It Started with Managing Knowledge
My starting point was simple:
How should we handle organizational knowledge?
RAG.
Search.
Wiki.
FAQ.
Knowledge Bases.
For a long time, we believed that accumulating knowledge would make us smarter.
Organizations possess enormous amounts of knowledge.
Past documents.
Meeting records.
Manuals.
Design documents.
Emails.
Experiences.
Know-how.
With the rise of generative AI, searching, summarizing, and utilizing that knowledge became far easier than before.
Yet, there was one problem I repeatedly encountered in practice.
“So, what should we do?”
That was the question.
Information could be found.
Past cases could be retrieved.
AI could even make suggestions.
But what people really needed was not information itself.
What they truly needed was the ability to understand the situation and decide how to act.
What should be prioritized?
How much risk should be accepted?
Who should take responsibility?
When should we move forward?
When should we stop?
Knowledge can provide answers.
But it does not tell us what to do.
The problem was not a lack of knowledge.
The problem was how to generate action.
In other words, the issue was not Knowledge.
It was Decision.
I Wanted to Record Decisions
This naturally led to the birth of the Decision Trace Model (DTM).
Event ↓ Signal ↓ Decision ↓ Boundary ↓ Human ↓ Execution ↓ Log
What mattered was not what people knew.
What mattered was:
What situation were they facing?
What did they observe?
Why did they make that decision?
Who approved it?
What actions were taken?
What happened afterward?
And how would that experience influence future decisions?
What I wanted to record was not knowledge.
It was decision-making itself.
What Is a Decision?
As I continued pursuing Decision, I encountered a deeper question.
What exactly is a decision?
Why do humans make decisions?
Why do different people reach different conclusions from the same information?
Why does experience matter?
Why does the meaning of information change when goals change?
Gradually, I realized that decision is not merely selection.
Perhaps decision is the generation of meaning guided by purpose.
External state.
Internal state.
Goal.
Together, these create meaning.
And meaning gives rise to action.
Decision, then, is the process through which meaning is transformed into action.
That led to another set of questions.
What is meaning?
What is the difference between information and meaning?
Where do goals come from?
Naturally, I found myself moving into cognitive science, philosophy, and biology.
Terrence Deacon.
Meaning is not merely information; it emerges from absence, constraints, and purpose.
Francisco Varela.
Life is a self-producing system.
Autopoiesis.
The Embodied Mind, revised edition: Cognitive Science and Human Experience
Evan Thompson.
Mind and intelligence are not confined to the brain.
They emerge through the relationship between body and environment.
Daniel Dennett.
Living systems and minds can be understood through the Intentional Stance.
From Bacteria to Bach and Back: The Evolution of Minds
Paul Davies.
The essence of life lies not in matter but in information and causal organization.
John Maynard Smith.
Evolution can be understood as processes of information inheritance and selection.
Encountering Schrödinger
Eventually, that journey led me to Erwin Schrödinger’s What Is Life?
The question he asked was profoundly simple.
What is life?
Is life merely matter?
Why can living systems maintain order?
Why can they preserve themselves?
Why do they interact with their environment?
Schrödinger attempted to understand life not as a collection of molecules, but as an information system capable of maintaining order.
I Had Been Chasing Life
And then I realized something.
I thought I had been studying Decision.
But in reality, I had been studying Life.
Life itself perceives the world,
generates meaning,
acts,
learns from outcomes,
and maintains itself.
Stimulus ↓ Perception ↓ Meaning ↓ Action ↓ Learning
The resemblance to DTM was striking.
Event ↓ Signal ↓ Decision ↓ Execution ↓ Log
DTM had been created as a framework for organizations.
But looking back, I realized that it was, in some sense, an extension of life’s information processing into society.
Can AI Become Life?
And then an even larger question emerged.
Today’s AI is evolving as an enormous knowledge generator.
But life does not exist merely to generate knowledge.
Life possesses goals.
It creates meaning.
It makes decisions.
It maintains itself.
It co-evolves with its environment.
If so, what kind of AI should we ultimately aim for?
A gigantic knowledge base?
A giant brain?
Or something else—
A new kind of system that possesses goals,
creates meaning,
maintains itself,
learns,
and evolves through interaction with others.
Perhaps something closer to an artificial organism.
Toward “What Is Intelligence?”
Schrödinger asked:
“What is Life?”
Today, in the age of AI, we are beginning to ask another question.
“What is Intelligence?”
Does intelligence reside inside the brain?
Inside a model?
Or is it something that emerges through relationships—
between humans and AI,
organizations,
knowledge,
trust,
experience,
and interactions?
Looking back,
Decision Trace Model.
Knowledge Infrastructure.
Trust Infrastructure.
Multi-Agent Systems.
Coordination AI.
Intelligence Fields.
Runtime Society.
These were never separate theories.
They were all stages of a long journey toward a single question:
“What is Intelligence?”
I thought I was studying Decision.
But that journey ultimately led me back to Schrödinger’s question:
“What is Life?”
And now,
that question itself seems to be transforming into a new one:
“What is Intelligence?”
Chinoba — Runtime Society and Coordination Systems
Chinoba
Intelligence as Relationship
Research Platform
founded by
Masao Watanabe
AI Systems Architecture
Decision Trace
Human–AI Coordination
Algorithmic Governance
Related Research
This topic is part of the Chinoba Knowledge Base.
コメント