EN, essay

Challenges and Prospects for Future Artificial Intelligence as Seen from the 2nd French-German-Japanese AI Symposium

French-German-Japanese Symposium
For five days from November 16 to 20, 2020, the Embassy of France to Japan and the German Centre for Research and Innovation Tokyo (DWIH Tokyo) co-hosted the Human-Centric Artificial Intelligence: 2nd French-German-Japanese Symposium with AI Japan R&D Network as a co-organizer. The first symposium was held in November 2018, during which I chaired the session on Ethical &Legal Aspects of AI; since then, I have been involved in its preparations as a member of the planning committee for the second symposium.

The event was originally supposed to be held at the National Museum of Emerging Science and Innovation (Miraikan), but COVID-19 forced a sudden change of venue to a virtual space called Laval Virtual. The reason why we decided to hold the event in a virtual space was because its purpose was to emphasize networking. Creating your own avatar, walking around the virtual space, and talking to people was certainly a valuable experience that could not be provided in a regular online conference. In the end, more than 1,000 participants from all over the world, including 150 staff and speakers combined, attended the event and experienced a novel way of holding a networking event in the COVID-19 era.

Looking back on the event after its conclusion, I feel that the discussion around AI has significantly changed compared to the first symposium in 2018. Through this event, I would like to reflect on the changes in themes surrounding AI over the past few years and the future direction of the field.

Changes in themes surrounding AI
AI is now beginning to be used in various areas such as automated driving, medical care, and recruitment. On the other hand, issues of fairness and safety are also being raised. Various issues have been pointed out in recent years, such as accidents caused by self-driving cars, false arrests due to misjudgments by facial recognition technology, and election interventions by deepfakes and fake news. These are no longer problems that can be solved by AI technology alone, but they require the creation of frameworks in different fields and industries from a variety of perspectives, including law, ethics, society, policy, and economics. In fact, in addition to the AI strategies of national governments, discussions on the development and utilization of AI are taking place in international organizations such as the OECD, WHO, UNESCO, and Interpol as well as in international networks such as the World Economic Forum.

Changes in topical themes have also occurred over the past two years. Table 1 shows the session titles of the first and second symposia, which were carefully selected to address several issues related to AI that could be shared among the three countries; 2018 focused on broad themes such as environment, mobility, industry, and safety, while some themes such as healthcare and education continued in 2020. However, as can be seen from the fact that many of them were in the form of “AI &” in 2020, considerable progress has been made in the discussion of each individual domain. As mentioned above, organizations such as WHO and Interpol are also working on best practice collections and guidelines.

Table 1  Themes discussed in Japan, Germany and France

In addition, the 2020 event did not include a session on “ethics,” which was covered in the “Ethical and Legal aspects of AI” session that I chaired in 2018. The reason for this is that the discussion of ethics related to AI has been included in all issues, and it is impossible to discuss ethical aspects alone. The fact that only “law” was singled out this time indicates an urgent need to develop laws related to AI in each country.

In addition, “human-centric” and “trustworthy” AI have coalesced over the past two years as concepts that can be shared between Japan, Germany, and France. The latter was also used as the title of a report released by the European Commission in March 2019, and human-centric was used in “Social Principles of Human-Centric AI” released by the Japanese Cabinet Office at the same time. These reports from Japan and Europe have many items in common in terms of policies on AI development and utilization. While the U.S. has seen initiatives by GAFA and other giant IT companies, and China has held strong national initiatives to develop AI, Japan and Europe are trying to steer the direction of AI development and utilization with soft laws and guidelines such as personal information protection and AI governance.

Directions that have changed over the past two years
At the first symposium held in 2018, Prof. Yuichiro Anzai of the Japan Society for the Promotion of Science (JSPS) proposed a final joint statement. It included the words “human-centric approach,” and in line with this direction, the title of the second symposium was “Human-centric AI.”

JOINT STATEMENT OF THE GERMAN, JAPANESE AND FRENCH PARTICIPANTS ON INTENSIFIED COLLABORATION IN AI

On the occasion of the first Japanese French and German Symposium on Artificial Intelligence, organised by the DWIH Tokyo and the Embassy of France in Japan, the participants from the fields of research, policy, and funding organisations express their willingness to intensify their future collaboration in AI research and innovation. The cooperation will be carried out in the spirit of shared ethical values for the commongood of our societies. At the centre of this collaboration we put a human-centred approach which will set common standards and a joint understanding of the potential of Artificial Intelligence.

We strongly support the creation and reinforcement of networks of individuals as well as networks of institutions. Based on the respective national strategies of Artificial Intelligence we see the need and the scope for intensified exchange of researchers, ideas and perspectives to face challenges in areas such as health care, mobility, environment, connected industries, or disaster risk reduction. The ultimate aim of Artificial Intelligence is to serve people and contribute to the improvement of the quality of life for the individual as well as for society as a whole.

Tokyo, November 22, 2018

During the second event, we had several meetings in the virtual space as well as by email and online to determine whether we could make a joint statement and hold a third symposium in 2022. In two years, the world should hopefully be a post-COVID-19 world. Then, we wonder what could be on the agenda for AI in the future; is it appropriate to treat individual fields such as healthcare and the environment in the same breath under the theme of AI? Is it desirable to limit the discussion to three countries even two years from now?

Discussions were held from a variety of perspectives, but the final consensus was to move forward from the 2018 theme of “human-centric” and to build a society that pays attention not only to people and machines but also to people, machines, and the environment that surrounds them. At present, global warming, pandemics such as that of COVID-19, and global fragmentation are social issues, and the question of how AI can provide solutions to these issues is one that the three advanced AI countries have in common as a proposal for the next two years.

The following joint statement was read out by Prof. Junichi Tsujii, Director of the Artificial Intelligence Research Center at the National Institute of Advanced Industrial Science and Technology (AIST), on the last day of the conference, and it was applauded by the participants:

Joint statement of French, German and Japanese participants on AI collaboration aimed at tackling planetary-scale problems in the Anthropocene

The 2nd French-German-Japanese Symposium, organised by the Embassy of France to Japan, DWIH Tokyo, and AI Japan R&D Network, brought together stakeholders from academia, industry, government and civil society from the three participating countries in virtual space to discuss the subject of “Human-centric Artificial Intelligence”. This human-centric approach was the central message of the first joint statement of 2018, and as a shared value of the three countries, the importance of human-AI collaboration was discussed in this event with respect to various topics such as health, agriculture, risk prevention, education, and democracy. Humanity is currently facing planetary-scale challenges, such as the COVID-19 pandemic, climate change, and community fragmentation.

In preparation for our third symposium in 2022, we will address the issues we face in the Anthropocene, the epoch in which humanity is having a profound impact on the planet, including its ecosystems, from a broader perspective that includes not only humans and AI technologies, but also the environment we live in. The three countries of France, Germany, and Japan, which share the same values and social challenges, will be at the core of these efforts, and will invite other regions and countries to exchange ideas on how AI can help to solve these challenges, not for a single nation or company alone, but for the benefit of all of humanity and our planet itself.

Laval Virtual World, November 20, 2020

Issues and perspectives to be addressed in the future
The society, technology, and environment surrounding us will continue to change at a dizzying pace. However, it is precisely because we live in such an era that human connections are important, and we must continue making efforts to form and maintain networks. Currently, we are at the stage where individual countries or international organizations are developing principles, toolkits, education, and research on the use of AI.

However, we should not forget that these discussions are being led by the U.S., China, Europe as well as by other advanced AI countries. In African countries and Southeast Asia, AI technologies from Europe, the U.S., and China are being exported and are penetrating the region first—a phenomenon known as leapfrogging. In many of these countries, the frameworks for personal information protection and quality assurance are not yet in place, and it is considered important to formulate discussions involving the Global South on the issues raised by AI technology. Thinking on a global or planetary scale will undoubtedly include these issues.

In 2022, Japan will be the host country of this event. In order to discuss with the speakers and participants, who will be even more powerful, we must think about the relationship between AI technology and society while promoting cooperation with other countries and organizations during the next two years.