Media Foreign Affairs and National Security 2019.01.29
My son is a musician and composer. He desperately endeavors, every day and night, to compose creative music. I dreamed of becoming a professional musician 40 years ago. Now I have full sympathy for my son because it seems increasingly difficult for artists in the 21st century to be original in music or any other form of art.
Take my son's example. When I was a teenager, electric instruments changed the music scene. There were no digital recorders, computers or databases then. No matter what you did, you could easily sound creative. Now everything is digitally recorded for reference and, for my son's generation, being original is very difficult.
Although he utilizes digital sound sources, my son composes music manually, meaning without computer-assisted composition software. AI-assisted composition, he argues, doesn't always sound creative and only humans can make great music. It is particularly so, he believes, in the case of live music. I agree with him.
Then more questions arise. If art is an intellectual activity to move people by transmitting feelings to them, and if people say they are moved by an AI-composed song without knowing it, should that piece be called art? If so, my son will lose his job. Can artificial intelligence really recreate human sensitivity in the first place?
AI may eventually outsmart humans. Pundits call that moment "the singularity." Some warn that will be the day when machines outperform human beings and finally start controlling them. Others, however, disagree by claiming that AI's current technological level is still low and will not surpass humans in the foreseeable future.
Maybe so. I have no expertise to challenge their propositions. Indeed, practical applications of AI technologies have just started. Moreover, unfortunately, Japan is only focused on business and commercial -- not political and military -- applications of AI technologies. The singularity may come later rather than sooner, at least here.
Having said that, the history of mankind looks like one of battles between machines and humans. Since our ancestors started using fire and tools a few million years ago, the holders of state-of-the-art technologies have grabbed power and the technologically inferior have been forced to obey them, whether they like it or not.
For the past few centuries, revolutions in mobility (steam engines), in energy (internal combustion engines) and in communication (data processing) have changed the winners and losers in the history of mankind. The Fourth Industrial Revolution with AI technologies will soon follow. The question is, what will then happen to us.
There are two schools of thoughts on the AI revolution. Optimists claim that advanced technologies will enhance productivity and lead to more new industries and employment. Pundits claim that even after the Luddite movement in 19th-century Britain, workers' living standards improved. Is that true?
Such optimism is based on the logic of macroeconomics by powerful winners who enjoy a technological edge. From the viewpoint of the weak losers, however, the result is the opposite. Once lost, it is not easy for them to find new jobs quickly. The voices of frustrated, angry, unemployed men and women should not be underestimated.
When the AI revolution takes place the world will dramatically change again. AI-assisted economization and rationalization will result in numerous unemployed people in both the manufacturing and services sectors. Many of them may join and even radicalize the politically intense nationalist and populist movements in the West.
Macroeconomic solutions could take a decade or more. By that time, the socio-political environment in the industrialized countries could easily deteriorate. A half century after the Luddite movement, the "First International" was established in London. If history rhymes, the AI revolution may have two serious consequences.
The first is a series of radical or anti-government political movements by frustrated and desperate unemployed people. Western governments may need to be ready to face such political challenges. The second is the possible rise of "neo-socialism," which gives priority to economic equality over liberty and democracy under capitalism.
All in all, political stability and economic prosperity in the future seems likely to depend on who can control the anger and frustration of the critical mass of the unemployed. I may be too pessimistic about AI technologies but the following is my take on how the AI revolution will change domestic politics in the next few decades:
■ There will be no "AI versus humans" battle in the foreseeable future.
It may be wrong to expect the singularity to take place soon in politics. What AI technologies will bring about in domestic politics will be intensified power struggles among human groups, each using different AI technologies to compete with one another in economic, political and social affairs.
■ In the "AI versus AI" battles, the superior AI will win.
Future battles in the AI field will be fought by AI technologies. But they will be battles between the people who wish to utilize AI for good and those who want to use it for malicious purposes. The singularity may eventually come, but in the foreseeable future what matters most is who uses AI and for what objectives.
■ It is not machines but humans that make AI evil.
From a practical point of view, AI or any other technologies are apolitical and fundamentally non-partisan by nature. While AI seems to be favored by authoritarian regimes such as China, for example, where it can be applied to control people, AI also can be a means to weaken such harsh control if properly utilized.
All in all, we should not be afraid of the artificial intelligence revolution. In fact, it already started in the past decade and will only accelerate from now. By the same token, we've started benefitting from AI even while we don't recognize its endgame yet. Again, it is humans, not machines, that make AI a force for good.