2018年12月25日火曜日

We need to dream of a super smart society


We are currently noticing articles in the media reporting that Japan has had a late start in the AI development race. American companies such as IBM, Google, and Apple are the current industry leaders for AI development, and while the Japanese government has appealed to advance development with the Artificial Intelligence Technology Strategy Council, there remains a concern that we have missed the boat.

In Japan, the AI industry died down after the original boom in the 80’s. While the AI faction was lively during my time at the MIT Media Lab in the late 90’s, Japanese companies had already begun to lose interest.

Though Deep Learning has regained popularity in recent years, it is too late for Japan to put its taxes into universities for research because American companies overseas are already investing trillions of JPY into the race. Japan’s late start was not caused by the lack of talented personnel in universities, but because of company investments.

Our one chance to succeed is not via the development side of things, but the use side. Take a look at how female high school students used IT/mobile phone technology to create the cultures of emojis and selfies, or how the video-game industry was born by using American technology.

I believe that we should take the lead and adopt new technologies to give birth to new industries and cultures with them, such as with pop culture and our other fields of expertise, or with welfare, nursing, and disaster prevention, areas in which Japan should exist as a global challenge forerunner.

The government should also guide us in the proactive use of AI. They should show a high level of commitment to its use, such as by giving an Assistant Vice Minister post to an AI.

However, while it is important to consider AI in terms of investment and sales, this is missing the big picture.

We are currently starting to refer to AI and IoT as Industry 4.0. 1.0 is light industry, 2.0 is heavy industry, and 3.0 is information industry.

However, we must ask ourselves: is this truly an industrial revolution? Of course, the industries will be the first to change, but is the effect had on culture and society not more significant?

Even looking back to IT, which is mostly linked to Industry 3.0, was that truly an industrial revolution for us? Was it not more of a cultural revolution? I believe that rather than the growth of the IT industry and adoption of IT by businesses, the majority of the population beginning to carry and transmit information to one another was a much larger change for us.

If the 18th century industrial revolution was a once-in-300-years change, then I believe that the information revolution brought on by IT is a once-in-1000-years change of expression. The AI/IT revolution will be an entirely new frontier for us, a once-in-10000-year revolution. I believe that by taking something of such grand scale and simply summing it up as “Industry 4.0,” the strategies that we for it will be too small to fit.

This is why the Japanese government has introduced their vision of a super smart society known as Society 5.0 to stand alongside Industry 4.0. Each society was a revolution, with 1.0’s being hunting and gathering, 2.0’s being agricultural, 3.0’s being industrial, 4.0’s being information, and 5.0’s being civilization. This is a better fit for AI and IoT’s massive scale.

This time, the government’s White Paper on Information and Communications focused on IT’s consumer surplus. The value of IT cannot be measured as GDP. Even if the industry does not have huge sales, it has a huge effect on the money, time, and physiological use of its users. We must ask ourselves how we perceive an industry in terms of society, culture, and civilization, not just monetary value.

What impact will AI have on investment and sales, and how will further reliance on AI change civilization? I would like an index to measure this with so that countermeasures can be formed. However, we are still on the tip of the iceberg.

2018年12月11日火曜日

We need investment strategies for new infrastructures


Professor Noriyuki Yanagawa submitted an article titled “The development of IT, and its effects on investment.” to Nikkei Shimbun.

The summary of the article is as follows:
1) Information technology such as the internet, phone apps, and social media are becoming the base, with sharing and blockchain technologies seeing more and more activity. Companies like Google and Amazon are making their AIs open to the public, and developing related services.

2) These new technologies are infrastructures that encourage investment. New infrastructures like these are vital. IT and AI are like engines that give birth to new infrastructure, meaning that one infrastructure leads to another.

3) New infrastructures are handled by private enterprises, and have the characteristics of change and expansion. The strategy of the modern age will be to invest more in superior new infrastructures in order to increase the amount of new businesses that will use it.

This theory encourages the reform of public economics. Infrastructure such as roads, airports, waterworks, and electricity are public goods, meaning that they cannot function for business purposes and are left to the country, municipalities, public-service corporations, or any other entities in the “public” sector. They are financed by taxes and use fees.
New infrastructures will be provided for free by pure private enterprises. By working economies of scale, they gain compensation from derivative businesses as well as obtain client information and other big data.

For this, Professor Yanagawa believes that large companies, not start-ups, will take the reins. Because this is a matter of new infrastructure, I too believe that for Japan, it is more realistic to make a large-scale investment in a major company than it is to invest in a startup.

The problem with these new IT/AI infrastructures is that they will inevitably cross borders, meaning that no restrictions on foreign capital investment can be made and that global companies are guaranteed to invade. We cannot compete with national power alone. This is yet another reason why we need to reconsider public economics, which uses the policies of sovereign nations as the base for its theory.

Furthermore, the concept of market failure, a part of public economics, is relevant here. How do we overcome the problem of insufficient infrastructure investment? Professor Yanagawa mentions that the decrease of capital investment caused by the application of infrastructure will bring down economy aggregate demand. I would like to see this analyzed from an economist’s point of view.

New infrastructures are known for becoming cross-industry. This is why they requires new rules, rather than vertical stove-pipe regulations for each individual sector.

To maximize the benefit for these new infrastructures, we must promote investment and form policies that allow the benefits of utilization to become widespread.

The government must break down the stove-pipes, and commence deregulation and the formulation of policies to promote use. One valid measure would be to introduce a tax system that promotes both the use and investment of the new infrastructures.

On the other hand, targeting specific sectors with subsidiary aid to promote growth is likely to fail. A better idea would be for the government itself to take the lead by being a consumer, which would encourage investment by the private sectors.

Our problem is that the IT and AI sectors are already controlled by the U.S., and we are unlikely to ever gain the lead. What we need to do instead if aim for the layers above, such as IT, robots, blockchain platforms, or application areas such as education, medical care, disaster prevention, agriculture, and urban management.

To make this a reality, a comprehensive strategy is essential. The IT Headquarters and Intellectual Property Strategy Headquarters is currently working on both intangible/tangible strategies. Of course, Professor Yanagawa and myself are participating as well. However, what we require is a different strategy, one to promote new infrastructure investment.