The Comparative Advantage Effect in Artificial Intelligence Adoption

A new discussion paper reveals that the use of artificial intelligence in workplaces depends not only on capabilities, but also on user costs and comparative advantage.

◉ 0 views
Beyond exposure: Predicting AI adoption based on comparative advantage

According to a newly published CEPR discussion paper, the adoption process of artificial intelligence in the workplace is shaped by comparative advantage and user costs rather than pure capability or exposure levels.

Discussions on Artificial Intelligence and Economic Impact

Where and how widely artificial intelligence will spread remains a central source of concern regarding employment, wages, productivity, and inequality.

Despite extraordinary advancements in artificial intelligence capabilities, the pace of technology adoption is following a slower trajectory than expected.

Inadequacy of Exposure Metrics

Standard economic impact metrics focus on the concept of exposure, which evaluates whether artificial intelligence can perform the tasks of a specific profession.

However, while exposure metrics show what artificial intelligence can do, they do not fully explain who will actually use the technology.

Research on German Data

In a new study conducted by Lindenlaub and colleagues, representative data from the 2024 DiWaBe 2.0 survey covering 9,835 German workers was utilized.

Research findings indicate that technological capability constitutes only one side of the equation and that the primary barrier to diffusion is usage costs.

Three Forces Determining Artificial Intelligence Adoption

The newly developed framework distinguishes among three main forces: artificial intelligence efficiency, usage costs, and efficiency relative to the employee's salary.

A model based solely on exposure can explain only 25 percent of the adoption variation across professions.

Usage Costs and Occupational Examples

Although the accounting field possesses massive technical exposure, it harbors high verification costs and privacy concerns.

In contrast, the teaching profession possesses lower artificial intelligence exposure while simultaneously exhibiting a strong adoption process thanks to lower usage costs.

Artificial Intelligence Projections for the Next Three Years

Projections for the next three years made using the model predict that the proportion of employees adopting artificial intelligence will rise from the 44 percent level to approximately 81 percent.

It is stated that the primary driving force behind this future growth will be declines in usage costs rather than productivity increases.