1. Human beings' understanding of the world is essentially a process of establishing connections.

The process by which human beings understand the world is essentially a process of constantly discovering differences, establishing classifications, and finding connections between things.

For early humans, the external world was initially presented as a large number of objects and phenomena that were not yet understood: the sun, moon and stars in the sky, plants and animals on the earth, the changing seasons, weather changes, and humans themselves, all mixed together in a world that was not yet understood.

The first step in human understanding of the world was attempting to categorize this complex whole: What are animals? What are plants? What are things in the sky? What are things on earth? What are observable and measurable phenomena? These seemingly simple questions actually represent a significant shift in the human cognitive process:Humans began to try to separate things that were originally mixed together.

However, distinguishing things is not humanity's ultimate goal. More accurately, it's a cognitive approach humans adopt to understand the world. Because the real world contains a vast number of complex things, the human brain cannot simultaneously process all unorganized information. If all things existed simply as independent, disordered entities, it would be difficult for humans to discover patterns or form stable understandings.

Therefore, humans need to find similarities and differences between things, group things with common characteristics together, and give them a common concept; this is classification.

From a cognitive perspective, classification has actually completed an important transformation: humans no longer simply receive and store various information about the world, but rather, through abstraction and induction, transform this information into knowledge that can be understood and used.

For example, our understanding of nature today doesn't involve dealing with thousands of unrelated individual life forms: cats, dogs, whales, eagles, roses, pine trees… but rather placing them into different categorization systems:

Through this classification method, humans no longer simply remember isolated things, but can further understand their place in a larger cognitive system. For example, when we recognize the animal "cat," we don't just remember the name "cat," but also know that it belongs to a larger category such as animal or mammal.

More importantly, as classification systems become more sophisticated, humans can summarize common characteristics from a large number of specific things and apply these experiences to new things. For example, when a new mammal is discovered that has never been seen before, although humans do not know all its details, they can make certain inferences about its possible characteristics based on past knowledge of mammals. It is in the process of constantly recognizing specific things and discovering commonalities that humans gradually form an understanding of the unknown world.

However, as our understanding deepens, humanity has discovered that simply knowing "what things are" is insufficient to comprehend the world. Classification can help us answer "what characteristics a thing has," but it cannot explain "why things are connected."

Therefore, humanity has further begun to explore the relationships between things and the laws behind these relationships. For example, the development of biology is not simply about discovering more species, but about constantly establishing deeper connections: the evolutionary relationships between species; the relationship between genes and biological traits; and the relationship between organisms and their environment. These connections have enabled life sciences to gradually develop from a "list of species" into a knowledge system capable of explaining life phenomena.

A similar process is occurring in other scientific fields. For example, the development of physics involves not only continuously recording more natural phenomena, but also searching for the unifying laws between phenomena and describing these relationships through mathematical models.

On the surface, it appears that humanity is constantly accumulating information about the world; but more accurately, what humanity is truly deepening is its understanding of things, relationships, and laws in the world, and this understanding is gradually formed in the process of constantly establishing connections.

2 From Information to Understanding: Why Do Humans Need Structured Knowledge?

2.1 From Information to Cognitive Structure: How Humans Understand Knowledge

Establishing connections is only the beginning of understanding the world. A further question is: once humans discover and understand these connections, how do these understandings enter their own cognitive system and be stored, retrieved, and reused long-term?

If the process of human learning is simply about constantly collecting and storing information, then learning should be a very simple thing: just keep reading, observing and recording, save all the information you come into contact with, and then retrieve it when you need to use it.

But reality is not like that. Many people go through a similar process when learning a new field: they are exposed to a large number of new concepts, terms and methods in a short period of time, and can even remember a lot of them, but after a while, they find that they still cannot truly understand the field.

The problem isn't necessarily a lack of information. Often, what people lack isn't more information, but rather that the information hasn't been understood, interpreted, and connected to form knowledge that can be used to understand and solve problems.

Computers can save data to storage media according to addresses and retrieve it accurately when needed; however, the human brain does not work this way. Humans do not save information received from the outside world in a completely independent manner, but rather continuously process, classify, interpret, and abstract it. In this process, what was originally just received information gradually transforms into knowledge that can be understood and applied, and this knowledge further forms higher-level cognitive structures through association—one important reason for needing to process information in this way is that the human cognitive system itself has processing limitations.

Research in psychology on working memory shows that the amount of information humans can simultaneously retain and process is very limited. Early research proposed "Miller's Law," which suggested that the capacity of human short-term memory is approximately 7 ± 2 chunks of information; subsequent research suggests that this capacity may be even lower in more complex cognitive tasks.

Regardless of the specific capacity, these studies all illustrate one fact:Humans cannot rely on processing large amounts of isolated information simultaneously to understand the complex world in the long run. When faced with dozens of unrelated pieces of information, the brain quickly experiences a huge cognitive burden, a limitation that is particularly evident when learning a complex field.

For example, someone new to computer networks might quickly remember:

TCP UDP IP DNS HTTP HTTPS NAT BGP CDN Firewall

This content itself is real and valid information; however, if it exists merely in the brain in independent forms, it will still be difficult for a person to determine how to use this knowledge when faced with a real problem. For someone new to internet knowledge, it may simply be an ever-growing list of terms.

However, when this information is no longer simply arranged by category, but rather structured with interrelationships based on those categories, the brain processes not just a set of isolated terms, but several interconnected wholes:

This is what cognitive psychology calls "chunking." Chunking is not simply reducing the amount of information, but rather organizing and integrating information to combine multiple independent pieces of information into a higher-level cognitive unit, thereby reducing the burden on the brain when processing complex information. When this information, after being understood, interpreted, and organized, forms a cognitive unit with clear meaning, it begins to transform from simple fragments of information into knowledge that can be understood and used.

This shows that human understanding of the world is not a simple process of accumulating information, but a continuous process of processing and organizing: information is understood, interpreted and abstracted to form knowledge, and knowledge is then gradually formed into cognitive structures by establishing connections.

Information, knowledge, and cognitive structure thus form a continuous hierarchical relationship: information is the raw material for human beings to understand the world, which can be recorded, disseminated, and preserved, but does not in itself represent understanding; knowledge is formed on the basis of understanding, interpreting, and abstracting information, and can be used to explain phenomena, solve problems, and guide actions; cognitive structure further describes the overall organizational network formed after knowledge is connected, which determines how a person understands new information and transfers existing knowledge to new problems.

2.2 From Information Connection to Understanding: How Cognitive Structures Generate Meaning

Through chunking, humans can compress vast amounts of complex information into more manageable cognitive units. These cognitive units, after being understood, organized, and abstracted, can further form independent knowledge units. However, a single knowledge unit is still only a partial understanding of a particular object or phenomenon; only when connections are established between different knowledge units can a cognitive structure be further formed.

So how exactly are these knowledge units connected? And how do they collectively explain a more complex phenomenon through these connections?

Taking computer networks as an example, someone who is new to computer networks may know that: DNS is used for domain name resolution; TCP is responsible for reliable transmission; HTTP is used to transmit web page content; and CDN can improve access speed.

These basic descriptions of the concepts are correct in themselves. However, if DNS, TCP, HTTP, and CDN exist only in one's mind as a few separate concepts, then when faced with a real problem, it will still be difficult to understand why these technologies appear simultaneously and how they work together to complete a network access.

Before encountering CDN, a person might have already formed this basic access model:

In this cognitive model, the user finds the target server via DNS and then communicates directly with the server. This model is not flawed and can explain a large number of basic website access processes.

However, after further learning about CDN, a new question arises: how should CDN be integrated into the existing access model? If we only remember that "CDN is a website acceleration technology," then this concept remains at the definition level and has not truly entered our existing cognitive structure.

A true understanding requires integrating CDN into the existing model and readjusting the original access process:

At this point, CDN is no longer just a "speed-enhancing tool," but has become an integral part of the entire website access architecture. Simultaneously, the concept of the "target server" has also changed—previously: users accessed the server, and the server returned content; now: users first access the CDN node, and the CDN, depending on the situation, directly returns cached content or further requests the origin server.

This is the change that occurs when a new concept enters the existing cognitive structure: it does not simply add a new knowledge point, but changes the relationship between existing knowledge units and reorganizes the original understanding model.

The "schema" theory in psychology describes a similar cognitive process. A schema can be understood as a cognitive framework that humans use to organize experiences and understand things; it reflects how concepts, experiences, and relationships are organized within the cognitive structure.A concept is truly understood not just because people remember its definition, but because it has entered into existing cognitive structures and established stable connections with other concepts.

This explains why many people, despite reading numerous materials and memorizing many concepts, still fail to truly master a new field. Learning isn't simply about increasing the amount of information in your memory; it's about continuously adjusting and refining your existing cognitive structures. Only when new concepts connect with existing knowledge can they truly integrate into your cognitive structure and become knowledge that can be used for understanding, reasoning, and application.

Experts are often able to make judgments more quickly than ordinary people when faced with unfamiliar problems because they have formed a large number of stable conceptual connections through long-term learning and practice. When new problems arise, they can map them onto their existing cognitive structures and use their accumulated experience for analysis.

Therefore, the process of human learning and understanding the world is not simply about storing more information in the brain, but about continuously building new connections based on existing cognitive structures, giving new information a clear location, meaning, and application context. In other words, humans do not merely accumulate information, but constantly process information into knowledge, and form and improve cognitive structures through the connections between knowledge.

2.3 How do humans actively construct their cognitive structures?

However, cognitive structures are not fixed once formed. For simple problems, humans can gradually build and adjust their understanding through the accumulation of experience; but when faced with complex domains, the number of concepts to be processed increases, and the relationships between concepts become more and more complex, making it increasingly difficult to rely solely on the brain to naturally form and maintain cognitive structures.

Therefore, humans have gradually developed various methods to aid cognition, which help them observe, verify, expand, and express existing cognitive structures.

Although these methods differ in form, they all essentially solve the same problem:How can we make the cognitive structures hidden in the human brain clearer, more stable, and continuously improved?

Visual structuring: Mind maps and concept maps

One of the difficulties in understanding a complex field is that even if some connections have been formed in the brain, these connections are often implicit and difficult to observe directly.

Therefore, humans have developed methods such as mind maps and concept maps to express the relationships between concepts through external graphics.

For example, in the previous section, we compared two website access models to show how CDN integrates into the original website access cognitive structure. In fact, these two models can also be converted into mind maps.

Basic access mode:

Access model after CDN is added:

Visualization is not simply "text rewriting" or "formatting changes," but rather making implicit cognitive structures explicit: it helps us see which knowledge units belong to the same module (classification); what kind of relationship exists between which knowledge units (associations); and how the introduction of a new knowledge unit reconstructs the original system like triggering dominoes (evolution).

Whether it's a mind map or a concept map, their essence is a "visual lens" for the brain's thought process.

Semantic structuring: Feynman Learning Method

Mind maps and concept maps address the question of "whether the relationships between knowledge can be seen," while the Feynman Learning Technique further addresses the question of "whether these relationships are truly understood."

Often, people can remember a concept and identify its position in the knowledge structure, but they cannot explain why it is in that position or why it has such a relationship with other concepts.

The Feynman Learning Technique focuses on this process. It not only helps learners discover whether connections have been established between pieces of knowledge, but also exposes gaps in understanding through active expression, thus driving further refinement of cognitive structures.

For example, someone might remember: "CDN is a technology that improves website access speed." This statement itself is not wrong, but it is just an isolated piece of knowledge.

If we delve deeper into the question, "Why can CDN improve access speed?", we need to understand why user requests don't necessarily access the origin server directly, but instead retrieve content through edge nodes deployed in different regions. If we continue to ask, "Why does DNS affect CDN node selection?", we need to understand how the DNS resolution process participates in the selection of user request paths. If we further ask, "Why does CDN sometimes still need to access the origin server?", we need to understand caching mechanisms, cache invalidation, and the relationship between origin servers.

These questions are not actually examining a single concept, but rather testing whether the concept of CDN has been linked to knowledge of DNS, HTTP, caching, distributed nodes, origin servers, etc.

If these questions cannot be answered, it means that although the concept of "CDN" has been memorized, it remains an isolated unit of knowledge and has not been truly integrated into the existing knowledge structure. Through active explanation, learners can discover: what content has been truly understood, what content remains at the memorization level, and what connections are missing from their existing cognitive structure.

Therefore, the Feynman Learning Technique is not essentially a simple memory technique, but rather a method of checking and refining cognitive structures through language output.

Network structuring: Card box note-taking method

As knowledge continues to grow, even with a certain cognitive structure already established, people will still face new challenges: new concepts constantly emerge; old knowledge continuously forms new connections; and existing structures require ongoing adjustments. At this point, relying solely on one-time reorganization is no longer sufficient to meet the needs of long-term accumulation.

Zettelkasten offers a different approach: it does not require knowledge to be fixed in a certain classification system, but rather allows the knowledge network to grow continuously through independent knowledge units and the connections between them.

For example, when learning about CDN, cards can be created for different knowledge units such as CDN, DNS resolution, caching mechanisms, and edge computing, and a knowledge network can be formed through the connections between them:

This approach differs from mind mapping in that it doesn't require knowledge to revolve around a central theme. Instead, it allows each knowledge unit to be an independent and continuously connected node. As learning progresses, new knowledge can be added to the existing network, rather than redesigning the entire classification system each time.

This approach is actually closer to the characteristics of human cognitive structure: knowledge is not a simple catalog, but a dynamic network composed of a large number of concepts and relationships.

Structured Expression: The Pyramid Principle

In addition to building and maintaining cognitive structures, humans also need to convey their understanding to others. However, the cognitive structures in an individual's brain are usually complex, and if output directly, they can easily become a large amount of fragmented information.

Therefore, humans have also developed methods to help express complex ideas, such as the Pyramid Principle, which emphasizes: first present the core idea, and then organize the reasons that support this idea according to logical relationships, thus reorganizing complex information into an easy-to-understand structure.

For example, after understanding CDN, a person may have already formed a large amount of related knowledge in their mind:

DNS CDN node caching origin server TCP HTTP access distance traffic scheduling

While these contents are related, explaining them directly to others can easily turn into a simple pile of technical terms. The listener can receive the information, but it is difficult to understand why these concepts appear at the same time, and how they work together to explain the role of CDN.

When organizing your expression according to the pyramid principle, you need to first identify a core question: "Why do websites need CDN?", and then organize the supporting concepts and reasons around this question.

The output in this way is no longer a list of knowledge, but a logical structure that can be understood. It does not change the cognitive structure itself, but rather reorganizes and presents the existing cognitive structure according to the purpose of expression, so that complex knowledge can be understood and conveyed in a clearer way.

These methods reveal that the various learning and thinking tools developed by humankind are not essentially designed to simply store more information. The common problem they address is how to make the connections between concepts more stable, clear, and reusable, gradually transforming them from implicit personal cognitive structures into observable and expressible ones.

2.4 From Cognitive Structure to Knowledge Structure

However, the external expression of an individual's cognitive structure does not equate to the fact that knowledge has been structured. While mind mapping, note-taking systems, and other methods can help individuals record and express parts of their cognitive structure, relying solely on personalized recording methods remains limited when the scale of knowledge continues to expand, or when knowledge needs to be preserved, shared, and reused across individuals for a long period of time.

Therefore, humanity needs a more systematic approach to explicitly represent the concepts in knowledge and the relationships between them, so that the structures originally implicit in human cognition can be expressed in a more systematic way.Understandable, reusable, and extensible The form is preserved—this is the problem that knowledge structuring aims to solve.

Unlike simple information organization, knowledge structuring does not focus on adding more content, but rather on clearly expressing the concepts, attributes, and relationships between concepts in knowledge, so that some relationships that are originally implicit in the human cognitive structure can be clearly represented.

For example, three independent knowledge units:

A B C

Although they were all recorded, they were not connected to each other. At this point, they were still just three isolated units of knowledge.

And once relationships are established between these pieces of information:

  A / \ B---C

They are no longer just three isolated points, but form a whole with a stable internal structure. In this structure, each node not only has its own meaning, but also gains a more definite position and significance due to its relationship with other nodes.

Therefore, the focus of knowledge structuring is not simply recording existing information, but rather further clarifying which knowledge units constitute a knowledge system of a domain, and how these knowledge units are related to each other.

3 From Knowledge Organization to Knowledge Engineering: How do humans construct structured knowledge?

3.1 Knowledge Engineering: Making Knowledge Representable Structures

When knowledge needs to be preserved, shared, or even processed by computers for a long time, understanding that exists only in personal cognition needs to be externalized and represented in a more explicit way.

In response to this need, humanity began to explore how to integrate knowledge...Concepts and their relationshipsThis exploration gradually led to the research direction of Knowledge Representation, and further developed into the research field of Knowledge Engineering. Knowledge Engineering focuses on how to acquire, represent, organize, and utilize knowledge so that the knowledge and relational structures that originally exist in human cognition can be preserved and processed in a structured form.

In the 1970s, the field of artificial intelligence began exploring how to give computers expert-like judgment abilities. Researchers found that general-purpose algorithms alone could not solve problems in all domains; computers also needed domain-specific knowledge.

However, computers do not possess the domain understanding that humans develop through long-term learning. Therefore, if we want computers to handle complex problems, we need to convert the knowledge that humans possess into a structured form that computers can represent, store, and process. This makes "how knowledge is acquired, represented, and utilized" an important issue in artificial intelligence research.

Against this backdrop, knowledge engineering gradually developed. Early knowledge engineering primarily focused on how to acquire and represent expert experience, and transform this experience into a form that computers could process. Among these, expert systems became the most representative application of this stage.

In expert systems, the judgment experience of human experts is usually represented as a series of rules:

If condition A is satisfied, then conclusion B can be derived.

With a large number of such rules, computers can simulate the knowledge and reasoning processes of experts in specific fields (such as medical diagnosis, financial risk assessment, and industrial fault analysis).

However, as applications have expanded, people have gradually discovered that knowledge in the real world is not simply composed of simple rules. Knowledge in a domain typically contains a large number of different types of objects and complex relationships between them. For example, in the field of computer networks:

• CDN is a type of network service;
• A CDN consists of multiple distributed nodes;
• DNS can affect the CDN nodes accessed by users;
• CDN nodes reduce the pressure on the origin server through caching.

This content is not just a few independent judgments, but a knowledge structure composed of multiple concepts and the relationships between them.

Therefore, modern knowledge engineering has gradually evolved from "how to store rules" to "how to represent knowledge structures." It has begun to focus on:

• What objects constitute a knowledge structure?
• What attributes do these objects have?
• What relationships exist between different objects?
• How these relationships collectively shape an understanding of a particular field.

In other words, knowledge engineering attempts to express the knowledge network formed by humans in the cognitive process in a clear structure, so that knowledge is no longer just existing in personal experience and understanding, but can be preserved, shared, and further processed by computers.

To achieve this structured representation, a fundamental problem needs to be solved first:What exactly are the basic building blocks of knowledge?

3.2 Basic Units of Knowledge Representation: Entities, Attributes, and Relations

To represent knowledge in a structured form, we first need to define the basic elements that constitute knowledge. From the perspective of knowledge representation, people usually organize knowledge around three basic components:

• An entity is a basic unit in knowledge that can be independently described and establish relationships with other knowledge elements. It can be a concrete object in reality, such as a person or a place, or an abstract object with a clear meaning, such as a technology, a software system, or a network component.

• Attribute: Describes the characteristics of an object;

• Relation: Describes the connection between different objects.

These three parts together form the foundation of most knowledge structures.

Entity: an object in knowledge

When humans try to understand the world, the first thing they need to do is identify "what is there".

In the field of computer networks: DNS, CDN, TCP, HTTP, origin servers… these can all be considered independent knowledge objects. Without clearly distinguishing these objects, knowledge remains only in vague descriptions. For example: CDN can improve website access speed. This statement contains some information, but what exactly is "CDN"? What is its relationship with the website, server, and network requests? Without clear object classification, further organization of knowledge becomes impossible.

Therefore, the first step in knowledge structuring is to identify various objects in the world and represent them as independent knowledge units.

Attributes: Describe the characteristics of an object.

Simply knowing what objects exist is not enough; when humans understand an object, they also need to know what characteristics it possesses.

For example, the term "CDN" is merely a name. Knowing that such a technology exists doesn't provide a true understanding of what it is. Describing its type, function, and characteristics gives this object a more precise meaning.

For example:

Object: CDN Attributes: Type: Network Service Function: Caching content, optimizing access paths Features: Reducing user access latency through distributed nodes

These descriptions are not explaining the relationship between CDN and other objects, but rather describing the characteristics of CDN itself. Through attributes, an object that originally only had a name gradually becomes a knowledge object with clear meaning.

Relationships: Enabling isolated objects to form a knowledge structure

However, even with a large number of objects and properties, it is still insufficient to form complete knowledge. This is because what truly determines the depth of understanding is the relationships between objects.

For example:

DNS | Select | CDN Node | Cached Content | From | Origin Server

This doesn't describe information about individual objects, but rather the relationships between them: DNS influences which CDN node a user accesses; CDN nodes can cache content from the origin server; and CDN nodes need to access the origin server when a cache miss occurs.

It is these relationships that enable originally independent objects to form a knowledge structure capable of explaining actual phenomena.

In knowledge engineering, entities, attributes, and relationships are not three abstract concepts defined by humans, but rather correspond to the most basic cognitive methods humans use to understand the world: we first identify what objects exist in the world, then understand what characteristics the objects have, and finally explore how objects are related to each other.

What knowledge engineering does is represent this way of understanding, which originally exists in human cognition, in a clear structural form. When a large number of entities, attributes, and relationships are organized together, a structured representation that can describe knowledge in a certain domain is formed.

3.3 From Knowledge Units to Knowledge Networks: How Knowledge Forms a System

As the scale of knowledge continues to expand, simply preserving individual knowledge units is no longer sufficient. This is because a single knowledge unit can only describe partial information, while a truly complete understanding of a domain comes from the overall structure formed by a large number of knowledge units. Real-world knowledge itself is not composed of a small number of isolated knowledge units, but rather contains a large number of interconnected knowledge units and relationships. In this structure, the value of knowledge comes not only from the individual knowledge unit itself, but also from the connections formed between different knowledge units.

Therefore, knowledge engineering needs to further address the following issues:As the number of knowledge units continues to increase, how can we connect these scattered knowledge units through the relationships between them to form a knowledge network that can be understood, queried, and utilized?

Through these connections, computers can further understand: what relationships exist between different concepts; where a certain knowledge unit is located in the entire domain; and how a concept, together with other concepts, constitutes a complete knowledge system.

Knowledge graphs are a method for constructing structured knowledge networks, developed based on this idea. They organize large amounts of knowledge units into a network of nodes and relationships, enabling previously scattered knowledge to be stored and utilized in a structured manner.

In this system, different nodes represent different units of knowledge, and the connections between nodes represent the relationships between them. In this way, knowledge is no longer just passively stored information, but becomes a structure that can be further queried, analyzed, and expanded by computers.

The applications of knowledge graphs are very broad and not limited to the technical field. As long as a domain contains a large amount of knowledge that needs to be organized and connected, a structured representation can be built using knowledge graphs. Examples include areas such as interpersonal relationships, healthcare, scientific research, business analysis, and cultural knowledge.

Although knowledge graphs in different domains contain different types of entities and relationships, their core idea is the same: to organize knowledge units, attributes, and relationships in a domain into a knowledge network that can be understood and utilized.

4. The AI Era: The Structural Challenges of Increased Knowledge Representation

4.1 From Knowledge Acquisition to Knowledge Generation: Changes Brought About by Large Language Models

In the preceding development of knowledge engineering, humanity primarily employed a method...Explicitly construct knowledgeThe approach involves first identifying the basic units of knowledge, then organizing them through entities, attributes, and relationships to ultimately form a knowledge network that can be processed by computers—knowledge graphs are a typical example of this approach.

However, this method also has a clear limitation:Knowledge needs to be actively discovered, organized, and defined by people. As the scale of knowledge continues to expand, relying on manual methods to continuously build and maintain a complete knowledge structure becomes increasingly costly. Therefore, artificial intelligence has begun to explore an alternative path:Is it possible for computers to learn patterns and rules directly from large amounts of data, rather than relying entirely on manually defined knowledge structures?

Towards this goal, artificial intelligence is gradually evolving from relying on manually defined knowledge to using data to train models, enabling computers to extract patterns and regularities from large amounts of data through statistical learning, and forming implicit representations of the data. This means that the way knowledge is acquired and represented is gradually shifting from "manually constructing explicit structures" to "machines learning from data."

The emergence of Large Language Models (LLMs) represents a significant stage in this development. By learning language patterns and statistical relationships from massive amounts of text and generating new content based on context, LLMs have given computers unprecedented capabilities in knowledge representation and content generation.

At the same time, the way humans acquire knowledge has also changed. Before the advent of large language models, people usually needed to actively search for information, read and understand it, and organize existing information; but in the era of large language models, people can directly interact with the model through natural language, and the model can quickly generate a complete answer on a certain topic.

This means that the process by which humans acquire information and express knowledge begins to shift from...Actively seek out and organize existing data, and gradually shift towards interacting with intelligent systems that can generate content.

However, new problems have arisen:Does the ability to generate clearly structured and fully expressed answers mean that computers have developed a cognitive structure similar to that of humans?

4.2 The Structured Challenges of Increased Knowledge Expression

Large language models reduce the cost of knowledge representation, enabling the rapid generation of large amounts of structurally complete and fluent content. However, the increase in knowledge representation does not imply the automatic formation of knowledge understanding. This is because the answers generated by large language models are typically generalized expressions geared towards a wide range of users—they can cover a large number of concepts and organize content according to certain logic, but this organization may not necessarily conform to the organization methods of each individual's existing cognitive structure.

For those with existing knowledge, AI-generated content can be quickly connected to their existing cognitive structures, helping them supplement information, discover connections, and further expand their understanding. However, for those lacking this foundation, even if the answer itself is complete, the numerous knowledge units and relationships it contains may still fail to establish effective connections with their existing knowledge. This is because, without the necessary knowledge base, new knowledge units often lack the concepts and relationships to which they can be linked, making it difficult to truly integrate them into the original cognitive structure.

For example, if someone new to computer networks asks an AI about CDN, the AI can generate a complete explanation that includes knowledge expressions such as CDN definition, workflow, caching mechanism, and edge nodes.

However, if users are not previously aware of how DNS affects access paths; how HTTP requests reach the server; why caching reduces pressure on the origin server; and why distributed nodes improve access efficiency, then establishing true relationships between these concepts remains difficult. The problem is not whether AI can generate complete knowledge representations, but whether these representations can be integrated into the user's existing cognitive structure.

This problem not only affects individual learning, but the entire knowledge ecosystem may also change when knowledge expression becomes easier to generate, while the processes of true understanding, verification, and structuring do not increase at the same pace.

In the past, much of the knowledge on the internet came primarily from people's experience summaries, professional creations, and long-term accumulation. Although this content was limited in quantity, it usually contained the author's understanding, judgment, and organization of the field.

With the large-scale emergence of AI-generated content, a new content cycle is emerging on the internet:

AI-generated content ↓ enters the internet ↓ is searched, cited, and disseminated ↓ becomes training data for future models

If a large amount of content mainly consists of recombination of existing information, lacking real-world experience, professional verification, and further structured organization, then the amount of information on the internet may continue to increase, but the knowledge structure that can truly help humans form cognitive structures may not grow in tandem.

4.3 From Knowledge Acquirer to Knowledge Organizer

Before the advent of artificial intelligence, the process of acquiring knowledge for humans typically involved a great deal of information gathering and organization. Due to the high cost of information acquisition, people needed to actively search for materials, read and learn, summarize experiences, and gradually form an understanding of a particular field in the process.At this stage, the process of knowledge acquisition itself is also an important process for humans to build cognitive structures.

However, in the era of large language models, this process is changing—when AI can quickly complete information retrieval, content organization, and knowledge expression, humans no longer need to spend a lot of time on simple information collection and organization, and the efficiency of acquiring knowledge is greatly improved.

However, the increased efficiency of knowledge acquisition has also altered the process of knowledge formation. Understanding, associating, and accumulating knowledge that used to occur naturally with knowledge acquisition no longer happens automatically: people can quickly obtain complete knowledge expressions about a topic from AI, but whether this content can be truly understood and connected with their existing knowledge has changed from a "natural outcome of the learning process" to a "process that needs to be actively completed."

Therefore, in the AI era, the problems that humans need to solve when facing knowledge have become: how to understand the content generated by AI, how to discover the connections between them, and how to integrate them into the existing cognitive system. AI can help humans quickly acquire relevant information, summarize existing knowledge, discover potential connections, and provide explanations from different perspectives, but how to express this knowledge, transform it into one's own knowledge, and further form a cognitive structure still requires active human participation.

A truly valuable knowledge structure is not merely the accumulation of information quantity, but more importantly: which concepts constitute a field; what relationships exist between different concepts; which laws can explain phenomena; and which experiences can be transferred to new problems… These factors determine whether a person truly understands and masters a particular field.

These elements that constitute cognitive structures do not naturally form as the amount of knowledge expressed increases. Whether in traditional knowledge environments or in today's world where AI-generated content is constantly increasing, knowledge needs to be understood, organized, and connected in order to transform scattered information into a structured system—and this is precisely the core issue that knowledge engineering addresses.

5. Returning to Personal Knowledge Engineering

In a previous article, I had ChatGPT and Qwen analyze blog posts to discover long-standing, stable concepts and relationships between different articles. Unexpectedly, they derived the basic units of knowledge engineering (see article: [link to article]).Personal Knowledge Engineering (Part 1): From Article Index to Knowledge Structure – A Practice of AI Collaboration in Exploring Personal Knowledge Systems).

Through this period of exploration, I have gradually discovered that knowledge engineering is not just a technical method for organizing and arranging content. It actually corresponds to the basic way humans understand the world: people do not simply memorize a large number of facts, but rather establish connections between concepts and form a holistic understanding of a certain field through continuous learning and practice.

This leads to the questions discussed in this article: Why does knowledge need to be structured? Why do humans need to understand the world through concepts, relationships, and cognitive structures? And why has knowledge structuring become even more important today, given that large language models can quickly generate massive amounts of content?

Therefore, this article is actually an exploration from practice to theory in the process of personal knowledge engineering: the first article focuses on how to use AI to help individuals discover the connections between content and gradually build a knowledge structure; while this article further explores why these connections exist, and the knowledge structure and cognitive basis behind them.

In this process, I also further realized that knowledge structuring is not simply organizing existing content, but rather reorganizing scattered knowledge units according to certain ways of understanding, establishing clear relationships between different concepts, and forming a knowledge system that can be understood, reused, and expanded. There is no single answer to this organizational method: the same set of knowledge may form different structures for different people due to differences in experience, goals, and cognitive styles.

This means that knowledge structuring is actually an outward manifestation of an individual's way of understanding: how a person organizes knowledge largely reflects how he understands the relationships between concepts, and the final result of this structuring also reflects the depth of his understanding of a certain field.

Through this article, I have further clarified the long-term development direction of blogs: personal accumulated information, practical experience and thoughts should not just be scattered content records, but should be transformed into knowledge by leveraging the concept of knowledge engineering, and combined with one's own understanding, to form a knowledge structure with personal characteristics and sustainable development through the connection between concepts.

In the future, the content of blogs should not just be a collection of more and more articles, but should gradually form a knowledge network that can be continuously maintained and expanded, reflecting the accumulation of personal long-term practice and thinking.

This is also the direction I continue to explore in my personal knowledge engineering series.