松易涅

Current Information#

Time period: 25/01/22-28

Keywords: Macau, social conflicts, humanities, deepseek, engineering innovation, US-China competition, philosophy of technology, information theory, military, systems theory, information networks, epistemology

Directory#

Information Flow#

1. 【Why is Macau much more peaceful compared to Hong Kong?】(Macau, social conflicts, labor movements) https://www.zhihu.com/question/577727842/answer/69180648117?utm_psn=1867652575219814400
2. 【Liu Jianjun｜The humanities will not disappear, but need to shrink】(humanities, professional cuts, division of labor) https://mp.weixin.qq.com/s/ICj50T1lnNtgFEB1lbn0Dw
3. 【The practical significance of deepseek】(deepseek, systems engineering, engineering innovation) https://boist.org/2025/01/27/2025%e5%b9%b41%e6%9c%8826%e6%97%a5%e5%91%a8%e6%9c%ab%e7%89%b9%e5%88%ab%e7%af%87%ef%bc%9adeepseek%e9%a2%a0%e8%a6%86ai%ef%bc%8c%e8%8a%af%e7%89%87%e4%ba%a7%e4%b8%9a%e9%93%be%e9%80%bb%e8%be%91%e9%87%8d/
4. 【Why deepseek surpasses openai】(deepseek, openai, AI large models, US-China competition) https://www.zhihu.com/question/10714706736/answer/88286561072?utm_psn=1867634403842351106
5. 【A technical historical perspective on 'The Wandering Earth': A post-humanist odyssey】(philosophy of technology, history of technology, The Wandering Earth) https://www.thepaper.cn/newsDetail_forward_23041075
6. 【Kobe: The Birth of a Legend Episode 1】(Kobe, inspirational growth, destiny) https://mp.weixin.qq.com/s/Y658W-SGmIftpAPs2Wlv2w

Thoughts#

1. Information and military
2. Power and information
3. Advice: Do not discuss definitions vacuously
4. A systems perspective on life
5. Build your own high-quality information network

Information Flow#

#1【Why is Macau much more peaceful compared to Hong Kong?】#

(Macau, social conflicts, labor movements) https://www.zhihu.com/question/577727842/answer/69180648117?utm_psn=1867652575219814400

Notes#

I wonder how many organizations/political parties in Macau can truly gain seats through parliamentary elections? How many of them have the ability to participate in governance? If there were truly free elections, how many could propose systematic political programs rather than fragmented political slogans? How many could present practical political solutions rather than blindly radical political actions?

A political group or organized party that can genuinely participate in governance over the long term should be able to combine social demands with the social power structure and transmission pathways, proposing actionable solutions to achieve political goals through struggle and compromise. However, small political groups and parties generally lack such capabilities, relying instead on radical slogans and measures to attract public opinion and gain political support. The ruling party, which has long occupied a dominant position in parliament, can continuously strengthen its electoral mobilization capabilities by summarizing, condensing, and inheriting internal organizational practices, thereby expanding its political advantage; dismantling the political offensive of competing parties, thus achieving long-term political monopoly.

#2【Liu Jianjun｜The humanities will not disappear, but need to shrink】#

(humanities, professional cuts, division of labor) https://mp.weixin.qq.com/s/ICj50T1lnNtgFEB1lbn0Dw

Mr. Chen Xianda said: “Most of us authors are ‘thieves’, stealing the thoughts of others. Our articles can only be considered reflections after reading, a rehash of previous or classic thoughts. True thought is original, it should express what has not been said by others. I do not have a single article like that.”
……
Take political science as an example, actually 90% of the books are a burden of knowledge, some have become the evil of knowledge. Based on my shallow experience, to learn political science well, one mainly needs to read the works of the following three types of figures, each category not exceeding 20 individuals. For instance, Tocqueville does not make it onto this list; at best, he is merely a clever and sensitive recorder of political observations. Mr. Chen Xianda said: “Most of us authors are thieves, stealing the thoughts of others. Our articles can only be considered reflections after reading, a rehash of previous or classic thoughts. True thought is original, it should express what has not been said by others. I do not have a single article like that.” This is not Mr. Chen's modesty; in fact, the vast majority of people do not have such articles. Given this, there are very few books that can genuinely engage us. Here, I list the three types of figures or books that need to be read:
(1) Social thinkers who have influenced the progress of civilization
There are very few social thinkers who have had a substantial impact on the progress of human civilization. Most of what people write is self-indulgent and has no relation to civilization or facts. The social thinkers who have had a substantial impact on the progress of civilization include Confucius, Jesus, Buddha, Muhammad, Adam Smith, Keynes, Locke, Marx and Engels, Dong Zhongshu, Cicero, etc. Thinkers who can have a substantial impact on the progress of civilization are as rare as stars in the morning. There are very few throughout history and across cultures. We can even say that almost everyone today lives within the intellectual legacy of these social thinkers.
(2) Politicians who have transformed the world
Transforming the world is more difficult than understanding or explaining it. Did not Marx say a very classic line: Philosophers have only interpreted the world in different ways; the point, however, is to change it. The maxims or principles that political scientists take pride in have little influence on politicians transforming the world. I have never heard of a politician governing according to any theory that reveals causal mechanisms, any concept, or any paradigm. Great politicians continuously enrich and develop their thoughts, theories, wisdom, and skills through arduous practice. Among politicians throughout history, there are not many who can create institutional legacies, produce long-term impacts, and effect substantial transformations in the political world. Representative figures include Qin Shi Huang, Mao Zedong, Deng Xiaoping, Washington, Lincoln, Roosevelt, Napoleon, de Gaulle, Churchill, Peter the Great, Stalin, etc.
(3) Thinkers (scholars) who influence the composition of knowledge
Setting aside those scholars who merely produce textual products, those who can enter our reading range are thinkers and scholars who influence the composition of knowledge. These thinkers and scholars are mostly academic, especially university professors since modern times. They are very narcissistic about their writings, believing that their words can influence social and historical processes. Although this is a huge illusion. Nowadays, students studying social sciences in universities deal with these figures every day. Their works are often referred to as "classics." I teach at a university and find that many students are tortured by these works, and after reading them, they are still at a loss. Many teachers who teach these works are the kind of people Mr. Chen Xianda referred to as lacking life experience outside of academia. Those who teach diplomacy have never engaged in diplomacy, those who teach management have never practiced management, and those who discuss governance have never been group leaders. Therefore, I often tell students: From the current perspective, the teachers teaching social sciences in universities are not much more capable than you! There are not many thinkers who can influence the composition of the knowledge world, and those who can enter our sight and reading lists mainly include Plato, Aristotle, Augustine, Aquinas, Laozi, Weber, Durkheim, Foucault, Rousseau, Liang Qichao, etc. Among the three types of figures mentioned in this article, the one that should be read the least is this type. Among the thinkers (scholars) who influence the composition of knowledge, most are also creators of knowledge troubles and burdens. They are inventors of concepts but not shapers of civilization. Of course, some people will disagree with this view. They always feel that their words can command history, influence the world, and shape others. In fact, many works and papers are nothing but a pile of words. In today's era of information explosion, a book or a paper may have only two readers: one is the editor, and the other is the author themselves. In the world of the internet, the endings of most papers and works are very simple: only those who like them, no readers! Nietzsche said God is dead; in the end, it is not God who dies, but Nietzsche himself. A phrase in an advertisement is very profound: "Nietzsche said: God is dead! God said: Nietzsche is dead!" To be honest, almost all scholars today have their influence largely confined within high walls and studies. Even their influence is often fabricated.
Finally, I would like to add that social science is always the servant of literature. Reading literature is much more interesting than reading social science! Years later, almost all social science theories, concepts, and paradigms produced in today's universities will be thrown into the historical garbage heap, but the four great classics, novels by Yu Hua, Liu Zhenyun, and Mo Yan will always be read. "Romance of the Three Kingdoms" has shaped the political cognition and evaluation of Chinese people in ways that "Records of the Three Kingdoms" cannot compare. Balzac said: "Novels are considered the secret history of a nation." The secret history of a nation is not hidden in social science theories but in literature. In novels, except for names and places, everything else may be real; in social sciences, except for names and places, everything else may be false. Literature is real, social science is false. This is my greatest feeling after studying political science for over thirty years. Although this assertion is overly extreme, the current development trend of social science undeniably proves this point. Many materials for scientific research conducted in the name of science are often "manufactured." Stepping back, from the perspective of aesthetics and art, the pleasure brought by reading Su Shi and Shakespeare is far beyond that of reading Weber and Rawls. From this perspective, Su Shi and Shakespeare are immortal and eternal, while the vast majority of social science theories are transient, not even making a ripple in the ocean of human knowledge, not even qualifying as a trivial interlude.

Notes#

In fact, Liu points out a problem: the academic world, detached from the masses, needs to produce "products" that society deems useful for its existence to maintain and expand reproduction. "Shrinking the humanities" aims to eliminate this portion of useless work, allowing Hayek's invisible hand to reallocate resources, at least liberating/driving away some scholars and graduate students who are not genuinely valuable, letting them do more meaningful things for society.

From the perspective of "reading," this text expresses the following core viewpoints and suggestions:

1. Be cautious when reading: avoid "knowledge burdens"

• Most books are merely "burdens of knowledge" or even "evils of knowledge," lacking real value.
• Many authors' works are merely repetitions and imitations of previous thoughts, lacking originality. There are very few books worth reading.

2. Suggested reading of works by three types of figures
   When studying political science, one should focus on reading the works of the following three types of figures, with each category not exceeding 20 representative individuals:
   (1) Social thinkers who have influenced the progress of civilization

• These thinkers have had a profound impact on human civilization, and their thoughts constitute the foundational heritage of our lives. Representative figures: Confucius, Jesus, Buddha, Muhammad, Adam Smith, Keynes, Locke, Marx, Dong Zhongshu, Cicero, etc.
• Reading suggestion: Carefully read the core works of these thinkers, as their thoughts define the direction of civilization.
  (2) Politicians who have transformed the world
• The practice of politicians is far more important than theory; they transform the world through practice, leaving behind institutional legacies and profound impacts. Representative figures: Qin Shi Huang, Mao Zedong, Deng Xiaoping, Washington, Lincoln, Roosevelt, Napoleon, Churchill, de Gaulle, Peter the Great, Stalin, etc.
• Reading suggestion: Focus on their biographies, governance philosophies, and practical experiences rather than purely explanatory theories.
  (3) Thinkers (scholars) who influence the composition of knowledge
• This type of scholar mainly influences the knowledge system, but many of their works are dull and difficult to understand, even becoming "knowledge troubles." Representative figures: Plato, Aristotle, Augustine, Aquinas, Laozi, Weber, Durkheim, Foucault, Rousseau, Liang Qichao, etc.
• Reading suggestion: Read less or selectively, as their thoughts lean more towards conceptual invention, lacking direct shaping power over civilization.

3. Critique of the current reading situation

• Limitations of academic works: Many scholars narcissistically believe their words can change the world, but in reality, their influence is often confined to the academic circle and even exaggerated.
• Problems of the era of information explosion: Nowadays, most books and papers are merely a pile of words, lacking readers and practical significance.
• Educational issues: Teachers in universities teaching social sciences often lack practical experience, leading to limited teaching effectiveness.

4. Summary reading attitude

• Less is more: Only read truly valuable classic books, avoiding wasting time on meaningless piles of words.
• Emphasize practice: Pay more attention to thoughts that can guide practice and shape civilization rather than purely for the sake of learning theory.
• Beware of arrogance: Do not mistakenly believe that knowledge or words themselves can influence the world; the key lies in the practicality and real significance of thoughts.

#3【The practical significance of deepseek】#

(deepseek, systems engineering, engineering innovation) https://boist.org/2025/01/27/2025%e5%b9%b41%e6%9c%8826%e6%97%a5%e5%91%a8%e6%9c%ab%e7%89%b9%e5%88%ab%e7%af%87%ef%bc%9adeepseek%e9%a2%a0%e8%a6%86ai%ef%bc%8c%e8%8a%af%e7%89%87%e4%ba%a7%e4%b8%9a%e9%93%be%e9%80%bb%e8%be%91%e9%87%8d/

Notes#

The impact of deepseek on the US stock market and its significant innovation lies in "systems engineering" innovation. Before deepseek became popular, developers (or was it the application side?) criticized the transition from the phantom model to deepseek-V3, not giving developers a transition period, directly switching the backend API to the V3 model without a buffer period for users of the old model; furthermore, the new model's API price was higher, equivalent to an "use it or not" attitude.

#4【Why deepseek surpasses openai】#

(deepseek, openai, AI large models, US-China competition) https://www.zhihu.com/question/10714706736/answer/88286561072?utm_psn=1867634403842351106

Notes#

But thinking about it the other way, isn't Sam Altman also the big boss who unites everyone? He directly shattered the existing talent team of openai, scattering like stars in the sky~ He overthrew himself: the reverse is the movement of the Dao. This prevents openai from controlling the flow of knowledge through controlling talent—controlling talent is more terrifying than simply controlling technology patents and other intellectual property.

#6【Kobe: The Birth of a Legend Episode 1】#

(Kobe, inspirational growth, destiny) https://mp.weixin.qq.com/s/Y658W-SGmIftpAPs2Wlv2w

Notes#

Some people seem to be self-important, but they actually know what they want to do. Man, what can I say? After watching it, I can only express my respect for Kobe.

Thoughts#

Information and military#

Some say the introduction of the transformer model will change the nature of warfare. What they are essentially talking about is how humans process information and make decisions. They are talking about AI-enabled armies. Before that, it was computerized armies. I can only say that from the perspective of information theory, one can see too much. Have you read "Ender's Game"? The future mode of warfare may be like this, where humans control machines to fight against machines.

As long as an organization is composed of humans, the efficiency of information collection, storage, processing, decision-making, and action always has an upper limit. It's similar to the yield rate in industrial production:
The results of 99%*99%*99%*99% vs 99.99%*99.99%*99.99%*99.99% are different. Traditional human command requires a staff team because one person can process very little information simultaneously. After introducing artificial intelligence, AI can process vast amounts of data simultaneously and extract information from it, using this as the basis for commanding warfare. In other words, the discontinuity and inconsistency of human decision-making can be improved by AI intervention, with the degree of improvement depending on the synergy and efficiency of the entire chain from sensors to transmission channels to processors (human brain or computer or AI) to transmission channels to effectors.

In this sense, Mao Zedong's guerrilla warfare was also an information war.

What I am saying can be illustrated by a single diagram from high school biology: in a situation of equal combat power, it comes down to who has the faster reflexes! Asymmetric warfare relies more on the speed of information flow.

This has inspired me: What is the essence of data? I believe it is electronic, structured information. It is a tool and carrier for dimensionality reduction of information from the real world.

Information - data - information. Encode and then decode. Decode into a language that humans can understand: auditory, visual, tactile, olfactory.

Although humans are also multimodal information processing systems, they are still inferior to computers. Computer systems can convert various dimensions of information into data matrices and then perform multidimensional parallel calculations, which is far more efficient than human processing. After external information is input, communication between computers is "information - data - data - …" If no translation is needed, then efficiency is at its highest level. Once human intervention is ultimately required, efficiency drops significantly.

Therefore, you will see AI training chips, besides tflops computing power indicators, also look at bandwidth rates. Humans are rushing to build quantum computers because they see the inherent advantages of quantum superposition states in parallel storage and parallel computing.

Once general quantum computers are built—or stepping back, military-specific quantum computers—humans may not need to intervene much in military operations. But the potential cost is whether real-time network communication can be guaranteed or deployed to combat areas for local area network communication, allowing war machines to operate autonomously. If a country fails to produce the talents, equipment, facilities, and resources for these war machines, it will have to revert to traditional non-AI military forces.

Power and information#

What is the essence of power? Why is there a distinction between violent power and non-violent power? Why does information constitute a part of power?

Is power the ability to make others obey one's will and act as expected? Information, as a component of power—does this mean that possessing information that others do not have, or that is inaccurate or incomplete, allows the holder to foresee the future and respond reasonably to upcoming events? Or is it to say that unchecked power partly refers to "the power exercised under opaque external information"?

Advice: Do not discuss definitions vacuously#

I asked the above questions in the group (What is the essence of power? Why is there a distinction between violent power and non-violent power? Why does information constitute a part of power?), and a group member bluntly stated not to attempt to discuss concepts and their definitions at too high an abstract level: "Because every philosopher and social science theorist has their own definition for every term; if you want to learn these, you will never finish in your lifetime. In the end, you will also come up with your own definitions like them. Read less philosophical dogma and more primary historical materials." I think he is very practical.

Another group member advised: It is possible to discuss definitions under the condition of situational cases, but "air-to-air" discussions without specific situations are difficult. Air-to-air definition disputes only serve one purpose in debates: to consume the opponent's time and energy, leaving them no time to think and respond to what you really want to say. Air-to-air discussions also take up your precious time. This is essentially a lose-lose situation.

Their comments resemble a confrontation between "knowledge starting from definitions" and "conceptual hierarchies." However, I partially agree with what they said: abstraction must ultimately be grounded in practical evidence.

A systems perspective on life#

Self-perception, self-organization, self-adaptation. I didn't feel the weight of these three adjectives before, but now I realize their significance: these are the three basic attributes of living organisms from a systems perspective.

Build your own high-quality information network#

Organizing a discussion or sharing session themed "Building Your Own High-Quality Information Network" should cover the entire chain of information acquisition, filtering, integration, and application, while combining personal and group practical experiences. The following content framework can help organize discussion points:
I. Core Issues: Why build a high-quality information network?

1. Challenges in the era of information overload

• The contradiction between information explosion and attention scarcity
• The dangers of low-quality information (rumors, clickbait, fragmented content)
• Avoiding information cocoons and cognitive biases

2. The value of high-quality information

• Enhancing decision-making efficiency and cognitive depth
• Supporting career development, knowledge innovation, and lifelong learning
• Establishing systematic thinking and cross-domain connections

II. Key Content Modules
Module 1: Filtering and diversifying information sources

• Evaluation criteria for credible sources
  • Authority (institution, author background), timeliness, transparency of data sources
  • Cross-validation: How to judge the authenticity of information through multi-source comparison
• Designing diversified information channels
  • Categorized by field: academic journals, industry reports, quality media, professional communities (e.g., Hacker News, Minority)
  • Categorized by format: long articles, podcasts, videos, offline events
  • Anti-information cocoon strategies: Actively introducing opposing viewpoints and cross-disciplinary content
    Module 2: Efficient tools and techniques for information acquisition
• Active search techniques
  • Advanced syntax for search engines (e.g., Google's site:, filetype:)
  • Academic resource searches (Google Scholar, ResearchGate)
  • RSS subscription tools (Inoreader, Feedly) and newsletter filtering
• Optimizing passive push notifications
  • Algorithmic training of social media (e.g., Twitter List, Weibo groups)
  • AI-assisted tools (ChatGPT information summaries, Perplexity real-time searches)
    Module 3: Information processing and knowledge management
• Information filtering and prioritization
  • Tagging classification vs. thematic tree structures
  • The 80/20 rule: Focusing on high-value information
• Knowledge accumulation tool practices
  • Note-taking tools (Obsidian, Logseq's bidirectional linking)
  • Card note method (Zettelkasten) and progressive summarization
  • Visualization tools (mind maps, knowledge graphs)
    Module 4: Dynamic maintenance of the information network
• Regular cleaning and updating mechanisms
  • Criteria for eliminating information sources (e.g., activity level, declining output quality)
  • Warning signals for information overload (e.g., anxiety, imbalanced time allocation)
• Automation assistance
  • IFTTT/Zapier for automated integration of information flows
  • Browser plugins (e.g., Pocket for later reading, Readwise for highlight management)
    Module 5: The application loop from information to action
• Matching information with goals
  • How to customize an information network based on personal/career goals (e.g., updating a programmer's tech stack vs. tracking industry trends for entrepreneurs)
• Validating input through output
  • Optimizing information quality through writing, speaking, and practical projects
  • Community collaboration: Knowledge sharing and collective wisdom (e.g., online reading groups, peer reviews)

III. Interactive Session Design

1. Workshop practice

• Group design of personal ideal information flow templates (tools + channels + management processes)
• Simulated scenarios: Quickly building a temporary information network for specific issues (e.g., "AI ethics")

2. Case breakdown

• Inviting guests to share their own information network structures (e.g., scholars, investors, content creators)
• Participants critique each other's information source lists and propose optimization suggestions

3. Debate session

• Example debate topic: "Does algorithmic recommendation inevitably lead to cognitive narrowing?"

IV. Recommended Extended Resources

• Reading list: "A Brief History of Information," "The Future of Knowledge Management," "Deep Work"
• Toolkit: Whitelist of information sources (recommended by field), list of browser plugins, RSS subscription templates
• Long-term practice community: Establishing online groups for regular sharing of quality information sources, organizing "information decluttering" challenges

Through the above framework, participants can gain methodological guidance and form actionable personal plans through practice and interaction. The key is to lower the cognitive threshold of abstract concepts through specific cases and tools, emphasizing "personalization" rather than absolute standards, avoiding the trap of tool worship or excessive systematization.