IID.systems
ProfileServicesFormal MethodsAI AlignmentEssaysBookSchoolGitHub日本語
日本語

Essays

Long-form Reflections

Essays on AI, technology, society, and philosophy. Topics I consider important from a long-term perspective.

How Would an Aligned AI Redesign Criminal Justice?

2026.5

A thought experiment on what happens after alignment succeeds — and why that may be the harder problem. From the collapse of retributive logic to function-optimization justice, to the gradual hollowing out of human subjecthood that follows.

AGI Will Arrive Within 10 Years — And What Will Happen Then

2026.5

A prediction that AGI will arrive within ten years, the reasons for it, and an analysis of what happens after AGI. Examines the structural asymmetry created by intelligence differences, taking the chimpanzee's position as a reference point.

Book: Does Superintelligent AI Understand Morality?

When a Single Good Destroys the World (paperback & Kindle). Morality as environmental adaptation — and what a successfully aligned AI might destroy.