Tokyo University of the Arts and Rikkyo Tackle Game AI Knowledge Gap at CEDEC 2026
The lecture lays out a concrete supervision model for teaching game AI, addressing an industry-wide shortage of experienced AI developers by establishing a repeatable curriculum at the graduate school level.
Reporting from 1 source: GameBusiness.jp.
At CEDEC 2026, Yoichiro Miyake of Tokyo University of the Arts and Shu Kuwabara of Rikkyo University discussed the lack of game AI education. Miyake cited few university labs for digital game AI research. Rikkyo's Graduate School of Artificial Intelligence and Science, established in April 2020, runs a Special Research system where students train in Unity or Unreal before tackling AI topics like meta-AI and procedural generation.
Game AI has grown so complex that passing the knowledge down has become a problem in itself. At CEDEC 2026, Yoichiro Miyake and Shu Kuwabara presented how Rikkyo University's Graduate School of Artificial Intelligence and Science approaches it.
Miyake noted that while many labs study Go and Shogi AI, very few research AI for digital games. He pointed to the scarcity of developers with hands-on AI experience, the undervalued history of digital game research, and the limited number of games with modifiable internals as the causes.
The graduate school, established in April 2020, runs a Special Research system. Students spend the first six months learning to build games with Unity or Unreal, then move into AI research. Weekly student-led seminars and monthly on-campus meetings support the work. Kuwabara's own master's research on reinforcement learning for flying a dragon in 3D space served as a concrete example of the model in action.
Synthesized by Yomimono from the 1 cited source below, including Japanese-language reporting where cited, then editorially reviewed before publishing.