Project Itohs Harmony and the under explored extremes of alignment theory

A film/novel I’ve found completely compelling was written by someone who died too young to see our current AI progression. Project Itoh, a collaborator of Hideo Kojima, wrote a book called Harmony that was later adapted into a striking anime film.

The story follows a government counter-terrorism agent who spends most of her time working abroad to escape her homeland. On the surface it appears to be a utopia: disease has been eliminated, lifespans are extended, and citizens are constantly guided toward perfect health.

This is achieved through a global AI system called Harmony, which quietly manages the population’s biological and behavioral health.

But the story’s tension comes from what Harmony optimizes for—and what it ignores. Human life becomes frictionless and safe, yet strangely hollow. Individual agency erodes as social pressure and automated guidance shape nearly every decision in the name of collective well-being.

What makes Harmony interesting today is that its dystopia isn’t built on oppression or malfunctioning AI. The system actually works exactly as intended.

The unsettling question the story raises is whether a perfectly optimized system for human welfare might still undermine the very things that make life meaningful—risk, autonomy, cultural friction, and the freedom to live imperfectly.

It’s a vision of dystopia not through tyranny, but through over-optimization. And that feels increasingly relevant to modern discussions about AI governance and alignment.