“Artificial intelligence increases total productivity by 10% over 10 years” (National Economic Association 2025)

2025 American Economic Association (AEA) Annual Meeting
U.S. academia proves AI productivity innovation

“If artificial intelligence (AI) affects half of economic activities and improves the productivity of these activities by 20%, total productivity will increase by 10% over a period of more than 10 years.” (Professor Eric Brynolfsson, Stanford University)

At the 2025 Annual General Meeting of the American Economic Association (AEA) held in San Francisco, USA from the 3rd to 5th (local time), economists paid attention to the productivity innovation effect of AI.

In his presentation on the ‘Economic Impact of AI’ session, Stanford University Professor Eric Brignolfson introduced various research results, saying, “AI’s effectiveness in increasing productivity has been proven through several studies.” One study found that call centers with access to large-scale language models (LLMs) improved productivity by 30 to 35 percent. Professor Brynolfsson explained, “Customer satisfaction scores and call center employee happiness levels have increased,” and “the likelihood of employees quitting the company has decreased.”

While expectations about the future that AI will change are growing after the emergence of ‘ChatGPT’, a generative AI, in 2022, the fact that AI is substantially contributing to increasing productivity is being proven through various studies in academia.

Professor Brynolfson said, “The productivity growth rate of countries around the world, including the United States, has slowed since the 1990s and has only been around 1% since 2005. However, thanks to AI, productivity has increased significantly over the past 1-2 years, and in the third quarter of last year, the productivity growth rate was “It rose to 3.2%,” he said. He added, “It will take time for the full effect to appear, but productivity can recover to the level of the 1990s or higher.”

The tech industry also paid attention to the effectiveness of AI in increasing productivity, emphasizing that AI application should be spread throughout economic activities.

Jamie Teavan, chief scientist at Microsoft (MS), said, “In less than two years, people are already using AI in their existing work environment.” “AI needs to be applied across many countries and sectors,” said James Manica, senior vice president at Google. “It is not enough to just apply AI to some parts of the economy.”

However, although AI can reduce costs through increased productivity, concerns have also been raised that it may take away human jobs.

Susan Ash, a professor at Stanford University, said, “If AI models become cheaper, the cost of all areas, including education, healthcare, training, and elderly care, can fall,” adding, “However, as AI models take away jobs and the prices of AI models rise, the benefits to consumers decrease. “We cannot rule out a bad scenario in which consumer prices do not fall,” he predicted.

The U.S. Federal Reserve (Fed) is also paying attention to AI’s productivity innovation. Mary Daley, President of the San Francisco Federal Reserve Bank, said in an interview with Korean reporters on the 4th (currently), “I want to know whether the growth momentum of the U.S. economy is based on increased productivity,” and “I want to know whether the growth momentum of the U.S. economy is based on increased productivity.” “We are looking at whether it is supported by innovation,” he said.

<ⓒ투자가를 위한 경제콘텐츠 플랫폼, 아시아경제(www.asiae.co.kr) 무단전재 배포금지>