Overview

Outline of JSAI

Contents

• Inaugural Address of President
• Address of Editor-in-Chief
• Past presidents
• Board members (JP)
• Representatives (JP)
• Fellow
• Main Activities
• Congress documents (JP)
• Articles of Incorporation of the Japanese Society for Artificial Intelligence
• Information from JSAI
• The Japanese Society for Artificial Intelligence Anti-Harassment Policy

Inaugural Address of President

Now Is the Time for “ALL Japan” to Spark the Emergence of AI

Satoshi KURIHARA
(Keio University)

　A heavy baton of responsibility was passed to me from President Shusaku Tsumoto when I succeeded him as the next president of the society. It is both an honor and a serious responsibility to take part in the operation of the society at such a symbolic juncture—my appointment as the 20th president. There is a palpable sense of expectation that I must exceed the achievements of my esteemed predecessors, and I take that expectation seriously. At the same time, I am eager to work together with the Board of Directors and all members to invigorate the society by actively engaging with the broader public and help realize a uniquely Japanese vision of a society where people and AI coexist in harmony.
　Technology has traditionally improved through the accumulation of innovations. However, the success of ChatGPT was not so much the result of a groundbreaking new technology as it was a “qualitative shift” brought about by the scaling of data and computational resources. This event, I believe, represents a major turning point in the future development of AI. Unlike conventional IT technologies, which are designed by humans to operate exactly as intended, data-driven AI systems are built through large-scale learning, and what humans design is not the behavior itself but the mechanism that produces the desired behavior from massive datasets. Just as brewing fine sake requires not only a refined process but also high-quality rice—and it is the sake, not the rice, that the user ultimately enjoys—AI requires both an abundance of data, which serves as the rice, and large-scale computational infrastructure, which functions as the brewing facility, in order to produce a high-quality model—the sake in this analogy. Unfortunately, the reality is that building such large-scale models remains difficult in Japan. There is widespread recognition that, from the perspective of economic security, Japan needs to develop its own large-scale models domestically. So, how can we overcome this challenge? I believe the key lies in the qualitative transformation that emerges through scaling.
　Even if each element is small in scale, a qualitative transformation—what we call “emergence”—may arise when many of these elements work together in concert. Japanese society, moreover, has a remarkable openness to the idea of highly autonomous and versatile AI coexisting with people, as reflected in long-standing cultural icons like Doraemon and Astro Boy. This cultural receptiveness could become a unique strength for Japan in future AI research and development. In contrast, Western societies generally tend to be cautious about AI that is capable of thinking independently like a human. The fact that companies like Microsoft and OpenAI are seeking to establish R&D centers in Japan, and that sakana.ai chose to launch its operations here, can be seen as evidence of Japan’s appeal in this regard. If Japan is indeed well-positioned to serve as a testing ground for next-generation AI research and its implementation in society, it would be a great loss if Japan failed to seize this opportunity. It would be deeply unfortunate if new AI technologies were born on Japanese soil but with no Japanese researchers on the teams involved in creating them. That is why, even at this late stage, I believe it is essential that we establish a solid foundation for ambitious AI research and development here in Japan.
　To that end, our society plans to devote even greater efforts to educational activities related to AI. In just two years, we will mark the 40th anniversary of the society’s founding, a significant milestone. The so-called third wave of AI shows no sign of abating. In fact, with the advent of generative AI, it is gaining even more momentum. However, this surge has been driven almost entirely by just a handful of major AI companies based in the United States, and the world is, in effect, being steered by them alone. Now is the time for us to come together and take action to realize a uniquely Japanese vision of AI—one that sparks a powerful wave of emergence. As a catalyst for this effort, our society aims to play an active and vital role as well.
　In order to promote research and development with a sense of security, it is also important to establish guidelines for AI research, development, and utilization. While some claim that Japan’s approach to AI regulation is too lenient compared to regions like the EU, I take quite the opposite view. I believe that Japan is engaged in careful, deliberate discussions aimed at formulating well-thought-out guidelines that look not to hit the brakes like the EU, nor to blindly race toward the frontier like the United States, but to advance AI research and development in a shrewd and strategic manner that suits Japan’s position. Our society, for its part, established an ethics committee early on and has been consistently sharing its perspectives from the standpoint of those directly involved in building AI.
　The Japanese Society for Artificial Intelligence is currently enjoying strong momentum as a community, for which we are deeply grateful. However, if we remain passive, we risk losing that freshness over time. Growth can only be achieved by continuously moving forward while preserving diversity. During my two-year term, I hope to contribute as much as I can to keeping the society’s activities vibrant and dynamic, but of course, there is a limit to what one individual can accomplish alone. I hope we can move forward by valuing the networks we have with the Board, our members, and society at large. Moreover, I would like to see the society evolve beyond its traditional role as a community for those directly involved in AI research and development and become a broader platform for exchanging information, one that includes the general public and connects all those who engage with AI as it becomes increasingly integrated into society.

Address of Editor-in-Chief

The Transactions and Journal as Forums for Enhancing Our Understanding of Artificial Intelligence

Hirotaka OSAWA
(Keio University)

　It seems I have assumed the role of Editor-in-Chief at a particularly challenging time. Recently, much attention has been drawn to ChatGPT-4o, and it feels as though the implementation of artificial intelligence technologies in society has truly entered a practical phase. Of course, unlike the so-called “professional astonishment artists” who appear in the media, I have no intention of sensationalizing technologies as if they were brand new when, in reality, they have only come into public view after being widely adopted. Doing so would be disrespectful to those who have long and steadily advanced research in this field. That said, despite being familiar with technological trends, I continue to be genuinely impressed by the way these technologies are being quickly adopted by the public and by the speed at which people gain the ability to solve problems through collective information exchange as these tools become more widespread.
　From the perspective of research environments, we are likely entering a period of dramatic change. Until now, it was considered rational to share knowledge and make it openly accessible among humans in order to make the most of our limited cognitive resources. Approaches to problem-solving through collective intelligence, as well as the principles of open data, were based on such an environment. Going forward, however, it is expected that a certain level of “intelligence” will become widespread and commoditized. In such a context, the value of knowledge itself may rise in relative terms. If organizations can no longer differentiate themselves based on intelligence, they may be forced to compete through knowledge instead. In fact, multiple social media platforms have already begun imposing strong restrictions on search functionality and data usage.
　In this context, the question of what this society, specifically its Transactions and Journal, should aim for has become more important than ever. After all, we are now at a stage where people are seriously asking whether researchers are even necessary for conducting research. Personally, I believe that human-run academic societies will continue to play a vital role as forums for discussing artificial intelligence. More precisely, I feel that the level of resolution required in discussions about AI has risen to the point where it is now needed by a broader range of people. For example, questions such as “What exactly is ‘intelligence’ in artificial intelligence?” and “What should the field of AI fundamentally be researching?” have long been essential and foundational, yet were mostly treated as matters of research direction among specialists. While it was certainly possible, even desirable, for researchers to keep such long-term questions in mind, we have also lived through a more pastoral era when research could continue more or less smoothly, simply by treating immediate problems as optimization tasks without necessarily thinking about the broader picture.
　Today, however, discussions around questions such as “What is artificial intelligence?” and “Where should AI research be heading?” have become urgent issues that the broader public must also engage with. At the heart of these questions lies the issue of which parts of the cognitive processes we call “intelligence” can be replaced, and which ones cannot. Our research direction increasingly connects with an individual’s sense of purpose in life, and ultimately, with the survival of humanity itself. In this context, I believe that one of the essential ongoing roles of the Transactions and the Journal is to serve as forums that bring up such wide-ranging discussions about artificial intelligence not only within the community, but also beyond it.
　Public scrutiny of every move made in AI research is intense, yet we are far from having sufficient resources to properly manage that scrutiny. This truly is a difficult time.
　That said, I remain reasonably optimistic about the future.
　As a researcher of interaction—that is, of mutual interaction—I believe that the essence of intelligent processing in problem-solving lies in interaction itself. It is precisely “the discovery of problems through interaction” that constitutes a truly intelligent function. Once we can articulate what the problem is, the rest can largely be left to automated processes. The act of making the unknown understandable is what scholarship is, and I believe artificial intelligence is poised to play a central role in that domain as well. To be honest, I often feel like I have no idea what will happen next. Still, I hold onto the hope that by continuing to think—even vaguely—and by continuing to interact with those around me, the answers will eventually emerge. I believe that the Transactions and the Journal will continue to serve as important meeting points for such encounters, and the strong interest from outside the community is, in that sense, a real asset when it comes to enabling mutual engagement. At the very least, I can say with confidence that, over the past few years, the Editorial Committee has steadily developed an environment that enables such productive interactions to take place. This includes the paper review system for the Transactions that was built with great effort by former Editor-in-Chief Fujio Toriumi, as well as the structural separation of the editorial committees for the Transactions and the Journal. Also noteworthy are the many outreach activities to society that have taken place since the Journal shifted toward a broader audience in 2014. I hope to enjoy the “scenery that can only be seen now.”
　Starting this fiscal year, the JSAI has begun operating the Transactions and the Journal as separate entities. I believe this change will further increase opportunities for interaction. Together with Dr. Youichiro Miyake, Editor-in-Chief of the Journal, Associate Editors Dr. Yoshinobu Kano and Dr. Yukino Baba, as well as the editorial committee members, student editors, and the administrative staff, I hope to help revitalize discussions about AI research and contribute to enhancing our collective resolution in understanding artificial intelligence. With that, I look forward to your continued support throughout the year, and of course, we welcome your submissions.

The Journal as a Reflecting Telescope into the Future

Youichiro MIYAKE
(SQUARE ENIX CO., LTD.)

　My name is Miyake, and I recently took over from Dr. Fujio Toriumi as Editor-in-Chief of the JSAI Journal Artificial Intelligence. I look forward to working with you. Although I refer to it as the “Journal,” starting this fiscal year, the JSAI has adopted a new structure under which the Journal and the Transactions will be managed by separate editorial committees. Dr. Hirotaka Osawa has been appointed as Editor-in-Chief of the Transactions, and in close collaboration with Dr. Osawa, I hope to develop the Journal into a resource that is valuable for both the members of the society and the general public.
　I joined the JSAI in 2003 to give a student presentation at the annual conference held at Toki Messe in Niigata. At that time, the field of artificial intelligence in Japan was in a lull between the second AI boom of the 1980s and the third AI boom we are seeing today. The sky over Niigata was cloudy with a light rain falling, but inside the venue, the atmosphere was lively, despite its much smaller scale as compared to now. I have nothing but the deepest respect for our predecessors who carried on AI research in Japan, regardless of whether the field was experiencing a boom or not. In 2004, I began working in the game industry, and Artificial Intelligence became the only link that connected me to academia. Each time the Journal arrived, I reminded myself that digital game AI, my area of expertise, was still in its infancy. I kept thinking that once I had produced meaningful results, I would love to contribute. That wish came true in Volume 23, No. 1, published in 2008, and I have continued contributing ever since. The interactions that began because of those contributions have become truly invaluable to me. Since 2016, I have also had the honor of serving on the Editorial Committee under successive Editors-in-Chief Dr. Satoshi Kurihara, Dr. Hiroshi Yamakawa, Dr. Ryutaro Ichise, Dr. Yoji Kiyota, and Dr. Fujio Toriumi, with my main focus being the Journal’s cover design and book review section. The Journal and my involvement with the society have brought me into contact with many wonderful people, and it has been an essential part of my professional journey. I sincerely hope it will continue to serve as such a space for others as well.
　JSAI’s official journal, Artificial Intelligence, along with the annual conference and joint SIGs, serves as a vital point of connection among those trying to conduct research in artificial intelligence and those with an interest in the field. In its pages, readers can discover the work of a wide range of individuals and gain a sense of where their own efforts stand within the broader landscape, as well as appreciate their own unique contributions. Each page of the Journal reflects the dedication of the Editorial Committee and the authors and is written with the awareness that it will be delivered to and benefit the thousands of JSAI members. Moreover, Artificial Intelligence is available at bookstores across Japan, through online retailers, and also in digital formats. The Journal occupies a truly unique position as a bridge between the society and the broader public, a position rarely found among the journals of other academic societies. In today’s world, where artificial intelligence is in such high demand by society, it is thanks to the tireless efforts of successive Editors-in-Chief, editorial committee members, and contributors that Artificial Intelligence has risen to meet the current moment so effectively.
　Artificial intelligence is a curious field of study. Unlike mathematics or physics, it lacks a solid, well-defined structure, and unlike biology or medicine, it does not have clearly delineated objects of study. AI research is, at times, a science that dives into the deep ocean of intelligence and at other times an engineering discipline striving to build towering peaks of intelligent systems, and at still other times, a philosophical endeavor that seeks to expand the very ground on which such exploration is founded. The path of research is by no means smooth, and it is through researchers reporting from their own individual perspectives that a fuller picture of the development of artificial intelligence begins to emerge. Academic journals serve as a forum for collecting the voices of these ongoing activities in real time. However, in this era of rapid advancement, particularly in areas like deep learning, the pace of publishing once every two months cannot be considered fast. Moreover, due to editorial logistics, manuscripts must be prepared several months in advance. Against this backdrop, the Journal of the JSAI is confronted with the pressing question of what role it can play for its members. Looking through past issues of Artificial Intelligence, one can sense not only its efforts to capture the present moment but also its aspiration to lay the foundations for the future. While the Journal cannot match the speed of publication found in paper archives, social media, or online news sites, its mission lies elsewhere: to gather the vast and scattered body of information and deliver it in a form that is as quickly and systematically organized as possible.
　The Journal crystallizes knowledge from the past to the present. The more we polish that crystal, the more clearly the future shape of artificial intelligence begins to appear within it. In this sense, the Journal serves as a reflecting telescope pointed toward the future. I hope to make Artificial Intelligence a journal through which one can clearly see the future—a Journal that truly reflects what lies ahead. At the same time, the Journal is also a reflection of the identity of our society. I will do my utmost to ensure that members can feel proud to be part of the JSAI. I sincerely ask for your continued support in this endeavor.

Past presidents