Could AI save the global economy?

SHIN MIN-YOUNG
The author is a visiting professor of economics at Hongik University.

How will the long-term prospect of global economic growth change after the Covid-19 pandemic is swept away? In general, growth slows down after a major economic or social crisis. Investment and innovation are delayed. As companies’ top priority is to survive by securing cash liquidity and reducing costs in times of crisis, new ventures are pushed back. In fact, for the five-year period after the 1997 financial crisis in Korea and the 2007-08 global financial crisis in the United States, Korea’s GDP growth fell by 2.0 percent and 1.2 percent, respectively, from the previous five years.

Of course, crisis response can also lead to innovation. The inventions of nylon and artificial rubber were the outcomes of desperation. During the global financial crisis, Airbnb grew rapidly, and during the Covid-19 pandemic, contactless technologies such as remote management and education spread fast.

But such individual cases cannot sufficiently offset the overall decline of growth. In a paper published in the Journal of Econometrics in 1995, Timothy Bresnahan and Manuel Trajtenberg proposed three conditions based on which new technology could bring drastic changes to the economy: The technology needs to be used in multiple industries; bring a series of innovations in such industries; and have the potential for continued improvement.

In the 2010s, new technologies such as tablet PCs and 4G mobile internet emerged, but the productivity growth rate of developed countries during that period was only about 1 percent. It seems that the aforementioned conditions were not sufficiently met.

Many experts agree that generative AI is a strong candidate to raise the growth of the global economy to the next level. Above all, generative AI has the high possibility of improving overall productivity thanks to its consumer-friendly nature.

The demand for generative AI spread tremendously within a short time. As it can be applied to almost all types of companies and can improve all kinds of jobs, AI is already integrated into the operation of many companies. AI is even used for the development and diagnosis of new drugs and for the efficient coding of programmers.

But it takes time to raise the overall standard of living, no matter how powerful a new technology is. Although the internet has been used since the early 1990s, it was only in the late 2000s that two-thirds of American companies had their own websites. It took 100 years for the steam engine — and 40 years for computer technology — to achieve full-scale productivity improvement.

There may be some resistance to the adoption of AI technology due to the fear of losing jobs. But the time it takes for AI to lead to productivity improvement could be much shorter than expected. Efforts to make AI technology cheaper and easier to use could reverse the declining trend of global economic growth when combined with the development of computer technology.