embracing change and flexibly reconstructing skills, as well as developing the perspective to decipher the new demand hidden within efficiency.