About AGI HUB — English
Communicating AGI — and AI safety — to Japan.
AGI HUB is a Japanese organization helping Japan prepare for transformative AI. We communicate the societal impact of artificial general intelligence — including AI safety, AI alignment, and existential risk — to the Japanese public, media, industry, and policymakers.
Why we exist
AI smarter than humans may emerge within years, and at latest within a decade or so. Within one or two years after that, it could progress into superintelligence, surpassing the collective intelligence of humanity. AGI could trigger an industrial explosion, lead to a technological explosion, and culminate in an irreversible intelligence explosion.
Whether that transition goes well depends in large part on AI alignment: the possibility that superintelligent systems diverge from human intent is discussed as an existential risk, up to and including human extinction. Japan’s decisions — in industry, policy, and research — should be made with this on the table. Most of the global conversation about AI safety happens in English; AGI HUB works to close that gap for Japanese society.
What we do
Most of our work is published and delivered in Japanese.
Lectures
Expert talks for companies, universities, and government institutions — from the basics of AGI to industrial and intelligence explosions, alignment, and what they mean for Japan.
Writing
Articles for media and corporate publications on AGI timelines, current developments, AI safety, and the open problems in alignment.
Joint research & policy
With companies and research institutions, we examine how AGI may affect industry, education, policy, and daily life, and turn that work into reports and recommendations.
Focus areas
- Timelines — when, and in what sequence, will AI beyond humans appear?
- Explosive growth — which fields will technological and industrial explosions transform, and how far will that change go?
- Alignment & safety — superintelligence diverging from human intent is discussed as an existential risk, including human extinction. What would it take to get this right?
- Post-AGI futures — what can we do to make AI alignment succeed, and what would a successful future look like?
Our explainers on these topics — the introductory guide, glossary, and research archive — are in Japanese.
People
bioshok — Co-representative
Studied semiconductors at university and natural language processing at the University of Tokyo Graduate School. Publishes extensively on AI impact, risk, and alignment. Author of Serious Risks Brought by AI and Their Historical Background, and contributor of the “AI Alignment” entry to the Japanese Society for Artificial Intelligence column “My Bookmark.” (X, note)
Yatima Kagurazaka — Co-representative
Physician. Co-founded and sold MinColle, then worked at the University of Tokyo before taking the current role. Led the Japanese translation of a leading work on evidence-based science and technology policy, published as Science of Science (Morikita Publishing). Organized “The Rise of AI Scientists” at the 2025 Japanese Society for Artificial Intelligence conference, one of its most popular sessions. (X, note)
Work with us
For lectures, writing, joint research, media inquiries, or collaboration with the international AI safety community, please reach out through our contact form. Inquiries in English are welcome.
Contact us