AI: Asia – The next frontier in AI development

YANO Makoto (President, RIETI)

Connecting Cyber space and Physical Space

Society 5.0 represents the current Japanese government's vision on the development of future technologies. It advocates the integration between cyberspace and physical space meaning that all economic activities that take place in the physical space will be digitized through the use of sensor and input technology, the data will be collected through the internet and IoT and, then, analyzed by artificial intelligence (AI), with the results being fed back to the physical space. The government is aiming to establish a new economy that is human centered, with comfort, vitality, and high quality of life through this effort.

I think that Society 5.0 makes a lot of sense. It however poses a big question to us; would such a society automatically come true?

There are many issues to be resolved before cyberspace and physicalspace can be integrated. A concrete and significant threat we are facing is data monopoly and data manipulation. The Facebook-Cambridge Analytica data scandal that occurred in the US and UK poses a serious problem. Cambridge Analytica, a political consulting firm, utilized big data from the Facebook system. They created a mechanism where you can collect a massive amount of data by asking a one-click question. According to an article in the Guardian, Cambridge Analytica collected personal data from 230 million people in the United States, which means that 80% of the entire adult population's personal data was collected. As reported in the Guardian half a year ago, Cambridge Analytica manipulated negative news about various parts of society and influenced people's views on Brexit and the U.S. Presidential elections to influence voters.

A recent research project at RIETI has confirmed that people tend to react more to negative information than positive information. If propaganda is designed in a very organized way, it can negatively influence people.

It is also true that many people are concerned with AI and the singularity. Issues such as data manipulation and monopolies, the AI singularity and so forth are actually arising. There are many issues that have to be overcome before achieving a new type of human-centered society.

Industrial Revolution and Market

According to market quality theory I have developed, every time there is a significant industrial revolution, the market faces trouble. After the First Industrial Revolution, we experienced labor exploitation, from which the Marxean economics grew. After the Second Industrial Revolution, we also had a bad experiences. During the early 20th century, the Rockefeller Company held huge market power over gasoline for then newly introduced cars. Just before the Great Depression, people manipulated the markets by using misinformation. We are currently facing a very similar problem; a singularity, data manipulation and monopoly. In order to overcome these problems, we have to create a better market.

History tells us that we have overcome labor exploitation, the great depression, and the formation of monopolies. The economy is equipped with internal mechanism through which the economy is pushed back to a proper track. After the first industrial revolution, we created labor laws. After the second one, we created antitrust and securities laws. Currently, we are still deeply confused on what to do with the problems that we are facing in creating the human centered society through our new technologies. As an economist, I blieve that it is important to link cyberspace and physical space through high quality markets. It is highly desirable to develop a sound market to actually connect these two spaces in a nice, integrated structure, so that we will be able to come up with the next stage of the human centered society.

Distributed Computing and Distributed Management of Money

Two issues are important in connecting these two worlds: one is data ownership and the other one is money and micropayments. With the Internet and with AI, we are able to structure distributed computing, where many different computers connected by network are being used for one particular purpose. Different servers can share one task, divide it into small pieces, and send information through the Internet to produce an outcome. To avoid data monopolies, distributed ownership of big data is needed both in the cyber world and in the real world. IoT, blockchains and distributed ledgers are important.

Distributed management and ownership of money is important in this time when the banking industry is more and more concentrated. The first idea for crypto currency was published in 2008, and it works like a ledger, a book of permanent record. Compared to currency deposits, which are centrally controlled, crypto-currency is held by a distributed mechanism called a block chain. The largest difference between the current deposit currency and the crypto currency economy is that in the current deposit currency, the governmental trust provides the value of money, where the value of the money is higher than the paper that it is printed on, whereas block chain is based on proof of work and various other mechanisms.

The introduction of crypto currency is desirable and will be beneficial to the economy if there is an efficient way to make crypto currency. If such currency could be developed, the demand for and supply of currency would determine the value through market mechanism.In order to make it a possibility for the real world, it is necessary that the cripto currency reduces the marginal cost for money significantly. We will also need to have a fundamental trust in crypto currency, just like we have in the real currency right now.

Conclusion

We are seeing the very beginning of a huge development of many and various types of distributed ledgers. We will need more stable support mechanisms to incorporate crypto currency into the economy. The use of crypto currency will be one of the key elements in connecting the cyber economy, cyber space and physical space.