A Proposed Session (3B1-KS-13) at the 2025 annual conference of the Japanese Society for Artificial Intelligence (JSAI 2025), held on Thursday, 29 May 2025, from 9:00 to 10:40, in Room B (Small Hall) of the Osaka International Convention Center (Grand Cube Osaka) and online. It was organized by Hiroshi Yamakawa (University of Tokyo) and Koichi Takahashi (Chair, ALIGN).
AGI (artificial general intelligence) refers to general-purpose AI with intelligence equal to or greater than that of humans, able to handle a wide range of tasks flexibly; some expect it to arrive before 2030. AGI could greatly advance science, the economy and other fields, but there are also concerns about risks such as behavior that departs from its designers’ intent and unexpected disruption caused by its huge demand for electricity. The session set out the issues to be addressed for a hopeful future in a society with AGI and discussed measures for safety and alignment from several angles: the foundations of AI research, AI alignment (including the role of brain-inspired AI), energy, and the redistribution of wealth through a basic income from a macroeconomic perspective.
Program:
- Opening, and an overview of AGI and introduction of ALIGN: Koichi Takahashi
- AI alignment, Post-Singularity Symbiosis, the NAIA vision and the role of brain-inspired AI: Hiroshi Yamakawa (University of Tokyo)
- Autonomy and spontaneity of AI: Yusuke Hayashi (Director, ALIGN)
- The electricity problem of AGI: Ryuichi Maruyama (freelance)
- Talk by Tomohiro Inoue (Komazawa University)
- Panel discussion with all speakers, and closing by Koichi Takahashi