I would like to recommend reading yesterday's interview with Mail Business by Eric Brignolffson, director of the Center for Digital Economy at Stanford University. It contains the latest analysis and concerns of world-class scholars who have led research in economics and management in relation to the impact of the artificial intelligence (AI) revolution. Last year, Professor Paul Krugman and his talks were a hot topic, and I was even more glad that the progress made since then has been updated well.

Economists have two main concerns. The first is when the productivity explosion will occur due to the introduction of a general-purpose technology called AI, and how to advance the timing. The second is how to minimize the mass unemployment and income polarization that the AI revolution will bring.

Introducing advanced technologies does not immediately improve productivity. It was like that when power appeared 140 years ago, and it was like that when computers first came into the world 50 years ago. Economic historian Paul David summarized this early in his famous paper 'Generators and Computers' in 1990 as the limit of simple technology substitution. Edison introduced electric power in 1882, but the actual productivity improvement began in the 1920s. At the time, companies simply took out steam engines from factories and put electric motors in the place. It missed the essence that the power transmission method has changed from centralized to decentralized. The steam engine is characterized by its nature, and all machines must be densely packed near the main power shaft. On the other hand, electricity can be placed freely in accordance with the flow of the process by attaching individual electric motors to each machine. It took me 40 years to realize this. The factory layout was changed and the organization was redesigned to match the electricity. The workers were also retrained. Only then did it switch to a Ford-style assembly line capable of mass production.

The same was true of computers. It took 15 to 20 years to convert society, organizational systems, and business practices into computer-centered ones. It wasn't until the mid-to-late 1990s that a strong IT-based productivity explosion began. Enterprise-wide resource management (ERP), supply chain management (SCM), and Internet innovation finally broke out.

AI is no different. However, the period is expected to be greatly shortened. There is a precondition. Likewise, it is a paradigm shift at the organizational level that re-adjusts existing jobs and completely reorganizes tasks. As if replacing fuel from coal to oil, productivity improvement is minimal. This is the background in which the timing of productivity explosion varies from company to company and country to country.

There are also side effects. It is large-scale unemployment and polarization that are repeated every time a new technology is introduced. This happened as technology simply automates and replaces human labor. However, if the new technology advances through human capacity enhancement or supplementation, the story will be different. In particular, if AI helps to "simplify jobs" that lower barriers to entry into complex tasks, it will lower the difficulty of the job so that even low-skilled workers can enter high-value-added jobs. In fact, in a paper published by the National Institute of Economic Research (NBER) in July, Lucas Althoff, an economics professor at Stanford University, proposed a scenario in which the average wage (productivity) of the country as a whole rises by up to 21% due to the job simplification effect of AI. Polarization is also largely resolved.

AI can be a blessing or a disaster. But it's our choice, not our destiny. This is consistent advice from Major General Brignolffson. "AI, created to imitate and replace people, concentrates wealth and power, but AI to expand what people can do widely divides profits and grows pies. Even the same technology can produce the opposite result." Humanity is now rapidly being sucked into a world that no one has ever been to. A blessing can turn into a disaster at any time if you indulge for a moment.

[Knowledge Director Song Sung Hoon]