Hello, everyone. My name is Hiromu Yakura. I started programming when I was in the seventh grade. I've worked as a programmer since then. I have some experience in attending the defense readiness condition, which Mr. Shinoda talked about in the second session. As a programmer, I am often asked, "What is your favorite subject?" It is a very common question. They often say, "You like math, don't you?" Actually, I like social studies such as ethics, politics and economics. I think very few people can't visualize clearly what these studies are, especially ethics. Put briefly, I like "history of ideas." What is "history of ideas"? It teaches us the history of our thoughts such as "what makes us human?", or "what makes the world?" By learning what human beings have thought in the past, we discover the rules of the human beings and the world. We learn how to to relativize the relation between humans and the world. That's what the history of ideas is about. Thales of Miletus said: "Water is the primary principle of nature." What he means exactly is that the world is made of water. According to Heraclitus of Ephesus, "Everything is constantly changing." You see, many people have thought many different things about the world and humanity. I've studied such history. I was especially attracted by the idea of Hegel, a German philosopher. His thought covered various fields from human spirit to patriotism, history or esthetics. I like his systematic modelization of different things. To modelize is to organize various things related by logical coherence, For instance, Newton elicited the "law of gravitation" by the fact that when you release something, it falls down on the earth, and that the earth goes around the sun. He made up a logical modeling. Actually the procedure of modelization is quite similar to programming. Take a car navigation system as an example. you car is most probably equipped with a navigation system. it finds your easiest and best route What makes this possible is the modelization of actual roads and traffic jams, in other word, a mathematical mechanism makes this possible. Computers will find routes successfully. Hegel has developed a modelization method called "dialectic". Let me explain very briefly. If there is a "thesis" and its "antithesis", the model doesn't believe any, nor criticize them. It reviews the contradiction and makes adjustments. It extracts more essential metaphysics. That is the dialectic. One idea occurred to me when I learn this Hegel way of thinking What happens if we apply the concept of Hegel to this era? We live in the world, in which computers will take the place of human. So I wonder what could happen. Let's take an example. This is a past problem in the entrance exam of an university. If we solve it by ourselves, we calculate on a paper like this and finally figure out the solution. But if you access to "wolframalpa.com" and enter the formula, you will receive the solution instantly. In addition, it provides a graph. No wonder you don't feel motivated to study. Another example is Larry Page, Google CEO. He said, "Robots and computers will replace our jobs due to rapid AI development." According to a thesis by the University of Oxford, robots will replace 47% of American jobs in 20 years. In the case of data input, the probability rises to 99%. For referee and insurance certification: 98%. Robots will replace all these people. I get back on track. Owing to the AI, machine learning or robot engineering, technology potential is dramatically spreading. What should we do under such circumstances? Should we be pessimistic about no meaning of existence or shall we be free from work alienation? None of them are correct according to me. I have a new way of thinking. So we need to compare the differences between human beings and computers. This copy android of Dr. Hiroshi Ishigiuro was made by him, the czar of humanoid robot research. He explained the significance to make this type of android that he can review "what human are" by spending time with close android. Another example is the theoretical experiment of virtual musicians. The experiment has the objective to test the assumption whether we can be moved by music composed by computers. What do you think of that? We are moved by music that we acknowledge has been composed by computers, it means that we are moved by the content of the music itself. But our very knowledge that the music was composed by a machine could prevent us from being moved by it. In this case, it could mean that we are moved by the existence of a human whose feelings are translated in the music itself. We are moved when Usain Bolt made the world record in the Olympic Games. But we are not moved when cars run at the same speed. We could say it's the same phenomena. However, we may be moved when we learn about the tremendous efforts and difficulties overcome by the computers to compose music. In this case, we could say that we are moved by the story of the composing machine. What is the meaning of comparison between human beings and computers? I think that we try to move our hearts with computers. Now I want to introduce one of my actual research. We have developed a program designed to find the theme of any music. Do you know what theme of music is? Briefly speaking, it is melodic climax. Generally most musics are composed by combination and repetition of "A melody", "B melody" and "theme". So how does the program find the theme? I first thought that finding the repetition would make it. The program finds out the repetitions. Based on the aforementioned rules, it detects melody A and melody B and infers that it could be the theme. That is how conventional programs are working. But this method has reached its limit. A factor that made it reach its limit is "Vocaloid", a software to compose music. It doesn't matter if you are a professional composer or not, this software makes composition easy, even for amateurs. The result is that much unique music is created. So many musics don't follow the rules. And most of them are posted on "Niconico", video sharing website. It means that there is no limit to the length of the music. We can compose various kinds of music from very short to over 10 minutes. As a result, I can't use the method I had been using until now. So I created a new method to detect themes more accurately by analyzing the comments posted with the musics on Niconico. Put simply, I extract the major elements from the comments, and look for the theme parts that are related to the comments. That's the idea. I attended a science and technology convention for high school students and received a "Minister award for science and technology policy" for it. It is a mouthful award. I could also attend an Intel convention as a representative of Japan, where I received the "award of Fondazione Bruno Kessler." That was the technical part of my talk. I would like to talk about the emotions I felt during the process. It's about the boundary between what human beings can do and what machine can do. Combining the comments to the music is grounded in the idea to link the music to a human being. Even when you combine it to the comments, it is not always possible to detect the theme. But usually, we find it. So for me, the theme opacity is filled with humanity. We are developing technologies more and more in our time. It is my conviction that there are things that only humans are able to realize under the well-developed technology society. I think that we find the answer to our survival in the very boundary between humans and machines. A philosopher said: "Alienation of labor causes the alienation of human by human," He thought "working" is what makes man worthy of the name. But I think "creativity" is what makes man worthy of the name. Thank you very much. (Applause)