[3rd] AI Action Plan Formulation Committee Report｜Education, Smart City, Manufacturing!

What is the AI ​​Action Plan Committee?

Currently, artificial intelligence (AI) technology development is actively invested in research and development worldwide, mainly in the United States and China, and each country continues to evolve to obtain cutting-edge technological capabilities.

Japan also needs to build a system for AI technology development that can respond to these trends of the times.

It is also undeniable that Japan is lagging behind the United States and China in the utilization of AI technology through big data.

Therefore, from January 2021, NEDO (National Research and Development Agency New Energy and Industrial Technology Development Organization) will conduct a “survey for formulating a comprehensive research and development action plan and extracting business in the field of artificial intelligence (AI) technology ” . It was started.

The AI ​​Action Plan Formulation Committee , which promotes this survey, has been held since February.

The AI ​​Action Plan Formulation Committee will form a committee of experts and consider a clear action plan based on overseas cases and domestic and overseas institutional policies.

The committee will be held six times between February 2021 and June 2021 in the following flow.

In addition, the committee consists of the following members.

ASCII.jp and AINOW will serve as media partners of the AI ​​Action Plan Formulation Committee to strengthen the utilization of AI in Japan.

This article introduces the discussion at the second meeting of the AI ​​Action Plan Formulation Committee held on February 26 (survey to formulate a broad R&D action plan and business extraction in the field of artificial intelligence technology). I will continue.

Experts who participated in this committee will hold the ” Action Plan Symposium on Overall Research and Development in the Field of Artificial Intelligence (AI) Technology ” on Tuesday, June 15th.

Overview of the 3rd AI Action Plan Formulation Committee

In February 2021, the AI ​​Action Plan Development Committee formulated 20 future societies based on the “Survey on future predictions by science and technology conducted by governments and institutions ” conducted by the Institute of Future Engineering, a public interest incorporated foundation in February 2021. We will list up important events and proceed with discussions based on them.

AI technical keywords that are highly relevant to each social event are listed.

In the 3rd AI Action Plan, while digging deeper into the keywords and opinions added in the 2nd, we discussed specific technical and social issues.

This time, I would like to introduce five topics that were particularly important points in the discussion.

education

Regarding education, one of the committee members expressed the opinion that ” VR will be used to expand learning opportunities and possibilities .”

When a learner does something, they can only judge whether they were able to do it or not.

For example, if it were possible to measure how far the racket and the ball were separated from each other when the player missed a tennis swing and display it as a tactile sensation, it might be possible to learn more efficiently.

Also, not only VR but also AI may be able to present ” how well it matches (is it off) “.

For example, when a student grades a proof question in a math test, it is important to know ” how much did you miss” and ” where did you go wrong “.

However, the grading side naturally has limited resources, so there may be only a few comments, or if it is a large-scale test, only information about 1/0 points may be returned. To do.

On the other hand, with a machine or AI system, it may be possible to create a system that gives detailed feedback on how far the item has been removed .

Learners should be able to progress more efficiently if they can see how close they are to the correct answer.

Living (living and working style)

A topic that has become a point of discussion in everyday life is the ” smart city .” Currently, its attention is increasing in Japan, and although efforts are being made, it has not yet been implemented.

From the committee members

There is a limit to the “smart city” that tries to “smartize” legacy cities, and it cannot be considered as basic smart. Therefore, I think I will create a city from scratch.

There was an opinion.

For example, just like a major home appliance manufacturer in Japan tried to do in Fujisawa City, it is an image of “ building all houses with the same design on a vast land’ ‘, and the city is fully equipped with surveillance cameras and LED lighting interlocking security systems. increase.

By world standards, Monaco is similarly well-equipped with cameras and is recognized as a very safe city.

In addition, the introduction of self-driving vehicles is an important factor in the transportation of public transportation. For example, Venice has the characteristic that cars cannot enter, but in cities like this where there is only public mobility, life is designed based on that.

Also, from the committee

Autonomous driving will become easier if movement within a certain urban space is controlled by spatial mobility.

Also, if we were to create a smart city from scratch, we would have a concrete ideal based on the existing technology, and then think about how we should build the houses and the city.

If such a city is to be created, I think it would be better to do it in a non-legacy location.

The opinion was also stated.

In fact, in Susono City, Shizuoka Prefecture, a project is underway to create a new experimental city where cutting-edge technologies such as robots, AI, and autonomous driving can be introduced and verified in people’s living environments.

Also, if we try to forcibly insert various things into human space, robot-like evolution will be necessary.

Thinking about the design of such cities is likely to become a major point of discussion in future smart city topics.

self-driving

Regarding automated driving, one of the committee members was of the opinion, ” Can it be socially implemented in Japan? “

Last time, rather than focusing on technical issues, we agreed smoothly on the idea of ​​developing social legislation, but it can be said that how to actually implement autonomous driving in society is a major issue.

There were two opinions from the committee members regarding the implementation of autonomous driving.

The first is the opinion of “ creating roads exclusively for autonomous driving ”.

If we create roads where only self-driving cars can run, we will not only be able to travel smoothly without accidents, but we may also be able to significantly increase the speed limit. If that happens, the logistics industry in particular will benefit greatly.

The second is the opinion that ” first implement semi-automatic driving “.

Specifically, the idea is to spread semi-automated driving in which a human operator remotely intervenes only when an unforeseen event (such as a landslide or traffic jam) occurs, rather than fully automated driving.

In fact, we’ve seen scenarios in the past where human operators remotely control 10-20 cars, and we’ve seen semi-autonomous driving in a potato harvesting tractor.

In the future action plan formulation committee, it is a policy to deepen discussions on ” spreading semi-automated driving “, which is highly feasible .

Manufacturing

Regarding manufacturing, one of the committee members expressed the opinion of ” strengthening the process industry .”

The process industry (equipment industry) is an industry that is considered to require huge equipment (systems) in order to provide a certain level of production or services, or an industry that requires sufficient equipment and facilities to achieve certain results and profits. It is a term that refers to an industry that is expected to

The process industry includes carbon, oils, and drug discovery.

In addition, regarding this, the committee members

Within the process industry, drug discovery in particular is one of the areas where the use of automated design and machine learning is growing, and the huge budget makes it uniquely capable of such enhancements.

However, in other equipment industries, each target area is fragmented, or, like the pharmaceutical industry, each company is complete and lacks resources.

Therefore, it may be better to consider activities that can be cross-linked from the standpoint of AI a little more, and the creation of a platform that supports such activities.

There was an opinion.

In fact, there are successful examples of such platforms in Europe, and although efforts are being made to follow suit in the United States and Japan, there is currently no successful experience comparable to the preceding examples in Europe. .

In order to succeed in this, the problem is how to share the parts that each company wants to keep secret . In other words, “Can it be put on a machine learning model?”

In Europe, the results of collecting various equipment and data are analyzed by researchers attached to the platform itself.

They get information from companies as a trade-off, but the system is such that they pay more money in exchange for that information being kept confidential.

However, Japan and the United States are currently unable to fully imitate this move. What kind of system Japan will use to create such a platform is likely to become an important issue in the future.

AI x Science

In AI x Science, the first topic of discussion was ” How can humans understand the new science created by AI? ”

Even in new science, it is possible to learn high-dimensional things that humans cannot understand, but the problem is how to explain them to humans.

In addition, from the committee

Like the inter-universal Teichmüller theory, it may not be possible to understand if it is different from the existing system, but if each part is connected and can be traced, it would be good if someone could understand it.

There was also an opinion.

An example related to this opinion is AI Go/Shogi.

AI Go and Shogi have created many new formulas, and even commentators cannot understand why AI plays there.

And the shogi players re-research and understand the AI’s moves after the game.

In the same way, in science, only the results come out, and although it is incomprehensible at first, it may become possible for everyone to understand it eventually.

The next topic of discussion was how to put AI and science into action .

Regarding this, one of the committee members expressed the opinion that “we will foster an industrial ecosystem that facilitates high-throughput experiments by government agencies .”

When you search for materials, you still have to create the data for the story (there is a way to collect data by actually creating experimental equipment instead of simulation).

However, many of such experimental operations are one-off, and it is very difficult to create them.

For example, even if a university researcher thinks, “First of all, I want to create a certain kind of machine learning in this field, so I want to create an experimental device like this,” it is not easy to do so.

Therefore, we have to look for small and medium-sized companies that will make machines (experimental equipment), but even if we do find them, it may be difficult with only the technology of that company.

Based on this, the committee member said, ” Actually, having an industrial ecosystem in place may be the base for promoting science (especially high-dimensional science) .”

Also, if this is realized, it may be used for vaccine development.

In a world where we do not know when we will need a vaccine in the future, the country may need to maintain such an industrial ecosystem.

summary

This time, we have introduced five topics (education, daily life, automated driving, manufacturing, and AI x science) that were the points of discussion at the 3rd meeting of the AI ​​Action Plan Committee .

Also, a common keyword this time was ” how to implement and realize it in the world “.

Expectations are high for future discussions on how the measures and research that came up in the discussion will be incorporated into a clear action plan.

The New Energy and Industrial Technology Development Organization (NEDO), a national research and development agency, held the “ NEDO Symposium on Action Plans and Symposiums for Overall Research and Development in the Field of Artificial Intelligence (AI) – The Direction of AI Social Implementation that Japan Should Aim for – ” will be held as follows.

In this symposium, we will introduce the results of the ” Survey for formulating an action plan for overall research and development in the field of artificial intelligence (AI) technology and extracting projects “, and will also discuss with the expert members of the action plan formulation committee what Japan is aiming for. We will discuss the direction of social implementation of AI that should be.