[Reporter’s Notebook] Making Sure the Call for an 'AI Powerhouse' Is Not in Vain

"I work at a government-funded research institute, but it is extremely difficult to obtain government data. The best I can do is ask someone I know inside the government to get it for me," the researcher lamented.

This was the complaint of a researcher in the field of social policy whom I met last spring. Despite working at an institution that advises the government, the researcher wore a bitter expression, saying that obtaining government data was just as difficult as when working in the private sector (at a university). Listening to this, I realized that if the quality of policy research depends on personal connections, there is a clear need for improvement. Still, I brushed it off, thinking that, given the nature of government data, some restrictions were inevitable.

It took only a few months for me to realize how complacent that thought was. When I recently visited another government-funded research institute, I heard the same complaint. The researcher there, who is developing an artificial intelligence (AI) model to assist in policy decision-making, said, "I requested past data from a government ministry to improve AI predictability, but my request was denied due to personal information protection concerns." For AI research, securing data is especially critical to success, so this situation struck me as quite serious.

This is the reality on the ground, yet government announcements make it seem as if the path toward becoming an AI powerhouse is already well prepared. The government has declared its intention to achieve an "AI Grand Transformation" and leap into the ranks of the world's top three AI nations. Massive support budgets and a 150 trillion won National Growth Fund have also been announced. However, in reality, we are not even properly utilizing the data we already have, and these problems are left unaddressed. It is also difficult to find statements from policymakers that emphasize data, the core fuel of AI.

On November 4, President Lee Jaemyung stated in his budget speech to the National Assembly for next year, "In the AI era, being a day late means falling a generation behind." Deputy Prime Minister and Minister of Economy and Finance Koo Yooncheol, the nation's chief economic policymaker, identified AI as "the only breakthrough to reverse the decline in growth caused by demographic shocks." If the government's sense of urgency is to be genuine, it must first clear the bottlenecks present in the field. Rather than touting AI as a magic wand, it is necessary to properly nurture the seeds-data. No matter how large a data center is built, it is meaningless if there is no data to fill it. It is time to face this criticism head-on.

Fortunately, the solution does not seem far off. Experts I have met in various fields have advised that the government must demonstrate strong will in managing, operating, and utilizing data. Since data silos exist both between government ministries and within departments, these barriers must be addressed, and highly useful government data should be anonymized and provided to the public and other entities. They say that if the government pays more attention to data utilization, these issues can be sufficiently resolved. Data is one of the three pillars of the AI industry-data, infrastructure, and talent. Neglecting this and still expecting success in the AI industry is nonsense.

This content was produced with the assistance of AI translation services.