2022.Jul.08

REPORTS

The 1st BAIRAL Research Meeting for 2022 Report on “How far Can VR Make People Empathize?: Its Limitations and How to Use It”

On March 27th, 2022, BAIRAL (a study group organized by research assistants of the B’AI Global Forum) held its first meeting for 2022 online. The guest speakers were Mr. Yuji Hatada and Mr. Ryu Kudoh from the Graduate School of Interdisciplinary Information Studies at the University of Tokyo. They introduced a diversity training program utilizing virtual reality (VR) that they are involved in developing. Based on the findings of their efforts, a discussion was held with participants about the uses and limitations of VR technology and the impact of the empathy caused by VR on people.

They are both participating in a project in which Associate Professor Arisa Ema (Institute for Future Initiatives, The University of Tokyo), who is a co-investigator of the B’AI, is also involved. Based on the idea that promoting empathy by using VR to simulate the experiences of others will raise awareness of unconscious bias and lead to appropriate workplace communication, the project is developing VR content that allows participants to experience the respective situations of a supervisor and a subordinate who is raising a child. The entire training program aims to consist of not only VR contents but also a set with workshops in order to link the awareness gained through VR contents to dialogue and discussion.

Mr. Hatada began by explaining the empathy that VR brings about, which is the underlying assumption for the development of this content. He explained that this content started from the question of whether avatars could be used to mitigate against unconscious bias and prejudice and that it is based on recent research which suggests that empathy through “perspective taking,” as in the field of psychology, can be achieved through “embodiment” using VR, feeling another person’s body as if it were your own, rather than just imagining it. On the other hand, he also stressed the limitations of this approach, such as how to handle the ambiguity of the definition of “empathy,” how to measure “prejudice” brought about by VR or unconscious bias, concerns that over-emphasizing the difficulties of childcare and making it too much of a personal matter may have rather undesirable consequences, and the issues to be aware of when implementing it in society.

Following this, Mr. Kudoh introduced the actual contents. In the office scene, the perspective of a supervisor can be experienced. Here, an urgent e-mail from a client arrives after the subordinate who is a working parent returning home, so the supervisor should send an e-mail to his/her subordinate requesting work. The home scene that follows is from the perspective of the subordinate, who receives an e-mail from the supervisor while having a meal with the child, in which the participants in VR can experience the situation faced by him/her being unable to concentrate on work because he/she is taking care of the child. He then presented the concept of a workshop that uses this VR content as a device to experience an example of child-rearing and working styles to lead to dialogue and discussion, as well as outlines for future development. It should be noted that users can choose from male or female avatars for both supervisor and subordinate.

After the presentations, a discussion was held with the other participants on various points, including the issues raised by the two speakers. Questions related to the significance and purpose of the content were asked, such as whether the more realistic the experience by AI become, the stronger the empathy it should bring, or whether the aim was to encourage people to change their behavior consciously, although it would be difficult to change the unconscious. There were also a number of ideas and opinions on future development and deployment, such as the possibility of changing the sense of urgency by using avatars of one’s own face, the possibility of having both supervisors and subordinates experience the same scene to enhance discussion, the possibility of sharing the background of the characters and giving avatars their own names like in a game to create a sense of familiarity, and the possibility that feeling one’s heart beat increase would lead to a sense of immersion and empathy. On the other hand, difficulties related to content development and effectiveness measurement were also raised, such as how to define empathy in the first place, how to indicate the difference in empathy and immersion between imagination and VR, and whether too frequent switching among multiple perspectives in the same scene could lead to confusion.

During the discussion, the limitations of avatar technology were shared openly. Although it is technically possible to share the physical sensation of the avatar, it is a challenge to make the users realize as they were in the same situation, so it is necessary to supplement this by presenting the background of the person or making the participants face events, etc. Also, the avatars’ shapes and the use of them can reproduce stereotypes and unconsciously send out messages, which is why the design of the avatars should be carefully considered. Given these limitations, the workshops are designed to cover areas that cannot be fully illustrated in VR. In addition, the idea was raised that it might be beneficial to use VR to experience hugging a baby, since it is now not possible to have a hands-on workshop on child-rearing under the COVID-19 crisis. In this meeting, it was learned once again that the key to better collaboration between people and technology is not to entrust everything to technology but to be aware of its limitations, especially when handling with things like empathy or unconscious bias, which are difficult to visualize and measure.