AI and Robot Law Virtual Workshop – The Moonshot’s Vision for Adaptable AI Robots

An online seminar on “AI and Robot Law Virtual Workshop - Privacy and Ethics Focus” was co-hosted by the National Yang Ming Chiao Tung University School of Law and The Frontier Research Institute for Interdisciplinary Sciences of Tohoku University (FRIS) on November 23, 2021. The seminar was a cross-disciplinary discussion of research results on robotics-related topics. One of the speakers was Yueh-Hsuan Weng, an assistant professor from FRIS, a newly established platform under Tohoku University’s Organization for Advanced Studies, with its mission to promote creative interdisciplinary research. Its final goal is to develop a new and promising research discipline across the legacy model fields. FRIS implements several measures to support interdisciplinary research, and one major measure is to assign researchers to work with other departments in the university.

The title of Professor Weng’s report was “The Moonshot’s Vision for Adaptable AI Robots.” His research focuses on moonshot projects and the concept of the next generation of adaptable AI robots. He introduced the program “Society 5.0.” The purpose of Society 5.0 is to establish a human-centered society—a society where humans and AI can co-exist. Society 5.0 is part of the Moonshot R&D program. Moonshot Goal 1 is the realization of a society in which human beings can be free from limitations of body, brain, space, and time by 2050. In addition, Moonshot Goal 3 is the realization of AI robots that autonomously learn, adapt to their environment, evolve in intelligence, and act alongside human beings by 2050. Co-existence and co-evolution of robots and human beings will create a vibrant society in which older people can participate. So, the vision of Society 5.0 in 2050 is that robots will be everywhere in our daily lives, and we will all be using them naturally. Their goal is to develop AI robots that can provide appropriate support and services depending on where they are used and the condition of the user. Professor Weng mentioned that many people are using shared bicycles to get around the city more easily, and when people go to a ski resort, they can enjoy skiing by renting skiing equipment. Therefore, robots will be shared and used everywhere in our daily lives in the future. In addition, their project team does not want to develop robots, their goal is to develop AI robots as partners which can provide appropriate support and services depending on where they are used and the condition of the user, so they would like to call such robots adaptable AI robots. This new type of robots will be used not only at home, but in many places as part of the infrastructure of society, such as commercial facilities, cultural facilities, sightseeing spots, sports facilities, nursing homes, hospitals, childcare facilities, etc. They want to create a smarter inclusive society where everyone lives a healthy life by coexisting with AI robots, cultivating a feeling of “I want to do or to be something.”

Professor Weng said that the project is a long-term research program. Their milestone for the year 2030 is to show that adaptable AI robots can be a driving force for creating a vibrant society in which everyone can participate enthusiastically not only in daily life but also in travel, sports, and other activities. If the milestone can be achieved, the adaptable AI robot will be able to be use in various fields. They aim to achieve this milestone through demonstrations in the field of nursing care, because Japan’s current super-aging society has a variety of related problems. For example, there were about 300,000 people who could not enter a special nursing home even if they wanted to in 2019. And it is reported that there will be a shortage of about 800,000 nursing care workers in 2035. In 2017, about 90,000 people left the workforce because of family care.

The moonshot project is a very huge project, so they have researchers from various backgrounds to participate in the interdisciplinary project, such as psychiatrists, safety engineers, and legal researchers. The front-line work is very interdisciplinary. To realize a new life with AI robots, they also set up three major research topics. The first one is human-robot coevolution AI, which is to design how robots should behave to provide appropriate services to humans. The second one is adaptable AI robots, which is to develop robot hardware to provide safe and appropriate assistance to humans. And the third one is social implementation of AI robots, which is a social implementation where AI and robot hardware mutually evolve, and AI robots can be used in everyday life. Professor Weng further explained the goal of the Moonshot R&D program. Moonshot Goal 1 is to develop an AI robot that encourages people to act on their own without providing excessive assistance, focusing on the concept of self-efficacy. In this project, self-efficacy is the expectation of how well people can perform the actions necessary to produce a certain result and the recognition of possibility of their own actions. If the self-efficacy is low, the person would hesitate to try new things and think that he or she cannot do it, or that he or she might cause trouble for others. On the other hand, if the self-effects are high, the person would try new things and think that he or she can do things by himself or herself. The end of the project is to find out a mechanism of the change in self-efficacy, feelings like “I may be able to do it” or “I will try it.” Therefore, the project team set appropriate goals for each task of improving self-efficacy and to design the behaviors of AI robots. In the end of his report, Professor Weng mentioned that in order for AI robots to be accepted by humans and society, it is important to obtain social consensus on how to control the robots from the perspective of ethics and law. Therefore, it is also necessary to establish safety evaluation standards for adaptable AI robots so that all people can use them.