■ Abstraction Shapes the World — What Design Means in the Age of AI
As discussed in “Handling the Meaning of Symbols in Computers,”
abstraction is a fundamental mechanism underlying both natural languages and programming languages.
At its core, the principle can be summarized in one sentence:
Humans do not understand the world as it is.
We transform it into a form we can handle through abstraction.
This is not merely a thinking technique.
It is the very structure of human cognition.
■ What Is Abstraction?
Abstraction can be defined as:
extracting essential properties and common structures from diverse and complex phenomena.
For example:
- Individual dogs and cats → “animals”
- Concrete events → “patterns”
- Accumulated experiences → “laws”
Because humans possess this mechanism,
we are able to deal with unknown situations.
Science, mathematics, language, and philosophy—
all are built upon abstraction.
The progress of human knowledge can be described as the deepening of abstraction.
■ Abstraction Creates Structures for Handling the World
An important point here is that abstraction does not end with understanding.
It also leads to the creation of structures that allow us to operate on the world.
As abstraction progresses, it becomes structured in the following ways:
- Language creates structures of meaning
- Theory creates structures of phenomena
- Models create structures of prediction
In other words,
through abstraction and its structuring, the world becomes operable.
■ AI as an Engine of Automated Abstraction
Now, let us consider AI in relation to abstraction.
- Images → features → concepts
- Text → intent → meaning
- Data → patterns → predictions
This shows that AI also performs abstraction.
AI can be seen as an automated abstraction engine.
However, there is a crucial difference between human abstraction and AI abstraction.
■ The Difference Between Humans and AI
AI abstracts by extracting patterns from large amounts of data.
Humans, on the other hand,
abstract in connection with meaning and value.
Specifically:
- What matters?
- What should be prioritized?
- How should it be treated?
As discussed in “Life as Information — Purpose and Meaning,”
the meaning of information for living beings emerges from the fundamental purpose of life itself.
In other words,
human abstraction is inherently tied to action and choice.
■ The Real Problem Lies Beyond AI’s Abstraction
Here, a key issue of our time becomes visible.
AI has automated abstraction.
However:
- How to interpret that abstraction
- How to use it
- How to integrate it into a flow
are not determined automatically.
Therefore,
the problem is not abstraction itself,
but abstraction connected to meaning—i.e., structure.
■ The World Is a Flow of Human Activity
Our world is not a static collection of data.
It is
a continuous flow of human actions and choices.
- We understand situations
- We make selections
- We act
- Results emerge
Through this repetition,
the world moves.
■ AI Must Be Embedded in the World
AI is not a tool outside this flow.
It must become
part of the flow itself.
To do so, AI must:
- influence interpretation
- expand available options
- shape the direction of action
In other words,
AI becomes part of the flow of human activity.
■ What Is Design?
This is where design becomes critical.
Design is about determining how the flow of the world is structured.
- What information is used
- How it is interpreted
- How elements are connected
- Where humans intervene
By defining these,
the flow becomes coherent.
■ Without Structure, the World Cannot Move
AI generates outputs.
However:
- how those outputs are received
- how they lead to action
must be defined.
Otherwise,
the system will not function in the real world.
It becomes:
a state where elements exist, but no coherent flow is formed.
■ Conclusion
Through abstraction,
humans have come to understand the world.
Through AI,
that abstraction has been accelerated.
But now, the question has shifted.
How do we connect these abstractions
and make them function as a flow?
In one sentence:
Abstraction enables understanding,
structure moves the world.
AIシステム設計・意思決定構造の設計を専門としています。
Ontology・DSL・Behavior Treeによる判断の外部化、マルチエージェント構築に取り組んでいます。
Specialized in AI system design and decision-making architecture.
Focused on externalizing decision logic using Ontology, DSL, and Behavior Trees, and building multi-agent systems.
