Awaji Conference Statement 2017

As a member of the audience, I remained excited throughout the event. I have learned many new things, and encountered many issues that provided me food for thought. It was a truly joyous and intellectual conference. I would like to express my sincere appreciation to all of you.

This year’s forum was very significant in that the event enabled us to learn about the reality and current situations surrounding robots and AI. To people of my generation, a robot is synonymous with AIBO released by Sony in 1999. In those days, it was sold for 250,000 yen. A new robot that I did not know until attending this forum is Azuma Hikari of Gatebox. I thought that it might be good if the hologram said to me “Have a good day!” and “How was your day today?” every single day. I also thought that making people feel comfortable in such a way might be one of the functions of robots. At the same time, however, I felt concerned widespread use of Azuma Hikari designed to provide psychological support to people living alone might lead to an increase in the number of people who would no longer need to get married, which would work negatively on the Japanese society facing a declining birthrate and aging population. I checked the product details to find that Gatebox had accepted orders for Azuma Hikari only online and sold the product from December 14, 2016 to January 31, 2017. Moreover, only 300 unites of the product were shipped in and after December 2017. It seems that my concern about the effect on the Japanese declining birthrate and aging population is groundless.

At the forum, I listened to a wide variety of speakers, including Prof. Ishiguro. I did not know that robots developed by him could not be used yet in industry. As a typical AI product, Bonanza, a shogi software program, is much talked about today, especially when the fourth-dan (grade) shogi player Souta Fujii is covered in news. In addition, AlphaGo is also attracting much attention as an excellent go AI program. According to Prof. Ishiguro, the company that developed AlphaGo tried to develop business programs based on the expectation of AI’s potentials, but their attempts totally failed. This episode suggests the reality that each function of AI and robots is truly excellent, but their general capabilities are still insufficient.

However, even that episode did not stop my excitement during the conference. At the international symposium yesterday, four speakers delivered lectures. Now, I would like to talk about what parts of the lectures interested and surprised me. First, Prof. Ishiguro spoke about the change of the times from PCs to personal robots. I was surprised to hear that androids were not well accepted in the West, but in Japan. Another thing that surprised me was that androids today have not been equipped with AI yet. According to Prof. Ishiguro, although future androids might have AI, the time has not come yet today. The most surprising thing was his remarks that his main purpose of creating androids was to know more about humans. I agree with his recognition that humans are above robots.

Second, Mr. Hamaguchi spoke about a four-quadrant taxonomy. The right side of the transverse axis means the concept of “A or B” (selecting either of them), while the left does the concept of “A and B” (trying to select both). The top of the vertical axis represents objectives, while the bottom does evaluations. According to Mr. Hamaguchi, the U.S. is in the first quadrant in the upper right, Japan is in the third quadrant diagonal with the first quadrant, China is in the second, and Europe is in the fourth. In the crosstalk between Prof. Ishiguro and Mr. Hamaguchi, they said that androids were categorized in the most ambiguous area on the coordinate plane, and that it was the reason why androids were popular in Japan. This explanation was very convincing to me. I hear that Hatsune Miku is so popular in Indonesia as well and that she held a concert there, making me feel that Southeast Asia is categorized in the same quadrant as Japan. I feel this type of effort launched in Japan will become popular in Southeast Asia.

Next, Mr. Tachikawa stated that “combining” was the keyword, and I was truly impressed with his idea. He said that combination generated innovation, and I was surprised to hear it from a designer. When economists consider value added to the iPhone, they disassemble the product, divide all the components into categories such as a processor made in the U.S., a monitor made by Sharp, and a camera made by Sony, and examine the product based on the fact that it is made by Hong Hai at its plant in China. In short, although the iPhone is manufactured in China, not all its components are manufactured in China, but almost all the components are imported. Then, what is the value added to the production of the iPhone in China? The answer is simply a labor cost equivalent to 5% of the product price. This is a typical analysis in terms of economics, but Mr. Tachikawa presented his own analysis of the product through a visual presentation. He indicated that the iPhone was an excellent result of the combination of a wide variety of components, and that its cutting-edge design was a visualized result of a new combination.

At the same time, Mr. Tachikawa also focused on “backing to the past,” which seems to be a concept opposite to a cutting-edge design. Referring to the example of Yamamotoyama, he spoke about the significance of going back to the basics in the past. He said: “The focus of design shaping and motifs is being shifted toward something universal and timeless. While seeking to enjoy new invention and technology, we have a longing for the past.” This was a truly profound explanation, suggesting that even if AI and robots become dominant in society, humans will keep their dignity as humans and cherish their past memories.

Mr. Ito spoke about Hatsune Miku, which I really had been looking forward to listening to him. He said that Hatsune Miku was very popular among teenage girls, more specifically girls aged 17 or under, and this phenomenon was found everywhere around the world. His explanation was truly convincing to me, because the explanation suggested that the popularity of Hatsune Miku was relevant not to the gap between developed countries and developing countries but to a generation gap or gender gap. In addition, I was truly surprised to learn about the license system regarding Hatsune Miku. As commented by Prof. Katayama, the system is similar to the one adopted by Prof. Yamanaka regarding iPS cells. As a result of the clearance of the copyrights, Hatsune Miku has become a huge global trend. I am very proud that as a representative of Japan, Mr. Ito has been working on such efforts.

At a keynote lecture today, Prof. Yamaguchi presented a truly logical explanation of interactive robots of Tencent. This was followed by an explanation on the misunderstanding of interrelationships and cause-and-effect relationships regarding the remarks “Those in their 40s living alone will destroy Japan,” made by AI in an NHK program. I think that Prof. Yamaguchi disseminated a truly significant message, indicating the danger of a conclusion spreading like wildfire in society without an explaining the process leading to the conclusion. Regarding what AI should be, I personally always have some questions, and I was impressed that his explanation covered such questions. As a result of deep learning using big data, although AI could not explain cause-and-effect relationships, AI found some interrelationship between those in their 40s living alone and Japanese economic trends. Consequently, AI proposed the hypothesis that “Those in their 40s living alone will destroy Japan.” Those who worship AI might believe in the hypothesis, and try to implement some measures to eliminate those in their 40s living alone, although this, of course, must not be done. In this regard, it is necessary to establish a system to control such reaction. My own understanding is that we should use the cause-and-effect relationships found by AI and robots, but we need to find how to use them from the perspective of humans.

Actually, in the field of economics as well, research conducted by young researchers today is becoming Greek to me. Prof. Iokibe called himself “a man from the Stone Age.” While reading the latest research papers on economics, I think that I am from the Bronze Age. In line with advances in computers, software programs have also developed, enabling us to obtain a result simply with entry of the necessary data. However, no explanation regarding the result itself is provided. In my early days after graduation from graduate school, I used to make econometric models. The past style econometric models focused on cause-and-effect relationships. For example, the production of a Malaysian economic model will reveal everything regarding how to move the country’s economy in a specified way. However, the recent econometric analysis in the field of economics has become the so-called “black box,” and seems to be following the path taken by AI and robots. I feel that it will remain truly important for humans to provide the necessary explanations.

President Yoshida also made some very interesting topics. For example, Spectee finds interesting news topics from among SNSs and sells them to newspaper publishers, TV broadcasting companies and so on. President Yoshida asserted that the time has gone in which no information can be collected without using SNSs. According to him, now that TV broadcasting companies themselves had begun to release some of their programs through the Internet, it was expected that the old-type, established media would face a lot of difficulties in the future, even though they strive to ensure prosperity together with new-type media. I feel that this discussion concerns how to treat fake news. Although this topic might seem to be irrelevant to robots and AI, a little more thought will tell you that both fake news, robots and AI share the same roots. After all, something unusual turns to news. Unlike the episode that I referred to earlier regarding the interrelationship found by AI, fake news is the result of finding something different from others. However, the final judgment should be made by humans, and humans need to confirm the genuineness of news by identifying the news sources. The key lies in that we remember this point.

Prof. Tsukamoto, who previously drew hot-button issues regarding ethics, said that none of the 20 forecasts that were made 15 years ago had been proven right. However, I feel that some of them have come true in different forms from what he expected. For a society that might be realized in 15 years, he referred to a “shift from PCs to wearable computers, and then to cyborgs,” in a response to Prof. Ishiguro’s comments on the “shift from PCs to personal robots.” I was surprised to hear Prof. Tsukamoto saying that he would be a cyborg within 15 years.

Prof. Tsukamoto also proposed that all the attendees of next year’s Awaji Conference should wear head-mounted displays (HMDs). Since Prof. Iokibe said that he liked the idea, I am now looking forward to seeing what will happen at the Awaji Conference next year. Prof. Tsukamoto also said that he was opposed to the idea of prohibiting the use of a smartphone while walking. I assume that he said so probably because present wearables might be an extension of walking while on the phone. I think that technology for handling this matter appropriately is also important.

A wide variety of perspectives are found in the relationship between culture and technology. Previously, Vice-Governor Kanazawa and Mr. Tachikawa engaged in discussion, and said that various things would run smoothly if there were no regulations imposed by the government. I totally agree with them. At this conference, I learned about the current situations of androids and Hatsune Miku, watched a hologram robot, and listened to the discussion of Prof. Ishiguro and Mr. Hamaguchi about that Japan and androids were categorized in the most ambiguous area of the fourth quadrant. These findings tell me the followings: the development of Hatsune Miku, androids, and holograms in Japan is based on Japanese culture; culture relates to technology; the connected technology further produces a different aspect of culture; and culture and technology are interdependent. In other words, culture and technology are closely connected with each other, and this sounds the alarm bell to discussion on technology and culture separately.

When explaining the purpose of the conference at the beginning of the event, I referred to the Fourth Industrial Revolution, and listed three of my concerns. One of them was about employment. I hear that the employment situation is rather tough in the U.S., and an estimate shows that it will be difficult to retain employment in certain types of occupations. Considering that AI and robots fulfill simple functions, however, I expect that employment will not be a very serious issue. When the Plaza Agreement was concluded in 1985 to plunge the yen from 240 yen to 120 yen, Japanese companies successively shifted their production functions from Japan to establish plants in China and Southeast Asia. The result was the hollowing out of Japanese industry. The loss of employment in Japan in those days was much worse than the one expected to be caused by AI and robots. Accordingly, I believe that there is no need to worry about employment very seriously.

The most successful AI is Watson developed by IBM. According to Prof. Yamaguchi, sales of the product have reached 1 trillion yen. However, no Japanese AI or robots sell as well as Watson, and I do not see that such business development will be possible in Japan. In the discussion on what needs to be done to achieve the development of robot and AI business in Japan like the case of Watson, language issues that Prof. Yamaguchi pointed out were a truly interesting. For example, one of the characteristics of the Japanese language is the absence of subjects. In addition, some Japanese sentences have confusing endings whether the sentences are positive or negative. With this background, it is expected that the strategy of forming big data and making AI engage in deep learning will not probably work. In this regard, I feel that to achieve commercial success in the field of robots and AI, we might need to eliminate ambiguity, although this relates to the Japanese culture. In addition, it is certainly necessary to educate human resources. I belong to the Faculty of Policy Studies, Doshisha University, and there has probably been no change in the courses provided by the faculty since it was established 14 years ago. Even if I think that students must know about the much-talked-about AI, there are no courses where students can learn about it. I believe that Japanese people must learn quick and flexible response to new issues. In the U.S., outdated faculties are soon terminated, and new faculties and courses are established promptly. Japanese people cannot act that way, and I feel that this is very serious.

At the forum, we had a talk about the followings: AI had not yet become almighty, but today’s fast speed of AI evolution suggested that an integrated-type AI might be developed; and a wide variety of attempts needed to be made before the development. I would like to connect this indication with suggestions from Mr. Tachikawa. I feel that one of the options is to combine individual excellent AI and robots. For example, it might be truly difficult to create robots and AI that can present policy recommendations. However, there are already very successful AI in the fields of go and shogi. If we begin to make efforts to combine such robots and AI from now on, I think that it is highly likely that Japan will become an AI superpower. If we do so, I feel that AI will be genuinely practical.

In the Asia-Pacific region, China is an advanced country in terms of the development of AI. While the country invests as much as 500 billion yen in research and development, Japan invests only 20 billion yen. Moreover, President Trump, who is has been sticking to employment reduction, is acting to boost up his employment policy. Believing that AI will deprive people of employment, he has considerably cut subsidies for the development of AI. Under these circumstances, while headhunting excellent human resources, China is trying to be the global AI center. This is also applicable to Korea, and Japan lags behind. Accordingly, Japan is now following China and some other countries in terms of the development of AI. However, I feel that if China becomes the only superpower in the development of AI and exerts dominant influence in the world, one of the strategies that Japan can take is to survive as an AI user country. In other words, it is a strategy of ensuring that all the citizens in Japan can use AI.

When explaining the purpose of the conference, I stated that AI and robots might solve the problem of a declining birthrate and aging population. In the previous discussion, Mr. Ito referred to the concept of “basic income.” Under the concept, it is believed that a production increase due to the introduction of robots might enable the government to provide all the citizens with an income high enough to make a living without working. If such a society is realized, everybody will belong to the leisure class, and can be dedicated to genuine creation in the field of painting, any other form of art, or whatever. Will such a Utopia-like society become reality? I sincerely hope that I can live until then.

I was truly stimulated by this year’s Awaji Conference. I have never attended such an exciting conference. I would like to conclude my summary by expressing my appreciation to all the participants, the MC, the simultaneous interpreters, the conference secretariat, the facilitators, and the coordinators. Thank you very much.