Message from the Dean

Welcome to the Graduate School of Informatics
Tatsuya Kawahara
Dean, Graduate School of Informatics

◆ Information as the Foundational Infrastructure of Society, Industry, and Science
  What do you think is the biggest difference in our society and lifestyle between now and your childhood? Most likely, it is the high availability of smartphones and social networking services (SNSs). While it is a truly revolutionary change to be able to communicate with anyone, any time, and anywhere during your waking hours, it all happened within the last 10 years or so. IT companies such as those whose services you use daily, like Google and LINE, are around 20 years old. I wonder how many people 10 or 20 years ago could have predicted where we would be today (anyone that did would now be wildly successful). I myself have been involved in the research of speech recognition and interaction for nearly 30 years, and I am excited that these technologies have found their way into many corners of society. It is safe to say that an advanced information-oriented society is here to stay, and it happened faster than anyone could have imagined.
  Information has triggered dramatic changes in the realms of industry and science, too. Some say that we are in the middle of the Fourth Industrial Revolution. All sorts of natural and social phenomena and production and distribution processes of goods have been digitalized (IoT), stored in large databases together with search, purchase, travel, and other human behavioral data (big data), and optimized by AI. Meanwhile, we are witnessing the advent of the fourth paradigm in science, where the methodology of pooling data to build a model is being applied across disciplines in fields as diverse as humanities and social sciences, such as economics and linguistics, as well as natural sciences, including medicine, pharmacology, engineering, and agriculture. There is no denying that information has thus come to form the foundation of industry and science.

◆ Kyoto University Graduate School of Informatics
  It was in the context of this cross-disciplinary spanning of information that the Kyoto University Graduate School of Informatics was established in 1998. Prior to that, there existed faculties and graduate schools that called themselves “computer science/engineering” or “information science” in Japan, but ours was the first graduate school to adopt the term “informatics.” Around the same time, other universities in Japan began to establish graduate schools focusing on information, but the Kyoto University Graduate School of Informatics distinguishes itself from others by the breadth of fields it covers (the study of information in a broad sense of the term). For example, we have a wide spectrum of courses and chairs on information and communications technology (ICT), some of which focus on computers and others on communication. Our courses and chairs on mathematical and data sciences, on the other hand, encompass wide-ranging fields from mathematics and statistical machine learning to physics and control theory. We even have courses and chairs on human intelligence, on top of those with a focus on AI. We also have cooperating chairs and courses that cover an even broader range of fields such as ecology and environment as well as healthcare and disaster management.
  What makes our graduate school even more special is that leading experts in all respective fields are in charge of these courses and chairs. When new professors are recruited, their achievements are read aloud in the faculty meeting, and I am always impressed by their outstanding qualifications. Our students also come from diverse backgrounds in terms of the undergraduate programs they attended, the universities/colleges from which they graduated, and the countries from which they hail, and they are well motivated to attend to their academic pursuits and research with passion.
  The Graduate School of Informatics has over 100 faculty members, making it the sixth largest among the 15 graduate schools at Kyoto University. Every year, we admit about 200 students for the Master’s program and about 60 students for the Doctoral program, a capacity that is more than double the size of the Kyoto University Undergraduate School of Informatics and Mathematical Sciences. Many international students are enrolled, too.

◆ To Prospective Students: What AI Can’t Do
  The Graduate School of Informatics opens its doors wide to prospective students who seek to learn “the study of information in the broad sense of the term.” This is evidenced by the fact that we accept both domestic and international students from a variety of backgrounds, unfettered by the conventional framework of sciences versus humanities.
  I hope that you will learn to do things that AI cannot accomplish. AI has already surpassed humans in terms of knowledge capacity and the ability to make inferences from large-scale data, with the result that what used to be regarded as intellectual professions are increasingly being replaced by AI. What then are the things that AI cannot do?
  First, it is the ability to identify a problem and formulate it. AI that we have now functions only when both the input and output are clearly defined. However, many of the problems in the natural world and in human societies are so complex that it is extremely difficult to clearly formulate them. For example, the problem to “develop a robot that is capable of having natural conversations as humans do,” could be reduced to questions such as how to generate natural backchannels and responsive actions, or to understand/express emotions. In other words, once you formulate your problem, you have only to collect data and leave the training and inference to AI.
  The second thing that AI is not good at is communicate. The ability to solve problems through dialogue is what current AI is lacking. As science and technology have advanced this much, most of the remaining tasks are hard to be resolved alone. For instance, it took cross-border collaboration among researchers from various fields to visualize a black hole. In many of the informatics research fields, code and data are shared within the international community of researchers, thus making rapid progress even as I write this. We are virtually “standing upon the shoulders of giants,” and we must have good communication within and outside of labs to continue to progress.
  This third thing that AI is incapable of is having a broad perspective. As I mentioned earlier, respecting diversity is crucial. Unfortunately, AI is blind to the deeper meanings that underlie data. For example, AI that has learned from data gathered in a male-dominated society will make decisions advantageous to males in recruitment and promotion. To find a good model and make a good decision, I would like you to take an interest in what is going on in the world and what persons around you are doing, rather than exclusively focusing inward on your own circumstances and activities. Coming from diverse backgrounds and conducting fascinating studies, the faculty and researchers at Kyoto University and its Graduate School of Informatics will surely stimulate your intellectual curiosity.

◆ For Potential Partners to the Graduate School of Informatics: What Private Enterprises Can’t Do
  The Graduate School of Informatics is most proactive in joint research with other institutions and industry-government-university collaboration. This is because, as I stated at the outset, information serves as the foundational infrastructure for society and industry, and “the study of information in the broad sense of the term” is interdisciplinary by nature. True to this belief, we have conducted numerous joint research and contract research projects and signed academic exchange agreements with universities outside of Japan.
  Needless to say, the Graduate School of Informatics is working on research projects that private enterprises find it difficult to do for one reason or another. Every year, we sponsor a joint seminar with overseas universities, and many of our counterparts say that the research themes our faculty members often choose are both “basic and long-term.” If you seek to collaborate with us, therefore, I would appreciate it very much if you would share our long-term perspective. We also promote open innovations and I would appreciate your understanding in this regard.

◆ For the Future
  As I mentioned before, information technology is innovating every second. Just as it was 10 or 20 years ago, it is extremely difficult to predict what the future will hold 10 or 20 years from now. It is even more difficult to predict what will come next after the fourth Industrial Revolution and fourth paradigm in science.
  Yet, one cannot help wonder if today’s world is better than how it was 20 years ago. Is our lifestyle truly more affluent now? Some of the major social problems, such as political division, economic disparity, and global warming, may be regarded as side effects of the advanced information-oriented society. Or it could be the case that more people suffer from stress, depression, and dependence because society as a whole has become busier and less generous and tolerant. I believe that it is the responsibility of informatics researchers to find clues to solutions to these various problems and shape a better society by taking these aspects into account. After all, this approach has much in common with Society 5.0, a vision for the future society which is being promoted in Japan, and with the Sustainable Development Goals (SDGs) proposed by the United Nations.