Intelligence Is Knowledge

Introduction

How do you perceive "intelligence" and "knowledge"? Many people consider "intelligence" and "knowledge" to be different concepts. In particular, people described as having "natural smarts" or "high thinking ability" are often recognized as possessing special thinking abilities or talents beyond merely having extensive knowledge. This perception is deeply rooted in educational systems and social evaluation mechanisms, reinforced by assertions such as "mere memorization is meaningless" and "thinking ability is important."

However, I believe that "intelligence is knowledge." In other words, high intelligence simply means having abundant knowledge and a robust network that can efficiently utilize it. In the human brain, knowledge isn't merely a static collection of information but exists as a dynamic network of interconnected elements. The richer and more structured this network is, the more likely we are to be evaluated as having "high intelligence."

Furthermore, "thinking ability" or "thinking" itself may not exist independently but could be merely the brain's post-hoc rationalization. The process we feel as "thinking" might actually be an automatic response from our existing knowledge network and a subsequent interpretation of the result.

In this essay, I will examine the "intelligence = knowledge" perspective from multiple angles, considering insights from modern neuroscience and cognitive science to argue why this view is rational. In particular, I will analyze in detail the differences between explicit and tacit knowledge, the relationship between simple knowledge and abstraction ability, and the essence of the process called "thinking," presenting a new perspective on the conventional view of "intelligence."

1. Common Misconceptions About Intelligence

First, let's organize the thoughts of people who consider "intelligence" and "knowledge" as separate entities. Many common misconceptions include:

• "Intelligence" is an innate cognitive ability
  Many people consider intelligence to be an innate ability, using expressions like "naturally smart." This is based on the incorrect premise that intelligence is fixed and unchangeable.

• "Knowledge" is a collection of individual pieces of information acquired through study and memorization
  The common view is that knowledge is merely a collection of facts and information acquired through memorization. However, this overlooks the essential structure of knowledge.

• "Natural intelligence" or "thinking ability" is a problem-solving skill independent of the amount of knowledge
  Problem-solving ability and adaptability to new situations are often considered "innate abilities" separate from knowledge, but these are merely forms of knowledge utilization.

• "Tacit knowledge" is fundamentally different from "explicit knowledge"
  Skills and intuitions gained from experience are considered completely different from verbalized knowledge. However, these differences are merely superficial.

While these ideas may seem reasonable at first glance, they arise from not correctly understanding the essence of knowledge and its utilization. Knowledge is not simply "memorized facts" but a network of information that dynamically interacts within the brain. And the quality and quantity of this network constitute what we call "intelligence."

To unravel these misconceptions, let's analyze in detail the concepts of explicit and tacit knowledge, simple knowledge and abstraction ability, and the process called "thinking."

2. The Difference Between Explicit and Tacit Knowledge Is Merely Verbalization

(1) Common Definitions of Explicit and Tacit Knowledge

Explicit and tacit knowledge are generally distinguished as follows:

• Explicit Knowledge

  • Knowledge that can be expressed in language or mathematical formulas.
  • Clearly defined and easily shared with others.
  • Can be described in textbooks, papers, manuals, etc.
  • Examples: mathematical formulas, historical facts, scientific theories, programming language syntax.

• Tacit Knowledge

  • Knowledge that is difficult to verbalize, based on experience and intuition.
  • Rooted in individual acquisition and sensation.
  • Expressed as "skills" or "knacks."
  • Examples: craftsmanship, physical sensations in sports, understanding conversational nuances, artistic sense.

This dichotomy was proposed by Michael Polanyi and popularized in the context of management studies by Ikujiro Nonaka and others. In many cases, tacit knowledge is considered more advanced and valuable, obtainable only through special talent or years of experience.

(2) The Difference Between Explicit and Tacit Knowledge Is Not Fundamental

At first glance, explicit and tacit knowledge may seem to have completely different properties, but their difference is merely a matter of "whether verbalization is possible." This is because both are represented similarly as neural network patterns in the brain.

From a neuroscientific perspective, knowledge is stored as synaptic connection strengths and neuronal network patterns. This mechanism is the same for both verbalizable and non-verbalizable knowledge. If there are differences, they lie only in the strength of connections to language areas and whether we can consciously access them.

Tacit knowledge appears special because it cannot be easily expressed in language or logic, not because the knowledge itself has a special structure. For example, when a professional shogi player intuitively feels a move is "bad," it's merely the unconscious operation of pattern recognition gained from years of playing experience. Even if they cannot explain the reasoning verbally, in their brain, it is networked as "experiential knowledge."

(3) Language Causes Misconceptions About Knowledge Characteristics

The reason that non-verbalizable information is considered "special" is because humans "perceive the world primarily through language." However, craftsmanship techniques and sports movements can be formalized as data using sensors and motion capture. Indeed, in recent years, there have been increasing cases of AI learning and reproducing the skills of craftsmen.

For instance, a skilled chef's "sense of proportion" can be measured as subtle changes in weight and temperature, and a professional pianist's "expressiveness" can be recorded as patterns of key pressure and timing. These are simply knowledge that is "difficult to verbalize at present" and not fundamentally different from explicit knowledge.

(4) The Substance of Tacit Knowledge Is Complex Pattern Recognition

Much of tacit knowledge is actually the result of complex pattern recognition. For example, a doctor can instantly diagnose a patient's symptoms because they have accumulated vast pattern information from years of clinical experience in their brain. This process is difficult to verbalize and is therefore expressed as "intuition" or "instinct," but in reality, it is merely the instantaneous activation of a highly structured knowledge network.

The development of AI supports this idea. AI using deep learning can acquire judgment capabilities equivalent to "tacit knowledge" without explicit verbal instructions, by recognizing patterns in data. This suggests that tacit knowledge is not a special ability but arises from sufficient data and pattern recognition.

3. The Relationship Between Simple Knowledge and Abstracted Knowledge

(1) The Difference Between Simple Knowledge and Abstracted Knowledge

Knowledge includes "simple knowledge," which consists of concrete facts and information, and "abstracted knowledge," which integrates and generalizes these.

• Simple Knowledge

  • Individual facts and information.
  • Concrete and highly context-dependent.
  • Examples: "Tokyo's population is about 14 million," "Water boils at 100°C"

• Abstracted Knowledge

  • Patterns or principles derived from multiple simple knowledge elements.
  • Highly generalized and applicable to different contexts.
  • Examples: "Urbanization increases population density," "Substances have boiling points"

Many people believe that the ability to manipulate abstracted knowledge indicates "high intelligence." However, this abstraction process is based on the accumulation of simple knowledge and pattern recognition between them.

(2) Abstraction Emerges from Knowledge Accumulation

Developmental psychology research shows that abstract thinking becomes possible after experiencing a sufficient number of concrete examples. Children first learn specific cases one by one. Later, as similar cases accumulate, they can identify commonalities and generalize or abstract.

For example, after experiencing numerous individual calculation examples like 1+1=2, 2+2=4, 3+3=6, they can discover the abstract rule that "adding the same numbers together results in twice that number." This demonstrates that abstraction naturally emerges from the accumulation of simple knowledge.

(3) Abstraction Ability Is Also Knowledge-Dependent

People evaluated as having high abstraction ability simply have more knowledge and can identify connections between them. For instance, outstanding scientists identify commonalities between seemingly unrelated phenomena and construct new theories. This is possible precisely because they possess extensive knowledge in their field, not because they have a special independent ability called "abstraction ability."

In fact, even people who excel at abstract thinking in one field struggle with abstraction in fields where they lack knowledge. A mathematician cannot make abstract considerations about literary theory because they lack knowledge in that field. This is evidence that abstraction ability is knowledge-dependent.

(4) The Relationship Between Abstraction and Knowledge Networks

Abstraction is the process of creating new connections within knowledge networks. The more individual knowledge nodes (simple knowledge) increase, the more potential patterns connecting them also increase. These patterns are then recognized as "abstract knowledge."

The connectionist model, a cognitive science theory, supports this view. In this theory, knowledge is represented as an interconnected network, and learning is the process of creating new connections and strengthening existing ones. Abstraction is nothing more than discovering new patterns within this network.

4. Thinking Does Not Exist: The Brain Is Merely Providing Post-Hoc Rationalizations

(1) Thinking Is "Post-Hoc Rationalization" of Conclusions

It's also important to note that people considered to have "high thinking ability" aren't actually engaging in a special "thinking" process. Recent neuroscience research has shown that human judgments and actions are determined instantaneously.

Libet's experiment (1983) and many subsequent studies have shown that decision-making in the brain occurs before conscious thought. For example, it has been observed that the brain's motor cortex is already activated about 0.3 seconds before we become conscious of "taking this action." This suggests the possibility that conscious decisions are the result rather than the cause of the actual decision-making process.

Michael Gazzaniga's split-brain research has also observed the phenomenon where the left brain (the hemisphere responsible for language) provides post-hoc reasoning for actions performed by the right brain. For example, after intuitively thinking "this is the correct answer," a process works where the left brain rationally explains that choice. This is a reversal phenomenon where "thinking does not lead to conclusions; rather, conclusions come first, and thinking is applied afterward."

(2) Knowledge Networks Instantly Make Judgments

The brain's knowledge network instantly derives conclusions, and the left brain merely provides reasons for them. A professional chess player can judge the next move in an instant because their knowledge network, built from years of experience, automatically activates and "intuitively" selects the optimal move.

Similarly, many judgments we make in daily life are likely the result of automatic processing by the brain's knowledge networks rather than conscious thought. Thus, the "thinking" process is rationalization that occurs after knowledge connections have yielded a conclusion, not a "special ability."

(3) The Reality of Thinking Is Knowledge Activation Patterns

If thinking is not a special process but an activation pattern of knowledge networks, then "thinking ability" can also be reduced to the quantity and quality of knowledge. People considered to have "high logical thinking ability" have abundant knowledge about logical relationships and can appropriately utilize them.

For example, mathematical thinking ability emerges from deep knowledge of mathematical concepts and relationships, and understanding patterns of their application. Considering this, thinking itself is utilization of knowledge and dependent on knowledge itself.

(4) The Relationship Between Intuition and Analytical Thinking

Daniel Kahneman's dual-process theory, which distinguishes between intuition (System 1) and analytical thinking (System 2), is widely known, but this distinction may also be merely a difference in processing speed and conscious access to knowledge.

"Intuition" is fast, automatic knowledge processing, while "analytical thinking" is slower, conscious knowledge processing. Both are based on knowledge networks and have no essential difference. If there are differences, they are only in the consciousness of processing, speed, and the type of knowledge used.

5. Intelligence Is Domain-Dependent

(1) The Evaluation Criteria for Intelligence Depend on Domains

Whether someone is evaluated as having high intelligence is greatly influenced by whether their knowledge has economic and social value. In modern society, certain knowledge domains are emphasized over others.

For example, people with abundant knowledge in programming or economics are easily considered "intelligent," while people familiar with trending TikTok content or insect ecology are rarely evaluated as "intelligent." However, this is merely a bias based on social values.

In reality, regardless of the knowledge domain, someone who has deep, broad, and structured knowledge in their field should be considered to have "high intelligence" in that field. In other words, whether someone is evaluated as having "high intelligence" depends on whether their knowledge domain is socially valued.

(2) The Concept of "Culture" Is Merely the Selection of Knowledge Domains

Evaluations like "being cultured" simply refer to familiarity with specific knowledge domains (literature, art, history, philosophy, etc.). These fields are emphasized as "culture" due to historical and cultural backgrounds, not because of their intrinsic importance.

For example, being knowledgeable about classical literature is considered "cultured," but being knowledgeable about the latest social media trends is not called "culture." However, both demonstrate expertise in specific knowledge domains. The concept of "culture" itself is merely a concept for selecting and hierarchizing socially valuable knowledge domains.

(3) Can Humans Be Said to Be More Intelligent Than Other Animals?

The claim that humans are "more intelligent" than dogs or monkeys is also a bias due to human-centered values. According to animal behavior research, each species has developed intelligence adapted to its ecological niche.

The olfactory knowledge possessed by dogs or the social rules understood by monkeys might be advanced knowledge systems incomprehensible to humans. For instance, dogs have an olfactory ability more than 10,000 times that of humans and possess vast "knowledge" about the world of smells. This is a cognitive ability unimaginable to humans.

Because the criteria for evaluating intelligence are human-centered, we are merely under the illusion that "humans are specially intelligent." Instead, we should consider that each organism has a knowledge system optimized within its evolutionary context.

(4) AI Intelligence Is Also Knowledge-Dependent

The development of modern AI also supports the view that "intelligence = knowledge." Large language models like GPT-4 have become capable of human-like responses by learning from vast amounts of text data.

These AIs do not have special modules for "thinking" but simply generate responses based on patterns and relationships (knowledge) extracted from large amounts of data. The "intelligence" of AI has improved because it has learned more knowledge and patterns, nothing more.

6. Intelligence Is Knowledge

(1) Enhancing Intelligence Means Enriching Knowledge

Synthesizing the discussions so far, intelligence is knowledge itself. The distinction between explicit and tacit knowledge is merely superficial, and abstraction ability also depends on the quantity and quality of knowledge. Additionally, the process called thinking is also the activation pattern of knowledge networks and not an ability independent of knowledge.

Therefore, enhancing intelligence means:

1. Acquiring more knowledge
2. Strengthening the connections between knowledge
3. Refining the structure of knowledge
4. Increasing patterns of knowledge application in new contexts

These are all interventions related to "knowledge," not enhancing separate "abilities."

(2) Intelligence as Understanding of the Universe

If there is an ultimate criterion for measuring intelligence, it might be the total amount of understanding, or "how much one knows about the universe." Each of us has a "miniature universe" in our brain, all of which are incomplete.

Theoretically, if Universe A completely encompasses Universe B (knowledge system), A could be said to be more intelligent than B. However, it is impossible to make such a complete comparison without a god-like perspective that can fully grasp all knowledge of individuals.

(3) Practical Approaches to Enhancing Intelligence

What we can do is gradually expand our "miniature universe." This includes methods such as:

1. Exploring diverse knowledge domains: Acquiring knowledge from different fields increases the diversity of the knowledge network.
2. Deepening knowledge: Learning deeply about a specific field elaborates the knowledge structure in that domain.
3. Applying knowledge: Applying learned knowledge in different contexts creates new connections between knowledge.
4. Utilizing metacognition: Consciously recognizing one's knowledge state and complementing deficient areas improves the overall balance of the knowledge system.

Through these practices, we can enrich our knowledge networks and consequently enhance what is called "intelligence."

Conclusion

The view that "intelligence is knowledge" is consistent with insights from modern neuroscience and cognitive science. It is more practical than the traditional view of "innate ability" and affirms human potential for growth.

The fact that intelligence can be reduced to knowledge means that anyone can enhance their intelligence through learning and experience. Additionally, by recognizing the social bias that only certain knowledge domains are evaluated as "intelligence," we can reevaluate the value of diverse knowledge.

Acquiring new knowledge, strengthening connections with existing knowledge, and becoming able to recognize more complex patterns—that is the essence of enhancing intelligence. By deepening our understanding of the universe, our knowledge networks become richer, and consequently, what is called "intelligence" will also be enhanced.