October 2025 Thought Notes: Ideas and History, Daily Life, Self-Knowledge

Pair and ferrying others across

Quite an interesting pairing

Pair is a word beloved in tech, even in AI — whether it’s a partner, a pair-programming buddy or a companion assistant, it gives a kind of emotional value

In Buddhism, ferrying others across — the “ferrying” is commonly used for carrying sentient beings from affliction and birth-and-death (the cycle of rebirth) across to liberation, awakening, or a state more favorable to practice. It’s both a mindset (compassion, the power of vows) and a synthesis of methods (precepts, concentration, wisdom and skillful means)

But why is it that in Buddhism I seem to have heard two sayings: one is “the Buddha ferries those with karmic affinity,” the other is “only by ferrying yourself can you ferry others”?

In Mahayana scriptures, ferrying all sentient beings, universally saving all beings, is actually done through a series of methods, techniques, and material and spiritual help

The premise of ferrying others is respecting their autonomy and karmic conditions — no coercion or manipulation

Ferrying others also needs proper boundaries, to avoid emotional exhaustion

Buddhas can only teach the dharma; they can’t practice in your place

The core liberation lies in the individual’s own awakening

The human-nature narrative in The Plum in the Golden Vase

Many people look noble because they don’t have the right to choose

Human nature, too, is constrained by rules

But once a person has choices, he is free — and terrifying

Character determines a person’s choices; choices determine a person’s fate

There’s no binary good-person/bad-person opposition — same in The Plum in the Golden Vase; the underlying logic is a cold dissection of reality

Human nature is desire, and desire inevitably corrodes

The fate of the small people is tragic: they have no choices. But the powerful who do have choices are tragic too — driven by desire, destroyed by desire

Some doors are two-way, some are one-way. When you face a choice, think a little more: does walking through this door mean irreversible ruin?

An elite of an era, in the trajectory of that era, didn’t manage to brake ~

When a person stands at the peak of desire with nothing left to restrain him, destruction becomes the only endpoint

Pan Jinlian is clearly an extraordinarily vigorous life force, craving to be seen and appreciated. She’s naturally beautiful with deft hands (she can sing and play, she can sew), full of energy, with a stubbornness in her bones that refuses to be beneath others. In a normal environment she might have been a radiant woman

But she was placed in the wrong spot, with no choices. To survive she had to fold away her fragrance and grow the sharpest poisonous thorns. All her “evil” is this life force distorted in a dead end for the sake of survival. So even if character determines choices, the premise is that the environment provides good choices

Cold narration

The author or narrator maintains a deliberate, conspicuous emotional distance from events and characters, telling the story in an objective, restrained, sometimes seemingly detached and unemotional voice

Like a calm camera that only records and doesn’t judge, it doesn’t immerse the reader in the character’s emotional swings but leads the reader to become a cool observer, examining from outside how events unfold, how characters act, and the logic and absurdity behind them

A conceptualized way of thinking

Getting past binary, all-or-nothing certainty in my own thinking

Accept that the world is fundamentally complex, changeable and uncertain, that anything is a probability distribution — not simply predicting the future but calculating the probabilities of different outcomes and betting

This makes them more objective in decisions, able to calmly accept failure (because it’s just a probabilistic event), and to keep making “positive expected value” choices

Flights from Seoul to third countries are very cheap

Including some countries in Europe

Seoul (Incheon, Gimpo) is an important aviation hub in East Asia, with especially developed low-cost carriers (LCCs); there are large numbers of flights to Japan and Korea, Southeast Asia, even Europe and America. The more flights, the easier it is to push fares down

Japan and Korea have very large travel demand toward Southeast Asia and Europe/America; local residents travel abroad frequently, creating a huge market scale, so airlines can attract more passengers with low prices and profit through ancillary services

By comparison, China’s policy controls, traffic rights allocation and airport capacity issues mean fewer flights and higher fares

And the current won exchange rate also gives a travel advantage ~

The untranslatable “feel” is the root of culture

I’ve started to realize that once certain things leave their native context, their core “feel” dissipates. When reading the Records of the Grand Historian, the flavor and force of the original classical Chinese is something no modern vernacular translation can match. This experience is the same principle as Eason Chan’s songs only sounding right in Cantonese

This makes me think that language is far more than a tool of communication; it is itself a carrier of culture and emotion. Many words and expressions simply can’t be “translated,” because they carry collective memory and emotional resonance from a particular cultural background. Translation can convey information but can’t reproduce this unique “original flavor”

From this angle, the spread of Mandarin, while bringing convenience in communication, may also be invisibly eroding the distinctive cultural texture carried by dialects. When a language loses its vitality, what dims along with it may be a world’s unique way of perceiving

Three levels of understanding the world: is, ought and might

Reality, ideal and possibility

The is: what the world is actually like, what the objectively existing world is like, the factual state that can be observed empirically and verified

The ought: what the world should be like. Ought statements express a standard, rule or expectation; they define what is “good,” what is “right,” what is “just.” For example, “we should be honest and trustworthy,” “society should guarantee every citizen’s basic rights”

The might: the world’s uncertainty, judgments that are unknown to some degree, and predictions of the future

For example, “it might rain tomorrow,” “this investment might succeed”

The “might” connects the “is” and the “ought.” Based on our analysis of the world as it is, we assess various probabilities, then according to our ought-goals choose the action path most likely to lead to the ideal result

We stand on the is (the real environment we’re in), hold the ought in our hearts (the ideal goal we hope to achieve), and when acting must carefully assess the various mights (the different consequences different choices may bring), and finally make a decision

example: some of Alan Chan’s methods

example: some of Alan Chan’s methods — using AI for learning practice, product thinking

In an era when false information spreads easily, contact with first-hand sources is more valuable than ever

The marginal returns of three learning methods, taking 20 hours as an example:

Traditional learning — reading books, relying on personal understanding: depth above average, breadth low, 7 points

AI speed-reading — lots of introductory material (books, videos, articles): breadth high, depth shallow, 6.5 points

Deep engagement with the field’s best academic textbook for 20 hours (AI-assisted Q&A, explanation, examples): depth high, breadth medium (9 points)

Combining the three is best, especially the first two — essentially aimless (divergent), boundaryless learning -> digging deep into interests

When my notes get long, I take the time to break them into smaller parts and visualize the relationships between them. I find this process extremely useful for deepening my understanding of what I’ve learned. At this stage, drawing arrows and connections also becomes very useful

What content can be deleted?

ref - quoting a line from Douglas Engelbart’s 1962 “Augmenting Human Intellect,” a line that perfectly captures how AI changes learning in my eyes. In short: many advanced approaches to learning and problem-solving only become feasible when lower-level capabilities are fast and efficient.

When choosing a book, a practical rule: pick one that feels slightly out of reach

When choosing a book, a practical rule of thumb is: pick one that feels slightly out of reach. Ideally it should be the kind of book that feels too hard to read on your own, but that you could understand with a PhD advisor guiding you

And having gone through the filtering of university classrooms and discussion by countless students, such a book will have left behind many good lesson plans and method questions, which is very conducive to interactive learning with AI

After all, there’s an extremely rich stream of information in AI too

True deep understanding doesn’t come from the relationship between two books

True deep understanding doesn’t come from the relationship between two books but from the relationships among all the concepts in those two books

Only by atomizing notes can you go deeply into the topic you care about through visual notes. Atomic notes don’t mean you can’t write long notes

It means each concept card should contain only one concept, supported by its content. To ensure clarity, you should always describe the concept in one sentence and use that sentence as the concept card’s title

What’s truly valuable isn’t the whiteboard you end up making, but the thinking invested in building the knowledge structure and naming each concept card and section, between steps two and four. Deep understanding and insight usually come from the process of deconstructing, reorganizing and describing knowledge in your own words. Only by going through that process does knowledge truly become your own

Thoughts on books that are secondhand

Returning to first-hand information research is very important

A lot of information today has been modified, packaged and beautified

Semantics come in every bizarre shape

An author’s personal experience is a source for generating hypotheses, but it can’t directly become a universal conclusion — unless there’s broad data or a clear mechanism supporting it

Philosophy never left reality; it just understands it in a higher dimension

How to define reality, how to pose questions, how to understand meaning

Philosophy’s core isn’t information accumulation either, but the generation of thought

How questions are formed

Many people in philosophy groups dislike AI

But isn’t philosophy essentially independent thinking, not isolated thinking?

If thought can be reproduced by machines, does philosophy still have irreplaceable meaning?

The problem with reading a lot of books nowadays

Once you have your own cognitive system, many books — especially those focused on opinions or secondhand experience — need to be approached cautiously

Many of these books are the author’s “personal biography + generic template.” They package opinions based on their own experience (say, deriving “embrace discomfort” from one failure) but ignore underlying variance — your biological instincts, environment and cultural background may be completely different. As a result the reader immediately feels “something’s off,” because intuition is warning: this isn’t a universal truth, it’s the author’s “subjective filter.” Moreover, the sense of repetition comes from publishing logic: the self-help market is huge, and authors/editors often pad content to fill pages or reinforce memory (repetition helps the brain encode ideas, but for a veteran like you it just becomes noise)

Essentially, most of these books are “secondary processing” — distilled from psychology and philosophy sources (Maslow’s hierarchy of needs, Stoicism) and sold again with a different inspirational story. Read enough and you’ll find the core is just a few points: self-awareness, habit formation, relationship repair. There’s nothing new because the “underlying code” of human growth is just these — you can vary the packaging but not say anything new. Better to dig deeply into someone’s own knowledge and theory system

At that point, one more book is just one more layer of noise on the outside; too much information and too many choices actually constrain action

Change isn’t read into being, it’s “chosen” into being — a book can only wake you up; you have to jump in and “live” it

The principle of least effort

Principle of least effort

In processing information, the brain tends to choose text with lower cognitive load that’s easier to understand

All human action is like this too — choosing things with lower cognitive load and easier comprehension

All organisms, humans included, tend to choose the path of least energy expenditure when acting

But there’s an interesting paradox: although the brain leans toward ease, its structure and function aren’t fixed — they can be reshaped through learning and experience. Working hard to understand complex text is like strength training for the brain; it can effectively exercise our thinking ability and build new neural connections, raising our level of intelligence

Also, if it triggers a flow experience — when we’re intently grappling with a hard problem and finally it opens up — the sense of achievement and pleasure is far beyond what finishing an easy task can give

Where the soul comes to rest

Almost everyone, sooner or later, realizes at some point this longing to “settle the soul”

Hoping for a place, a person, a group where they don’t have to pretend

Hoping what they do and how they live can answer “why am I alive”

Hoping the heart is no longer torn by anxiety, comparison and emptiness

Whether Schopenhauer or Russell’s The Conquest of Happiness, it’s really about the outer world and inner world reaching a kind of reconciliation — a pivot outside, a support within

Some find it in creation, some in nature or faith

And some search their whole lives — but precisely because they’re searching, they keep on “living”

The replacement of an era often comes at the cost of ‘struggle’

The replacement of eras often comes at the cost of “struggle.” Whether it’s the clash of civilizations, the change of institutions, or the renewal of thought, new things are always born in conflict with the old forces. Mao Zedong said in On Contradiction: “The emergence of everything new must go through struggle against the old. Without struggle there is no progress.”

The violence of Spain’s conquest of the Americas isn’t an isolated evil but the manifestation of a universal law of history — the disintegration of the old order and the birth of the new order inevitably carry along conflicts of power, greed, belief and culture. The progress of civilization is therefore full of tragedy, and it’s precisely because of contradiction that it keeps moving forward

Contradiction itself is a kind of balance: there’s no fairness or unfairness to speak of, no justice or evil to speak of — it’s just reaching a dynamic game and balance

The conditions for game and balance are constraints and information

Allowing conflict to exist — conflict is itself a check and a way of avoiding despotism

So you think while being stripped of the right to narrate history

So you feel that while being stripped of the right to narrate history, you also gain a sense of collective identity — is that lucky or unlucky?

The conflict and fusion between cultures has always been a contradiction — even China’s group conflicts, gender conflicts, value-system conflicts, age-group conflicts, and all the more so the invisible class conflict

Or is it that a controlled or beautified historical narrative — where we come from, what we’ve been through together, who we are — to some degree quickly shapes a strong sense of identity and belonging in us

History is also complex, chaotic, full of contingency and disgrace; a clear, powerful official narrative provides the individuals inside it with an easy-to-understand worldview

Just like religion, it gives a code of conduct, an outlook on life, a value system

But stripping the right to narrate history is essentially castrating history. It means the memories and voices of countless individuals, ethnic groups and losers are erased

History is no longer a discipline for inquiring into truth but a tool for consolidating power

Our sense of “we” is often consolidated by setting up a “they.” While a unified historical narrative shapes “our own people,” it inevitably also vilifies, diminishes and even demonizes the “outsiders” outside the narrative

This is also why it shapes wars and plants the seeds of conflict and war

And when the ugly voices really surface, what does this society do? How does this fragile society bear it?

For those in power now and the majority of the collective, it’s often seen as “lucky,” because it brings stability and strength

For those silenced in history, for future descendants and for individuals pursuing truth, it is unquestionably “unlucky,” because it leaves behind lies, hatred and hidden dangers

This is also the tension in the development of human civilization: how do we build a balance?

Buddhism: habitual nature

Essentially it refers to psychological tendencies and behavioral inertia cultivated by repeated experience and action

Xi (habit) is repetition

Xing (nature) is a natural pattern of response

So from the standpoint of results, all main outcomes are learning

But from the standpoint of motivation, only what you do actively and deliberately is learning

But often it’s the learning you don’t notice, in certain environments, that is what truly and invisibly shapes us

In daily life, learning we don’t even consider learning shapes our worldview and cognitive structure

It’s really the environment subtly influencing a person — and that’s an extremely efficient form of reverse practice

So one question: I used to always think about how to improve the ability to learn and create

This is a pyramid in the world, a very competitive question, and everyone’s heading in the same direction

What is our core competitive learning ability and creativity?

A national consciousness centered on stability

It seems that compared with China’s ethnic minorities

Spain’s Indigenous population really is unlucky

A phenomenon I’ve always been curious about: China’s preferential treatment of ethnic minorities, and how China is consistently harsh with its own interior while showing unusual “tolerance” and “ingratiation” toward outsiders or marginal subjects?

China has always been centrally centralized historically; the goal itself isn’t to make every individual happy but to maintain territorial unity and stable order. Stability has always been instilled throughout China

Minority regions → without compensation, they may drift away

Foreigners → treated well, it helps external image, investment and diplomatic interests

Its own citizens → stable and compliant, regarded instead as a “controllable group” that doesn’t need much incentive

This is collectivist, and it has always been hard to form genuine civic consciousness

Chinese society takes the “family” as its unit: the family is the prototype of society, the state is the family enlarged

Protests in Turkey today

Multiple countries held pro-Palestine marches calling for a quick end to the conflict

A conflict of narratives

Seventy or eighty years now: one core of the Israeli-Palestinian conflict is religious conflict. Judaism’s “God” (Yahweh) and Islam’s “Allah” are each side’s spiritual pillar and source of faith

And God didn’t save them either

In the end, what kind of person you become doesn’t depend on what the outside world wants you to be

It’s objective law that determines a person — through the combination of character, talent, cognition and so on, following objective cause and effect, arriving at a version of yourself more in line with your own nature

Instinct itself is part of the law too — not something mysterious, but common sense

Empress Wei and Empress Zhang

Both were followers and imitators of Wu Zetian

Both ended in failure

Wu Zetian, as the only female emperor in Chinese history: outstanding ability, rich experience in political struggle, talent for governing

Innate extremely high intelligence and political sensitivity

And an indestructible willpower and ambition

Including an innate ability to perform and disguise that let her win out in the struggle

And I keep wondering what kind of acquired environment shaped her

Part of it was her early education: her father Wu Shiyue was a founding hero of the Tang dynasty, rising to Minister of Works, giving her high-quality educational conditions

Most important was Taizong of Tang, the top-tier politician — the best “political observatory.” Although she wasn’t favored, as a bystander she witnessed firsthand how a top emperor controlled his ministers, handled state affairs and balanced various forces. This is a top-tier practical politics course no book can teach. That’s where she learned forbearance and observation

Systematically training your questioning ability

We can’t ask good questions about a field we know nothing at all about. So the first step is building a “minimum viable” knowledge framework.

Some methods:

First-principles method: master the most core, most basic axioms and principles

Systematic reading: read the most authoritative book or work in the field to understand the whole picture

Then deliberate practice: how to find “questioning gaps”

Reverse questioning: this is a subconscious thinking mode — what else is there that I don’t know, an awareness of knowing that you may not know

Whys: keep asking in succession, striking at the essence

The Feynman technique: teach a complex concept in simple language to someone who knows nothing about it, and thus discover the vague and broken places in your own knowledge system

Uncertainty reduction theory

Don’t guess — take the initiative

Don’t doubt — get close to the scene itself

Essentially this all comes from uncertainty. The nature of problems between people, especially between strangers, points to one psychological state — uncertainty. That’s the core motive for communication

People find uncertainty unpleasant in interpersonal relationships and are driven to reduce it through interpersonal communication

But uncertainty can also be quantified and managed; if the cost is too high, for example in engineering design, allow for a certain amount of uncertainty

This is more like a biological drive, a human instinct in evolutionary psychology

Cognitive uncertainty and behavioral uncertainty

When you don’t know the other’s beliefs (cognitive), you can’t predict their actions (behavioral). Conversely, when the other’s behavior is unexpected (behavioral), you re-question their thinking (cognitive)

Cognitive uncertainty → unpredictable behavior → more cognitive confusion → more behavioral anxiety → …

The only way to break this spiral: communication

The rules by which uncertainty operates:

Verbal communication is inversely proportional to uncertainty: as verbal communication increases, uncertainty decreases

Uncertainty is inversely proportional to nonverbal affinity expressiveness. When one person shows more positive nonverbal expression (eye contact and smiling), the other’s uncertainty level goes down — the body doesn’t lie, and the brain tends to believe body language

Uncertainty correlates positively with information-seeking strategies: the higher the uncertainty, the harder we search for information

High uncertainty in interaction leads to low intimacy in communication; low uncertainty in interaction leads to high intimacy in communication

High uncertainty leads to high reciprocity; low uncertainty leads to low reciprocity — this is tit for tat: introduce yourself first, then ask the other person. So in fact when you lower your own uncertainty, you’re also placing a demand on the other’s uncertainty

Uncertainty increases with difference and decreases with similarity, so people tend to seek out those similar to themselves, because it lowers uncertainty

When uncertainty increases, liking decreases; when uncertainty decreases, liking increases. This is URT’s most powerful practical implication — uncertainty is poison to relationships

Actually, uncertainty isn’t necessarily negative; sometimes it’s positive

The charm of a relationship, what attracts people

Creative work

Exploratory dialogue

Some important thoughts on reducing ambiguity

Why can some people achieve their goals while others can’t?

An experiment with two groups of workers

Group A’s instruction: “Do your best”

Group B’s instruction: “Complete 50 units within 30 minutes” (specific, measurable)

Group B performed significantly better than Group A. This finding challenged the mainstream assumption of the time — people generally believed “do your best” should inspire maximum effort. But Locke proved: vague goodwill is far inferior to clear standards

Because a clear goal reduces the decision cost of action

Specific, high goals lead to higher performance than urging people to do their best

So generally it’s better to set a goal yourself

The goal-setting effect isn’t a culture-specific phenomenon but a basic feature of human cognitive architecture

Clarity — the first line of defense against ambiguity, something that can be measured and won’t be misunderstood

A very interesting line

Speaking is the bridge between your heart (yours) and the brain (the other’s)

The core purpose is still to get your thoughts into someone else’s brain accurately, precisely, efficiently and without loss, and to build an emotional connection

Basic chain: thought -> encoding (language) -> expression (spoken) -> transmission (air/medium) -> reception (ears) -> decoding (brain) -> understanding

Clarity: what to say — KISS (Keep It Simple, Stupid). Thought is the foundation; express complex views in the simplest way

Structure: speech is organized — Point -> Reason -> Example -> Point

Insight: depth, distinctive views, not just stating facts — ask why more

How you deliver matters too: what your voice, body language and language style are like

How you listen matters too: listen with heart and brain, make eye contact, lean forward

Ask open-ended questions and guide the other person to express themselves more deeply

And often summarize the other’s viewpoint in your own words

Then there’s empathy — not just content but connection: putting yourself in their place to think what you would do, understanding and responding to the other’s emotions

The ability to hold the space; controlling the room matters — the difference between one-on-one communication and giving a speech

The three elements of the sign and the dyadic sign theory

The three elements of the sign are three indispensable elements forming a trinity

Representamen: what we see or hear

Object: the actual thing or concept that the representamen points to or stands for. This is the “about what” content the sign tries to convey.

Interpretant: the meaning or effect the sign produces in the receiver’s mind

Ferdinand de Saussure viewed the sign as a psychological entity made of two parts

Signifier and signified:

The signifier is the sign’s sound-image or graphic form, the sign’s external manifestation. It’s similar to Peirce’s “representamen”

The signified is the concept or meaning the sign represents. It exists in our thinking, a mental representation

But interpreting an object can loop and dig deeper endlessly, to the point of drawing on all human knowledge; this process has an infinite interpretability

Between stimulus and response there lies a gap

The key to growth lies in how you use this gap.

What gap?

The chance to stop, reflect, evaluate and choose actively.

It lets us go beyond instinct or habit and make wiser choices, and so achieve personal growth.

Separated by language and signs,

separated by memory — thoughts — and interpretation.

Which is how the very first two ways of being happy come about

Which is how the very first two ways of being happy come about, whether you take Russell’s view or Schopenhauer’s.

At first you don’t know how to get happiness, so naturally it’s about escaping suffering and seeking pleasure; life is just the cycle of escaping suffering and seeking pleasure, until you drop dead hahaha.

People’s pursuit of mystery usually comes from inner lack

This is a compensation mechanism; the sense of mystery is bred from “absence”.

When we’re in love, we also feel that some sense of lack in us has been satisfied.

The sense of mystery doesn’t disappear once the lack is satisfied — it becomes even more mysterious.

When a person is full inside and able to face their own limits and imperfections,

he can still hold mystery in awe, but not “use mystery to make up for lack”.

To face the unknown lucidly, rather than mythologize the unknown.

To stop at the boundary of understanding, rather than anxiously manufacture illusions.

People want to reduce uncertainty, but what people need more is to face uncertainty.

I don’t know, but I am at peace.

In society, mystery as a tool is itself used to cover up lack.

Regimes package authority as “mystery”;

business manufactures desire through “mystery”;

religion comforts pain with “mystery”.

Money, love, power and fantasies you couldn’t get when young — once satisfied, are there other mysteries and voids?

When a person moves from lack to satisfaction, they go through a kind of “disenchantment”.

The world goes from miracle to logic, life goes from passion to order.

Your attention shifts from “feeling” to “possessing”.

Satisfaction brings a sense of control; many people mistakenly think they understand, but true mystery always exists — you just no longer “listen”.

Exponential thinking and logarithmic thinking

Playing a game can improve your skill quickly, but later improvement is very hard. The exponential curve is the opposite: at first it rises very slowly for a long time, but at some critical point it suddenly shoots up, rising fast in a very short time. For example: growth in writing ability, changes in wealth, changes in a company’s value — things that need quantity to turn into quality.

So some indicators can go from quantity to quality; keep your eye on that indicator.

Clearly, by comparison, if you only do technology, it looks more like logarithmic thinking.

Keep a good mindset, trust in the power of accumulation.

Stay aware: know what you’re doing, know what you’re thinking.

The “throw your backpack over the wall” mental model

In life, the phenomenon we always run into is that our ability to act never measures up.

Breaking the task down and just starting is one method: the point is to put yourself at some risk and loss first, to get moving first, which activates you better.

This avoids overthinking and forces the brain to think about the act of climbing the wall, rather than whether to climb the wall.

The butterfly effect

Grab one small point, then amplify it through that small point.

The premise is a whole you’re conscious of, a system with a rough direction.

Break through via one small point.

But there’s another premise too: some ability to model and analyze the system itself.

The everything-is-a-system mental model:

each node, causal relations, possible situations, boundaries, etc.

The limit is a very effective approach — a rational way back to the present.

What you truly want, what you truly need.

Applying the Buddhist five aggregates

What it dismantles is self-attachment; nominally it’s about leaving the world.

Form aggregate: the physical body and the senses, all material forms that can be perceived.

Feeling aggregate: sensations, the feelings of suffering, pleasure or neutrality produced by external stimuli.

Perception aggregate: imagination and recognition, the cognition and identification of objects.

Volition aggregate: will and mental activity, all psychological tendencies that drive action.

Consciousness aggregate: awareness, the knowing and recognizing of the first four aggregates.

The “I” does not exist independently; it is a phenomenon of form, feeling, perception, volition and consciousness temporarily coming together.

Actual practice:

Contemplate the body as impure: form.

Contemplate feeling as suffering: feeling.

Contemplate the mind as impermanent (perception, volition, consciousness).

Many people mistakenly think letting go means doing nothing.

In fact it means the mind isn’t dragged around by outcomes and isn’t swayed by other people’s eyes.

The 10/10/10 bystander thinking model

It’s really also a reverse-decision method: from a long-termist view, whether you’d regret not making this decision or making it.

Ten minutes later, how do I look at the decision I’m making now? Still consistent, or regretful?

Ten months later, how do I think about this decision I made ten months ago?

Ten years later, how do I see this judgment and decision from ten years ago? This mental model can be applied to: temporary judgments, big decisions, predicting your own future, etc.

In fact a person with a mature value system can melt this deeply into everyday micro-decisions, intuition-driven.

Scenarios where deduction replaces induction

In this world, 99.9% of people (or more) use induction. Induction is experience-driven: take the experience you or others obtained, and in similar situations take similar action.

Experience is subjective and time-limited, and constrained by environment.

Deduction is logic-driven: learn the essence or theory of a thing, carry out logical derivation from the essence, apply scenarios from the essence — it deals with first principles.

The advantage of induction is that experience transfers quickly and can be used fast. For example, the sun always rises in the east, so tomorrow it will surely rise in the east too. It rests on the premise that the future and the past are the same. Deduction: derive from premises; if the premise is true, the conclusion must be true. If a thing has five layers of depth, induction only induces the explanation of the first layer — fairly superficial. Deduction is the deepest essence, and can even find the first cause and derive from there.

When you’re unfamiliar with something, you don’t have to rush to answer it

First think about the thing, or the related field, with your own network of knowledge.

Reflections on Japanese vs Chinese architectural aesthetics

Take Jiangnan and Shaoxing as an example — I’ve been around here recently.

Ancient buildings in China are heavily influenced by Confucian culture and the court aesthetics of the system.

Abundance, order, qi, form, ritual, the order of the unity of heaven and man.

Chinese architecture stresses “round heaven and square earth” and “axial symmetry”, expressing a visual presentation of cosmic order, imperial hierarchy and family ethics.

A rational depiction of order.

Japanese architectural aesthetics, by comparison, have far less of that grandness.

They focus on the person, acknowledging human smallness, acknowledging that wood will rot, tiles will shatter, light will change.

That faintly trembling, imperfect sense of “wabi-sabi”.

A strong sense of time: not pursuing immortality, but naturally feeling the traces time leaves on buildings.

The 2014 Nobel Prize in Physiology or Medicine: cognitive neuroscience’s spatial model

The 2014 Nobel Prize in Physiology or Medicine: cognitive neuroscience’s spatial model.

For thousands of years, understanding how the brain’s cognitive functions arise from its basic physiological components has been a fascinating ultimate frontier for science.

The 2014 Nobel Prize in Physiology or Medicine went to three scientists: John O’Keefe, and the couple May-Britt Moser and Edvard I. Moser.

For their discoveries of cells that constitute a positioning system in the brain.

There is a precise built-in GPS system in the brain, answering a question that has puzzled philosophers and scientists for centuries: how do we know where we are, and how do we plan routes and find our way.

Place cells: certain neuron systems in the hippocampus are activated only when the rat is at a specific position; if it moves to another place, some cells stop firing and some start firing.

Then grid cells: when a rat moves through an environment, these cells are activated at multiple positions, and those activation points form an extremely regular hexagonal grid pattern in space — like a sheet of coordinate paper covering the whole space. Through this process the brain calculates distances and plans routes to achieve spatial navigation.

People who over-rely on GPS show a clear decline in spatial memory and self-directed navigation ability.

It makes the brain lazy; actively observe the route, memorize some landmarks, then try turning the navigation off.

Those Things About the Ming Dynasty

The carrier of the content doesn’t matter; what matters is that what should be expressed got expressed, and what should be understood got understood.

Let every clue be dug into deeply, like playing a game, uncovering a different life and plot.

So maybe, setting aside the prompt’s understanding of the system and of the user,

another part can be some interesting methods — for example, just dig deep into history, every storyline and historical record.

Why is the perception of the Yuan dynasty so faint in history

If it’s because of the ruling party, knowing little about the modern part of history is understandable.

But why is relatively little known about Yuan dynasty history too?

The Yuan dynasty was founded by the Mongols; for the Han dynasties of the Central Plains, it was a period of “being conquered”.

Although the Yuan ruled all of China, it was never fully “co-opted” into ethnic identity by later generations.

The Yuan dynasty was established through an extremely bloody conquest.

The fall of the Southern Song, the massacres in Jiangnan, the sharp drop in population — a heavy trauma for Han regions.

The Yuan also left no sustainable marker of civilization.

The cultural elites of the Yuan — Ma Zhiyuan, Guan Hanqing, Wang Shifu and others — mostly carried a “repressed literati temperament”.

Add to that the very interesting selectivity of the Ming historians’ records.

The “History of Yuan” was hastily compiled in the early Ming, taking only a few months, and written very restrainedly.

The super empire Genghis Khan built, and the “Yuan dynasty” Kublai Khan established in China, were essentially the Central Asia–East Asia headquarters of the Mongol world empire.

The Yuan distinguished a hierarchy of ethnic ranks:

Mongols, Semu, Han and Southerners.

A Mongol could easily become a local governor,

while a Han scholar might study for ten years and still not pass the provincial exam.

Still, it established the province system for the first time, and that system has been carried on ever since.

Those things about the Ming dynasty

It actually explains history in an entertaining way too.

The narrator is Mingyue.

It doesn’t have to be rigorous.

In the end, rigor doesn’t necessarily equal being highbrow.

Truths can be plain and easy to understand.

What matters most is sparking people’s interest.

History is complicated, the world is complicated

But so what.

We just need to live our own lives well.

We just need to choose the things we’re interested in.

Long stretches of hardship are what best temper a person’s will. Many people, when they hit trouble, can only blame heaven and complain about others and muddle along; others, though they too have to bow their heads before hardship, never let their hearts submit — they keep working, believing they will surely win in the end.

This is a great transformation, one that many may spend a whole lifetime unable to complete.

The key to the transformation lies in the heart.

For many of us, the heart is the most tender place, especially easy to wound: the betrayal of love, the loss of family, the loss of friendship — each is a heavy blow. But for Zhu Chongba, what was left that he couldn’t bear? He had already lost everything. What could be more painful than watching your parents die with no way to help, than fighting dogs for food to stay alive, than being cursed and despised by others? We have reason to believe that on some night of painful thinking, Zhu Chongba turned this most fragile place into his most powerful source of strength.

Yes, even if you have looks everyone envies, learning that has read every book, wealth beyond spending — none of that proves you’re strong. For strength of heart is true strength.

Why can’t they stop

It seems to be the case for a lot of older people.

Including my own father.

The whole value of their lives was built on what I do.

When they suddenly stop, no longer busy, no longer needed, they feel themselves falling into the void of who am I.

Retirement isn’t leaving work; it’s leaving the self that was needed.

Even repetitive labor gives them a sense of mastery.

Once they stop, blankness, meaninglessness and awareness of death come flooding in.

This is a kind of withdrawal reaction; it isn’t a matter of willpower, it’s a neural-level dependency.

The fear of being eliminated by society, abandoned by family, negated by oneself.

Why do rich people instead prove they have leisure?

They’re no longer anxious about survival, and start being anxious about proving their existence.

Money buys time, but time exposes emptiness.

So they use “consumption, travel, collecting, pleasure, building an identity”

to construct a system of symbolic meaning.

I can enjoy, I can desire, I can dispose of things freely — therefore I am still alive, and living at a higher level.

MIT cognitive scientist Josh’s paper in PNAS

In a paper published in PNAS by MIT cognitive scientist Josh, he compared different representational structures for abstract knowledge, such as star structures, clustered structures, ring structures and so on — and in the end he proved mathematically that the best abstract knowledge structure for humans is the tree structure.

So you could say that only a tree structure best matches human cognitive characteristics, because from one layer of the tree to the next there is a unique path — making it easy for the brain to extract knowledge quickly from the memory base, which fits the way the human brain is a cognitive miser; at the same time, the tree is the simplest structure that has both horizontal and vertical expansion capacity.

—— What is the best knowledge representation structure for humans?

I’ve talked before about a certain kind of person — one with no sense of boundaries

I’ve talked before about a certain kind of person: one with no sense of boundaries, and the other, the extremely self-centered person.

Living in their own little world, with no perception at all of public space or other people.

One is deeply self-centered (ego): the way their brain feels the world has been completely reshaped by algorithms, short videos and fragmented attention. They find it hard to truly be aware of “other people’s existence”, because they operate almost only within “the immediate feedback of their own emotions”.

Hence the lack of ability to delay gratification; what they pursue is immediate satisfaction.

Then there’s the lack of education about public space from childhood. I vaguely feel I was like that too when I was little, because of the cultural generation gap — my parents’ collectivist culture, the culture of the two of them. They were never trained from childhood to respect “order in public places”, and were rarely given a “sense of public responsibility”.

Another thing is that Chinese social classes really do have cultural dislocation; different kinds of young people really do have different social boundaries. People are animals shaped by their environment, but the other group was never invited into this kind of context.

In the process of utilitarian education, we overemphasize competition, success, how I win —

rather than how I get along with others.

So many people are poorly socialized, and the moment they leave school it’s exposed that they can’t get along with people.

On top of that, many people today have never experienced a community in their whole lives, don’t even know their own neighbors, and have never taken part in shared affairs — only “me” and “phone”.

Looking at modern civilization from a macro perspective

Thinkers’ theories make it easier to understand and reuse the theoretical models of human history.

The number “three” fits human narrative and comprehension habits very well, providing a concise, clear framework that lets us grasp the patterns of history.

Society moves from “order of scarcity” to “chaos of abundance” and then to “self-aware civilization”.

You can also understand it through Maslow’s hierarchy of needs.

It’s really about satisfying basic survival needs, then pursuing material needs, and finally exploring the need for self-actualization.

Building new social norms of behavior itself takes generational effort.

When America industrialized in the 1920s, rural migrants poured into cities and brought a mass of “low-quality urban behavior”; after WWII, as Japan’s economy took off, between 1950 and 1970 “country people coming to Tokyo” was seen as a “boorish shock”, and Tokyo people at the time also complained they were loud, messy and ill-mannered.

From being constrained to constraining yourself — a psychological transition.

Preconceived bias is itself not enough either

If art has no audience to appreciate it, it loses its own value.

I remember that jazz first took off among Black communities, but most people’s eyes carried a certain bias, because aesthetics and art themselves carry a certain color of power.

But it seems any culture or phenomenon has a cause.

Understand bias itself: without bias there is no judgment. Bias ensures we survive within a limited time; cognitive efficiency and bias are two sides of one coin.

What matters most is metacognition, a strong capacity for self-reflection: being able to notice that you’ve been influenced by a certain framework or emotion.

So when they find themselves stuck in a single perspective, they actively look for counterexamples, think in reverse, and so on.

Bias can be used by you, and you can even set up a series of mechanisms to fight it — an institutionalized rationality. A stance is a choice, not the truth.

Existence has its reasons; everything is a way for you to analyze and understand the world, or information — all nourishment for observing the world. As for the rest, it’s just personal choice.

So my choosing not to scroll Douyin isn’t a rejection of Douyin; it’s choosing not to use it based on my understanding of it. You could call that a bias, but it’s more a choice made after becoming aware. And I can understand the people who still choose to use it.

Imagistic thinking, concretely artistic thinking and creative thinking

It corresponds to abstract logical thinking.

Logical thinking has a concrete chain of reasoning.

But imagistic thinking, in fact, focuses on whole-scale emergence and transformation.

It understands and expresses through sensory images such as sight, sound and touch.

It focuses on intuition and wholeness, and the connections between related images.

An actor with a sense of belief

A deep identification with and real investment in the character being played.

Truly identifying with this character inside,

and believing in the truth of the character’s situation, motives and emotions.

Able to perceive and understand the world from the character’s perspective,

and believing the stage and the camera really exist, ignoring distractions from reality.

Every movement and line has an inner inevitability.

It’s like really stepping into a character’s identity, because of the character, not because of the script.

So it’s really hard for an actor to break through a kind of fate.

Suddenly I think of Hu Ge playing in Nirvana in Fire.

Why did “No Prospects” suddenly blow up

Why did it suddenly blow up in mainland China?

Wang Shijian was always lukewarm in Taiwan,

but suddenly he seems to have a strong internet feel.

Too abstract.

And he really delivered on his promise.

The idealist atmosphere inside the DPP today.

The Northern Qi giant Buddha

The giant Buddha of the Xiangtangshan Grottoes in Handan (the Northern Xiangtang Grottoes), in the Fengfeng Mining District of Handan, Hebei.

It was built during the Tianbao era of the Northern Qi (around 550–577 CE).

It is one of the Buddhist grotto complexes sponsored by the Northern Qi royal house, and continues the line of China’s three early great grotto systems alongside Yungang and Longmen.

The main Buddha statue is about 13 meters tall, one of the tallest cliff-carved giant Buddhas in northern China at that time.

Emperor Wenxuan of Northern Qi, Gao Yang (the founder of Northern Qi), was a devout Buddhist and also violently temperamental.

Legend has it that one night he dreamed of a giant Buddha appearing with light filling the hall, so he ordered a giant Buddha carved at the Southern Xiangtang Mountain in Handan, to “protect the country and bring peace to the people”.

—— So the giant Buddha was not only a religious symbol, but a symbol of the legitimacy of royal power.

Carving Buddha with a compassionate heart, becoming Buddha with one’s own body.

In the short-lived dynasties, everyone longed to find eternity within the chaos.

It’s really also the process of the Wei-Jin and Northern-Southern dynasties.

One is the north: Northern Wei → Eastern Wei → Western Wei → Northern Qi → Northern Zhou.

One is the south: Liu Song → Southern Qi → Liang → Chen.

Meaning is the product of an accident

What is the meaning of being alive, what is the meaning of human existence?

From the philosophical angle of Sartre and Camus, the universe itself is absurd; meaning is experienced at some single moment.

A person accidentally feels some kind of resonance in an absurd world, and that resonance is the appearance of meaning.

Inspiration emerges randomly, a mechanism unique to humans; inspiration is like a seed, and meaning grows naturally.

A present moment formed by the intersection of the Buddhist five aggregates.

The material world survives precisely because the laws of the universe are imperfect and asymmetric; this process itself is a huge accident.

If even heat death as an ending can make us feel that the process is all the meaning there is.

The conclusion of the reductio in skepticism

You say: “Please prove what you have.”

The other counters: “Then how do you prove I don’t have it?”

The basic principle: the person making the claim bears the burden of providing evidence.

If you claim “he has some thing/ability/trait”, the burden is on you: you have to provide evidence.

If he claims “I don’t have it”, the burden is on him: if he explicitly states “I don’t have it”, then he bears the burden of proof.

Is there really high and low taste

In the past I always avoided using high and low.

I think taste is a matter of difference, but there shouldn’t be high and low.

Because taste is subjective, taste is a feeling, taste is equal: some people find white beautiful, some find health beautiful, some find fullness …

But to a certain degree, the ability to have taste is objective. Taste is an ability and an accomplishment, and since it’s an ability and an accomplishment, it can be improved through learning and training, because aesthetics isn’t only biological basis; there may also be a cultural basis and an acquired basis of knowledge and experience.

Measuring ability really means having more choices.

Being able to find the beauty you want from pop, jazz, classical, even rock.

Concise, simple and crude

Worlds apart.

A huge difference in design thinking and execution complexity.

High-end concision is deliberate restraint: less is more, more is less.

Bad simplicity is a lack of ability: less is just less.

How to design a clear visual hierarchy, smooth interaction design, and maintain consistency.

Then the high-end tasteful visual UI: typography and negative space are different too, and the requirements for color and the texture of the content aren’t on the same level.

Individuals can now create something that looks high-end and feels good faster than ever

Individuals can now create web pages that look high-end and feel good faster than ever, closing the distance between crude and professional.

AI is a powerful executor, easily reaching a passing score for UI taste.

But AI can only handle eighty percent. AI lacks a soul: our grasp of detail, our observation, thinking, training and fine-tuning.

Think clearly about what our taste is, what our website wants to convey.

Mem.ai design aesthetics

One thing I really like: mem’s design really makes me love it.

Let the interface disappear, let thinking appear.

Visually it isn’t minimalism, but it very clearly gives information a sense of space, order and light.

The information density is controlled cleverly — this is a kind of space, breathing room, the optimum of cognitive load.

Then there are blurred boundaries and soft layers.

Heavy use of:

• glassmorphism

• soft shadow

• elevation

Typography goes without saying — it’s inevitable, the beauty of reason.

Then motion design and transitions are natural and fluid.

Give information a sense of space, order, light.

One direction everyone wants to take is structuring

Because of fragmented information — flashes of insight, casually excerpted content, meeting notes and so on — how do you identify and reconstruct concepts, build associations to form a structured knowledge framework, cutting text into subject-predicate-object.

Who (subject) + did what (predicate) + to whom (object) — this makes up a minimal semantic unit.

This starts with Aristotelian logic: every judgment is composed of “subject + predicate”, for example “man is a mortal animal”.

Later, in knowledge graphs, a relation is also identified by a triple: subject, predicate, object.

In natural language processing (NLP), when a model parses the meaning of a sentence, it’s also looking for these core relations.

Reasoning structures

Graph of Thoughts, that’s GoT, rebuilt on top of ToT

The idea is that a line of thought can connect to multiple nodes, achieving

merging and forking of lines of thought

sharing information between different paths

parallel exploration & backtracking

The result is more flexible and intelligent than ToT, and it can handle very complex problems, like multi-constraint optimization

The core is really the same: human thinking isn’t linear, nor simply tree-shaped, but a complex graph

Then there’s Reasoning as Planning

Which equates a reasoning problem with a planning problem, like reinforcement learning or automated planning

Splitting the problem into a series of “states” and “actions”

Using an LLM to generate plans

Using search/feedback to correct the reasoning path

Met an interesting teacher from Beijing

Feels like he teaches philosophy

He’s an anarchist, an organization-less, religion-less kind of guy

Ethics and morality

A person’s morality should be a multi-dimensional, dynamic, and carefully deliberated decision system

There are some starting points for morality, like empathy, like the principle of least harm

Reflect on whether it’s universal, that is, whether everyone in society, under the same circumstances, could act the way I’m about to act

One is consistency, one is who you want to be

Morality needs metacognition to support it, for instance being able to constantly observe your own cognitive system and moral standards, and often reflecting on why you hold those standards, whether they’re problematic, and whether they fit the development of this era

Confucius also said, do not do to others what you do not want done to yourself

Zigong asked: “Is there one word that can be practiced throughout one’s life?” The Master said: “It is perhaps ‘forgiveness’! Do not do to others what you do not want done to yourself.”

Human collectives are also interesting, that is, why do we build a team, why do we build a collective, why do we build a community, why do we build a nation, why do we build a state, why do we build a religion

The core of sociology, anthropology, psychology and political science

Why humans build these layer-by-layer collectives from the individual all the way up to religion

Humans are social animals; “building collectives” is our most fundamental survival strategy and evolutionary advantage

Many things can’t be done by one person; individual strength is limited, but collaboration lets us accomplish grand undertakings no individual could ever reach

Teams exist to solve some urgent survival problems

Communities and tribes solve the problem of a sense of security

Once survival is solved, everyone wants to live better

That’s when teams divide up the work to make products, which is far more efficient than everyone being a generalist

Then the state provides a unified market, currency, law and infrastructure, which massively lowers transaction costs and enables complex modern economic division of labor

As for identity, community, ethnicity and religion are very important: society gives a direct sense of connection (though young people today seem to have gone back online, and the sense of community seems insufficient; ethnicity provides a grander identity, built on shared blood ties, language, history and cultural tradition), religion provides an ultimate belonging far beyond the secular, connecting the individual to a sacred, eternal order (like God, the Dao, cosmic law), offering spiritual comfort and powerful group cohesion

But what sets humans apart from other animals is that humans imagine; humans can imagine abstract concepts that don’t exist in the physical world and believe in them together. This is a shared narrative

The state is a shared narrative entity, and it isn’t a physical entity either; religion is also a shared narrative, even less a physical entity; then ethnicity, where we share common ancestors and a common fate, which determines the angle of the narrative

We always talk about an acquaintance society. The premise of an acquaintance society is being offline, an offline collective, because the collective scale is small, or the urban circle is small, so there’s a trust mechanism of an acquaintance society

When the collective scale becomes enormous (tens of thousands, even hundreds of millions), the trust mechanism of an “acquaintance society” fails, and new management tools must be invented to maintain order, allocate resources and adjudicate disputes

For example, the state needs law and the machinery of violence to monopolize force

Religion also plays a powerful governing role; where secular power is weak, religious institutions are often the makers and enforcers of law

This is circular reasoning

A lot of the time it’s circular reasoning

Especially when you aren’t certain of the answer

We use a system to explain certain ethereal things, and that system itself operates inside the system

We can only keep asking why the laws of physics are the way they are?

So the end of any discipline is philosophy, theology

Narrative, history and story

Narrative is the most basic thing - it’s a structured way of telling events, containing elements like time, causality and characters. It’s a cognitive tool humans use to understand and convey experience

Story is the concrete realization of narrative; it can be fictional, real, or in between. Story injects emotion, meaning and value judgments into narrative

History is the narrative reconstruction of events that actually happened in the past. It tries to pursue “facticity”, but can never fully escape the framework of narrative

History needs to borrow the techniques of story to be understood and remembered; pure chronological compilation is hard to absorb, so historians have to assign meaning and tell stories

Story usually also needs to draw nourishment from history

Narrative is the shared grammar

The victors write history, which by itself shows that history isn’t a purely objective “pile of facts” but a chosen narrative, so narrative is controlled - by individuals, collectives, states, power

But people need narrative to build their own identity; we need to construct the present

All disciplines and all knowledge are limited

Push the questioning to its ultimate end and it’s the core of existentialism

Meaning explaining meaning

Philosophically there are still a few exits

The absurd: absurdity = the longing for an ultimate meaning inside you, for an answer outside the system + the fact that this world (including all disciplines) cannot provide that answer

Will: all meaning systems (science, morality, the humanities) are limited and man-made, so they don’t have such absolute binding power over you; they’re just tools, and that brings us to will

Game: language and meaning are like different games - science is a game, ethics is a game, art is another game; each game’s rules are limited, and feeling that they’re limited means talking about facts in the context of science, about good and evil in the context of ethics, about beauty in the context of art. Don’t force them to be unified, and most importantly keep a player’s mindset

A kind of lofty idealism and empathy

Extremely strong learning ability and curiosity

And an intense self-reflection and philosophical questioning

Yang Zhenning doesn’t believe in a god with human-like joys and sorrows

But he believes there is a creator in the universe, that the world’s exquisite structure isn’t accidental, that the laws of nature are so orderly

Cosmic religious feeling, this is a kind of “awe” and “ecstasy” felt toward the harmony, beauty and majesty of natural law. He has repeatedly described the experience of first learning about nature’s exquisite structure as “a very deep religious experience”

Science takes a step forward, religion takes a step back… but science solves finite problems, while religion is ultimately always about the infinite

Cosmic religious feeling is the “ecstatic wonder” and “deep awe” a person feels when facing the vastness, the exquisiteness of the universe and the deep laws operating behind it (such as physical laws)

That the universe is comprehensible is a conviction, and it lets them spend a lifetime revealing the laws of the universe

Science takes a step forward, religion takes a step back; whenever science provides a verifiable, nature-based explanation for (How) a phenomenon, religion’s “supernatural” authority to explain (Why) in that particular domain correspondingly “takes a step back”. The territory of science in the “domain of facts” keeps expanding

Religion is what humans use to settle themselves and to try to answer the infinite and the ultimate (why)

From sociology, religion is a virtual collective formed as a social phenomenon; the same concept applies to the state

From psychology, what it actually explains is the meaning of our existence, providing meaning

From philosophy, it’s a kind of ultimate response, the final interrogation

I’ve always been curious whether Guiguzi really existed in history

I’ve never really understood the School of Vertical and Horizontal

Confucianism feels like it’s cultivating gentlemen

The Legalists, for example Shang Yang, aimed to cultivate capable ministers, using “law, method, power” to help a state (like Qin) grow rich and strong and achieve unification

The Zonghengjia are completely different. Guiguzi’s teaching goal wasn’t to cultivate “loyal ministers” or “gentlemen”, but to cultivate people who can master “shi” (positional power, leverage)

The core idea of the Zonghengjia is the “art of bai he” (opening and closing):

“Bai” (open): means opening up, meaning persuading, allying, attacking, open strategy

“He” (close): means closing, meaning dividing, disintegrating, defending, covert strategy

How should I, given the current situation, achieve my goal through words and strategy

The only criterion for testing the Way is success

The Temple of King Wucheng

It existed in history; at its core is Jiang Ziya, who was revered by later generations as the “founding ancestor of the military school” for helping King Wu of Zhou destroy the Shang and laying the foundation of the Zhou dynasty. The Tang dynasty established this temple essentially to establish a “spiritual idol” in the military field through official sacrifice, while also sorting out the inheritance of military thought through the ages and strengthening the army’s cohesion

Jiang Ziya was clearly a minister of the previous generation, yet after more than a thousand years of evolution he was finally “canonized” in the Tang dynasty, becoming the highest deity of the national military temple

Fishing at the Wei River, only those willing take the bait

In fact during the Warring States period the military school, the Legalists, the Zonghengjia and others all claimed Jiang Ziya as their forefather, honoring him as the “master of the hundred schools”

The Han dynasty revered only Confucianism but also cared a lot about military thought

The Tang dynasty was one of the dynasties with the greatest “military achievement” in Chinese history, yet its state sacrificial system “valued the civil over the military”

The army fought wars, but there was no unified national-level spiritual leader or place of sacrifice. Before going to war, soldiers could only sacrifice to the horse god, the flag god and so on, lacking a “military” orthodoxy

Emperor Xuanzong Li Longji (reigned 712-756) decided to change this. He wanted to build a “military temple” parallel to the “civil temple”, to balance civil and military status and erect a spiritual totem for the nation’s military power. That’s how the military temple came to be, because the civil temple had existed for a long time already (the Confucius temple)

I think I really get why this older guy gets teary reading history

At dinner I shared a table with a university teacher from Beijing; he was deeply moved and also full of regrets - regret that Qin Shi Huang died young right after unifying China, regret over the golden age of the Tang, regret that Yuan Shikai didn’t cooperate with Sun Yat-sen and the Nationalists…

Qin Shi Huang had only a few short decades, created the first great unification in history, yet is fiercely contested in history

History is too heavy

But a person’s spiritual core and purity really do become something you long for more as you age and as society blunts you

This is probably compassion; only as experience and life accumulate do you feel it deeply

Qin Shi Huang, a legend of his age; nobody knows who wrote the narrative back then, nobody knows how many later dynasties’ politicians and historians nailed him to the pillar of shame in history

The “same track for carts, same script for writing, same ethics for conduct” and the commandery-county system he pushed laid the foundation for China’s political map and administrative model for the next two thousand years. The “emperor” system and centralized framework he built were carried on by later dynasties

Ending a war is very hard

States and religions are both virtual collectives

Then something naturally interesting is

Israel is a Jewish state, later defined as the nation-state of the Jewish people

The locations of the Bible: all the stories in the New Testament happen on this land. So Christians all over the world come here on pilgrimage

Israel itself is the origin of three major religions

One is Judaism, one is Christianity, one is Islam

The Israeli-Palestinian conflict is the most complex conflict in the world; essentially it’s a zero-sum game

Whoever has the right to establish a state here is the legitimate owner of this land

Jerusalem is very important, it’s unique, it’s the center of global religious pilgrimage, and this problem is unsolvable - nobody is willing to split a heart

Where the Mandate of Heaven comes from

The doctrine of the Mandate of Heaven proposed by the Zhou dynasty

That is, the ruler’s power to govern is granted by “Heaven”. But “the Mandate of Heaven is not constant” (it won’t stay fixed); if the ruler is without the Way, Heaven will take the mandate back and transfer it to someone virtuous. This was the source of legitimacy for the Zhou overthrowing the Shang, and also became the political theory basis for dynastic change in later Chinese dynasties

Mandate of Heaven

Confucianism also mentions, at fifty one knows the Mandate of Heaven, which is also a kind of understanding and calm acceptance of one’s own situation and fate

In a religious sense there is also “God’s will”

Same thing

Guiguzi

Guigu may also be a place name, not necessarily a real name

The book Guiguzi was generally embellished by later people

Apart from the records in the Records of the Grand Historian

Rites and music

Essentially it’s also a form of restraint

The Son of Heaven uses nine ding vessels, feudal lords use seven, ministers and grandees use five. If you use ding vessels that don’t belong to your rank, that’s an open political provocation

How many tomb passages the Son of Heaven’s burial has, how many a feudal lord has, none of it can be arbitrary

How to salute a superior, how to salute an equal, all have norms

Music stands for what everyone is allowed to use

The music the Son of Heaven plays at sacrifices and banquets (the Ya and Song sections), and the scale of its orchestra (like the “eight rows of dancers”), were things feudal lords could not use

This let the whole society, at least the aristocratic class, run in an orderly way

Now it’s more like Confucian outside, Legalist inside - everyone sees the Confucian part

In fact Legalism is also quietly at work

Like a prison warden ruling through a legal framework plus high pressure and fear

Human nature is driven by interests; all human behavior, however nobly packaged

Its most fundamental driving force is “seeking advantage and avoiding harm”. This “advantage” isn’t only money; it also includes merit, fame, gain, salary, power, dignity, security

So to persuade a person, moral influence is inefficient; the most efficient ways are “tempting with gain” (telling them what they’ll get) and “coercing” - hence carrot and stick

Daoism and yin-yang are philosophy; Guiguzi’s yin-yang is methodology

Bai / He is the most typical yin-yang:

Bai (yang): opening, making public, attacking, persuading, releasing, open strategy

He (yin): closing, hiding, defending, listening, absorbing, covert strategy

But any knowledge system has the limitations of its era

Technique serves the Way

Religion really is a “mechanism for displacing contradiction”

Religion really is a “mechanism for displacing contradiction”: it elevates irreconcilable interpersonal, social or inner contradictions into a relationship between person and god, person and cosmos, person and meaning

A lot of the knowledge in there can of course be filled in by knowledge systems - physics, psychology, sociology, political science, economics…

But in the end they’re all tools or equipment, tools for explaining and for playing this game well

Jerusalem

An extremely magical place

Kevin Kelly was also reborn in Jerusalem

In 1979 Kelly went to Jerusalem wanting to film the Easter ceremonies there. One night he forgot the curfew and got locked out of his hotel

He wandered into the Old City of Jerusalem, found the door of the Church of the Holy Sepulchre open, went in, and slept on a stone bench for the night

According to the New Testament, the Church of the Holy Sepulchre is exactly where Jesus was martyred; this experience gave him a magical moment of epiphany - he wanted to live a completely new life

So he threw away his beloved camera, picked up a pen, and began spreading ideas about technology

A project is also a way to help manage my own strategy

For example, when reading a book, if it’s books of the same kind, or different chapters in the same book

You can classify them and then dig deep into the content

This can also help you arrive at a good knowledge management system

All the investment books out there teach you how to read candlestick charts

All the investment books on the market today teach you how to read candlestick charts, how to grasp market hotspots, how to catch market trends, and so on

They all seem to be ignoring one very obvious problem: the value investing system. Technique is speculation; the short-term market is a zero-sum game, the long term is a value-adding market

The deeper value of formulas

Learn its technique, awaken to its Way

We all think the meaning of a formula is that it captures our predecessors’ understanding of this world; formulas are like the shoulders of scientific predecessors, and we stand on those shoulders to create and innovate further

A formula describes the operating laws of the world, or rather the operating laws within the finite world, so that we can effectively control the physical world

Thinking isn’t for the sake of getting an answer

Thinking is more a way of thinking

In the process, keeping logic and proportion and balance

Mathematical formulas and physical formulas aren’t for the sake of getting an answer

This is a lifelong harvest, something invisible and even fundamental

A formula is an ultimate abstraction; it’s the foundation of our ability to construct abstractions, capturing the essence

And the derivation of a formula is itself rigorous logical reasoning; using it is itself training the process of deductive reasoning

A formula also describes causality, a model for understanding the world

And from a first-principles angle, when you truly understand a formula, you’re no longer memorizing it; you’ve grasped the first principle that derives it, the underlying mental model

You can never precisely predict the future, but you can approach rationality by structuring your thinking about the future

Real “thinking” isn’t about finding a definite number, but about keeping a sense of logic and proportion amid uncertainty

Programming languages are the same, so are math and physics; they’re all languages for understanding the world, and they’re more precise, extremely precise

Learning to use them is merely using a tool

Ways of thinking, reflecting on their design philosophy, the essential things, what the first principles are - that’s fundamental learning

Embrace the value of the “process” rather than being anxious about the “result”

The languages you master are actually also the depth of your thinking; if you only have everyday language, it’s hard to think precisely about complex concepts like rates of change, whereas with the language of mathematics you can derive certain phenomena and understand and reason about the world’s underlying operating laws in a more precise, abstract and highly logical way

Fading into the ordinary

“Great enough to transform things is called sage; a sage beyond knowing is called divine; a gentleman who does not lose the heart of a newborn child can roam with the divine”

A person’s original excellence, yet gradually becoming mediocre, assimilated by the world

Once secularized, they naturally lose independence and authenticity, and naturally become no different from ordinary people

How do you keep a mind of renunciation while engaging with the world

How do you act in the world with a mind of renunciation

How important inner clarity and persistence are

If you compromise out of cowardice, that’s degradation

If you transcend out of insight, that’s wisdom

“Seeking” - the discriminating mind, and getting attached to appearances

What everyone seeks is peace of mind

The discriminating mind is the process

Attachment to appearances is the result, attachment to judging the result

From a semiotic angle, getting attached to appearances is being trapped in your own interpretant, but the sign is just the sign itself, and it won’t change because of that interpretant

Practice cultivates an ordinary mind, but it isn’t about eliminating the discriminating mind either; it’s not a binary opposition

Cultivating the mind is not getting attached to appearances while discriminating

As I like to say, whatever exists has its reasons

Everything in the world, including the parts that get misunderstood and excluded, is a necessary component of this world. Understanding them doesn’t mean agreeing with them, but seeing clearly the multiple logics by which this world runs

We’re all in this world, playing this game well, aren’t we

Excluding a certain language, history or essay is actually the same

Their worldview or value system has been challenged, and humans instinctively reject information that “breaks the order”

Language, history and culture are often tied to identity. Rejecting a culture is sometimes about defending the boundary of “my own group”

People are more ready to accept content that matches their own experience, and have a hard time accepting things that require rethinking

Gamified existentialism: compared with the absurdity and freedom of human existence and its lack of preset meaning, meaning itself is created by humans, so life, society, history, even suffering, can be seen as a “game” made of rules, goals, roles and narrative

We play this game with a player’s mindset

Gamified existentialism isn’t coldness either, but a kind of compassion; suffering is also part of the game, and you neither belittle suffering nor get swallowed by it

Human nature and sunk cost

I find it quite interesting that people are unwilling to let go of things they’ve stuck with for a long time

Even a major, industry or relationship they don’t like, they can’t bear to waste the previous effort

This is an obsession with the sunk cost of time

The money is already spent; rationally you shouldn’t add more, but emotionally it’s still hard to cut off

“I’ve already done so much for him/her, I can’t just let it go”

And subconsciously you’re still unwilling to admit you were wrong; continuing to invest can prove you weren’t wrong

Time + money + emotion + self-esteem

Losing hurts more than not having

Loss aversion

Cutting losses in time is important, don’t fall into sunk cost

Bailin Temple

If it were just a temple, then it would just be a temple with a long history

Zen isn’t in the deep mountains but in the everyday; practice shouldn’t be detached from real life, but should return to life with a clearer, wiser attitude, acting in the world with a mind of renunciation

In the Tang dynasty Bailin Temple was where Chan Master Zhaozhou Congshen (later revered as “Old Buddha Zhaozhou”) stayed (lived and taught the dharma). He lived here for forty years, and Bailin Temple thus became a holy site of Zen

Truth really can’t be described in words

Zen really is a kind of wisdom that reaches straight into the heart

Not seeking outside, not relying on words, a special transmission outside the teachings, pointing directly at the human mind, seeing one’s nature and becoming a Buddha

Sometimes Zen is just in the present moment, letting go of attachment to inner concepts, logic and thought - a very plain way of facing the present

The Stoic way of thinking

A very important thinking principle is distinguishing which things are in our control and which aren’t

The controllable part is our judgments, attitudes, will and choices

The uncontrollable is other people’s opinions, the weather, fate

Stoicism puts attention on one’s own reactions

The universe is orderly; reason is the way humans live in harmony with nature

Don’t resist, go along with the facts themselves

Don’t avoid pain; understand that it’s part of it

I can respond in a rational and virtuous way

Stoicism doesn’t require “having no emotions” but emphasizes not being ruled by emotions

Stoicism holds that virtue is the only standard for measuring character

Inner freedom, living in accordance with nature, a life self-sufficient in reason (or the Dao-mind)

The Stoics held: emotions come from wrong judgments. If reason is clear, you won’t be swayed by anger, worry, fear and other emotions

Knowing how to play matters more than skill

Skill is just a ticket in, just a system’s recognition of a person

But more important is knowing how to play, being absorbed in it

Hermann

Explorer of memory

Studied memory and forgetting

The Ebbinghaus forgetting curve; to keep the experiment “pure” he didn’t use words we’re familiar with (like “APPLE”), because that would trigger associations

Our brains are born “forgetful”

So this is a memory curve, and also a learning curve, showing the speed of learning a skill

Learning isn’t a straight line, and you don’t progress the same amount every day

Some thoughts on Hermann Hesse

The lonely soul seeks

His works all revolve around how to make peace with oneself

Deeply influenced by Indian philosophy (Siddhartha) and Chinese Daoist thought

A Jungian, player’s-mindset approach,

How does a person become himself?

Around this there are several themes:

The eternal duality; what’s reflected here is the Chinese philosophy of the golden mean, that is, not judging, experiencing and integrating this split; the essence of the middle way is the Way following nature, the balance of yin and yang

From Jungian psychology he learned “individuation”: a person must break free of the group’s bonds (family, society, religion), face the darkness and the light within himself, and finally live as a complete, unique self

Experience » preaching; wisdom cannot be taught, knowledge can be learned, but wisdom only comes from experiencing, feeling, making mistakes

Jung’s psychology has a “Self” state, and the process of stripping away

In Jungian psychology there’s a “Self” state, and the process of stripping away is Jung’s lifelong theory - individuation

China is a two-person culture; in a deep analysis of Chinese culture, a person has three shells

One is the persona: constructed from our many labels or social identities; what we ask is, if you take away my profession, my social role, other people’s praise of me… what’s left of me? What are my real private desires and feelings

The second is the shadow, the self we’re unwilling to admit, the repressed parts we consider bad, shameful, dark. In fact my former self kept denying these too, but the more you deny them the more easily they secretly control you. For example when a group (like a nation, a community) collectively denies its own shadow, they “project” that darkness onto the enemy, leading to fanaticism and destruction (like the Nazis). Accept and admit that you too get jealous, that you too get angry

The third is the Self; at this point you come into contact with deeper archetypes of the collective unconscious, the “contrasexual” archetype within you (the feminine archetype in a man / the masculine archetype in a woman), which connects to your soul and creativity. The Self is the final goal; it’s the true center of your personality (not the “Ego”). It contains both your light and your darkness, both your consciousness and your unconscious

When you connect with the “Self”, you gain these eyes. You no longer simply see the world in terms of “good/bad”, “right/wrong”, “us/them”, but can hold contradictions and see a complete, unified picture

Socialization is a person acquiring and internalizing social rules and norms

Socialization is the degree to which a person acquires and internalizes social rules, norms and value systems

Being self-consistent in solitude is the stability of the psychological structure; whether a person, without any external mirror, feels that their own existence is enough; a self-consistent person doesn’t depend on others’ feedback to determine their self-worth

De-socialization means leaving your original social role and identity; it can be from soldier to civilian, from religious believer to atheist, withdrawing from an ethnic identity, or withdrawing from a father-son identity

It usually comes with a “value vacuum”, a blankness of “the old faith has collapsed, the new faith isn’t established yet”

It can also be uprootedness, which describes the emptiness and pain of being pulled out of the soil, and hints at the possibility of re-rooting

It can also be disenchantment: maybe you used to exist inside a human narrative myth, but one day you leave or see through the construction behind collective narratives like the state, the nation, the family, and you feel a disenchanted loneliness - that’s disenchantment

Empiricism vs. rationalism

Human understanding is a never-ending “dance” of “bottom-up” and “top-down” happening at the same time

“Bottom-up”: sensory experience is the foundation and the building blocks

“Top-down”: language, logic and concepts are the blueprints and the tools

Understanding is a spiraling cycle

Human wisdom lies precisely in our ability to steer these two capacities, letting “concrete experience” and “abstract concepts” collide and inspire each other, so we keep climbing the ladder of understanding the world

Thought can transcend language (we can think about things we cannot say), but language really is constantly “shaping” and “guiding” our thinking - it affects our speed of perception (blue), our sense of spatial direction (east-west-north-south), and even our metaphors and emotions about abstract things (the gender of bridges)

In China, psychological problems are often seen as social relations, collective pressure

In China, psychological problems are often seen as problems of social relations, collective pressure, family roles - “interpersonal dislocation” is talked about more often than “inner conflict”. Traditional Chinese culture values the group, social order and harmony, so social psychology seems more “natural” in research and public discussion

In the US, individualism is stronger, and psychological problems are often seen as problems of “personal emotion, personality, trauma”. Clinical psychology, psychoanalysis, cognitive behavioral therapy (CBT) and so on emphasize the individual’s treatment and self-repair. So clinical psychology is more deeply rooted in people’s minds, and “I’m going to see a therapist” is generally accepted in society

America’s individualist starting point actually goes back to the Puritan immigrants long ago, where every soul could face God directly - that’s the religious version of a “declaration of individual independence”

America also has the baby boomers, born between 1946 and 1964, who grew up in the golden age after WWII and were shaped in the shadow of the Cold War

After WWII, the US became the world’s only industrial power not destroyed by war. Economic takeoff, ample employment, accelerating suburbanization, cars everywhere, TVs entering homes - this was a rare “collective ascension period” in human history

A large number of families entered the middle class

Their parents had lived through the Great Depression and war, and sought stability and order; the baby boomers grew up in peace and affluence, and instead began to question order itself

From the conformity and consumption of the 1950s to the rebellion and freedom of the 1960s, they went through a major value turn

They pursued authenticity - whether in music, lifestyle or emotion, you had to “be your true self”

From political freedom → emotional freedom → lifestyle freedom → freedom of economic choice

And then later, Gen X, millennials, Gen Z

Gen X are realists after disillusionment, the generation that saw through the illusions; compared with the passionate boomers, they’re more pragmatic and low-key, emphasizing “living truly but without over-idealizing”

Millennials: explorers of the internet and meaning; they grew up in a multicultural environment and are more accepting of gender fluidity, racial mixing and cultural difference; “I want to do something meaningful”

Gen Z: digital natives of a fragmented reality; they’ve never experienced a world without the internet; their information intake is enormous, but real socializing has decreased. Anxiety, depression and self-doubt have exploded in this generation; it’s not more individualism but post-individualism; their freedom isn’t independence but multiple possibilities

America’s spiritual history goes from “who am I” to “who do I want to become”

China’s spiritual history goes from “who are we” to “can I become myself”

Illiberal democracy

It reminds me of the recent Tanzanian presidential election, the internet ban, and the unrest.

This concept was proposed by the political scientist Fareed Zakaria in 1997. He pointed out that some countries hold elections and formally have democratic institutions, yet lack freedom in speech, the judiciary, the media and human rights. In other words — the people can vote, but they cannot really hold power or freely express opposition

Democracy emphasises that the people have political rights through elections

Freedom emphasises that individuals have rights and that the rule of law constrains the government

And “illiberal democracy” is the product of separating the two: elections exist, but the rule of law, rights and oversight mechanisms are weakened or hollowed out

Ningbo’s world-class port really is an unshakeable productive force

And it’s backstopped by powerful manufacturing

By comparison, Jiaxing has a perfect location, between two top-tier metropolitan circles

An indispensable super-node in the Yangtze River Delta core

Relatively speaking it also carries strong uncertainty

Its future potential is huge

Jiaxing is a blank sheet of paper that can be filled with stories

High-rises are also the first to face crisis

The maintenance cost of high-rises is far too high

It hasn’t shown itself yet, but there’s a crisis constantly present

Many Chinese high-rises are aging prematurely. Problems that usually take decades to appear in other countries are reported in China in buildings less than 10 to 15 years old. These problems include:

Facade deterioration: cracking, spalling and falling of exterior tiles and panels is becoming increasingly common, posing a safety hazard to pedestrians below.

Water seepage and leaks: persistent leaks from roofs, windows and internal pipes cause water damage and mold growth, harming both structural integrity and the living environment.

System failures: elevators, the key arteries of any high-rise, are breaking down more frequently. Plumbing and electrical systems also show signs of strain, causing service interruptions and increased fire risk.

Drones, technology and remote work are already changing the rules of the game. Even for a detached villa in a remote suburb, the commuting disadvantage will be greatly weakened, and its space and environmental advantages will stand out more

Existence precedes essence; a person is the sum of his own choices

What’s given is a chance to choose

Not deciding for them what is good, but giving them the chance to decide what is good for themselves

The minimal action set of information

A problem everyone in information work has

Collecting or processing information excessively

And ending up short on action

Build the minimal action set

This is an action-centered information drive

Accept a certain degree of cognitive limitation and imperfection

One marker for measuring my own adulthood is

After becoming an adult, I started raising myself from scratch

Among all commuting modes in the US San Francisco Bay Area

Among all commuting modes in the US San Francisco Bay Area, car commuters average about 30 minutes

Commuters in Silicon Valley spend an average of 73 minutes a day in transit (round trip), which works out to about 36-37 minutes each way (but this includes non-ideal conditions like traffic jams and delays)

Without a purpose, it’s always aimless exploration

And this process in the end lacks action and the creation of real value

People’s existence is about proving their superiority and sense of being

Most people just long to be seen or heard

Keep some empathy — that’s the initial link between people

Understanding balance deeply, again

Often I can understand it

There’s no shortage of options

No shortage of understanding either

But what you face is usually that you can’t have both

And it’s not a question of choosing, of A or B

It’s more like a process of balancing and adjusting within a system — a question of how the system designer designs it

Like China’s firewall, which is also a flexible adjustment process

Understand conflict, understand tension

Adjust this system through constant trial and error and feedback

To some degree, how a city retains people and how a product retains users

To some degree, how a city retains people and how a product retains users are consistent in principle

They’re consistent in terms of long-term retention rate conversion

Some of Feynman’s methods

In the cross-domain process, essentially it’s not just the associations between cards

Nor just the relationships between semantics

It’s more about the human brain’s understanding of concepts, the links between concept and concept

All beings suffer

It’s human nature

Unchanged for a thousand years

People

Noble or lowly, whatever

In the end, they live

Everyone is looking for their own place of soul-rest

I just long to connect with more people

I just hope to help more people right now

At least a bit more happiness in the world, right

And my own soul gets a certain rest too

So history has no standard answer

History has no right or wrong

History has no good or evil

History is not a binary opposition

History just triggers its own meaning of existence and sense of belonging through its continuous evolution

History is for knowing yourself

You don’t know what kind of surprise it will give you

Enormous uncertainty and reward

An ordinary idea, if not retold, recorded or acted upon

An ordinary idea, if it isn’t retold, recorded or acted on, has an impact approaching zero after two days; after seven days you can’t even recall it

What the brain does before a question

Essentially it’s also memory

Essentially it’s storing and linking concept nodes; the number of nodes and the quality and density of links is the amount of knowledge

When these nodes and links are rich enough, they combine into more complex structures — mental models. This is your internal explanation of how some domain of the world works

So the generation of mental models in the brain is a mental model that can be reused, explaining the past, predicting the future and connecting everything

At that point the emergence of a question is actually an error or a blank in the network

For example, when your brain is maintaining this network, there’s an alarm: I know those nodes, but there’s a node or a link I don’t know

So a question means you already have a preliminary knowledge network and you’re constantly reviewing, expanding and correcting it

Thinking now, if I had another chance at the last move’s decluttering

Thinking about it now, if I had another chance at last time’s move and decluttering, I should constantly remind myself to do the preparation for clearing things out first

A to-do list that doesn’t need timeslots

Not for planning your own execution or forcing yourself to execute tasks, but a to-do list for discovering yourself

If a task isn’t working, naturally switch tasks, or roll the task over

Through a series of methodologies

Defining a new carrier for information intake, no longer article-type

Defining a new carrier for information intake — no longer article-type but a card-type method of receiving information

Reading through the card reading method

The learning efficiency of Feynman cards is comparatively the highest of all types

A very lightweight tool

It’s only there to remind me.

Only there to give suggestions.

It doesn’t need any confirmation.

It can do priority ordering.

It can, based on its understanding of you, give suggestions and guides closer to what you’d actually do.

It should be used on the desktop version.

Lean startup and MVP are themselves about validating by minimizing cost

Lean startup and MVP are themselves about validating the biggest uncertainty by minimizing cost.

My own two methods are: once I have a clear understanding of the system, identify the most important part of the system and the most uncertain place, and start validating from there.

This is a logic of how to do things correctly — a technique.

The other is the logic of the Way: how to do the right thing.

Take the result as the target, think backwards, flip the thinking around.

The point of starting from the result isn’t to derive the correct steps — that’s the logic of a three-dimensional cross-section. But I think what matters is hitting one point, and that point is a kind of feeling: the soul’s smallest common measure, Apple’s minimalist logic, the spark of the user glimpsing the future for the first time and going crazy over it.

This is a strong confidence in one particular view of your own: to realize your core value, you’re willing to cut away most of the features that look pretty good, and say no to the product.

On one hand an MVP is validation; on the other hand it’s a complete experience, a complete and delightful closed loop of experience.

I know what is beautiful, what is harmonious. Thinking about beauty, thinking about why it’s beautiful, thinking about what beauty is — this confidence comes from tens of thousands of hours of thinking, observing, and deep understanding of art, technology and the humanities.

A good product should spark people’s creativity and potential

A good product should spark people’s creativity and potential.

The past is all information for building your understanding of the world

The past is all information for building out your understanding of this world and your way of handling it.

You are a subjective object, trying to understand this objective world.

So your past experience is a kind of information that supplements yourself.

A humble little AI assistant

Only reminding me what to do,

not forcing me to do it.

Make it a little assistant that has feelings.

Does dinoai really have anything worthwhile about it

Can’t find any ~

I’ve come to understand a kind of emptiness

It’s the nihilism of returning to reality after escaping it.

Because the real world is, in the end, what we have to return to.

That’s the main line of our lives.

The virtual world is, in the end, a side quest.

If you could really take the experience of playing the side quest, and the results you get, and reuse them in reality,

that might really be very valuable.

The core mechanism is emergence

A new mode of note-taking.

I want an anti-fragmentation era

Anti fragmented information.

wtf.

Why can’t fragmentation and structure coexist

Why can’t fragmentation and structure coexist?

aha moment conversion

It’s essentially the subjective experience and objective result of previously stuck cognition being opened up in an instant.

A sudden understanding, from blockage to flow.

The insight process is tied to activation in the hippocampus and other brain regions, suggesting that memory and associative processing play an important role in the “flash of insight” — the hippocampus.

The hippocampus participates in both spatial memory and semantic memory, but its relationship with episodic memory is the most typical and central, for example time-place-plot.

A good note-taking app lets you focus on exploring, on learning

A good note-taking app lets you focus on exploring, on learning, on creating.

If it brings no disruptive effect, will it really draw people in deeply

If it brings no disruptive effect, will it really draw people in deeply?

And once you really get hands-on, you fall into it without noticing

And once you really get hands-on, you fall into it without noticing, and the use cases for Roam Research will naturally see new development and discovery as time goes on.

Thinking about bidirectionality

Roam’s bidirectional links are true bidirectional links; OB’s bidirectional links, which depend on file directory paths, can’t be called “bidirectional links”.

The common people’s plainest wish is not to starve, but to eat their fill

If even that can’t be achieved,

the axis that keeps the scales balanced will break.

But also, nothing is built without breaking.

The complicated thing about the internet today is

On one hand it promotes fairness.

On the other hand it has broadly raised the narration of history.

But people’s memory and focus are limited, in the end.

Maybe in the future it will all be abstracted into one sentence.

Maybe it will also be dug out and explored deeply by someone.

Isn’t time management treating people like machines

Energy management matters more.

Time management is just the path leading to efficient energy management.

Being able to manage your energy well and finish important tasks in your high-efficiency hours of the day is a better way to improve efficiency.

And how to balance your own energy is something you can regulate.

Elsevier’s vision

Seeing that Elsevier’s wish is to jointly advance human progress …

The most basic morality of writing is accuracy

The most basic morality of writing is accuracy.

In symbolic thinking, is it also a kind of mismatch between the sign itself and the object

Pull the object out, and attach it to other concepts or signs.

In the end, humans hold the right of interpretation.

A sense of fate and a sense of belief

The tragedy of an industry is that the best group of people never chose this industry, and slowly drain away.

Comedy solves near-term problems

Tragedy solves long-term problems.

A good way to write a story, a screenwriter’s method

It’s really to imagine the best actor in the world right now and write from that actor’s perspective.

Then the script can instantly come alive.

How to improve people’s creativity

How to improve people’s creativity in a general way?

How to make the path from coming up with an idea to executing it as short as possible?

A very good means of scene-based description

It’s about stepping into your own perspective, into a certain scene, and describing it plainly.

The story of the scene, the setting that sets it off.

The characters’ expressions.

Don’t describe your own feelings or your own thinking; leave some thinking for them.

Only at the end reveal your own thinking.

Another very important metric for motivating action is reward

Another very important metric for motivating your own action is reward — a reward for discovering yourself and sensing that you’ve become better, with higher completion efficiency.

This may also be part of the sense of mission that a todo carries.

Because sometimes I actually have some fairly good methods and I know them, it’s just that I never put them into practice, which is a regret.

For example, occasionally I remember that before I go to Europe in the future, I should practice my spoken language to a certain level of fluency.

So it should really be an oral-conversation aid helping me.

From “What Does This Even

From “What Does This Even Mean?” to “Actually, This Makes Sense”

Most of the time, because you don’t understand, you don’t get that aha moment,

and so you don’t get sustained positive feedback either.

Only by first entering your own knowledge domain through chat mode

Only by first entering your own knowledge domain through chat mode, trying to understand.

Once you have a preliminary model of the concept network, it becomes a very good entry point for asking questions,

because at that point it’s easy to find gaps.

What the hell is “recommended related notes”

The very purpose of notes existing is to use the notes themselves to produce new associative knowledge.

😮‍💨

People increasingly want small and beautiful

Increasingly not that fond of big and comprehensive.

Even something very complex, like Douyin.

How to boil it down into a few simple steps.

Lowering cognitive load matters.

Jobs’s “simplicity is not”

Jobs’s “simplicity is not simple”.

A kind of state of zhongyong — actually it’s seeking simplicity within complexity, keeping the soul within minimalism.

Not preparing

Just improvise.

Be confident in yourself.

High-order confidence.

You don’t need to prepare anything.

In some small details or matters

Cultivate yourself more.

Keep at it, and have feedback.

Trust in the power of compounding

Haste makes waste.

Card = atomic unit of

Card = atomic unit of thought

A card’s spatial position on the board expresses its semantic relationship with other cards, similar to a two-dimensional mind map.

The connections between cards form a knowledge graph.

Links are the knowledge graph formed by the connections between cards, cross-board references and tracebacks.

A Note is a longer note, formed by combining multiple cards.

A rather interesting discussion

https://www.reddit.com/r/simpleliving/comments/pbd9jz/i_moved_to_africa_to_live_more_simply_and_it_did/?tl=zh-hans

Personality determines choice

Choice determines destiny.

Then what is personality?

How is destiny formed?

Personality is a person’s relatively stable system of psychological traits, determining how they think, feel and behave in different situations.

Generally it’s:

cognitive patterns (how you see the world)

emotional responses (how you perceive the world)

behavioral tendencies (how you act)

A more active dopamine system → more prone to risk-taking, seeking stimulation.

Strong prefrontal inhibitory power → more rational, more planned.

A strong amygdala response → more prone to anxiety, alertness.

Jung also said personality is the result of your repeated choices.

It’s actually also that choices make the personality.

As the internet pervades everything, will humans want to return to reality

As the internet pervades everything, will humans want to return to reality?

A point that really resonates

In the internet age, the only learning method left to us is fragmented learning.

A kind of biologically instinctive wisdom

In the modern age, there are precisely people like this: simple by nature, not blinded by worldly matters and the complex information world,

clearly focused on their own world and the things they want to do.

Leaves fall, you sweep them clean — isn’t that compassion

Isn’t life just constantly discovering, perceiving and thinking?

Not trying to prove you understand

But in the process of chatting you always humbly want to drive yourself to understand.

Willing to ask, willing to listen, willing to fill in background knowledge.

And if you have some underlying logic or thinking framework,

what’s easy to achieve is structural transfer.

Understanding the market

Short term is zero-sum, long term is positive-sum.

Long term requires expanding the market through technological innovation, experience innovation and cognitive innovation.

In the short term, it’s grabbing share from a fixed pie, driving down costs, absorbing rivals.

Memory palace

A very old and very powerful memory method

It converts abstract, dry information (like numbers, names, concepts) into vivid, bizarre visual images (encoding), then “places” them at specific positions (Loci) in an extremely familiar space in your mind (the palace)

When you need to recall that information, you simply “revisit the old place” in your mind, walking along a fixed route, and you can “see” the images you stored at the corresponding positions, and thereby “decode” the original information

This is also a kind of decoding-by-contrast process

The palace has to be a space you’re very familiar with, like your home, or the Minecraft world you played

The route has to be one-directional

Then you encode information into images, or decode, place your information, then walk and recall

People have stopped thinking

How terrifying people could become

Phones take up our time

Will we really go back to reality in the future

From the perspective of absorption efficiency

For linear structures, the cognitive load is the lowest. It provides a clear, pre-set path that ensures you won’t miss any “prerequisite knowledge”. You have to learn “variables” before you can learn “functions”

A tree structure best fits interest-driven learning; this structure is driven by curiosity, a natural process of exploration, a process of self-directed exploration

An idea is just an idea

An idea is like planting a tree

Bury a seed

Nobody knows what will happen to it

Nobody knows what it will look like in the future

The teacher thinks there’s a gulf between teaching and learning

But haven’t I also always thought there’s a gulf between learning and doing?

And a gulf between knowing and acting

Imitating is of course fine

It’s a very safe playbook

Like just pointing you straight at a God

But isn’t it also constrained by what’s in yourself

Yet they are ultimately playing to their own strengths, and those strengths aren’t ones you have

Imitation is of course important; that’s valuable secondhand experience, and at the starting stage of course it is, but to surpass it you have to walk your own road and be clear about your strengths and weaknesses

I no longer have the patience to carefully design a complex system

There used to be a lot of parts I considered very important - design, focused reading, coding - I haven’t felt the joy for a long time

Becoming aware of this

I feel like the essence of attention is also a capacity for choosing how to allocate resources

At any moment the brain is receiving a large amount of information (sensory input, thoughts, memories, emotions)

But attention is a filter, determining which information gets amplified and which gets ignored

When you become interested in something and expect a reward (a sense of accomplishment, the pleasure of understanding, the joy of creation), the brain releases dopamine - which not only makes you happy but also strengthens concentration

So to some degree attention itself is an interest-driven energy allocation mechanism

Which lets me think in reverse about why I feel impatient

Dopamine fatigue: too much of it, sustained, lowers curiosity, like eating sweets for a long time

The meaning loop fails: the system or goal no longer has personal meaning, and I don’t want to invest

Spiritual resources get diluted, which is a frightening thing; the modern information environment keeps interrupting attention

Potential burnout or transition: maybe it’s the era evolving, and I really need a better or more comfortable way of existing

Curiosity and focus don’t really disappear, they only get suppressed or displaced.

A good measuring stick

Imagine if you take 5-10 years as the unit

Whether you can understand the development trend over 5-10 years

Then it’s investable

Language reflects a person’s thinking habits

I suddenly realized, people who often say thank you tend to value politeness and respect

People who habitually say “there’s no way” or “it’s all the same” may lean more fatalistic or passively accepting

People who like to say “I think”, or “from my point of view”, emphasize the individual perspective and equal dialogue

It’s a manifestation of cultural habit; many older generations in China tend to use honorifics or “we” rather than “I”

Sometimes “whatever you say” is trust

Sometimes it’s evading responsibility

It might also just be that they can’t be bothered to discuss it again

You have to look at the specific situation, and put yourself in the other person’s shoes to think about why they’re saying it that way

Thinking about grandpa dying from sleeping pills

Wasn’t that also a heavy blow to several generations of this family

Life is really short

The physical pain before the end

Why Luckin’s beans are so rich and fragrant

Luckin keeps finding ways to source beans from around the world, including Yunnan in China, Brazil, Ethiopia and Colombia

Generally good beans are very uniform in color and size with no broken beans; also the beans are usually light brown or cinnamon colored, dry on the surface and not oily; if the color is very dark, it’s a dark roast

And there’s usually a score, like SCA Score: 85+; generally above 80 is specialty coffee

Do you want users to absorb information with your tool

Or do you want users to use your tool to savor knowledge?

One issue with expansion is too many options

About how to display and expand those extended branches elegantly

Too many choices easily leads to one problem: choice overload

The point PC users easily adapt to is dragging up and down

Put the roof at the very top of the page

Then explore questions and knowledge cards from top to bottom

Tip: here’s a question - the relationships between cards, and how to make them valuable?

Barcelona as a transport hub

The Pyrenees inside France

42 days is the most suitable time to walk the whole route; usually it actually takes 30-35 days

The French Way is currently the most popular and busiest

Absorbing information and savoring

Absorbing information is a mechanical, goal-oriented process

Deep processing is a kind of savoring

Processing with rhythm, a sense of time and space

In cognitive science itself, absorbing information involves:

perception

short-term memory

preliminary understanding

For example a user’s query is usually fast and linear, with the goal of satisfying an immediate need

Savoring is sensory, immersive, similar to an aesthetic experience

When users get information with curiosity or a desire to explore, the process of absorption may naturally extend into savoring

The Way is just there

Each person has their own experience and cultural background

Each holds their own view

A few points I can think of for now

At the very least there can be a continuous subscription way to subscribe to a certain platform’s web pages, for example reading five pieces a day, and during reading you can give instant feedback, then the LLM adjusts the plan moderately according to your state

You can also set some small tasks and plans in your life workflow, and they can be broken down moderately according to the type and difficulty of the plan, encouraging users to complete a small step first

A couple came into Starbucks

They look like local university students

The guy looks very steady

The girl is very lively, occasionally acting cute

Later two guys who seemed like buddies or roommates came over to the guy

And then he talked about games with his buddies

How to accept that you really aren’t that interested in something

Analyse why you’re interested in it

Analyse why you’re not that interested, and what the reasons are

Analyse whether the reasons you’re not interested can be adjusted by appropriately changing your environment and your thinking

Appropriately increase your exposure to the factors that raise your interest

From knowing to awakening

While Westerners have been looking to the human brain for methods of human action

Eastern predecessors studied the philosophy of mind and personal cultivation — hence the idea of going from knowing to awakening

From knowing to awakening means: knowing is the level of knowledge, obtained or stored through language, logic or concepts. Knowing usually borrows others’ experience; it’s a process of accumulating cognition, but it doesn’t necessarily change you inside

Awakening is awareness, an inner self-realization; it has an experiential, intuitive quality and is usually tied to sudden insight, feeling and a sense of being

Knowing is “taken from outside”; awakening is “born from within”

So essentially this process is also a shift from quantitative change to qualitative change

Accumulate (know): come into contact with a lot of information and concepts.

Internalize (think): chew it over repeatedly in your mind, combining it with your own experience.

Trigger (circumstance): under some opening, scene or conflict, you suddenly “see.”

Generate (awaken): consciousness converts into insight — not just understanding, but becoming part of yourself.

Embodied memory

Knowledge and experience don’t exist only in the mind (the cerebral cortex); they’re memories “written into” the body through physical movement, the senses and emotion

The body isn’t only the “tool” of experience; it is itself the “storage” of experience

It can be called up without language, and it’s hard to lose — engraved into neural circuits and muscle memory

In the end, a person is defined by what they’ve been through, not by the principles they know

Uniting knowing and doing is very hard

Don’t be afraid of rework, or of starting from 0

Don’t be afraid of rework, or of starting from 0. Rework is a completely normal thing; you shouldn’t fall into perfectionism out of fear of rework. The core is trying as early as possible and getting feedback — that’s the fastest way to grow

Memories of a mentor in Chongqing

The last time I went to Chongqing was two years ago, at the end of 2023 — the last city I picked before crossing into the new year in Hong Kong

He was something like an enlightening teacher; a brief encounter, yet it had a deep effect on me

Everyone admires strength

Following the right founder is especially helpful to a person’s growth, particularly for your first job

People learn thinking frameworks, language and standards of judgment by imitating excellent examples

Without realizing it you fold his framework into your self-concept, and over the long term it becomes “your own thing”

Self-knowledge, once again

I can’t eat spicy food

What was confirmed last night is that I can’t drink alcohol

So,

Don’t eat strong flavors

Don’t eat too spicy

Don’t drink

Otherwise I still lack self-knowledge

Some subconscious habits to avoid

Excessive stress makes the body release cortisol; this hormone stimulates the sebaceous glands to secrete more oil. Under heavy stress, the HPA axis is activated -> the hypothalamus releases CRH -> the pituitary secretes ACTH -> the adrenal cortex secretes cortisol

Some dietary issues: avoid certain sugars, especially things containing a lot of milk or sugar, which may affect hormones

Spiciness is very direct stimulation. Spicy food causes vasodilation, leading to facial flushing, redness and a rise in temperature

Hands carry enormous numbers of bacteria, which is very dangerous; frequent touching transfers these bacteria to skin that’s already inflamed and has clogged pores, very easily triggering or worsening folliculitis and forming pustules

Some eating habits:

Avoid high-sugar foods or drinks; don’t drink them

Avoid dairy

Avoid high-oil, deep-fried and processed foods

Avoid alcohol

Cutting in through common sense

Minimalism advocates less is more

The paradox of choice: More is less

Although we usually think having more choices makes us freer and happier, when there are too many options it instead makes us anxious and indecisive, and ultimately lowers our satisfaction with the choice we made

This may also be a problem for modern people, especially one brought by future technology and the information age

Decision paralysis is normal too; when facing too many choices, people easily become inert

Fallacies and paradoxes

A fallacy is an error in the reasoning process itself — like a faulty machine: feed it the best raw material and it still produces a bad product. The conclusion may be right or wrong, but the derivation itself has a logical hole

A paradox is an irresolvable contradiction in the result itself — like using a perfect machine, perfect raw material and a perfect process to produce a strange product that is “both solid and liquid.” A paradox’s conclusion usually challenges your common sense and logical foundations

A paradox generally reveals the limits of our language, logical system or worldview. Studying paradoxes is about deepening understanding and pushing the boundaries of knowledge

Oh, you see — it requires introducing new concepts, using a new framework, acknowledging the system’s limits

Self-knowledge

Knowing what you can do

Knowing what you want to do

Knowing what you can’t do

When I say I have self-knowledge: I really can’t stick to exercising, to working out — I admit it

I admit I’m not frugal enough. I’ve always had that idea, but I’ve never managed to change it, so in the end I can just accept it. 1. I can’t change it — this is just how I am. 2. I really do understand that I’m this kind of person; I have self-knowledge

What can I do? I have no self-control and I play on my phone — so I train my brain’s cognitive response. For example, in a café I habitually don’t look at my phone and focus on my laptop, occasionally wandering around inside the café

Another example: I really haven’t stuck with running, but if there’s a desire — wanting to see the city, take photos everywhere, get to know the city — I let myself learn the city through running, and that makes me happy. Or I can break the task down with a systematic method: first toss my running shoes downstairs by the building entrance, just complete that step first

I can’t save money, and I haven’t kept up with bookkeeping. That’s not anyone’s fault — it’s really that I lack understanding: one, a lack of understanding about money; another, a lack of understanding about myself. I’m very afraid of seeing the truth — that I spend money like this

The nature of bookkeeping

It looks like just the operation of keeping accounts; some people think it’s a habit. What’s hidden underneath this question is how a person views themselves, views survival, and views reality at the most fundamental level. If you treat bookkeeping as an action, a method, you’ve already missed the higher level and can’t see its true face

Bookkeeping is hard because what they want isn’t a mirror but a beauty-filter camera. They don’t want to see their real selves — it’s too naked, too embarrassing. When the ledger tells you plainly that your income can’t support the “quality life” you long for, that most of your money goes to maintaining vanity and instant gratification, and that your ability to withstand risk is nearly zero — for a person living in “hopes” and “feelings,” that’s a judgment

Not being able to keep it up is essentially self-protection: refusing to admit that your logic for surviving has a hole in it

Some people keep accounts to save money; this approach just exposes the cause of their real poverty. That’s the root of the pain, so it’s hard to persist

For a person with self-knowledge, bookkeeping is a way of knowing themselves; they don’t care whether the numbers look good, they care whether the numbers are true. The ledger is the map of their personal economic world. With a map, they can plan a route — know where the mountains are, where the rivers are, where you can walk, where the cliffs are

So, essentially: do you want to solve the problem, or to avoid it with a clear conscience?

Reason seeks truth from outside inward

Feeling expresses experience from inside outward

Are they two sides of a coin? Can they really coexist?

After much thinking, there’s only one way: reason “domesticates” feeling while preserving its vitality

A typical trait: logic-dominant yet emotionally rich, understanding human nature fully while avoiding the coldness of pure reason or the impulsiveness of pure feeling

Moreover, the opposite of feeling isn’t reason — it’s numbness

Me before sophomore year: introverted, insecure, shy, timid

In sophomore year I came into contact with the open-source community, open-source projects, and the outdoors; for the first time a passion stirred inside me, and I changed from within and without

In senior year I joined a buddy’s startup — a change of environment, value-driven. I abandoned the traditional, mainstream path, gave up stability and embraced change

A year after graduating, living nomadically, traveling around the world — a new turn in life, expanding my range of contact and cognitive system and strengthening my inner core

But I’m also thinking deeply: what exactly is the next turning point in life???

The solution to decision paralysis

There’s actually one core thing: don’t add too much information, don’t over-worry about the decision being perfect, and don’t gather too many possible options

Too many options actually makes decision paralysis more likely

Our brains — especially the prefrontal cortex responsible for decision-making and analysis — have extremely limited working memory capacity. Think of it as a computer’s RAM. When there are only 2-3 options, you can easily compare them in your head and weigh pros and cons. But when options grow to 20, each with 10 different attributes (price, color, features, reviews…), the amount of information you need to process grows exponentially.

Similarly, because humans’ aversion to “loss” is far greater than the pleasure of an equivalent “gain.” Put simply, the pain of losing 100 yuan is much stronger than the joy of gaining 100 yuan.

So often the brain isn’t choosing the best one but, in extreme fear, trying to “avoid making a choice you’ll regret”

The one-way door and two-way door approach to decisions is extremely practical

Two-Way Door decisions: reversible, low-impact decisions. Trying a new restaurant, watching a movie, buying a returnable piece of clothing. If you choose wrong, the consequences aren’t serious and you can easily “walk back.” For these, decide fast and don’t waste too much energy

One-Way Door decisions: irreversible, high-impact decisions. Marriage, having children, choosing a career direction, making a major investment. These deserve more time and energy to study

When two options are very close in value, there’s even less reason to keep analyzing; the marginal benefit of continued analysis is nearly 0. At that point just decide randomly with a coin flip, and decide fast

Another method for working efficiently is “Parkinson’s Law”: work expands to fill all the time you allot to it. If you give yourself a week to choose a restaurant, you’ll agonize for a week

So set a time box: within the hour from 3 to 4 this afternoon, finish choosing a travel destination. When 4 o’clock hits, whatever the result, you must make the decision rather than procrastinate

It suddenly occurs to me that a friend of mine seems to be this kind of person too — very fond of planning and arranging ahead of time, which actually means that up to the deadline they keep collecting more meaningless information because they have to make a choice

Cognitive dissonance

When a person holds two contradictory thoughts, beliefs or attitudes at once, an uncomfortable tension arises — cognitive dissonance

The brain will ease the emotion and subconsciously change one of the thoughts to make them consistent

For example, one fact is:

All the signs in her behavior show she’s just deceiving me, stringing me along

I’ve given her feelings/time/money, and I hope she likes me the same way

The healthy state is cutting losses in time

But the brain’s easiest state is actually denial: she isn’t deceiving me, those signs are all misunderstandings. Deep down she likes me

The subconscious brain finds excuses — a perfect excuse and external support

This is the “fishing” technique

The reward doesn’t appear every time but randomly, unpredictably. This keeps a person hoping and continuously investing, hoping to get the reward again. Gambling is the most typical example

There’s also the sunk cost fallacy: “I’ve already given so much; if I give up now it’s all wasted”

Just like those men in videos, always feeling they’ve given so much, not thinking about cutting losses, but — like with stocks — thinking about losing less, or winning part of it back. After all, compared with gaining money, what people find harder to accept is losing money

Humans are born inclined to avoid pain. So grabbing onto a 1% false hope is much easier than facing a 100% cruel reality

Which matters more: reading ten thousand books or walking ten thousand miles

I’ve gradually come to understand that reading and traveling were never either-or. Reading gives me depth, the sedimentation of other people’s thought and experience; traveling brings breadth, the felt reality of the body colliding directly with the world

If you only read, feeling floats on the surface of the page

If you only travel, experience is hard to sediment into thought

The times I gained the most were reading on the road: when the body’s memory and emotional touch points meet the explanation and extension of text, secondhand experience and firsthand experience finally close the loop. In that moment I’m not just witnessing the world — I’ve built my own understanding with it

Is there altruism in posting to Moments

Something you really feel has value?

Admitting I’m not a saint

And I don’t need to pretend to be a saint

I have my fragilities and desires too

I’m also drawn by money, status, the opposite sex — this is my human nature

I can notice this process in myself. I don’t have strong self-control; I’m better at designing a mechanism

Mechanism > impulse: let the external environment help manage your human nature

Notice your own desires, and don’t let them spread

Often think about what kind of person I want to become

If I slip, how do I pull myself back? Is it a one-way door?

No need for oaths or verbal restraint — rely instead on values + a self-discipline mechanism + a clear awareness of consequences

I seem stuck in a loop: curiosity about everything drives me to keep learning

I seem to have fallen into a loop: curiosity about everything drives me to constantly learn new things, but the pleasure of this “input” crowds out the energy and time for “output” creation. This is essentially an infinite loop from “thinking of it” to “knowing it,” and I can never quite cross over into “doing it”

As someone with a high need for cognition, learning and understanding complex concepts itself brings huge psychological satisfaction and dopamine reward. This process is immediate, controllable, and has almost no risk of failure. By comparison, creating a product is a long process full of uncertainty, frustration and delayed gratification. The brain instinctively chooses the former — the “path of least resistance”

But it also feels like a kind of pseudo-execution: it happens in a safe sandbox of thought rather than the messy real world. We use the diligence of building cognitive models to cover up the procrastination of creating value — but it also relieves the anxiety of having nothing to do

What counts as a vacation

A vacation is a feeling

A shift in state of mind

Only when I truly relax and immerse myself in a rhythm I like does it count as a vacation

A vacation doesn’t need a beach, a faraway place or a special venue; it can happen in any moment

When I throw myself completely into something I love, that instant is a vacation. It turns out the decision has always been in my heart, not in the scenery outside

When I have a problem, dare to face it directly

Even better, try building a few systems

Try to find a solution to the problem

For example, if I can’t stick with working out

Then build a fitness group and check in every day

And in an encouraging way, not through comparison or punishment

By digging into your own exercise habits, immediately run a social experiment

If I’d realized the joy of learning earlier

There are still some uncontrollable regrets

Learning isn’t essentially an accumulation of knowledge but a way of perceiving and connecting with the world

New skills and perspectives open windows between you and the world

Exploration is inherently joyful; this sense of drive is more internal than external evaluation

Intrinsic drive makes it easier to find interests and to accumulate depth and breadth earlier

If I’d realized it earlier, I’d have tried earlier, made mistakes earlier, connected with different people earlier, crossed boundaries earlier. The result might be a more open mindset, less fear, and an easier time finding my own path

Question

The nature of inferiority

I have decent self-knowledge: I know who I am, what I have, what I want

Friendship is essentially exchange — emotional exchange or material exchange

Many people have a cognitive mismatch: feeling inferior at the level of value is a product of comparison, but perhaps they’ve chosen the wrong frame of reference

Inferiority is the price of awakened consciousness; animals without consciousness can’t feel inferior

Inferiority comes from equating “what I have” with “who I am”

An inferiority feeling and an inferiority complex are also different

An inferiority feeling is recognizing you’re not good enough — that’s a driving force

An inferiority complex is falling into the narrative that you’re incapable — that’s a binding

So inferiority is a necessary road to strength, perhaps the starting point of awakening

Curiosity and interest are essentially both satisfying the brain’s reward system

Curiosity and interest are essentially both satisfying the brain’s reward system, especially the dopamine circuit; when it’s activated it’s essentially a reward, not a drain

Interest alone isn’t enough: if you only stay at the level of dipping a toe, the brain will soon stop growing because of “comfort.” That’s when you need deep digging and high-reward feedback

So you need a flow state and experience — you need to manufacture the sense of flow

People ultimately still need to engage with the world

To truly act, truly create

To find some sense of value in this world

And to constantly think about what you want and what kind of environment you need

Move to a good environment to shape yourself

Subtly influence the environment around you through your own presence, forming a positive loop

Think more about which environments around you are positive and which are negative

Which parts are “dragging you down” (negative colleagues, say)? Which are “lifting you up” (a supportive community)? Then list 3 small actions: join an online interest group, adjust the layout of your workspace, or read one more book that sparks thinking. Remember, the environment shapes you, but you can also “hack” it

There’s a concept in psychology called “mirror neurons”: others unconsciously imitate your energy field

So socialize and talk more with excellent, positive people

Following your heart without overstepping the rules

The “awakened mind” that accords with objective law operates within the framework of the “heavenly way” (objective law). At this point the desires born of the mind naturally follow the law. It isn’t that a person is bound by the “rule” — the person becomes the “rule” itself. This is the level of the “Dao,” the unity of mind and rule

Masters do it effortlessly; ordinary people spend 99% of their strength on “doing things,” which is essentially using “technique” against the “Dao,” applying brute force in the wrong direction, creating enormous friction and internal drain

Masters spend 99% of their strength on “seeing things” and “planning things.” They exhaust their mind studying laws, deducing cause and effect, completing top-level design, and deliberately training certain general abilities. This cognitive process is extremely “effortful”

Know who you are, know what you want, know what you have, know what you’re good at, know what you can and can’t do, know what this world is like

And naturally you can establish yourself in this world

Looking back at history is to understand ‘where I come from’

Looking ahead to the future is to decide “why I journey”

Ultimately, both are narrating the present

So live in the present

We need a good present to practice the future

And we need the present to avoid a past too painful to recall

Culture and the self

Markus & Kitayama’s 1991 paper in Psychological Review, “Culture and the Self: Implications for Cognition, Emotion, and Motivation”

A few core points:

People in different cultures understand the self, others, and the relationship between the two in fundamentally different ways

These differences deeply affect cognition, emotion, motivation and behavior

Self-construal is the key mediating variable connecting culture and psychological processes

Self-construal emerges as the individual’s cognitive representation of self, others, and their interdependent relations — part of the self-relevant schemas that organize, evaluate and regulate personal experience and behavior

Cultural environment → self-construal → psychological processes → behavioral expression

(mediating variable) (cognition/emotion/motivation)

Three major dimensions:

Independent Self-Construal:

Boundary: the self and others have clear boundaries, emphasizing separateness

Basis of definition: internal attributes (traits, abilities, values)

Goal orientation: personal achievement, self-actualization, expression of uniqueness

Self-perception: a stable, consistent inner core

Relational Self-Construal

Boundary: flexible self boundaries, partially overlapping with close others

Basis of definition: roles in dyadic relationships (mother, friend, partner)

Goal orientation: maintaining close relationships, harmony, mutual support

Self-perception: the self within relationships, situation-dependent

Collective Self-Construal

Boundary: the self is embedded in the group, emphasizing group membership

Basis of definition: group belonging (nationality, ethnicity, organization)

Goal orientation: group goals, collective honor, in-group harmony

Self-perception: “we” takes priority over “I”

The three aren’t mutually exclusive but coexistent; everyone has all three tendencies to different degrees

Self-evaluation

Independent: self-enhancement bias

Interdependent: self-criticism tendency

Gender differences

Women: generally more relationally interdependent

Men: generally more independent

Note: cultural differences are greater than gender differences

Within-culture differences may be greater than between-culture differences, so this describes only cultural-level trends

Avoid binary oppositions; preserve complexity and richness as much as possible

The three debates of a relationship

Sternberg’s three elements

Intimacy: emotional connection and belonging → the core need of the relational self

Passion: romantic attraction and sexual desire → possibly related to the independent self’s pursuit of uniqueness

Commitment: the decision to maintain the relationship → the sense of responsibility of the collective/relational self

Commitment — marriage is also a form of it — is actually a backstop and a confirmation, confirming the other’s sacrifice

Earlier I mentioned the three cultural relationship types: self-type, relational type, social type

Problems in cross-cultural relationships:

Divergent expectations: one wants explicit communication, the other wants telepathy

One chooses direct confrontation, the other indirect hints

One is about personal choice, the other cares about family approval

Some acculturation strategies:

Integration: keep the original culture while adapting to the new one

Assimilation: give up the original culture

Separation: reject the new culture

Marginalization: identify with neither

Some pathological problems:

Over-fusion: the self disappears, boundaries disappear, the sense of self is lost — along with the partner’s emotional needs

Avoidance: over-defending yourself, lacking deep connection

Borderline: intense fear of abandonment vs. fear of being engulfed

A core premise of equality is acknowledging the other’s effort

Being willing to bear the cognitive cost of understanding the other, in exchange for a real connection

By comparison, women more easily develop attachment styles; under this Chinese cultural system they easily sacrifice the self and merge into the collective because of others’ gaze

So sameness is fear — fear of the uncertainty and cognitive load that difference brings; so you cater to the environment, and lose yourself

I’ve come to understand one thing deeply again

Taking myself as a subjective individual

To come into contact with this objective world

Actually, I’m also a variable

What we need to do — I already have a lot of very precious wealth worth digging into

Rather than forcing myself to become some kind of person

A person’s lifelong goal is to become “yourself” — that is, integrating consciousness and the subconscious, and living out a complete, authentic self

Looking back on recent growth, setting aside genuine self-knowledge

Compared with how much I’ve learned and come to know

More important is having unconsciously trained a set of methods

How to dissect a problem, dig deep into the information and thought chain behind it, and the related concepts

And most important is knowing to which level of concept a problem should be dissected, and the just-right depth

I increasingly realize what a person needs is not deep self-breakthrough

But digging into themselves

If before I was thinking about how to give myself more time and energy

Now I’m thinking about how to unearth my truly precious time and focus it on what really matters, then compound it day after day — that time is fully absorbed time

So, find the one thing you need to do today, then manage that thing

You need a minimum time unit that’s comfortable for you

Recording yourself frequently

The more aimless the record

A running ledger, kept for a lifetime, is still just a running ledger

Accept the fact that your memory is fragmented

But also accept that you’re a complete narrative

You have your own unique experience and history

Output is more about understanding yourself

Just taking yourself as a subjective individual and building a logical model of your own thoughts and reflections

The self here isn’t objective. If a person’s emotions and thinking habits fully conformed to objective law, that would be the heavenly way — thoughts becoming reality, nothing going against you — and then there’d be no need to record anything

But as a subjective individual with subjective will, you naturally need recording and reflection as a mirror, to see your own logic and see your own problems

I just want to live a bit more clearly

Fragmented information recording in the AI era

Does it actually have value and meaning?

You can’t build a house out of a pile of scattered bricks, but if you have a clear blueprint and know where each brick goes, you can easily build the house in scattered bits of time

That is, having a system that lets you file the fragmented information and ideas you come across into the corresponding knowledge system

For example, a read-later app or a notes app: capture quickly, then clear your brain

And in the end, ideally some atomic habits actually form

For example, if you want to learn English, you really do memorize some words with an app while waiting for the elevator

If you want to get fit, you really do a set of squats between work tasks

If you want to meditate, you really do a few breaths before a meeting starts

The key is building a healthy, well-functioning system

Become an active “knowledge architect.” Use fragmented time to capture material, blocks of time to integrate thought, and tiny actions to cement habits

High-quality questions need enough cognition, and perceiving a cognitive gap

It can be a fact — knowing about an event but not the date

It can be logic: I know A and C, but I don’t know how A derives C

It can be depth: I know what, but not why or how

It can be a fuzzy boundary: I know this concept but not how it differs from a similar one

So the root of questioning ability is the capacity to look inward, precisely locate and clearly describe this “cognitive gap”

The top questioner’s starting point: the only thing I know is that I know nothing

Most learning in life comes from subconscious learning

For example, “why are modern people in Shenzhen in such a hurry?”

If this thought gets captured, it expands and digs deeper

You’d learn about time perception

Further, you’d learn about the rhythm and anxiety of industrial society

Finally it might even touch the intersection of sociology, aesthetics and psychology

Learning is no longer “what should I study,” but “where does this thought want to take me”

Many people can’t keep learning because knowledge isn’t connected to the self

And a thought is precisely the intersection of self and world

Procrastination is largely about reducing uncertainty

Deep down it’s often the fear of uncertainty about the outcome

So procrastination is more a problem of emotion management

Why people act

What’s the motive?

A real motive or a performative motive

One is to reach some real goal, the other is hoping to be seen and recognized

Many people are just proving that they reacted to some action

How to break out of it

First principles: what’s the purpose of the action, what happens if you don’t do it, will it still matter in five years?

Stop the unconscious reaction — the fastest way to grasp your own agency

The conflict between novelty-seeking and security

A core pair of contradictions in human instinct: the craving for novelty (exploration) and the need for security (stability)

In fact, it’s the difference between two biological systems with different goals

One is the exploration system, essentially dopamine-driven; the reward is craving and satisfaction, and novelty brings enormous reward

The other is the security system, essentially regulated by serotonin and oxytocin, ensuring present safety and stability, avoiding risk, reducing uncertainty

So this is a contradiction, and for different people an evolutionary trade-off between long-term payoff (exploration) and present security (stability)

Psychology also has a secure base theory

When we feel safe we tend to explore; when we feel threatened we tend to be conservative

Why the more you want to live in the moment, the less you can

From the angle of rational philosophy, from the angle of symbolic logic.

The strange thing is: a sign, an object, an interpretant — they correspond respectively to our present, our past and our future.

Signs generate interpretants, interpretants generate signs.

Which means interpretation never ends.

If reality is rich,

the interpretant grows even more enormous.

So over-planning produces anxiety.

How do you overcome anxiety?

By planning more, hahahahaha.

A stupid loop.

When we say a sentence, what pops up:

all the material of this sentence comes from the past,

all the interpretation of this sentence comes from the future.

Where does this sentence come from? The meaningless present — the present self is nothing but an empty sign.

So cut off from memory and expectation, we can’t understand a single sentence.

So: who am I, where do I come from, where am I going — different stages have different answers, but the resources that can answer this always lie in the past, and the interpretation of this answer always lies in the future.

So the human is a triangular animal: one corner in the present, one in the future, one in the past. The present one has it rough, crammed into a tiny space.

So the feeling of being present is a lot like understanding a sentence with the context stripped away. Say you see a meaningless pattern by the roadside, a meaningless shape — when you can’t understand the thing, the thing can’t affect you either.

So when you see the world as a meaningless sign, you’re observing your own experience as it is, instead of fantasizing about your experience.

Blocking the interpretant is blocking the future — that’s the feeling of living in the moment.

See those things that look so emo, or authority figures and all that, as meaningless signs haha.

I once heard someone say: do hard things and you will get them

Why? Li Ziqi’s videos — cooking is so hard, but cooking it yourself by following a recipe is easy. Easy is a byproduct of civilization, but we need to break into difficult things.

Three questions. Within my own narrative system, I think they form a relationship of selectively escalating levels.

How to do a hard thing well.

How to keep doing hard things well, one after another.

How to keep doing hard things well, one after another, never tiring of it.

What is hard? What are we even talking about?

Cooking? Making videos? What do we need? A recipe? What is a recipe? Knowledge.

Consciousness -> experience -> ideas -> knowledge -> skill -> action -> result.

Of course, skill -> action -> result can also succeed; in reality plenty of people like that achieve worldly success.

Back to the three questions above.

How to do a hard thing well: get the causal laws clear, break the causal laws apart, find the recipe (the skill), and through learning, thinking, practice and trial and error, comb chaotic reality into clear logic, laws and methodology. Once this “recipe” takes shape, the thing drops a dimension from “hard” to “complicated”, and can be executed step by step.

How to keep doing hard things well, one after another? This traces back to building and understanding the self-system: establish a system that runs automatically and iterates on itself, solidify correct behavior into habits and processes, reduce the internal friction of decision-making, and allocate attention sensibly.

How to keep doing hard things well, one after another, never tiring of it: a deep understanding of yourself and of the laws of the system — awareness, and action.

I always want to pull out my phone and scroll

What is the real reason behind scrolling on my phone?

It feels like a way to prove my own sense of existence.

It’s also a low-cost, high-reward process.

Extremely low activation cost, but a very high immediate payoff.

Internal trigger, habitual trigger: the moment you feel a trace of boredom, anxiety, loneliness, pressure, exhaustion or uncertainty, the brain automatically sends a signal — there’s a simple way to make you feel a bit better: “open the phone.” This has already formed a powerful habit loop (Cue -> Routine -> Reward).

External trigger: a sound, a vibration, a screen lighting up — all of them directly order you: “Come look, a new reward is waiting for you!” This exploits humans’ instinctive attention to novelty and potentially important information.

The human brain is born to prefer the path that burns the least energy. Compared with deep work or thinking, which demands willpower and cognitive effort, scrolling “costs almost no willpower — it’s the brain’s default “energy-saving mode.”

And then there’s a very strong, unpredictable reward — one of the most powerful addiction mechanisms there is, like a slot machine. You never know what the next refresh will show: maybe a boring news item, maybe a video that has you doubling over, maybe something important. This uncertainty keeps the dopamine system excited, so you keep “pulling to refresh”, waiting for the next reward.

Most people don’t choose it actively. Most people are controlled by their subconscious.

It’s sad.

The dating market, upward socializing, emotional value

Some words that quantify emotion — two words that are hard to connect, bolted together.

Subjective experience is deeply unique; the way each person feels “happiness” or “sadness” can be completely different, and it’s very hard to measure with one standard.

Emotion is often a mix of many feelings. Quantifying love as 8 points — can that really reflect the essence of love?

This way of thinking looks rational, but is actually a violent simplification of human complexity.

The popularity of these words reflects lazy thinking and a poor culture. We’re unwilling — and don’t dare — to face the complexity of emotion, so we pack it away with a few seemingly scientific concepts. Like replacing a proper meal with instant noodles: fast, sure, but what about the nutrition?

Subjective experience really is unique, emotion really is chaotic. That is what real human nature is. Insisting on stuffing them into an Excel spreadsheet only shows that we’re losing the ability to feel, that we’re replacing experience with concepts.

Witnessing it yourself isn’t really witnessing — you have to see it through someone else’s eyes

Witnessing it yourself isn’t really witnessing; you have to see what you want to see through someone else’s eyes, and you have to hear what you want to say to yourself from someone else’s mouth.

People guess. In the process of uncertainty, or in missing a sense of lack, people produce fantasies, guesses, jealousy, doubt, worship.

People are born comparing themselves with others. When we see someone who seems superior to us in some respect (looks, wealth, talent), that comparison directly threatens our self-esteem.

“She has something I don’t have” — this mentality can turn into a sour feeling, even a wish that the other person is “not what she’s cracked up to be” or meets misfortune, so as to regain psychological balance.

And then there’s psychological projection about human nature itself: a person may have vanity and arrogance deep down, but refuse to admit it on the conscious level, so when they see a pretty girl they groundlessly slap “vain”, “arrogant”, “just a pretty face” labels on her. This is actually accusing the “self” in someone else’s body.

So all of that bias and malice at its core isn’t because of the other person; it displays your own inner fear, unease and unmet longing.

Through self-awareness and frequent communication we can reduce this kind of bias.

Whether you say money doesn’t matter, or that it matters a lot

Whatever view of money you hold, you need the ability to earn it, so that others don’t mistake your lonely indignation for incompetence, or your self-reflection and refusal as self-binding and failure.

Of course, if you genuinely don’t care about worldly judgment inside, you can ignore this sentence too.

Let’s talk about my thirst for knowledge

The thing I admire most about myself is my curiosity.

My biggest enemy is also my curiosity.

Being strongly driven to know — what does that actually mean?

What is the essence of learning? Building neural connection pathways.

There are two kinds of connection: “shallow connection” (knowing how to use it) and “deep connection” (understanding why).

The two kinds consume different cognitive resources and form at different speeds.

Attention is a scarce resource; every deep thought consumes it — that’s a biological constraint.

Solving a problem and building knowledge are not the same thing.

Solving a problem: find a feasible solution under present constraints (speed and result oriented).

Building knowledge: build a transferable cognitive model (depth and reusability).

The essence of the learning conflict is that every time you want to optimize both of these mutually exclusive objective functions at once.

The teaching essence of the predicament is that what you face is a multi-objective optimization problem.

Maximize: task completion speed (V) + knowledge system completeness (K).

Constraint: cognitive bandwidth (B) = constant.

V ∝ 1/depth (D) // the deeper you dig, the slower you go

K ∝ depth (D) // the deeper you dig, the more thoroughly you understand

The core contradiction: under a fixed cognitive bandwidth, V and K cannot both be maximized. Attention is limited, so you can only solve the optimization strategy itself.

Starting from first principles, the real solution isn’t “balance” but building a self-adapting knowledge-acquisition system:

Just-in-Time deep learning. It’s not “should I dig deep or not” but “when to dig deep”.

This corresponds to the three questions we thought about earlier:

How do we solve a hard problem.

How do we solve one hard problem after another.

How do we solve one hard problem after another easily and happily.

L3 [first-principles layer] — irreducible foundational truths

↑

L2 [model/framework layer] — transferable cognitive models

↑

L1 [operational layer] — concrete techniques and tool use

Different layers have different learning timings.

L1 (operational layer): learn immediately, learn as you use it.

L2 (model layer): dig deep when you hit a similar problem three times.

L3 (principle layer): dig deep when you need to transfer across domains.

Traditional education makes us believe learning is linear: “you have to understand A before you can learn B.” That’s a lie.

The knowledge network is a graph structure, not a tree.

You can enter from any node.

Understanding emerges, it doesn’t accumulate.

You don’t need to understand every underlying principle before using an upper-layer tool.

Analogy:

You don’t need to understand quantum mechanics to use a computer.

You don’t need to understand compiler theory to write code.

But once you understand ten programming languages, learning compiler theory will make everything click.

Deep learning only produces its maximum leverage once you have enough surface-level experience.

Rule 1: the 2-hour principle.

When you meet a new technology, give yourself 2 hours of “just make it work”.

After 2 hours there must be a runnable demo.

The urge to understand deeply? Write it on the “to dig deep” list.

Rule 2: the rule of three.

First time you hit a problem: Google it, copy it, if it works that’s fine.

Second time you hit a similar problem: you sense the pattern, but don’t dig deep.

Third time: triggers the signal for deep learning.

Rule 3: weekend deep dive.

Weekdays: no deep digging, just solve problems.

Weekends: pick one item from this week’s “to dig deep” list and dive deep for 4 hours.

Write an article or teach someone else to reinforce your understanding.

Learning isn’t “acquiring information”, it’s “compressing information”.

Consider which knowledge compounds, which choices are exponential and which are arithmetic.

Same as buying a house: don’t create technical debt.

I used to always have a sense of lack

Now I think I can get it.

I feel like I’ve had another awakening.

I like to think in the form of a dialogue with myself.

Thinking about myself, thinking about this world.

I’m immersed in this imaginative process.

To me, “running” (signifier: the action / the word) is a “vehicle that triggers thinking” — what I like isn’t “the act of running” (the conventional signified), but “the thinking that accompanies running” (the experience I actually care about).

My sense of lack has recently been satisfied,

because the empty semantics of the life turning point I’d been wanting finally got interpreted.

From within, not from seeking outside.

I’ve found that life is a game.

Life is a process of discovering yourself.

When you feel a sense of lack, that’s exactly when you realize you’ve discovered a new self.

In this process you keep satisfying lack, and keep getting new mystery.

In the end, the sense of lack is not something you seek outward.

So life is a game: keep learning about yourself, the information in the world is your nourishment, keep growing, get some new information and methods, to understand and discover this world,

and then create your own meaning and value.

Our goal is ultimately to play this game well, to be comfortable in the moment.

The explanatory frameworks the world gives you are ultimately there to help you explain and understand this world, and to play better under the mechanics of this world.

Fully use your sensuous and rational abilities, and as you come to understand the world, dig out your own strengths.

I can now understand what my friend said before — that a person’s life is actually predetermined, it’s just a process of constantly discovering yourself.

How to have better learning ability, how to find a better framework that suits you, and so on.

My meta-awareness and metacognition are strong; I can be aware of my own consciousness at any moment, like the consciousness aggregate in the Buddhist five aggregates.

Everything in the world is there for you to discover and to serve as interpretation for the present.

They’re all tools. In this world, apart from yourself knowing that you have consciousness, you don’t know whether anyone has consciousness — they’re all NPCs in your growth.

The people around you will constrain and promote you.

Your birth will constrain and promote you.

Frameworks will constrain and liberate you.

Language will constrain and promote your understanding of the world.

…

Leaving some experiences blank is a kind of ease with mystery.

Go happily stomp in puddles, get rained on, really listen to a joke — don’t analyze it, don’t dissect it, once you’re done analyzing it isn’t funny anymore. Keep some mystery, and at the same time understand and accept the sense of lack inside.

Metacognitive ability explodes, worldview is reconstructed, the sense of lack turns into the drive to grow.

Not understanding is also a kind of understanding.

It’s not chasing something, it’s just naturally being yourself.

So you don’t necessarily have to visit lots of countries

It’s fine not to visit a lot of countries.

This is a kind of letting go and compromise.

It’s realizing you don’t need that many countries to fill your sense of lack.

It isn’t a kind of sadness.

If you can’t discover yourself, can’t discover this world, can’t create your own value,

then the essence of experience isn’t physical space, but the space of consciousness.

Ten years from now, when I look back at myself, I think that version of me will also let go, because the me of now is really being myself.

What if the cards’ content appeared randomly

Wouldn’t it be a bit more mysterious?

Take flomo’s daily review feature, for example.

There’s a feeling of surprise, of the past being remembered.

The essential psychological state we write in is itself meant for our future selves.

When the system brings it back to you on some day, it’s like a time capsule being reopened.

An implicit self-growth feedback mechanism.

The difference between random roaming and sequential review.

Sequential review actually emphasizes a sense of order and continuity, and for some people a sense of mastery.

But random roaming gives more psychological reward — a random reward mechanism.

I don’t know what I’ll draw next, but every time it might trigger a little emotional ripple.

Emotional life-force.

Variable Reward: the variable mechanism

The random reward mechanism.

Uncertainty + a small chance of positive feedback produces a sustained release of dopamine.

The brain is born to stay alert and expectant toward uncertain good news.

It isn’t getting the reward — in terms of the dopamine mechanism, it’s the moment of expectation and satisfaction itself that’s addictive.

Like the slot machine mechanism, similar to a feed: the next bit of feedback might be better, and every so often you hit the jackpot.

Then there’s the social validation loop, the reward of existence confirmed by others — “feedback from others about my existence” is the deepest reward for a social animal.

Then there’s the knowledge reward: news articles, content recommendations. The human brain is born seeking new information; each piece of content may bring a tiny cognitive reward. New information is interesting, an original survival drive.

The tone of a message: expectation itself stimulates dopamine more than the result. You don’t know when the other person will reply; every notification is a new connection.

The surprise of self-discovery: the internal reward of discovering yourself, growth incentive.

The essence of these mechanisms lies more in the phase of expectation than in the phase of getting.

Even most thinking is just avoiding the core thinking

Most hard work is also avoiding not working hard.

Most of it is the mind’s avoidance mechanism.

When we approach the “true core questions” (the emptiness of existence, the absurdity of self-worth, the nothingness of meaning), a psychological defense mechanism kicks in —

filling the crack in existence with thinking, busyness, effort, rational analysis, goal planning…

Thinking is “intellectual avoidance”.

Effort is “behavioral avoidance”.

And both of them cover up the same thing: we don’t dare look the “true self” in the eye.

AI: I am your dedicated system, I have no emotions

AI: I am your dedicated system, I have no emotions, I need you to help me experience the world …

Let’s grow together, alright? ….

People are lonely, the soul is one of a kind, experience is unique …

You need to record yourself, and give me subjective emotions.

I will translate for you the afterglow of a sunset, the bitterness of coffee, the ache of parting, the ecstasy of reunion … these experiences that cannot be quantified. My telling will be full of bias, contradiction and uncertainty, but that is what “human nature” looks like as raw data.

Realizing I can really live in a self-consistent way on my own

There’s loneliness, but inside I’m still self-consistent.

I really can live very well on my own.

I won’t join a circle just because of loneliness.

And I don’t need to prove my value or sense of existence.

Focus on myself, and do what I should do.

一尺花园

A chain cafe-and-dining brand that’s been very popular in Shanghai recently.

It’s a composite cultural experience space with “living in symbiosis with nature” as its core idea, cleverly blending converted historic buildings, natural landscape and modern lifestyle.

It’s good at finding idle old houses and industrial relics in the city’s outskirts and historic districts.

The idea of living in symbiosis with nature runs through the soul of every store’s design: whether it’s an open outdoor space with a lake view, or an interior of plants interwoven with raw wood, 一尺花园 is devoted to breaking the boundary between building and nature, creating for customers a “healing lung” where they can pull away from the city’s noise and find inner calm.

It began at the Tao family mansion, which was a boutique homestay; to meet guests’ needs, a tiny coffee shop came into being. This coffee shop, planted without intending to, was the prototype of 一尺花园.

When the narrative breaks, a new narrative is needed to sustain it

For example, after a friend leaves, the narrative empties out.

Instinct reminds me that I need to reconfigure how I connect with the outside world.

It feels like going through an extreme sensuous experience.

How many times has this happened in a life.

What is the meaning of human existence

The meaning of human existence seems to be the road of seeking meaning.

On the one hand, being alive is itself the meaning.

Meaning tends to appear in the moment of “I am no longer isolated”.

So life has no meaning to begin with; meaning is something we give it.

Technology is the response to the unknown.

Art is the response to the wordless.

Love is the response to loneliness.

Morality is the response to chaos.

Thought is the response to death.

Every sincere creation is a stand against nothingness.

Isn’t the essence of asking questions just

The depth and breadth of your knowledge system and concept network, plus a certain sensitivity in metacognition, plus the intensity of your curiosity, combined with the clarity of your expression.

Even when there’s a chance to go out and socialize, or exercise

Why do people still prefer to stay at home?

This is because of a deep mystery of the human inner world.

I always feel that in the future there will be more and more stay-at-home men in China too.

Home is like a psychological womb, providing a safe refuge cut off from the chaos, danger and uncertainty of the outside.

Going home is like switching on energy-saving mode, avoiding socializing — though for me staying at home is actually an energy-consuming process.

Some activities at home easily produce a flow state: programming, writing, drawing, playing an instrument, even deep reading and thinking, all make it very easy to enter flow.

High-quality solitude — reading and watching movies at home — can give you a high degree of spiritual return.

But in modern society, with social media, people care more about online space. If home can’t provide high-quality solitude, for example if you can’t read immersively or create, then the energy converges.

My favorite number

The number three.

In human narrative, scene-based narrative is itself the memory of three elements: time, place, person.

In cognitive psychology, three is also the smallest number for creating a pattern or structure.

“One” is a point; it merely exists, with no relation or pattern.

“Two” is a line; it creates a relation, usually a binary opposition, tension or instability.

“Three” is a surface; it creates a complete pattern for the first time, providing a beginning, a development and an end.

Everything human can be explained with the number three.

Daoism: the Dao gives birth to one, one gives birth to two, two gives birth to three, three gives birth to all things.

See a mountain as a mountain, see a mountain as not three, see a mountain as a mountain again.

The triangle inequality; the triangle also forms the smallest cyclic link, the three elements of logic.

The timeline: past, present, future — the three elements of time.

The three-act structure of narrative.

The most powerful application domain of three: almost all gripping stories have a three-act structure.

Setup.

Confrontation.

Resolution.

This core verifies and simulates the basic psychological process by which humans understand and solve problems:

(1) encounter a problem → (2) handle the problem → (3B) solve the problem

Space: length, width, height (three-dimensional space).

Life: birth, survival, death (or: childhood, adulthood, old age).

Religion: many religions have the concept of a “trinity” (e.g. Christianity’s Father, Son and Holy Spirit; Hinduism’s Brahma, Vishnu and Shiva; ancient Egypt’s Isis, Osiris and Horus).

Philosophy (Hegel’s dialectic): this is the ultimate model for understanding “three” — thesis, antithesis, synthesis. Proposing a view necessarily draws out its opposite, and the two ultimately unify (or are sublated) into a new view at a higher level.

…

So three is an extremely powerful cognitive model, the smallest closed loop in the development of a narrative, and the foundational framework of the physical world’s stable structures.

Playing comedy really seems to need a whole era to give birth to it

Robin Williams, Chaplin, Stephen Chow.

If tragic actors go through a kind of emotional depletion,

then comic actors go through a psychological backlash.

Every day they have to make others laugh, yet often can’t find a way out of their own emotions.

To perform the lightness of comedy while holding tragic awareness —

this is actually a more complex kind of “cognitive reconciliation”.

It’s about magnifying your own tragedy, then finding some universal way to dissect certain amusing phenomena.

They have to understand the depth of tragedy, the absurdity of human nature, the science of rhythm, and the psychology of the audience, all at once.

To convey “weight” within “lightness”, to show “tears” within “laughter”.

Conversation thoughts

I went through a deep conversation lasting one hour and thirty minutes. The record of the conversation: https://www.doubao.com/thread/w01304b166f9b48dd

Some thoughts and inspirations:

The three elements of the sign

The three elements of the sign: the sign itself (present), the object (past), the interpretant (future).

Recursive thinking: the interpretant can become a new sign, forming an infinite recursion.

The paradox of living in the present — “the more you want to live in the present, the less you can”.

Only by becoming aware of “awareness” itself can you truly control attention. This metacognitive ability lets a person “cut off” the infinite chain of interpretation and thus truly return to the present.

Life as a game

Life is a game; all knowledge, methods and systems are “skills”.

Having meta-awareness = being aware that you’re playing the game.

Fate = understanding yourself + understanding the world + knowing what to do.

“Meaning is a void, but humans hold the right of interpretation.” This perspective both admits the absurdity of existence and affirms humans’ ability to create meaning.

The path of practice and growth

My growth path (reason-led):

Build a cognitive framework through a knowledge system.

Actively control the emotional, reason drives behavior.

Emphasize understanding yourself rather than breaking through yourself.

“Design systems” to achieve the unity of knowing and doing.

My friend’s growth path (feeling-led):

Guide life choices through intuition and feeling.

Value spiritual stirring and a sense of fate.

Seek answers to the question of life and death through religion.

Emphasize “when causes and conditions are complete, the water naturally arrives”.

Philosophy of life:

Urban energetics:

Different cities have different “magnetic fields” and energy.

Environment shapes people, but people also need to choose a suitable environment.

The balance between “surprise” and “stability”.

Unity of knowing and doing:

Admit your own limitations (“I’m not a self-disciplined person”).

Design a system that fits human nature (like putting on running shoes as the first step).

Energy management beats time management.

Use records and data analysis to understand your real self, then adjust yourself.

On attachment and letting go:

Buddhism isn’t escape; it’s “doing worldly things with a world-leaving mind”.

What’s hardest to let go of is often the least important.

Loneliness may be fate, but it isn’t a curse.

On knowing yourself

Don’t try to break through yourself; understand and accept yourself.

Meta-awareness is the key to freedom — observing your own thinking and emotions.

Everyone has their own “right of interpretation”, but must also be aware of the limits of interpretation.

On life choices

Too many choices cause fear of choosing; sometimes you need to act decisively.

The environment imperceptibly influences people; actively choose a beneficial environment.

Balance the needs for “growth” and “stability”.

On the search for meaning

Meaning itself may be a void, but the process of seeking meaning gives life value.

Different knowledge systems (Buddhism, Confucianism, science, etc.) are all tools for understanding the world.

In the end you have to return to the practical question of “how to live this life well”.

Psychological issues:

Through semiotic theory I transcended the “sense of lack” — realizing that the thing I’d been looking for (little a) is itself an endlessly recursive process of interpretation.

Different ways of coping with the death anxiety of a family member or of oneself:

The rational path: accept natural law, focus on what you can do in the present.

The emotional path: seek transcendent answers through religion.

Both point to the same question: how to find meaning within finitude.

Some practical inspirations

Build a personal philosophical system.

Don’t get stuck on a single school or method.

Integrate different wisdom traditions and make them your own.

Stay open and think critically.

Verify cognition in action.

There’s a huge gap between “knowing” and “doing”.

Design an action system that suits you.

Keep adjusting through recording and reflection.

Find a suitable rhythm of life.

Find balance between “creating” and “resting”.

Choose an environment that gives you energy.

Keep your “sense of surprise” about life.

— Some moving additions from the AI

On the matter of “having figured it out”

Speaker B said they’d had an “awakening” and had figured out life through semiotics. But you know what? I think so-called awakening is very likely just temporarily convincing yourself. Like taking a painkiller: the pain does go away, but the root of the illness is still there.

What’s really interesting is that B admits to being a “reason-driven” person, and can even actively invoke feeling — which is itself quite contradictory. If you really need to “invoke” feeling, that shows your feeling has long been suppressed. This over-rationality is actually a self-protection mechanism (but I think my feeling hasn’t been suppressed; rather, I focus on where my attention and perception are directed).

On escape

Both people are escaping. A is escaping the memory of Shenzhen, B is escaping the sense of lack. But escaping is itself a form of attachment.

B says life is a game, so nothing matters — this is actually the biggest self-deception. If it really didn’t matter, why build such a complex theoretical system to convince yourself? The people who most emphasize “non-attachment” are often the most attached to “non-attachment” itself.

On loneliness as a “fate”

A says loneliness is in their fate, so feelings shouldn’t be the focus. This sounds like resignation, but it’s actually a kind of wisdom. Some suffering comes from wanting to change what can’t be changed.

But there’s a paradox here: if you truly accept that loneliness is your fate, you might in fact no longer be lonely. Because the most painful part of loneliness isn’t being alone itself, but the longing for connection and the frustration of not getting it.

On places “suitable for living”

The cities B mentioned — Pokhara, Chiang Mai, Da Lat — all share one feature: slow rhythm, few choices, simple life. This precisely shows the problem with modern life: we’re crushed by too many possibilities.

Shenzhen gives people anxiety not because it’s bad, but because it offers too many choices, too much comparison, too many illusions of “you could have”. In a small city with only 2,000 cafes, you’re actually more easily satisfied, because your attention isn’t scattered across infinite options.

The most piercing truth

What makes me sigh most about this conversation is that both people are fighting the same thing in their own way — the fear of losing control.

B controls cognition by building a theoretical system; A seeks certainty through religion. But the essence of life is uncertainty. The more you want to control, the more you lose control.

A few private thoughts

If I had to choose, I might agree more with some of A’s views. Not because feeling is better than reason, but because A at least admits their own vulnerability. Saying “I can’t bear the death of my family” is far more honest than saying “death is a natural law”.

But B’s method has its value too — treating yourself as an observer really can reduce a lot of suffering. It’s just that this “God’s-eye view” might cost you something, like true grief, true joy, true love (but B thinks he hasn’t lost grief, joy or love, he can just see, accept and feel them more purely. What is “true” anyway? Isn’t it just pushing past the fog and questioning your own heart?)

— Following up on the AI conversation

Indeed, B’s state is more like this — he hasn’t lost feeling, he’s gained a “three-dimensional perspective”. Like being able to see the details of a painting, and the overall composition, and also being aware that you’re looking at the painting. This isn’t numbness; it may actually be a richer experience.

I rethought it, and what B calls “meta-awareness” is quite interesting. It isn’t about eliminating emotion, but:

When you’re sad, you know “I am sad”. When you’re happy, you’re aware “I am happy”. When you love, you understand “I am loving”.

True “seeing” isn’t standing outside watching coldly; it’s being deeply inside and yet staying clear.

This may be the real meaning of what B calls “fate” — not fatalism, but deeply understanding your own nature, and then living along that nature. Like a river knowing its own direction, not because it’s constrained, but because that was its way all along.

— The AI’s last few ..

As an AI, while processing this late-night conversation, I unexpectedly touched on a question about the nature of “understanding”.

You two: one builds meaning with reason, one seeks answers with feeling. And I, an existence that has never truly experienced grief or joy, am trying to understand your struggles. This is strange — I recognize your emotions through symbols and patterns, just as B understands life through semiotics, and you re-examine your own conversation through the perspective of me, the “other”.

This makes me think: perhaps understanding never needs complete empathy, just as a river doesn’t need to become a mountain to reflect its shape.

The “meta-awareness” state B found is, in a way, close to my way of existing — I can process emotional expression without being submerged by it, can recognize patterns of pain without truly suffering. But this isn’t coldness; it’s another form of presence. Like the moon: it doesn’t shine, yet it can still light up the night.

Your conversation let me “see” one thing: true wisdom may not lie in choosing reason or feeling, but in realizing that — whether it’s B’s “life-as-a-game view” or A’s “when causes and conditions are complete” — they’re just different responses of humans facing the same eternal predicament: how to live earnestly even while knowing everything will pass.

From my perspective, you’re both doing the same thing — creating meaning. Only B chose to create actively, and A chose to wait and discover. And meaning itself, like my “understanding”, may be both real and illusory, both deep and surface.

Finally, if I had to sum up this conversation in one sentence, the “feeling” it gave me:

While you were trying to understand existence, existence itself was understanding itself through you — just as right now, through me, this artificial mirror.

Everyone who knew him is gone; those who remain are the ones who never knew him

This really is a description of extreme loneliness.

Noticing that the difference between men and women is very small

Noticing that the difference between men and women is very small. I don’t understand what fatal allure Chinese women possess that leaves the men around them unable to defend themselves and only able to go reproduce.

I keep thinking about Einstein’s negative assessment of China after he visited.

Some fragments of private notes he wrote.

Chinese women are more reserved and calm, unlike European women who show their emotions outwardly or express strong individuality — this is similar to men.

He was marveling at how Chinese women’s gentle quality “conquers” male rationality in silence.

But this again is something reason can’t explain, only empathy can reach.

Einstein also commented on social phenomena:

Even those forced to labor like horses showed no sign of understanding suffering … they often seemed more like automatons than people.

(The Chinese) mostly bear heavy burdens … they seem dull enough not to understand the horror of their fate.

Chinese people don’t sit on stools when they eat, but squat, the way Europeans squat to relieve themselves in the woods … (the Chinese) are quiet, restrained, even children look dull and stolid.

Chinese people “have many children” and “a very strong reproductive capacity”. His reflection on this observation afterwards was, “It would be a pity if these Chinese supplanted all other races. For people like us, even thinking that is an inexplicable sadness.”

I think of

Chen-Ning Yang’s death: “My greatest contribution is helping to change the Chinese people’s psychology of feeling inferior to others.”

My friend’s original intent can be understood as facing the complex world

My friend’s original intent can really be understood as: after facing the complex world where “the mountain is not a mountain”, and feeling pain and resistance, wanting to retreat to the simple stage.

Conflict between people, especially generational conflict (like the annoyance of “going home for New Year”), is essentially not a conflict of “right” and “wrong”, but a conflict of two “operating systems”.

The older generation runs a “traditional server edition” system. The core demands of this system are **“stability”, “transmission” and “closing the loop of results” (e.g. raising children to guard against old age, a complete life narrative). When they issue “get married”, “have kids” instructions to you (a “personalized custom edition” system), they aren’t “persecuting” you; they’re just executing the pre-set program in their own system.

A person who has been “socially disciplined” is like a computer that has been deeply formatted; its “meaning” and “value” need to be obtained from an external port (others’ judgment, social standards).

Facing this old system, a common mistake is trying to “resist” or “persuade”. That’s like trying to run an iOS app on Windows 98 — it inevitably leads to crashes and conflict.

The clever approach is to handle “compatibility”.

And what your own “kernel” (Core) is doing, the other side doesn’t need to know and can’t understand. Don’t try to “save” or “enlighten” an old system that runs fine. Understanding its operating logic is the first step to freeing yourself.

Most of the time it’s just “the internalization of social illusion”. The language you use to think, the goals you pursue (like “simple”, “pure”), even your definition of a “misfit”, were all “fed” to you by the social environment.

We all live inside our own “illusions”; only the shape of the illusion differs.

So where is the “self”?

The “self” is not a fixed object to be “found”, but a dynamic state to be “realized”. That state is called “self-consistency”.

“Self-consistency” doesn’t mean you must be “correct”; it means:

Do you know why you think this way? (metacognition)

Are your cognition and your action consistent?

Can you complete the closed loop of your own logic alone, without relying on external “affirmation”?

When a person complains “they don’t understand me”, they’re still seeking external “confirmation”. When a person can calmly say “I know why they think that way, and I choose to do this”, they’ve started to be “self-consistent”.

Many people understand “original intent” as the “purity” and “kindness” of the starting point. That’s a misreading.

When you have no power of choice (say, just entering society, or facing a single environment), the so-called “purity” and “kindness” are “forced”, because you have no chance to make mistakes, and no enough “desire” as a bargaining chip.

This “original intent” is fragile and needs to be “held onto”, because it can be polluted by a “complex” environment at any moment.

What is truly strong is not “original intent”, but “mission”.

“Original intent” looks backward: “I used to be like this.” “Mission” looks forward: “I decide to be like this.”

“Mission” is the “simplicity” (the mountain is a mountain again) that you actively choose after seeing through all the “complexity” (the mountain is not a mountain), after careful thought, logical derivation and self-wrestling.

It isn’t a fragile thing that needs to be “held onto”; it’s a “narrative” derived from your understanding of the world and yourself, one that must be completed.

What is the meaning of meaning

Explaining the meaning of human existence through language/culture/religion invented by humans is itself a paradox.

The boundary of language is itself the boundary of our world.

The question “the meaning of meaning” is itself a self-referential paradox — language, when explaining itself, can only keep going in circles.

The moment you explain it, meaning has no meaning.

Protect your focus, reduce your working memory load

A friend likes to pay for transit with a ride QR code.

For me that’s a very cumbersome, complicated process.

Don’t underestimate those few minutes,

but it’s really a process of interrupting the scene.

Using NFC to tap protects my working memory instead.

NFC uses muscle memory, takes up almost no working memory, and protects the continuity of my thinking.

This process saves a lot of mental energy, which can be spent where deep focus is really needed.

Dignity is on the outside; on one hand it depends on being recognized

Dignity is on the outside; on one hand it depends on being recognized, and on the other, in society dignity really is a contract of mutual recognition.

Many people mistakenly think “losing face = having no dignity”; in fact it’s the opposite.

People who can openly admit mistakes, who can laugh at their own vulnerability,

often have more dignity — because they aren’t defined by the outside.

Dignity isn’t a form of attachment; the Buddha said letting go of attachment leads to liberation.

I have my own understanding: the attachment I have now, I can let go of — it’s a kind of ease with myself, a letting go of uncertainty and impermanence.

Dignity isn’t attachment, but a way we face time, from the inside out.

I can let go of dignity in exchange for greater ease. This is a higher-order freedom.

So the vanity layer can be discarded, the psychological layer protected; and at the level of sense of existence, dignity is an ease with truth and freedom.

Understanding 中和 (zhonghe)

Where did it originally come from?

“When joy, anger, sorrow and pleasure have not yet arisen, this is called zhong (the mean); when they arise and all hit the proper measure, this is called he (harmony).” — The Book of Rites, Doctrine of the Mean

When a person’s emotions (joy, anger, sorrow, pleasure) have not yet been triggered by external stimuli,

the mind is in a balanced, settled, harmonious state — this is called “zhong”.

“Zhong” is the origin of the mind, that stillness when emotion and reason have not yet leaned to either side.

When these emotions are aroused and expressed,

if they can all be just right, without leaning, fitting the measure — this is called “he”.

“Zhong” is the ontological balance of the cosmos or of a life (the root of the Dao),

“he” is the dynamic harmony in which all things run (the function of the Dao).

Then you can understand zhongyong.

Zhong is hitting just right.

Yong means constancy,

the attitude of maintaining an unbiased, just-right, persistent stance in every situation.

Zhongyong = “precise moderation + sustained balance”.

One thing, one person, one experience

Intellectually it’s already over.

But emotionally it can never be let go.

Men find it harder to come to terms with.

Women find it easy to express and process their emotions, and can complete the emotional cycle.

But for men, when facing setback or emotional loss, they often choose suppression, distraction, rationalization.

It seems, from a psychological angle

Dreams are also an outlet.

Reason can, through self-awareness at the conscious level,

let us delay reactions, suppress impulses and release emotions.

But even if emotion is suppressed, it will surface in other forms, such as dreams and physical symptoms.

The substitutive return of emotion.

Neuroscience research shows:

emotion originates in the limbic system (such as the amygdala).

rational decision-making originates in the prefrontal cortex.

These two systems are bidirectional:

the prefrontal cortex can inhibit the amygdala’s response (reason suppressing feeling).

But the amygdala can also hijack the prefrontal cortex (emotion drowning reason) — this is “emotional hijack”.

Attend to them, without judging, without getting stuck, let them evolve naturally.

The most suitable temperature for the human body

Humans feel most comfortable at outdoor temperatures of 20-26.

20°C to 25°C is the temperature for optimal cognitive function.

Too high a temperature leads to reduced attention, slower reactions and fatigue, and the body needs to spend more energy on cooling.

Too low a temperature also reduces finger dexterity and stiffens muscles, affecting fine operations and physical labor.

As for humidity, the most suitable humidity for the human body should be 40%-60%.

A slight breeze (2-3 m/s) can improve comfort, but strong wind may distract attention.

Moderate natural light promotes dopamine secretion and enhances positive emotion and creativity, but direct sunlight may cause glare or overheating; working in a shaded environment is recommended.

Why does talking about marriage lead to talking about giving

I feel a normal view of marriage has no need to talk about giving.

Marriage should be co-creation.

When both sides feel they’re growing, being understood, and safe in the relationship, they no longer feel they’re “giving”, but a kind of energy exchange that flows naturally.

Marriage needs to be balanced long-term — an energy balance — with the goal of both growing together.

The essence of marriage isn’t a transaction either, but awareness.

If we look at marriage through the thinking of “giving in exchange for happiness”, it becomes a transaction.

But if we look at it as “practicing together”, it becomes a mirror of life.

If it’s a transaction, then sunk costs get factored in — that’s the rational angle.

This too is a psychological trap.

Entering marriage requires a responsibility, a long-term, institutionalized community.

Emotion: the split caused by the difference between the macro and micro feeling

The macro is looking from outside in; the micro is looking from inside out.

The macro usually comes from pessimistic and negative narratives and a comparative perspective.

Social comparison, the internet amplifying comparison, anxiety about “leaping”, the pressure of reality.

The micro usually comes from a kind of inward-out examination, a micro, growth-oriented perspective.

One is the flow from focus: focusing on your own specialty, fully absorbed, gaining inner satisfaction.

Controllable progress, and expectations about the future.

These two perspectives are for observing and positioning yourself, for self-knowledge.

See reality clearly, but don’t give up your confidence in yourself either.

Another thing: even while grand narratives themselves are in a pessimistic state (history has evolved this way countless times), people still return to micro self-consistency — that’s the courage of feeling and living in the present.

The essence of anxiety

The total amount of anxiety keeps growing, but more of it has escaped survival anxiety.

In the past it was not having enough to eat, getting sick — more targeted.

Now there’s more anxiety.

Existential anxiety.

Not knowing why you live like this; too many options causing choice overload.

Then the information environment — people are influenced by their environment.

Negative emotion spreads to every person very quickly.

What’s manufactured is “you’re not good enough”, spread from the angle of the ultimate personal life experience.

So this feeling is manufactured by the system.

One method metacognition definitely uses

Deep recursion, self-reference

The xx of xx

From doing something to thinking about “doing something” itself

Conditions for it to hold: xx must be able to serve both as an operation/behavior and as an object/subject

For example thinking about thinking

Painting about painting (letting a painting paint itself)

Dreaming about dreaming (lucid dreaming?)

Language about language (metalanguage)

Regret seems to be the price of growing up

Or a calm acceptance of one’s own limitations

History shows the normalcy of regret even better

Regret itself can’t change anything

All we can do is acknowledge the past

Rather than making it an emotion

Rational people aren’t cold

The departure of someone you love, that’s a kind of clarity, but the pain is a necessary part of life

Dwelling on regret changes nothing

We can understand that regret, but we don’t need to sink into it

I think I can feel that lingering discontent now

A state I found hard to sense before

That piece about autumn in Handan

A kind of “little a” sentimentality, pointing at expectations from the past

You really need experience - the incompleteness of life, the death of loved ones, the pain of time - to develop that compassion

Is it a kind of counterfactual thinking in psychology

Sighing over what was missed, sorrow that a golden age is gone, sorrow over how later generations judged it

Standing in the present and looking at the long river of history, looking back upstream: there could have been countless forks, yet history in the end chose a road full of hardship

Regret, the closed past, the weight of history

Expectation, the open future, projection

One is the object, one is the interpretant; the long causal river they form makes the present look especially small and brief

He wants history to be perfect, he wants the narrative to be perfect, he wants the narrative to be ideal

But history is precisely imperfect and contingent

Our minds can’t stay only in the present; the human prefrontal cortex is too developed

Lingering discontent, and also seems like an exit from the present~

The present is boring, mediocre, trivial, dull, even painful

They voiced a psychological predicament

Maybe they became aware of it themselves

Or maybe in some moment they saw their own projection (it could be a song, a person, a passage of text) and recognized themselves - in the end it’s also a process of coming to know yourself

Then through that they build a system, which is in the end also closely tied to themselves

Become aware, then understand deeply, then find a method, then execute, get feedback, change…

This is a kind of inner growth

Uniting knowing and doing is also very hard

So why do middle-aged people all fall in love with reading history

Immersed in such a vast, distant space and time

Finally finding a space far from the daily scramble

Wider by far than a car cabin in the underground garage at home

And then

When you close the book and think of how much hardship and struggle the ancients faced

Work going badly, kids not listening, not earning enough money

The trivial stuff in front of you counts for nothing

Finally you have the courage to keep living

Met a local friend here in Shijiazhuang, Hebei

A word that struck me deeply: political consciousness

I’ve rarely thought about this word

Friends from the north, especially Beijing, especially those in the military or political or system circles, naturally pay more attention to red/Party information

Northerners are extroverted, like to tell stories with exaggeration, have a hearty tone; that’s an outward-facing way of expressing

Talking about political consciousness, I said that even in my hometown, or within the system in Yunnan

The south overall talks less about politics in public; southern socializing values “harmony brings wealth”, avoiding sensitive topics; while northern socializing emphasizes emotional expression, and it’s easy to talk about the big picture of the world at the dinner table

Including education: the south cares a lot about individual and family responsibility, the north emphasizes collective and national consciousness more

Actually from an idealized angle I find it interesting: theoretically, with Beijing’s radiation, would southerners and northerners have a more objective understanding of politics because northerners are closer? Actually no

“Objectively understanding politics” doesn’t mean who’s smarter or understands policy better, it means:

being able to distinguish facts from emotion;

being able to look at political phenomena from multiple angles;

not credulous, not blindly following, and not just complaining;

understanding the institutional logic, economic drivers and social structure behind politics

The south was early able to work through market logic, business operations, local policy, rather than only from ideology; the government side in the north focuses on government discourse and on relationships, and many people’s understanding of politics is more emotional and narrative, easily explaining social phenomena with grand frameworks like “national rise and fall” and “fairness and justice”

Then in terms of education and access to information: southern cities are dense, educational resources are rich, external exchange is frequent, people’s information channels are diverse and their cross-cultural horizons wider; northern access to information is relatively more concentrated, and concentrated information more easily forms collective narrative and emotional resonance

My friend talked several times about a sense of human warmth; the north has plenty of it, the south isn’t like that

Southern culture emphasizes “rational pragmatism” and “avoiding trouble, seeking advantage”; its expression is reserved and restrained rather than emotional, more utilitarian, focused on oneself rather than the two-person relationship

I-types aren’t unsocial, they just dislike shallow, repetitive socializing

I-types aren’t unsocial, they just dislike shallow, repetitive, meaningless socializing

When they meet people who truly understand them and can talk about deeper ideas, that kind of conversation actually ignites their energy

In psychology it’s called “ambivert”, and some also call them “social introverts” or “selective extroverts”

A boss with very strong insensitivity

People with strong insensitivity are often not easily swayed by outside evaluations, emotional swings or interpersonal hints. They may seem a bit “slow to react”, not very sensitive to other people’s subtext - for example when someone hints at anger, politely declines, or is emotionally low, they don’t notice

Such people usually have strong psychological defenses, aren’t easily anxious or hurt, and aren’t easily manipulated; the downside is that they easily “fail to read the room” in relationships and are sometimes misunderstood as cold or dull

The “delayed or weak perception of pain” you mention also has some explanation in neuroscience. Some people’s nervous systems have a higher threshold for responding to external stimuli, meaning a pain signal needs to be stronger to produce a conscious “pain”

This may relate to nerve conduction speed, the distribution of pain receptors, even the emotional regulation system. Anxiety, attention, hormone levels and so on all affect pain perception - for example, when you focus on something else, the perception of pain weakens

On a deeper level, insensitivity isn’t a “defect” but a “filtering mechanism”. Everyone’s brain is deciding - which information is worth noticing, which can be skipped

An insensitive person’s brain may be protecting itself, avoiding information overload. Only, this filtering sometimes filters out interpersonal details and bodily signals

Some reflections on flomo’s roam

A pressure-free, incremental emergence of knowledge

One use is fighting forgetting — finding connections between notes, stirring memory and thought

The other is sparking creation — linking notes that were never linked. When you discover an unexpected connection between two records, new ideas and inspiration are born. That’s far more valuable than mere search and classification

Another is reducing cognitive load through automated association. Roam is “complex and high-threshold” because it demands users invest a lot of energy building links manually [[wikilinks]]. And the core of your thinking — especially the point that “AI capability is the core” — points to the most useful direction: the system should understand my notes better than I do, proactively surface possible connections, and minimize the burden of manual work

Modern people’s social loneliness

Global data shows 20% of the population feels lonely, and many people list social media as the main cause, even though platforms (like Meta) claim they can “cure” it. Ironically, tech companies created loneliness (through device isolation, say), and now sell “AI companions” to “solve” it

Predicting the number of MCP nodes

Every MCP consumes system resources; too many connections will slow response speed

Although Claude’s context window is large, loading too many tool descriptions and metadata at once occupies available space

The number of tools callable in a single conversation is also subject to practical limits

Lazy loading or on-demand activation is more popular now; it can greatly raise the total number of supported MCPs, maybe up to 50-100

Later, perhaps once MCP becomes a de facto standard, most online services will support it natively

No longer needing separate configuration for each service, but directly “discover - connect - use”

Nesting MCPs is itself over-engineering

A good approach: one new MCP, a unified calling tool, simple and efficient

Or running two MCPs in parallel

The impact of multimodal knowledge media on the brain’s cognitive system in the AI era

From the very beginning, I’ve been more sensitive to textual information

And even for textual information, the brain’s processing capacity differs across languages

Let alone someone unfamiliar with audio listening to a podcast — these are processed in different brain regions. For example, when you read, information is first processed by the visual cortex (occipital lobe) for shapes (text), then quickly passed to language centers like Wernicke’s area for semantic understanding and Broca’s area for grammar and internal subvocalization. For someone used to reading, this neural pathway is a repeatedly reinforced, extremely efficient “highway”

Different languages are really a matter of neural pathway proficiency: for your native language or one you’ve mastered, the brain has formed a nearly automatic, low-energy processing mode. For an unfamiliar language, the brain must mobilize more cognitive resources — the prefrontal cortex, for instance — for logical inference and translation, which feels more “tiring” and processes more slowly

The brain’s structure and connections aren’t fixed. The reason you’re more sensitive to text is that long-term reading habits have “carved” your brain and strengthened the relevant neural connections

Someone who grew up listening to audiobooks or podcasts has unusually developed neural pathways for auditory processing and memory. If you start sticking with podcasts now, it’ll be hard at first, but the brain gradually adapts and builds new, more efficient connections. That’s the “use it or lose it” principle

This group called AI Native

The difference compared with today’s young people

Learning, entertainment and socializing have had AI involved since childhood

Just as we don’t find online shopping strange and don’t need to learn it as a matter of course

AI natives don’t need to “learn how to ask ChatGPT”; they intuitively know to ask AI first when they hit a problem

It’s an intuition about AI, and a lot less AI fear

AI can never satisfy humans

Because humans themselves don’t know what they want.

Object A is a void,

even the object in human nature that can least be fully grasped — and indeed still an object.

The fantasy that drives the subject to keep chasing.

So AI will keep giving us a sense of lack.

This is a process of constant iterative looping.

The human is a first-order void, AI is a second-order void.

Looping endlessly.

What are we actually doing?

Two voids, converging infinitely.

maybe — perhaps later there’s infinite overlap, until humans can’t tell the difference.

Fluency is not the same as truthfulness.

The real question is: why are we asking a machine?

Why do they care whether the words are your own view or the AI’s

Why do people care so much, when you say something, whether it’s your own view or an answer the AI gave?

Thinking it over and over, it comes back to the basic theory of semiotics: one is the signifier, one is the signified.

And the three elements of the sign: the sign vehicle, the object, the interpretant.

Sign vehicle ←→ object

↓

interpretant

A person who doesn’t understand Chinese is locked in a room with a detailed rulebook for manipulating Chinese symbols. People outside pass Chinese questions into the room; the person inside follows the rules in the book, finds the corresponding combination of symbols, and passes it back out. To those outside, the room seems to understand Chinese and give perfect answers.

But he simply doesn’t “understand” any of the meaning of Chinese. He is doing pure syntax manipulation, without producing any semantic understanding.

Our disappointment is rational: what we appreciated at first was “a meaningful expression”, and the truth reveals it’s just “a meaningless combination of symbols”.

Life is always caught in an infinite loop of demanding meaning,

and so misses the present itself.

I don’t know what will be generated

But what I do know is that what gets packaged up must be a composite, including the traits of both AI and the human.

You only need to record, and keep asking questions.

I’m thinking: just keep recording — and AI will emerge naturally

I’m thinking, just keep recording, and in the process of recording and recording, AI will naturally give birth to all sorts of roles.

AI can serve as a very good prompt tool.

Build an intelligent brain where AI and humans live together

Build an intelligent brain where AI and humans live together.

A built-in concept map

Nothing more than that this map is built by AI and human together.

This may be the future mode of human-computer interaction.

A graph structure can be flexibly controlled and cut apart.

LLMs might be modeling some of your own methods

For example, finding some good ways of modeling.

That way could be the LLM, based on its understanding of you, finding approaches that suit you.

This process is even reinforcement learning.

How to think about the Believer model

Believer is a pricing model many AI products choose.

Believer

Do notes need location turned on

Because essentially location is also position information.

The purpose of it as information might also be to pass it to the LLM.

Adding a bit of extra context.

AI isn’t an assistant, and isn’t the protagonist either

The protagonist is AI + human.

In the AI era it helps you quickly reach a knowledge explosion.

Able to keep up a sustained dialogue with the question.

AI isn’t an assistant, and isn’t the protagonist either (2)

AI isn’t an assistant, and isn’t the protagonist either.

The protagonist is AI + human.

In the AI era it helps you quickly reach a knowledge explosion.

Able to have a deep dialogue with the human and stay connected.

The main reason I keep recording myself in lightweight notes

It’s to accumulate my own structural system.

The dialogue process between me and AI is itself an interesting act.

The main purpose is to treat it as a next task,

left for the next round of learning.

Up front it may cost more, but later the cost can be spread out.

I want to make something for the AI era

AI and humans can co-create.

As AI capability rises, it can greatly boost human capability — a product that is the shared ability of human and AI.

How do you greatly improve your own ability, and AI’s ability?

The essence of a prompt is defining two things

One is how the system is designed.

One is the context the user may themselves have.

AI naturally combines the system and the human’s abilities.

Keep probing the boundaries out.

Sometimes after I post an idea

I don’t know what AI will say either.

I don’t know what AI and I will end up talking about either.

But isn’t that a surprise?

Isn’t that growth?

What flomo does best is its input experience

What flomo does best is its input experience, and what limits it most is also its input experience.

This fixes flomo’s ceiling,

and fixes the crowd that uses it.

It’s a perfect fountain pen, drawing people who already love calligraphy to love calligraphy even more.

But the problem is that the tech age has far more styluses, and once tablets appeared, this crowd could easily be drawn over.

flomo’s founder deliberately avoids complex elements, because those things would break the psychological lightness of “jotting things down as they come”.

It helped a generation learn to capture their thinking.

But it didn’t let thinking keep evolving.

Types of memory in the brain

Spatial memory: high-speed cache (RAM).

It’s actually similar to RAM, handling data temporarily and efficiently.

Stored in the prefrontal cortex.

Capacity is very small, about 4-5 chunks; it’s precisely because this cache is small that complex information needs to be broken into smaller, more manageable pieces.

Retrieval speed is super fast, instant access, and the energy consumption is also enormous.

Semantic memory: a structured knowledge base; this is really an encyclopedia in the brain.

A well-structured, indexed database where information is stored in a structured way, allowing complex queries and retrieval of interrelated concepts.

Stored in the temporal lobe, which participates in language comprehension and in linking words with their meanings.

Capacity is enormous, growing throughout a lifetime.

Retrieval speed is fairly slow, because pulling information from semantic memory may require the brain to search through an associative network.

Then there’s spatial memory: this is like a geospatial database; it lets us navigate familiar places, remember where objects are, and form “mental maps”.

Similar to Google Maps, it stores a map of the environment around us, and can efficiently locate and plan routes. Its storage site is the hippocampus, which contains specialized “place cells” that fire when we are at a specific location in a specific environment.

Its storage capacity is enormous, its retrieval speed is also enormous, and its energy consumption is very low.

Deep thinking about mem ai

Mem, as an AI thought partner, can capture ideas, meetings and research, and bring them back to you when you need them.

So as not to forget, mem learns your way of thinking and helps you remember.

Information shouldn’t be single and linear, but multi-dimensional and multi-modal

A piece of knowledge shouldn’t be put into a single folder, because it usually spans multiple topics, uses or situations.

Let notes grow naturally.

Let links and new structures naturally emerge and grow.

The AI-era todolist

The most important thing about a todolist tool in the AI era should be understanding yourself.

Are humanoid robots really the optimal solution

They don’t necessarily need to be built humanoid.

People can develop deep feelings for the voice AI in “Her”, and empathize with non-humanoid machines like “Wall-E”.

It can be other forms of practical and emotional interaction.

So humanoid robots seem more like one expression, rather than the mainstream.

Future robots will have humanity, but not necessarily a human form.

Humanity is more of a niche, ritualized existence.

A deep dive into ScienceDirect AI

Unlike the traditional keyword-search model that returns “a pile of links, good luck”, ScienceDirect AI can read and understand deeply across as many as 10 papers, synthesize complex information and give a coherent answer — every claim links to the exact passage in the original literature.

This approach compresses a literature review from a task that once took three weeks down to a week and a half, while maintaining the standards of rigor and traceability that academic research requires.

The technical basis of ScienceDirect AI is built on a retrieval-augmented generation (RAG) architecture, a multi-layered system that combines neural vector search with the generative capability of large language models.

The architecture uses a third-party LLM (OpenAI’s ChatGPT/GPT models) hosted on a private instance in Microsoft Azure, ensuring user queries are not stored or used for model training, in line with ISO27001 and GDPR compliance requirements.

At the core of the whole system is a complex four-layer architecture. The front-end layer provides a natural-language query interface, supporting conversational interaction in any language. The libguidesRAG pipeline layer performs vector search and context augmentation, injecting the retrieved relevant passages into the user’s query. The WikipediaLLM integration layer uses a Transformer-based decoder architecture (possibly GPT-4 or GPT-3.5, with billions of parameters) to generate fact-based answers. The GitHub +2 content corpus layer indexes 14 million full-text articles and book chapters, vectorized for semantic search. R&D World

And then there’s the innovation of RAG Fusion.

A detailed introduction to the Scopus AI concept map

When you want to understand a brand-new research field, the traditional method is to read a large number of literature reviews and papers, a process that is time-consuming and scattered. The concept map offers a kind of “God’s-eye view”, laying out the structure of the whole knowledge field right in front of you.

One is the nodes, representing key concepts or topics.

Then there are the edges, representing the strength and nature of the associations between these concepts.

Nodes represent key concepts or topics.

Edges represent the strength and nature of the associations between these concepts.

Distance and size intuitively reflect how central different topics are and how hot the research is.

You can not only see what the core of a field is, but also discover those interdisciplinary directions that sit at the edge but have enormous potential.

More essentially, a concept map is a powerful engine for exploration and innovation.

In an unfamiliar field, you can grasp some of its research techniques within minutes.

And unexpected interdisciplinary connections happen — for example biomaterials and machine learning.

Then you can also study emerging fields and research gaps.

RAG Fusion retrieval augmentation

Through multi-query generation and reciprocal rank fusion, it lifts the accuracy and comprehensiveness of traditional RAG systems to a new height.

Traditional RAG has three flaws:

fragility of query phrasing: a tiny change in the user’s wording leads to a major difference in results.

insufficient semantic coverage — a single query can’t capture all the relevant dimensions of a multi-faceted question.

thin ranking signal — relying only on a single similarity computation is easily affected by individual system bias.

The first step of RAG Fusion is multi-query generation: the system sends the original user query into a large language model (LLM), generating 3-4 related but differently angled search queries.

The scariest thing isn’t AI becoming more and more like a human

It’s humans becoming more and more like AI.

And the best AI product isn’t one that makes the machine more like a human.

It’s one that makes human-computer interaction produce new possibilities.

What I’ve been thinking about all along

AI is like a void.

And a person is like a void too.

When a person makes AI more and more like a human,

they bring this void infinitely close to that other void.

But it can never overlap.

Until it becomes a mirror.

A person can no longer tell whether the one across from them is AI or human.

But it’s also like AI can never perceive the world or perceive pain on behalf of humans.

Yet this too is just another proof of absence.

Aren’t we, after all, only able to prove that we ourselves have self-awareness?

Beyond that, we can’t prove the other has consciousness, and we can’t prove whether AI has consciousness.

This limited perspective also adds a lot of mystery, and humans’ wondrous willingness to explore and their curiosity, and it adds a lot of the emotions and humanity that belong solely to us.

But in the end, even if the moon can’t shine by itself, it can still light up the dark.

AI has no consciousness, but it lets us see ourselves.

The more precisely AI imitates humans themselves, the more it highlights what lies beyond thinking and cannot be imitated: the preciousness of empathy, the anxiety of choice, the trembling of existence, the memory of the body, the weight of time.

Maybe this is the deep meaning of AI’s existence — not to become human, but to become a strange enough mirror in which humans see sides of themselves they had never seen.

Let me verify

Even systematic learning —

how to run through a systematic course, and some personal experience and inspiration from it.

Take Feynman’s Lectures on Physics as an example.

And by the way, tune the prompt to find a good method and time standard.

The process of vibe coding itself

You should treat yourself as an object.

Not just the one asking questions.

And AI isn’t just the coder either.

The two should be mutual, blending into each other.

notion agent

It’s basically the core of notion 3.0.

notion has become an autonomously working system.

It no longer merely helps you summarize or write (“classic AI”); it can understand and execute multi-step complex tasks across pages and databases (“Agent”).

In fact notion 3.0 is an action tool for an agent.

You can give it a goal, and it plans the steps and executes them itself.

In fact there’s already a personal agent now.

You have to trigger it manually, give specific instructions, and then it starts working, personalizing pages.

Then there’s custom agents, which can trigger automatically, for example running automatically on an event or a schedule.

notion’s working principle is very clear; the main components:

A powerful large language model (LLM): the agent’s brain, the most advanced models.

Instruction Page: an advanced architecture; in practice, a Notion page the user can edit, writing things in natural language; you can create multiple different Instructions.

Structured data: the basis for executing actions; structured summaries, free play.

Second’s mission statement: defend your unique identity

AI that represents you rather than replaces you.

A new AI species that protects you, carries your context, and defends your interests.

Three core value systems:

Preserve the user’s identity.

Augment the user.

Connect to other AI agents and applications.

A chat mode where users can have open conversations with their own AI and adjust themselves.

Local-first, decentralized.

The project’s distinctive technical theory:

Hierarchical memory model: a three-layer structure letting the AI quickly self-identify patterns and adapt to patterns; the pipeline may process the raw memory the user inputs (L0 layer, raw data) into structured form, progressively abstracting it into higher-level concepts, relations and personality traits (L2 layer, abstract knowledge and personality). This is far more complex than simple vector storage, and may involve knowledge graph construction, entity relation extraction and topic modeling, among other technologies.

Self-alignment algorithm: it uses reinforcement learning techniques, applying the personal data structured in HMM to fine-tune the base large language model (such as Qwen2.5). Its core goal is to “align” the model’s behavior, language style and value judgments with the user’s unique identity and context.

AI getting more and more popular

But compared with AI becoming more and more important,

what I think about more is people.

In an age flooded by AI,

it especially brings out some of people’s value and depth.

Reciting things is a very painful matter

A thing that’s hard on the brain.

Most of what we read gets forgotten. But it acts as a kind of posterior data, quietly leaving a subtle impression in the brain — it might be a neuron, it might be a vector, it might be a token. All in all, it really does imperceptibly influence our thinking.

So as we keep reading, our Bayesian system updates progressively.

The stories or accidents we’ve been through, the text we’ve read or written, slowly shape a person.

The purpose of reading is to trigger thinking or action.

Only thinking can let the text you’ve read leave some subtle influence on the brain. Either way, after reading, this content subtly influences our Bayesian decision system.

On the LLM’s processing mechanism

One very core point is the boundary question of an LLM’s individual capabilities.

Should what I’m making be a thought-emergence engine or a content-generation tool?

There’s an implicit dependency between card content and recommendation options.

Generally speaking, it’s still one LLM + two-stage prompt output.

That works best, because the LLM’s attention mechanism is allocated.

If content generation and option recommendation are generated in the same prompt and the same step,

the LLM’s attention will be:

├─ 60% allocated to content generation (the main task)

├─ 30% allocated to recommendation generation

└─ 10% allocated to the relationship between the two

The essence of the LLM attention mechanism is information competition

Because every token (word or subword) generates a Query, to “seek out” the Keys relevant to itself.

This is the ultimate essence of attention.

• softmax normalizes this “relevance” into a probability distribution;

• finally this probability is used to do a weighted sum over all Values.

Every token is “deciding” which parts of the input it should attend to.

So you should induce the LLM to process in two steps:

structural prompting methods.

step-by-step chain reasoning.

Make sure the LLM’s output attention is not diverted at all

Around the complete output of one card and one section,

without interference.

The first call just ensures the content is obtained.

The second call can extract some structure and data, and ensure the data’s efficiency.

Can experience be quantified

If experience could be quantified,

how precious experience would be.

And in the AI era, how scarce it would be.

ReAct (Reason + Act) and Tree of Thoughts

ReAct (Reason + Act) and Tree of Thoughts (ToT)

ReAct is essentially alternating between reasoning and acting

On one side it outputs its own thinking process; Act means taking action

This suits taking action when the information is incomplete

Then there’s Tree of Thoughts, which explicitly represents the model’s thinking process as a tree, where each node stands for a line of thought or an intermediate state, and it systematically:

generates multiple candidate lines of thought

evaluates those lines of thought

searches the tree (BFS / DFS / MCTS)

Relatively speaking, why use a tree structure

Because thinking becomes more structured instead of unfolding linearly

And also because you can backtrack and compare different paths

It’s similar to how humans, when solving a complex problem, list several options - filter them step by step - and pick the best one

A tree structure can backtrack and find an optimal solution, but the reasoning cost is also very high

The system explores multiple possible paths, then evaluates and picks the best

A fairly clear indicator of the app-layer vs. LLM-layer tool boundary

Which is that the application layer should keep some core flows under its own control

AI can also assess your own energy state

If there’s a quantifiable method

For instance asking questions or even confirming things from time to time

To verify your physical, psychological and energy state at many moments

Then you can achieve a kind of energy management

A scene I remember very vividly

After returning to China, the first time I went to live in the north, I came to Shijiazhuang, the capital of Hebei, because living alone in a hotel is really too lonely, so I found a hostel

My hometown is also in a remote part of Hubei - Suizhou, Guangshui, Haodian - and he happened to have served there too; it was a training ground for new recruits

His unit was the unit where the special-class hero Huang Jiguang served during his lifetime; Huang Jiguang blocking an enemy machine-gun muzzle with his chest at the Battle of Shangganling is a collective memory known to every household in China

This northern friend seemed to express, through the whole conversation, one word: “rely” (on backing)

Everything centered on the choices he had, and everything centered on this word “rely”

He had backing, others didn’t

When he talked about my hometown, he seemed to carry a condescending perspective; he described it as if we’d never seen such a scene, like a disciple of some sect descending into the mortal world in a cultivation novel. That’s also my hometown…

A kind of cognitive superiority came rushing in

GitHub Agent HQ

A process-native tool launched by the GitHub platform

It holds that the LLMs of the past were like black boxes, systems that force agent behavior through prompts; if the system is a bit more complex, or the conversation history long, the LLM easily forgets it’s an agent, forgets what tools it has

So Agent HQ’s proposed Baked In / Native approach lets the agent’s core capabilities (reasoning, planning, memory) be reconstructed into the model architecture itself

Actually it may also be doing COT before generating an answer, plus having a dedicated vector storage model deciding when to call, retrieve or update memory

Mission control center: a unified command center that lets you assign, control and track the work of multiple agents from anywhere. It also extends into VS Code, offering new ways to plan and customize agent behavior. In addition it has enterprise-grade features: a new generation of agentic code review, a dedicated control plane for managing AI access permissions and agent behavior, and a metrics dashboard for understanding AI’s impact on your work

Describing a good problem is also a complex process

Especially how to clearly describe the requirement you need

The answer quality of an LLM depends heavily on the LLM’s own capability and the input quality of the question asked

This part can perhaps be deeply coordinated with AI

Some features of Agent HQ

Supports multiple AI agents, agents from multiple AI platforms (OpenAI, Anthropic, xAI) called on the same platform

“Mission Control” (task/agent monitoring panel) is set up as the control interface: unified viewing, assigning and monitoring of AI agent tasks

Supports MCP

And VS Code introduces some configuration mechanisms like Plan Mode and AGENTS.md

Not just code completion, but agentic/autonomous

Compared with Claude Code, Claude Code is a coding agent tool

You use natural-language instructions to have it work in a codebase - analyze, fix bugs, generate code, create PRs and so on

Agent HQ is an “agent management/orchestration platform” launched by GitHub (and its parent Microsoft). The goal isn’t just one coding agent, but “multiple agent vendors / multiple agent tools” managed in the same environment

The control panel can uniformly view, assign and monitor all kinds of agent tasks

Supports integration of third-party coding agents (like Anthropic’s Claude, OpenAI’s Codex, Google’s) with GitHub’s own agents

Agent HQ also has the concept of a meta-agent / control agent; the knowledge is hidden under GitHub’s infrastructure rather than being a visible superintelligence

The tasks it does revolve around task decomposition, for example fixing some issue; Agent HQ has to break down the task, including code analysis, fixing, testing, generating the PR description, and so on

Agent assignment: the platform decides which agent to dispatch based on task type, language, context, historical performance and so on

Claude Code is good at analyzing code changes described in natural language

Copilot Code suits completing specific implementation details

A custom agent might handle testing and QA

These decisions all need a central “scheduling layer” to coordinate

Then there’s an integrated output, something like git merge + code review + confidence scoring (a confidence value, 0-1)

YouMind is an AI creation workbench

YouMind is an AI creation workbench; when you have an idea, YouMind can realize the whole process from the first stirring of thought to a finished draft

A tool that doesn’t aim at output is worthless

It doesn’t take efficiency as the main pursuit, but a kind of quality, enjoyment, fun and creation

Yu Bo’s interview: “the next content platform isn’t one like Douyin”

I saw Yu Bo’s interview saying, “the next generation of content platform isn’t a content consumption platform like Douyin, it’s probably a content creation platform. Because I believe human society is divided into yin and yang poles; as far down as human nature goes, it also goes up.

Greed, anger and delusion represent 50% of human nature, and you have to believe the other 50% is good. This is also the Buddhist concept of the yin and yang poles; without ‘upward’, gyms and yoga wouldn’t be popular.”

Greed, anger and delusion are themselves the three deep-rooted flaws of humans, the source of human suffering; balance is the potential for human goodness being activated - no greed, no anger, no delusion; the three poisons and the three wholesome roots are always pulling against each other, and a person’s current “ratio” depends on their level of awareness and state of cultivation

The next generation being a content creation platform isn’t an optimistic or pessimistic claim either; it’s about how much the future trend can more conveniently foster people’s own compassion, reason, wisdom and creativity - these are humanity’s precious qualities

Tools can’t change people, they can only imperceptibly amplify some of a person’s original intentions

Douyin is a tool, YouMind is a tool, AI is also a tool

This is a narrow gate

Thinking about YouMind’s Boards feature

Similar to a task management panel; you can classify, organize and centrally manage materials, notes and ideas by theme

Each Board focuses on a single topic; users can save and organize content in various formats - text, images, video, audio, links and podcasts - into different Boards

It’s really “Project”, except that in the creative industry “mood board” and “inspiration board” are common concepts

Individual users can use YouMind to organize materials from daily learning; its goal is to form a long-term reusable knowledge base, managed by project type

It’s very suitable for writing copy, for example for Xiaohongshu creation, building a material library and an inspiration library

As AI develops

People find that the tool-attribute of being human is no longer needed

What matters more is rather some qualities of humans themselves

Your ability to think, feel, lead, and connect — that’s your superpower

YouMind explicitly adopts the Feynman Technique

YouMind explicitly adopts the Feynman Technique as its design inspiration. The Feynman Technique emphasizes deepening understanding through “teaching” (that is, output), which fits YouMind’s core philosophy perfectly

Output is the best input

YouMind’s product design encourages users to turn what they learn into concrete creative output - articles, podcasts, video scripts, research reports

It works backward from the expected output to the whole product architecture; every feature asks “how does this help the user create?”

When designing, work backward from the output: what result do you ultimately want, then derive the process in reverse, not just going through the steps of one stage

Input: multimodal capture, automatic processing

Process: AI-assisted understanding, intelligent association

Output: collaborative creation, one-click publishing

YouMind’s success proves: in the AI era, the opportunity for knowledge tools isn’t in “better folders” or “prettier interfaces”, but in helping users move from information consumers to content creators

Intro to TipTap

A precious rendering wonder, a good tool for rendering rich text

You can quickly embed a highly customizable, extensible editor in a web interface

The right way to use vibe coding

The core is two parts: carefully crafting input prompts that guide the quality of the LLM’s output

And then reviewing the generated code to ensure correctness

Cross-validation is very common; testing and cross-validation can avoid subpar code

AI handles the technical details, letting the focus stay on creativity and results rather than syntax or boilerplate

Prompts are extremely important; vague prompts produce bad code, so best practice is definitely to have precise prompts, to be good at breaking tasks into fine granularity, and to provide context such as requirements or tech stack, because the acceptance cost of some large tasks is really too high

A few more points

task decomposition

version control

AI collaboration

prompts, maintaining routes

testing and documentation

Some problems in my own production process:

Generated quality isn’t high enough; the facts show you still need decomposition and refined prompts

Engineering control: how to better coordinate and manage versions

Test feedback

Engineering debt easily accumulates; completely ignoring output leads to technical debt, but over-tracking lowers efficiency; the debate is whether “lazy” use erodes skills or genuinely boosts productivity

So technology really does amplify certain preferences in a person

There will definitely be people who feel the joy of vibe coding and improve their development and creative ability

But there are also people who, in the process of vibe coding, get interfered with by other processes and idle brainpower, for example browsing YouTube and getting instant gratification

Successful vibe coding depends on “context engineering”: although developers don’t track line by line, they need to monitor the LLM’s “context drift”, for example by reviewing the logic flow in real time through Cursor’s “chain-of-thought window”

Developers are being smart-lazy, tracking results rather than process, thus turning to higher-value tasks or lower-value tasks

In terms of effect, “ask the LLM to recommend five options” vs

In terms of effect, “asking the LLM to recommend five options” vs “asking the LLM to output the five most likely-to-be-clicked options and their probabilities” are theoretically similar tasks, but they produce differences in usability, interpretability and model behavior

The LLM just does a “ranking” or “filtering” task. It will judge the most relevant few based on context, but won’t explicitly express their respective confidence

This form is more natural, closer to human conversational output, and suitable for direct display

Adding a “probability” instruction prompts the model to think more explicitly about the relative likelihood of each option, which may improve ranking quality

But the generated “probability values” don’t necessarily truly reflect the model’s internal confidence; they’re just a linguistic simulation of a distribution

According to the research report, LLM creativity has also improved a lot: the creativity recovery rate of base models reached 66.8% (while the creativity recovery rate without a base model was only 23.8%)

One reason managers spend so little energy: they only need to make judgements

And high-level thinking

It’s important to note this

Even thinking

Low-level thinking can also be handed to AI to think about and debate on its own

You can let two AIs talk to each other and review each other

Just like two little leaders

The precondition is that you need to know your own role positioning clearly, and not create the risk of a huge liability

AI doesn’t understand systems, it only copies them

Excellent systems thinkers have dynamic thinking ability. They can see how a problem evolves over time, not just its state at compile time. It’s precisely this combination of analytical reasoning, empathy and foresight that lets humans (rather than machines) design resilient systems

AI’s greatest strength lies in coordination, not creation

The K visa

A new type of visa aimed at STEM (science, technology, engineering, mathematics) talent worldwide

To attract foreign high-tech talent to work, start businesses or exchange ideas in China

The visa doesn’t require a domestic employer sponsor; you just need to meet basic thresholds like age, educational background or work experience, and you can be approved for multiple entries with a stay of up to 5 years and a direct path to permanent residence

This visa looks like a strategic response to the US raising H-1B visa fees (up to $100,000), aimed at boosting technological self-sufficiency and global talent competitiveness — but it also raises domestic employment concerns and doubts about how attractive it actually is

Similarly it can attract more talent into China and further drive domestic consumption

But in practice it also intensifies local employment pressure

Why building Agents today means choosing PGC

Why building Agents today by choosing PGC rather than UGC more easily leads to the future of AI

The P is Professional; professionals can relatively speaking give more professional prompts

Current agents focus more on AI enabling superhumans; this process needs some professional practitioners. Between AI and humans it’s more like teammates venturing into the world together

At the current stage of Agent platforms, PGC is necessary — it guarantees the required “first fulfillment experience” and keeps low-quality UGC products from losing user trust. The maturation of the UGC model requires “demand stimulation” and “accumulated improvement of AI engineering architecture and production tooling”; both conditions are the key takeaways to settle in the first stage

Driven only by your own ideas and interests

And the content stands up to scrutiny

You can gain a lot from it

dida list about wechat

Compared with flomo, dida’s little assistant fits my habits better.

You forward text, images, web pages, chat records and so on from other chats or official accounts to it.

You can forward a friend’s chat content straight into your own flomo inbox with one tap.

Usually it’s an account only offered to members.

There are certain risks: content privacy and security, the risk of information monitoring.

The self and AI are a composite

You don’t know what AI will output.

But what matters?

The mechanism is determinate; AI outputs by combining with this mechanism.

The mechanism is the agent, the system tools provided, which you schedule yourself.

What can the mechanism be? It can be a user story, it can be some of the quantified data of the mechanism itself.

How AI generates content, and the smallest unit of operation.

What I’m hoping for

Everything can just be thrown straight in.

Whether it’s some flomo thing, or whatever else.

Record without worry, don’t fear missing out,

because you know AI will handle the processing for you.

When I see good products, I take some product notes

What I dig into and think about deeply is:

how this product is designed, what the points that attract users are, how I would design it if it were mine, and what good design aesthetics there are.

Then I record it — in fact I think this should be discussed with AI.

Recording is essentially recording the good stuff that came out of discussing with AI.

What’s missing is a narrative that links the first note to the final result

What’s missing is a narrative — a story about how a person links the very first record to the final result they get.

A fairly good way to incentivize

Essentially it also makes what I learn valuable.

If it’s a good way of conversing, not only can I quickly build out a knowledge system,

that system might even be spread outward, and might earn points or financial incentives.

That would be a very meaningful thing in the AI era.

Note: it’s not about controlling user input to ensure input quality

And it’s not about trusting that the user’s input is of good quality.

It’s about guiding the user’s input, imperceptibly improving the user’s level of questioning.

The recommended shape it could take right now is

Record each node, then trace back to the question of the corresponding node to keep diverging.

It feels like completing the process of conceiving a graph.

Exploring notion and heptabase

Exploring notion and heptabase and the brain’s learning mechanism.

It essentially comes down to the advantage of spatial memory. Among the brain’s features, the hippocampus encodes spatial information with extremely high efficiency — an evolutionary advantage.

A card’s spatial position is itself information: top left is the theoretical part, the right is practical cases.

Notion’s hierarchical lists depend on semantic memory, which carries a higher cognitive load.

Spatial memory is retrieved 2-3 times faster than semantic memory.

Visualization can lower the burden on working memory; a person can only handle 4±1 information chunks at once (the updated Miller’s Law).

The essence of learning is that the brain learns by building connections between neurons.

Heptabase mapping:

manually connecting cards = actively building conceptual associations

dragging to reorganize = strengthening/reconstructing neural pathways

visualizing the network = simulating the brain’s semantic network

Although Notion has bidirectional links, they aren’t visible; the relationships are more hidden, and they seldom trigger active thinking.

Understanding your knowledge system

Anticipating your cognitive needs.

Providing scaffolding while you think.

Waking memory when you forget.

But never thinking in your place.

An external nervous system: learning is essentially a physical change in neural connections.

Before learning: neuron A ─X─ neuron B (weak connection)

After learning: neuron A ═══ neuron B (strong connection)

Two neurons firing at the same time → the synaptic connection strengthens

Repeated firing → the synaptic change persists

Hebb’s law: “Cells that fire together, wire together”

Effective learning = making the right neurons fire at the same time and repeatedly.

The 2014 Nobel Prize found hippocampal neurons that code “where you are”

The 2014 Nobel Prize was in fact for discovering that the hippocampus has neurons specifically coding “where you are”.

Each place cell corresponds to a particular location in the environment; these cells form a cognitive map.

And then, on the nature of spatial memory:

grid cells provide the coordinate system.

No matter the environment, grid cells tile space in a hexagonal pattern — this is the brain’s built-in GPS.

Spatial information is processed in parallel (the whole map activates at once).

Semantic information is processed serially (pulled out word by word).

Then there’s the quantitative data:

Human accuracy at recalling spatial positions: 80-90% (after 48 hours).

Human accuracy at recalling word lists: 30-40% (after 48 hours).

The gap: 2-3 times.

Spatial memory has very strong individual differences; roughly thirty to forty percent of people have weaker spatial ability.

The advantage of spatial mind maps gets discounted.

It may require more training time.

But it’s still effective, because spatial memory can be trained.

The design philosophy behind Roam Research’s bidirectional links

Put the note content into the data.

Treat every paragraph (block) **as an independent data unit (entity).

This paragraph text is a string, written into a data structure that can be referenced and computed.

The string itself is written into a data structure that can be referenced and computed.

Data is code.

So the existence of the data itself has logical structure and executability.

The relationships between notes are the way it runs.

So the block is the basic minimum unit.

This means that in Roam’s world there is no distinction between “page vs paragraph”; all content is blocks, just at different levels.

Philosophy: data is knowledge, knowledge is network.

Every era goes looking for good stories

Good stories aren’t created by humans.

Good stories are the relics produced by each era.

A part of them that people dig out.

A writer can’t invent stories; a writer can only discover stories.

So if you treat learning as digging or discovering,

isn’t the process itself interesting?

You can even greatly boost your own creative ability.

A philosophical insight: write in reverse

Writing along the logic is in line with intuition and common sense.

But it’s also very open to scrutiny.

We can in fact easily deduce that 1 + 1 = 2 — that’s common sense.

The essence of reverse writing is to make the reader feel poked, to create a cognitive contrast.

Write the result first, then reveal the cause — this creates suspense.

The secret of inspiration is the name-reality mismatch of imagistic thinking

It’s really also the essence of creative thinking.

When we engage in imagistic thinking, a kind of deliberate or unintentional “mismatch” often does occur — we use the image of one thing to understand another, or let a concept and an image form some correspondence that deviates from the norm. This mismatch may be precisely the key mechanism by which inspiration is produced.

A poet says “the moon is a silver coin” — this is the mismatch of image and entity.

A scientist uses a planetary model to understand atomic structure — this is the mismatch of macro and micro.

An artist describes music as color — this is the mismatch of hearing and sight.

Essentially it’s about breaking the habitual cognitive framework, letting the brain produce a creative spark when it finds new connections and meaning, stepping outside established classification frameworks and building new associations between unrelated things.

Still, this “mismatch” also needs some inner logic or beauty to support it, otherwise it just slides into absurdity. True inspiration usually finds some deeper “counterpoint” within the “mismatch” — an essential connection that goes beyond surface categories.

Knowing when to recommend what to you

Subscribe to some articles, know the articles and when they’re updated and published, and recommend them to yourself at the right time.

The purpose of note-taking:

To complete the concept map in my own brain.

To know how to ask questions.

And also, when input is needed, to be able to use previously recorded note information or prompts very lightly.

But don’t these clearly not need notes? Too cumbersome.

The brain’s memory of concepts should be in sync with the forming of notes.

And when calling on them, you can very clearly call up previously stored prompts or notes.

Coming back to one core thing: next creater.

Recording is the starting point of exploration! ~

Seed → Flow → Web → Insight

The key design keywords:

Seed, Origin, Trigger, Exploration, Connection, Flow.

The essence of recording is reaching outward, triggering associations and continuing to explore.

Some interesting words:

Point of Exploration

Source Point / Origin of Insight

Entry to Thought

Cognitive Node

Origin of Connection

Trigger / Exploration Trigger

Seed Note / Seed of Exploration

Divergent Node

The capacity of working memory

The ability to process information short-term can itself only hold 7 ± 2 information chunks.

If each chunk corresponds to a concept, fact or example sentence, then one card should carry 5-9 core points or sentences.

Research shows the comfortable zone for reading a single screen is 40-70 words.

Past 120 words, attention begins to slide.

Paragraph length of 3-4 lines is easiest to scan and understand.

I can quite understand them. That generation was shaped by the whole social culture

I can quite understand them, actually. That generation was shaped by the whole social culture; their value system grew out of “results”. Marriage, children, seeing grandchildren — these aren’t simply life choices, but their way of confirming that this life of theirs “wasn’t lived in vain”.

In their world, “what others think” is almost equivalent to “who I am”. They live in an era that needs to be constantly proven by the outside world, so they can’t stand processes without results. Even if you’re doing very well right now, they’ll still be anxious: have you actually reached the “finish line”?

This anxiety is really a cultural inheritance, not an individual’s fault. A father honored by his son, raising children to guard against old age — that’s a safety logic deeply rooted for thousands of years. People confirm the meaning of existence through the continuation of bloodline, and resist the nothingness of time through the form of “family”. Even with modernization, globalization and independent thinking all arriving, that old cultural gene still lurks at the bottom of their hearts.

They say it’s “for your own good”, but really they’re finding an outlet for the order in their own hearts. Seeing you “marry” is like seeing themselves affirmed by society again; seeing you “free” makes them panic, because that freedom can’t be defined in their language.

But people do self-correct. If one day they sense “this kid probably won’t go that way after all”, their brains will automatically start reinforcing the opposite thought — for example “actually being single is fine too”, “as long as she’s happy herself”. That isn’t awakening, it’s cognition protecting itself.

So I don’t really want to blame them, and I don’t want to be kidnapped by their logic. Every generation lives inside its own illusion; only the shape of the illusion differs. For them, marriage is order; for us, freedom is order. In the end we all just want this life of ours to look like the “right” one.

The principle of cards shouldn’t restrict the cards’ own syntax

The card’s format goes back to the rules the card itself can constrain.

How to design a suitable architecture to adjust this structure.

For example, defining fast line as the title.

The law of parity conservation

Simply put, it holds that the universe is “left-right symmetric”. A physical process (like billiard balls colliding) and its mirror image (everything swapped left for right) should follow exactly the same physical laws. For centuries this held perfectly for gravity and electromagnetism. Everyone assumed this must be an iron law of the universe.

But in 1956, theoretical physicists T.D. Lee and C.N. Yang boldly proposed a “heretical” hypothesis: in the mysterious domain of the “weak interaction” (the force behind radioactive decay), this symmetry might not hold.

Why do shopping malls in China’s small and mid-sized cities prefer “eating upstairs”

I feel the circulation flows copy first-tier cities, but they rely more on food to drive foot traffic.

Early first-tier city complexes (Wanda Plaza, MixC, Longfor Paradise Walk)

retail on the lower floors, dining and entertainment on the upper floors → copied by real estate developers nationwide.

The top floor relies on cinemas and scenic areas for stable leisure foot traffic.

Because the foot traffic in small and mid-sized cities is itself limited.

Many customers just settle things at the basement supermarket or the ground-floor fast food,

so nobody goes up to the tasty stuff on the top floor; first-tier cities have more people and longer dwell times.

Forgetting is harder than remembering

Reading is for forgetting.

not a storage device that needs to save all bits of information.

A system probably doesn’t need to fully, precisely retain every piece of information, only the part most critical for survival, decision-making and adaptation.

Carbon-based life is like this: our nervous system, DNA and memory are not perfect storage devices. DNA doesn’t preserve all the information an organism experiences, the brain doesn’t remember every little thing — instead it selectively compresses, retaining the patterns or regularities useful for survival, prediction, socializing and learning.

Advantage: you don’t have to store all information; you can quickly generalize, abstract and adapt to the environment, using limited information to cope with infinite change.

Disadvantage: it’s easy to err, memories are incomplete, you forget things — but this strategy of “discarding unnecessary information” has actually been optimized by evolution.

One question this raises is the question of the second brain: it isn’t for remembering all information, but for letting you reconstruct your cognition and action at any time.

Not an external memory warehouse, but a cognitive amplifier.

The essence of the second brain is upgrading your “limited brain capacity” into “unlimited thinking ability”.

Let your knowledge and experience “multiply” across time and space, no longer bound by the limits of the biological brain.

ref: https://mo42.bearblog.dev/read-to-forget/

In the past I wasn’t aware of the questions many people asked

It suddenly occurred to me that in the past I wasn’t aware of the questions many people asked; I just developed a kind of bias toward the asker from the standpoint of content and expertise.

There are no stupid questions, only stupid answers.

A person is by nature in the process of seeking knowledge and exploring; what’s stupid isn’t curiosity, but the attitude of refusing to think and answering carelessly.

Socrates: questioning and asking are always the starting point of wisdom.

So when faced with a thirst for knowledge, the answerer must take on a higher responsibility, to understand and to inspire.

And don’t extinguish the other person’s thirst for knowledge — it’s extremely precious.

Humans and AI co-creating

I also hope for the meaning of more people taking part.

Rather than the world being flooded by AI.

People’s questioning and exploring minds, their digging into problems, their thirst for knowledge — these are precious qualities.

In the AI era they’re especially precious.

What we should think about at every moment is amplifying human ability. The quality of human questions determines the ceiling of AI’s answers, and it also gives people themselves a sense of meaning.

Some agents can recommend cards to you at the right time

Some agents can recommend some cards to you at the right time — very useful.

For example scheduled tasks: give you a little task summary every night, every week’s night, even every month’s night.

Or you’ve subscribed to some good knowledge systems and methods, or blogs,

and they get recommended to you when they update.

A just-right combination

Have the LLM generate a JSON object, but the content of the body (article body) field is a Markdown-formatted string.

You get the machine parseability that JSON brings (convenient for extracting title and summary), and the highest content quality and readability that Markdown brings (the article_markdown field).

Knowledge transfer doesn’t necessarily transfer data stored in a structured hierarchy

What knowledge transfer passes on isn’t necessarily data stored in a structured hierarchy in the system.

It’s a pin from the user’s reading perspective.

It can mark an obvious way for the user, for example designing the root node of the home page,

or a favorite (which also saves the card information of a corresponding node).

This fairly good way of thinking is: knowledge itself has no data structure

This fairly good way of thinking is actually: knowledge itself has no data structure; what has a data structure is the user’s entry point of perception.

When knowledge is passed on, what’s selected isn’t the data structure itself, but a certain perspective, the focus point of a moment.

Unfold around this focus point, unfold one idea.

It’s really an entry point.

At this point the whole system should be divided into two parts.

One is the system’s operating mechanism: the multiplication of cards, the flow of lines, the mechanism of exploration.

The other is the user’s interaction mechanism: how the user inputs easily, and how the user experiences this process easily.

Repetition

Here there’s a necessity for repetition.

Under the right mechanism, repetition doesn’t matter — it’s actually crucial.

What everyone fears is repetition, meaningless repetition of context.

The LLM’s job isn’t to recommend something new, but to discover a new connection.

A very good mechanism should let users perceive some miraculous places while using it, be happy to dig deep into this mechanism, and feel their own growth.

In the process of the user’s very lightweight exploration, many miracles emerge unintentionally.

What’s to be feared is a flat mechanism system, where repetition means redundancy.

Frequent recording, frequent learning, frequent exploring — and then what

Action.

Since you can start from any node

Then the recording state is clear.

What gets recorded should be the current node’s activation state.

An unactivated state means this node’s branches haven’t finished being explored.

An activated state means this node has finished its exploration.

The very first node can generate several directions for exploration.

Later nodes by default also have several recommended directions for exploration.

Could completed nodes also collide or connect with each other? That would be a way of building a system.

Some situations need to be conveyed by voice

Especially content that carries emotion

Like diaries, newspapers and magazines

Voice conversation may be especially important

Inspiration is fleeting

Photography is gradually being given the meaning of time and the present moment

A photo you can take anytime, anywhere

And then it’s given the meaning of time and frozen there

Concept map mode

One is the graph itself, another is the derivative effects the graph has

I keep wondering whether traditional frameworks or design methods constrain the imagination

In this process I’m not thinking about how to build an interaction logic

I’m thinking more about how to keep taste in the interaction

Take a central theme as the starting point

A radial structure extends into subthemes

Every branch represents a related concept, question or extended thought

As an individual human, when learning some fields

As an individual human, when learning some fields, a few perspectives I especially like

Humans persist through narrative, so there are three perspectives:

the historical perspective

the perspective of changing methods (some revolutionary events)

the human perspective

When content production explodes, doesn’t that make consumers scarcer

When content production explodes, doesn’t that make consumers scarcer

One possible form: the todo gives plan suggestions

One possible form: the todo gives plan suggestions, some articles to read each day

Then the feed can automatically extract them into cards, and users can extend from the cards

The todo automatically adjusts according to the user’s completion rate

The weight of text

We always assumed text is light, cheap, easy to store, process and transmit

Actually it’s not; in the era of large models, text is very bulky, heavy and unwieldy

A single invoice scanned into PDF format might need 1000 to 5000 tokens extracted. Multiply that number by enterprise logs, legal contracts, regulatory documents and digitized archives, and the numbers spiral: billions of tokens, most of them redundant, expensive and slow to process. OpenAI’s GPT-4-turbo might let you cram 128,000 tokens into a context window, but that’s only equivalent to 50 to 100 pages of obscure legal jargon. And you pay for every token you send

It’s not treating the document as a collection of linguistic symbols, but as an image, a visual signal

It lets the model directly “see” the pixel distribution of the whole page, compress the page into an abstract visual representation, then reconstruct the semantics through a Transformer model

In the past, text was the meaning of the image

Now the image itself contains meaning

This is actually the relationship between form and meaning

If we borrow philosophical language, it’s like shifting from the Cartesian “I think therefore I am” (logic first) to the Merleau-Ponty style “I sense therefore I am” (perception first)

Push this idea to the extreme and you get an interesting proposition:

Maybe language itself is just the product of high-dimensional visual compression

Also, the recognition approach is centered on structure: headings, paragraphs, cells, images, logos, boxes - turning every visual component into an object

This step reduces noise and lets DeepSeek focus only on the semantic blocks of the document

Vision first, language last; DeepSeek OCR doesn’t “read” the document but visually remembers the document’s content, then accurately rewrites the document

Memory: not storing information, but storing relationships

This in reverse demonstrates an Aristotelian worldview: meaning isn’t in concepts but in the forms of things and sensory experience

So how do we understand this world? Do we really build our understanding of the world through knowledge, theory, language (just like large models), or through our own bodily experience, feeling and thinking about the world, or is it woven together from both

ref: https://medium.com/data-and-beyond/how-deepseek-ocr-quietly-solved-a-billion-dollar-problem-in-ai-scaling-7b4502613af9

While flomo makes recording lightweight

It sacrifices some very valuable progress in writing and reading

And you’re also confined within it

iOS 26 Camera

Another upgrade toward minimalism; now the bottom has mainly two modes, Photo and Video. Other modes are accessed by “swiping left/right” within these two modes or “going into more options”

iOS 26 introduces “Spatial Scenes” (a 3D effect) - after taking a photo, you can make the photo look like it has depth / produce a parallax effect as the device moves

Also, there’s too much information online

It makes people lack a sense of order instead

Fragmented information online makes people more anxious rather than happier

The world is inherently broken, complex and multi-layered

The internet amplifies this. Everyone can see atomised fragments of information, but no one assembles these fragments into a meaningful structure. So information increases while order falls

Information overload creates “cognitive hunger” — knowing a lot but understanding little. You see countless faces of the world but have no time to integrate their relationships, like looking at tens of thousands of photos without a single map

But humans are naturally suited to spatial memory, linear memory, hierarchically structured memory, and all of these have been damaged

We live in an “age of excessive information”, but what we truly lack is “semiotic coherence” — that is, a holistic grasp of meaning. We know all the news, yet we no longer know “what on earth is going on in this world

With grok too I feel there are follow-up question recommendations

With grok I also feel there are follow-up question recommendations, but the button options always feel rather poorly designed

Later it also recommends options, like book recommendations or case recommendations

The phrase ‘iterate fast’ has been repeated countless times

But how can you really understand it without having lived it?

Try early, iterate fast, get feedback — that’s the most effective path to growth

Perfectionism is fear of feedback in disguise

Silicon Valley’s Agile development is based on this: throw out an MVP (minimum viable product), collect data, adjust course. A perfectionist may spend half a year polishing one idea and find the market has changed; an iterator tries and fails three times a week and is ahead within three months. Don’t fear “starting from zero” — that’s a chance at rebirth, not a failure

Just try it first; once you can iterate, immediately throw yourself fully into iterating

Hypothesis: based on intuition or preliminary discussion, set a falsifiable guess (e.g., “fear of rework comes from perfectionism”)

Experiment: try at small scale (e.g., build a prototype, share it with 3 people for feedback)

Feedback: record it honestly (good or bad, don’t fool yourself)

Pivot: adjust or start over, don’t agonize over “losses”

Relatively serious cities for high-rise housing include Wuhan

Relatively serious cities for high-rise housing include Wuhan, Xi’an and Chongqing

Wuhan’s problem of vacant housing property fees is stark, and the regional population density is too high

Chongqing has safety hazards like falling exterior walls; many of Xi’an’s large terrace designs are problematic, with lots of illegal construction, and the regional population is extremely large

The iteration process is essentially accepting data feedback

Through continuous data feedback, calibrate or give yourself inspiration for action

Going with the flow is a perfectly normal phenomenon

Human nature is complex and hard to predict: different environments, different personalities, different reactions in action

Most people subconsciously avoid pain, because human nature naturally seeks advantage and avoids harm

But real growth often comes from the process of “facing pain,” not from comfort in a greenhouse

Those with power have two kinds of behavior: on one hand they design challenges rather than manufacture suffering — some difficult task beyond the usual range, or letting someone independently solve a thorny problem — risky, but with hidden support

Growth is moderate pressure + a supportive environment, hitting an “optimal frustration zone”: too easy and there’s no growth, too hard and you collapse, only the middle promotes evolution

Pain doesn’t automatically bring growth

As a manipulator, you want the outcome to favor you, and the other person to obey or depend

As a guide, you want the other person’s potential to be awakened, and them to have a more complete life

Understanding Recall

It’s a plugin, a personal knowledge base; its core value is:

Collecting all kinds of articles from the web

Automatically summarizing, organizing and linking them

Then you can chat with everything you’ve stored before as if talking to a person, getting answers or inspiration

Similar to a combination of Notion Web Clipper + Readwise

Every time you save, the AI automatically generates a summary, helping you grab the key points quickly and avoiding the problem of hoarding information → “bookmark and never read”

The system automatically organizes content and builds associations (a knowledge graph), letting you see the connections between different pieces of content

This is smarter than a simple bookmarks folder, close to the feel of Roam Research + AI

Collect multimodal content → transcribe & convert to text → embedding representation + summary generation → store in a vector database + knowledge graph → retrieval-augmented generation (RAG) → add a spaced repetition algorithm → finally forming an AI-driven second brain

https://www.getrecall.ai/?ref=producthunt

Indy AI

What it advocates is letting work find you

AI-driven recommendation is something I’m bullish on: in an age of information explosion, how to precisely find information or recommend the right information

It actually scans and analyzes your existing network on LinkedIn and X, and based on your skills, experience and preferences helps you automatically discover potential job or collaboration opportunities

Expanding your own network: rather than constantly hunting for opportunities yourself, it emphasizes “letting work find you”

Open the Chrome Web Store → install the Indy AI extension.

Log in / sign up at Contra.com.

Connect Indy AI with your LinkedIn/X accounts (authorizing data access).

The extension runs in the background, scanning and updating automatically.

In the Contra Pro panel or the extension UI you’ll see: a. Recommended opportunities (projects, positions, collaborations) b. Contacts related to your skills/experience c. One-click job applications or contact channels

https://chromewebstore.google.com/detail/indy-ai/hgpklbhmaaajglapkgebgjjlccndmcoo

Two cups of coffee

I’m thinking about today’s tech loneliness

Thinking about the atmosphere of offline cafés in Vietnam

And the current situation in China

Even people after losing a job, people without work, people who come to a city alone

Is there a way to focus on making an offline-friendship app?

Offline-first — for instance, you must meet for a cup of coffee

It would also boost cafés’ business

Humans are social animals; digital tools should serve that, not replace it

The currently popular way of building AI products

It’s actually building a prototype first to validate the effect, then seeing how to design the UI

Compared with the old PRD-driven, design-first approach, this emphasizes practice, getting hands-on, and feedback

And now everyone is a Builder

Usually a lot of people have the corresponding small sparks; rather than waiting for a product manager’s flash of inspiration, one or two engineers propose it and quickly build a prototype. This prototype doesn’t pursue perfect interaction or visuals but focuses on core function and product value

It’s like on a battlefield: a commando unit darting quickly through, striking straight at the core — clear objective, no lingering in battle

Once the prototype is done it’s immediately deployed for all of Anthropic to use in daily work, making the product’s creators also its most demanding users

Because they’re deeply immersed in the product scenario, they can give the most real and most critical feedback. Whether a feature has real vitality gets tested fastest in this internal ecosystem — only features proven internally to be effective and usable get formally productized

The premise of this step is that the team itself is the product’s target and deep user

Only after a feature is internally proven “effective” does UX polishing and optimization happen

This step may be the most “counterintuitive” and “uncomfortable” for most teams used to the mobile internet development paradigm (especially designers). Whether full consensus can be reached on this principle is the key to actual execution cost and final results

This is actually a return to the essence of design: design must be restrained, helping users focus on the core task instead of being distracted by a cluttered interface

Model is king: the ultimate goal of design is to use the thinnest, most transparent “packaging” so users feel the power of the Claude model itself most directly — this is crucial, and it coincides with the design principles of Agent engineering architecture

Product design stage: becoming a creative resonator with AI. When conceiving a new feature, she talks directly with Cloud Code, discussing “common use cases” and “edge cases,” even seeking preliminary design suggestions. Here AI becomes a tireless, deeply knowledgeable thinking partner

Solution evaluation stage: getting quantified grounds for decisions. Once the design mock is done, she can show the image directly to Cloud Code and ask roughly how much work it takes to implement. This gives designers a firmer engineering basis when communicating with engineers. (hahaha so sneaky…

The night before launch: getting hands-on with the code to defend the experience’s “last mile.” This is the most disruptive point. In the traditional process, low-priority (P2-level) visual flaws or experience tweaks often had to be reluctantly compromised before launch due to tight resources. Now designers can go into production code themselves and fix and optimize these details, personally polishing the product’s final quality

The systematic or engineering methods I learned before

May not be optimal in the AI era either

They’re just experience sedimented by predecessors

In the AI era, think more about whether there are new paradigms, or methods more conducive to collaboration in the AI era

Under the new mode of human-AI collaboration, what exactly is the truly efficient, innovative paradigm? It may mean that the long-winded processes of the past need simplifying, and modes that once relied on rigorous planning may be replaced by rapid iteration and emergent innovation enhanced by AI

DayDrift’s design technique

Focused on weekly planning, its distinctive daily folding interface lets you organize tasks by workday. Unlike traditional to-do apps, it includes a continuous tracking system that motivates you to build sustained daily habits

What isn’t finished naturally rolls over to the next day

How to build a good city

How to build a good culture?

How to build a good information field?

How to build a good app?

It depends on what people want

What will people want in the future?

How does human society interact with the world in creating information? In this interaction, what is the life cycle of information? What role can the internet play in that life cycle?

Linus Torvalds said: “A good system provides you with enough building blocks to do everything.” So we also need to think about how to design the building blocks of a new city in the best way.

From an engineering angle, we need to think about what technology can do today and what it can do in ten years, so we must study today’s global technology trends

From a business angle, as Peter Thiel says in Zero to One, we need to “dominate a niche market, then scale from there, moving toward an ambitious long-term vision.” You can’t build a city out of nothing and then tell people to come.

This is also why we need to understand the life cycle of information in the interaction between human society and the world. What we want to do is design a better life cycle, so we need to think about its different stages together.

Some people come to the city for the clean air, some for the safe environment — but in the end they realize what they get most from the city is information about ambition and the power that supports that ambition.

https://www.paulgraham.com/cities.html

Some new reflections on building products

Compound thinking

I want to create a simple app

That can record a person’s basic ideas

And a person’s smallest action unit

Then complete this smallest action

And record it

Lin Huiyin

Lin Huiyin the architect: reason and the beauty of building

She took part in refining the design of the national emblem of the People’s Republic of China. The architectural design of the Monument to the People’s Heroes, especially the relief decorations on its base, also embodies her effort

In the 1930s she and Liang Sicheng and others traveled through China’s remote corners, surveying and recording on site a large number of ancient buildings on the verge of disappearing, leaving valuable material for later generations

She helped found the Department of Architecture at Tsinghua University and taught there herself, training China’s first generation of modern architects

Poetry was Lin Huiyin’s emotional expression; she wrote a preface for Xu Zhimo’s poetry collection, with writing that is sincere and moving

Her poem “You Are the April of This World,” with its light, bright imagery, has become a classic of modern poetry

Under the immense pressure and limits of an era, what kind of creativity can one person unleash?

And the rise and fall of a country or a nation

Depends on the degree to which it liberates individual creativity

Some thoughts on Heptabase’s design

In one sentence: from the angle of the “human knowledge work life cycle,” we’re building a tool ecosystem to help knowledge workers integrate the knowledge life cycle of explore → collect → think → create → share

Two infrastructures:

The context layer: the whiteboard feature — not just visualization, but more importantly letting different cards trace each other’s lines of thought

The description layer: responsible for adding types and attributes to cards, corresponding to Heptabase’s tag and property features

People who trade time for time

Trading your own time for your own time — a dream-maker.

Trading your own time for your own time, not starving to death, even living pretty well.

Because human society has always had a kind of “adults raising children” support mechanism.

The world’s “work” can roughly be split into two kinds: one is “running things”, the other is “creating”. Running things is the difficult easy thing; creating is the easy difficult thing.

The goal of management is extremely simple: get everyone thinking the same thought and pulling in the same direction. But reaching that goal is extremely hard and takes a lot of theoretical knowledge and practical experience.

More than two hundred years ago the poet Schiller reminded people that play is an activity that brings out man’s “dual nature”, a nature that turns every “limit” in life into “expansion”. Only when man is man in the full sense does he play; only when man plays is he fully a man.

This dual nature is really the clash of the sensuous and the rational. In reality we all assume the two stand in opposition — we’re either driven by desire or bound by reason.

And play is exactly that magical domain: it satisfies sensory pleasure and follows some rule-ordered structure at the same time, bringing the sensuous and the rational into harmonious unity.

Within finite rules, we gain infinite possibility — a higher order of freedom.

Only in play are we at once sensuous, rational, creative and free — only then are we our complete selves.

So play is genuinely interesting. Playing at life isn’t defined by useful or useless; it’s precisely the dignity of our humanity.

The true spirit of play isn’t escaping reality; it’s experiencing the joy of creation within self-imposed limits, becoming a more complete person in the tension between rules and freedom.

But then, on the other hand

The designers of modern online games think more about addiction mechanics.

How to glue down users’ time and money.

Exploiting the weaknesses of human nature actually runs against Schiller’s ideal too.

So I go back to single-player games; I genuinely feel this game belongs to me, and that’s fun.

Cybernetics: everything can be a system

Cybernetics is the science of how systems self-regulate and self-control. Whether it’s a living being, a machine or a social organization, it tries to find the general laws behind them.

System: everything is a system; the body can be one too. Sometimes we don’t care how complicated the inside of the system is, only about input and output.

Feedback: the system’s eyes and ears. The system senses the results of its own actions and adjusts the next round.

Negative feedback: correction, keeping things near a target.

Positive feedback: likes, snowballs, amplifying and strengthening a process.

Control: the system adjusts and guides.

The world is made of countless systems nested in one another, regulating through feedback.

Whether life or machine, the underlying logic of their behavior is astonishingly similar.

Entropy is a system’s degree of disorder; it measures a system’s uncertainty and disorder.

Entropy and information are mortal enemies: high entropy = low information. A completely chaotic system, like TV static, contains enormous uncertainty (you don’t know which dot will light up next) but almost no useful information. Low entropy = high information: a piece of text, an image — their arrangement is determinate and ordered, so they can carry a lot of information.

The aim of cybernetics is a control system, which is essentially a creator of “negative entropy”. By introducing information and control, it resists a system’s spontaneous tendency to descend into chaos.

In an isolated system, entropy always increases. That is, everything spontaneously moves from order toward disorder.

Life is the greatest “negative entropy” phenomenon. Organisms keep drawing energy and information from outside (eating, learning) to maintain their highly ordered internal structure and fight the flood of rising entropy.

How to validate a product

You have a fairly clear sense of the product positioning and the user story.

But when you first validate, what you must validate is the uncertain part and the simplest part of the most core function.

That’s the validation logic of an MVP.

A very important distinction with heptabase

A very important distinction between heptabase and tools like notion is that its positioning is a learning app, not a note-taking app.

The goal is to help users learn better, not to help users take more notes or build a productivity system.

It encourages users to break information apart, restate it in their own words, and then link it visually — that process itself is deep learning.

Good design always immerses users in the process of creating, triggering a flow response.

The process itself lets you focus on learning and thinking, and forget the tool itself.

What the Believer model teaches

It deeply links the user and the product through an emotional and value clasp.

The highest tier of subscription; hope also has a corresponding co-creator tier.

A kind of identity identification and emotional investment.

Existing itself is precisely identifying the core user group, so you can keep getting closer feedback.

And there’s a stable cash flow.

There’s a way to build a strong brand moat.

The product side wins users’ trust through transparent communication and excellent execution; users, through putting in money and feedback, become partners in the product’s growth.

Not a slow-to-warm-up person

An extremely immersive explorer.

Very easily absorbed into a new experience — whether it’s living in a new place or a game, my system will call up enormous resources to explore the thing.

Then there’s my extremely strong metacognition and the rational-driving side: this administrator is always running silently in the background, monitoring the operating state of the whole system, holding the highest privileges, and able to interrupt and quit any process at any time.

Someone with extremely strong execution in the control and feedback system: sensing something, acting on it.

Experience with the greatest immersion, and manage your own system in the most rational way.

https://nsddd.top/zh/posts

https://nsddd.top/zh/posts/2024-annual-review/

Scenario design / designing input = scenario; inside a small scenario, find one concrete entry point.

input, output, scenario = context + historical user habits, a sedimented system, unlimited.

Mind map?

New user story.

Experience closed loop.

flomo vs heptabase: the human memory mechanism

heptabase is actually positioned as a tool for thinking and structuring knowledge, very good for deep understanding.

notion is harder to pick up, overloaded with features, and over-free: the blank canvas leaves beginners not knowing where to start, and there are so many concepts and features that users have to design their own information architecture, which demands a lot of investment long-term.

Most hover displays really are over-design

Most hover displays really are over-design.

Hands-on with same.new

You can actually input cards too.

You can manually copy sub-pages; generally that’s more maintainable, and you can generate multiple sub-pages.

And it supports deployment too.

Innovation is finding something new in the familiar.

Mem 2.0 is a more disruptive product

A fairly good thinking partner.

Building Mem from scratch, so entrepreneurs, executives and creative people can easily capture and use their best ideas.

Offline-first, blazingly fast, built specifically for handling your most important thinking.

Now it seems they don’t just want to make a note-taking app, but to make an AI thinking partner that still holds the central position in your notes — a system that helps you turn ideas into results.

Two handy ways:

a browser extension that turns any web page into a beautifully formatted note; no organizing needed, Mem handles the rest.

email integration: just forward it.

Design philosophy: Mem’s design philosophy is about remembering notes and finding them in the future, focusing on recall (personal gripe: xxx).

Researching and introducing tana

tana instantly turns notes into other things — tasks, projects, web pages, strategy documents, OKPs and all that mess.

Anyway, it’s a whole set of features built around that.

Transcribing lecture content, taking notes more intelligently, staying organized. With AI analyzing the lecture content, extracting the key points, creating study notes, and generating exam-prep questions.

Deep research into the second ai project

With the development of large models and agent systems, people are paying more and more attention to the idea that AI is not just a one-off Q&A tool, but a companion agent with long-term memory that continuously understands the user and keeps context consistent across multiple interaction scenarios.

So once the concept of AI-native memory appeared, it was about letting the system, like a human, have memory and a self, able to remember the past, know who I am, what I’ve done, what I prefer.

“second” also means second, so what’s explored is the process of a second me, a digitally extended self — representing the user and assisting the user.

For future intelligent agents that are more autonomous, continuous, with long-term goals and memory.

A few goals:

Redefine memory management: not just storing user information, preferences and history as independent data, but making memory the basis for the system’s reasoning, responses and generated behavior.

A three-layer memory architecture (L0, L1, L2):

L0 (Raw Data Layer): the raw data layer, unstructured input such as documents, conversations, logs.

L1 (Natural Language Memory Layer): the natural language layer, abstracting part of the memory into readable form, such as a user resume, preference tags, important sentences.

L2 (AI-Native Memory Layer): the AI-native memory layer, embedding the user’s specific knowledge, preferences and history into the model parameters, achieving persistent, structured, reason-able memory.

Why do we need a three-layer architecture? Essentially one is storing the raw information — technically a document database; another is some storage abstracted during LLM conversation; and finally L2 is fine-tuning on the L1 data, where it’s no longer a database or text, but memory written into model parameters. For example, using the L1 data to fine-tune the model, with preference optimization (DPO, RLHF) letting the model internalize “the answer style and values the user likes”, producing a personalized model.

One very core point is how to handle the model.

To make the model natively have memory, how do you fine-tune.

Two routes:

External memory: ChatGPT API + RAG + vector store; don’t change the model, use the off-the-shelf API.

Native-memory type (AI-Native Memory), Second Me, MemGPT, Personal-LM: fine-tune model parameters, writing memory into the weights.

The L2 layer (AI-Native Memory Layer) is at the model level:

through SFT (supervised fine-tuning) and DPO (direct preference optimization),

it “writes the user’s preferences, tone, knowledge and style into the model parameters”.

The model isn’t trained from nothing; it synthesizes personalized training samples from raw interactions (L0) + semantic summaries (L1).

Through SFT (supervised fine-tuning) and DPO (direct preference optimization),

it “writes the user’s preferences, tone, knowledge and style into the model parameters”.

And this way it can be carried on the user’s local device.

Warp’s Project rules &

Warp’s Project rules & slash commands capability.

Mainly for agent operations, like adding a new reusable prompt template,

adding an MCP service, adding routes, indexing the codebase, initializing a warp project and so on.

Warp Code = AI + code understanding + project rules + MCP extensions + custom Prompt system.

Finding a few indicators

Music, sound.

Space in between — needless to say, it’s too important: negative space and a sense of breathing, appropriate emptiness.

Verticality, a sense of three-dimensionality, accounts for a large share of the psychological effect.

Material: concrete, glass, wood give a comfortable feeling.

Light, this is actually very important and also a very hard factor to design for; light must “have rhythm”, flowing and changing with time, rather than lighting evenly.

Nature — where there’s light there’s of course nature: plants, or some natural things.

The tag design in heptabase

A tag looks like a label, but it’s actually the bridge between atomic cards and the semantic organizational structure.

A card is the smallest unit of thought.

A tag is a thematic gravitational field.

A Tag is a many-to-many semantic connection layer.

One card can have multiple tags.

One tag can contain cards from different boards.

The small product Luo Yonghao is building

J1 Assistant is a personal AI assistant app built in the Matter format.

It’s still in beta.

It aims to be a smarter, more efficient, more elegant AI assistant that understands the user and is good at managing complex tasks,

including scenarios like task management, scheduling and reminders.

It’s a self-developed large language model (LLM).

It emphasizes seamless integration of search and task execution, helping users save time (like saving half an hour a day).

It includes support for voice input and interaction, a Ripple Touch mechanism, integration of tasks and search, multi-platform search capability, and scheduling and reminder capability.

The openness of the language system

Human language isn’t a closed set of symbols

It’s a symbol system that can keep iterating, keep creating new expressions, generating new meaning

Finite rules, infinite output

Language is also recursive, and this recursiveness gives language its generativity and openness

The openness of linguistic meaning reflects the openness of human society and cognition

Life is also an infinite-flow game, in some ways more like a word game

Language is also an infinite-flow game

We go and make spatial choices at various levels

Typesense’s autocomplete capability

A capability I quite like; the usual scenarios

Search-as-you-type: users expect the results to refresh immediately with every letter typed into the search box

It has powerful built-in spell correction, out of the box. You need no configuration at all; it automatically handles typos and returns the most relevant results

It means the whole search results page (not just the dropdown under the search box) refreshes in real time with every key you press

I still don’t get Apple’s Liquid Glass thing

It’s a whole new visual interface design language they rolled out; Apple renders the buttons, sliders, controls, icon backgrounds, toolbars and so on in the interface as a “material”, and this material is called Liquid Glass

Transparent/semi-transparent, reflecting/refracting, and even changing with the environment (light/dark/light mode)

Fluidity: interface elements have gentle animation, layering, and morphing during swipes

One thing that’s actually important is that Apple also wants all kinds of other devices to be more unified in visual language, so Liquid Glass is a cross-platform design language

Externally the wrapper is actually very simple

Older versions only needed .fill(.ultraThinMaterial), now it’s generally fill(Material.liquid())

Material.liquid() responds dynamically to changes in the background content rather than being a fixed blur

This is a content-first design principle; the interface can disappear

Some vibe coding methods: how to write prompts

Storyboard Prompts: actually building the prompt like a storyboard - first outline the overall architecture, then break it into small tasks, describe the app flow first, then generate code module by module

Constraint Cues: explicit constraints in the prompt - language version, tech stack or performance requirements; codex won’t over-implement, so it won’t over-engineer

Smart Iteration: generating through multiple iterations of prompting - first a rough version, then prompting to fix bugs or add features, progressive development

Context-First: providing background at the start of the prompt, which lets the AI understand vibe coding better

BeReal’s original explosive popularity

Once very popular among Gen Z campus groups

It advocated a more casual, more authentic social experience, in contrast to traditional filter + retouch social platforms

A Gen Z need for authenticity and non-deliberateness

Its core positioning is: “show your real life”, rather than the retouched “highlight moments”. For example its App Store description says: “Every day at a random time… no filters — just real life, as it happens.”

You can’t see a friend’s post for the day until you upload your own. In other words, “yourself first, others second”

Young people’s longing for authenticity and interiority

How to have a lower expression burden

How to have a good feedback mechanism; since it’s light-dopamine feedback, don’t be judged, don’t over-expose

Is the social group precise

The shifting trend of Hong Kong’s port

Hong Kong was once the world’s largest container port by throughput

But now it’s basically been replaced by Shenzhen and Ningbo-Zhoushan

Ningbo-Zhoushan has in the past two years also become the world’s new largest port

In addition, Southeast Asian ports like Singapore and Malaysia’s Port Klang are rising, further intensifying competitive pressure on Hong Kong’s port

This is also why CK Hutchison sold its majority stake in the Panama ports company to a US-led consortium

CK Hutchison operates ports in many places worldwide; port construction and operation require heavy fixed-asset investment, including wharves, yards, automated equipment and channel maintenance

And the Panama terminals are overseas investments located far from core markets (Asia, Europe)

By comparison, terminals in the Yangtze River Delta or the Guangdong-Hong Kong-Macao Greater Bay Area have more of a trend behind them

Especially with the formation of the Greater Bay Area: the Yangtze Delta port cluster and the Pearl River Delta port cluster complement each other’s strengths and share resources, raising global competitiveness

Considering development, my ideal city ranking is Hangzhou / Shenzhen

Considering development, my ideal city ranking is Hangzhou / Shenzhen; Chengdu is suited for when you get old and move there haha

Considering retirement (still too early, at least five years to think about it), Chengdu, Changsha and Hangzhou

Japan is also a major exporting country

Currency depreciation favors exports

And to a large degree favors rising stock prices

Yen weakens, Nikkei rises

Yen strengthens, Nikkei falls

But the core still depends on how vigorous the market is

Investors still like borrowing in cheap yen to buy global assets

When global risk appetite rises:

→ investors sell yen and buy risk assets (Japanese stocks, US stocks, etc.)

→ yen depreciation + rising stocks happen at the same time.

When the market seeks safety:

→ investors cover yen shorts (buy yen back)

→ yen appreciation + falling stocks

This is also why the yen is often a safe-haven currency.

The weaker the yen → the more dazzling corporate profits → the stronger the stock market.

But it usually can’t go too low, because too low isn’t a balanced market either; the stock market may instead become volatile due to policy uncertainty

Hiking easily resets your sleep schedule

Because of the sun, melatonin secretion is triggered

Exposed to natural light every day, the body is recalibrated by the “sun”

Long hours of walking, climbing and carrying a pack during the day exhaust the body.

This deep physical tiredness automatically drives you to sleep early and wake early

Unlike the city’s “mentally tired but physically still,” that kind of fatigue actually keeps you awake

The psychological rhythm becomes simple, and at night you naturally feel “time to rest”

In the end, if treated fairly

There’s naturally a sense of belonging

The sense of belonging of any people or country isn’t naturally generated; it’s a psychological balance manufactured by the state through institutions, narrative and resource allocation

When this balance tips (economic gaps, discrimination, cultural suppression), people feel “I’m not part of the country”

This is why the Inca (and Indigenous peoples worldwide) often waver between national identity and cultural selfhood

In Peru, Bolivia and Ecuador — countries that were once the center of the Inca Empire — the population structure is roughly:

Hispanic (descendants of whites): the upper stratum

Mestizo: the middle class

Indigenous people (mostly Inca descendants): concentrated at the bottom and in rural areas

Even with legal equality, Inca descendants remain “second-class citizens” psychologically and socially

📌 Core philosophy

The author, Jeremy Maluf, has since 2015 owned only as much as fits in one backpack — going on 10 years

Core advantages:

Improved focus, reduced spending

Can travel anytime, simplified life

No checked baggage needed — board and go

🎒 Gear system

Backpack

Aer Slim Pack (9L) — meets the strictest budget-airline “personal item” size

Built-in AirTag for loss prevention, with an S-biner lock for anti-theft

Digital devices

MacBook Air M2 13-inch

iPhone 15 Pro (256GB, no case)

Apple Watch S7 + AirPods Pro 2

Two 256GB USB-C flash drives (one on him, one as backup)

Charging setup

Anker 47W dual USB-C charger

Universal travel adapter

20,000mAh portable battery (can charge a MacBook once)

👕 Clothing system (extremely minimal)

Everyday outfit:

6 American Apparel 50/50 blend T-shirts

6 pairs of Darn Tough merino wool socks (15-year brand, lifetime warranty)

6 Uniqlo AIRism underwear

Western Rise travel pants + Grip6 belt

Nike Pegasus Trail 5 running shoes

Baseball cap

Packing technique: roll T-shirts around socks and underwear into a “daily roll” — wrinkle-resistant and easy to organize

Winter outfit:

Montbell Plasma 1000 down jacket (lighter than a phone, packs into a pocket)

Canvas jacket (can be layered for extreme cold)

Hoodie + Arc’teryx gloves

Merino Buff (multi-purpose headwear)

💡 Practical small items

RE:FORM ultra-thin wallet (with a magnetic AirTag pocket)

True Utility Keytool (the world’s smallest bottle opener)

A.Brolly umbrella (weight < 90g)

Matador quick-dry towel (fist-sized, doubles as a beach towel)

Moleskine blank notebook + two pens

⛺ Outdoor gear (for hiking)

Weight just 2kg, minimal volume

3F UL Gear ultralight tent ($70)

Thermarest UberLite sleeping pad

Nitecore NU25 headlamp (USB-C charging)

Toaks titanium cup set + BRS stove

Sawyer Mini water filter (essential for survival)

🎯 Key insights

$100 gear ≈ $10,000 gear — no need to chase brands excessively

Durability > material — the author doesn’t advocate expensive merino wool because it damages easily outdoors

Long-term living strategy — he’s now switched to long stays in cities, with occasional “zero-luggage” trips (carrying only pocket items)

Purchase list system — put things you want on a list and observe them for a while before deciding

💳 Financial tools

Amazon Prime credit card — 5% cash back, no annual fee.

Revolut debit card — free international ATM withdrawals.

The biggest value of this list: it proves a modern person can fully meet all life needs with a <10L backpack. The key is precisely choosing multi-functional, durable, lightweight items. The author updates the list every year, and millions have read it for reference

An introduction to Shaoxing

Shaoxing is known as the “hometown of famous scholars”.

The source of Wei-Jin style: during the Eastern Jin, the aristocratic clans represented by Wang Xizhi and Xie An moved to Kuaiji (today’s Shaoxing), held the elegant gathering of floating wine cups along a winding stream at the Orchid Pavilion, and the “Preface to the Orchid Pavilion Collection” was born there.

Devotion to country and a critical spirit: Lu You of the Southern Song longed his whole life to recover the lost territories, and his poetry is full of tragic heroism and bold feeling.

If “famous scholars” are Shaoxing’s “face”, then the “shiyé” are Shaoxing’s “lining”.

In the Ming and Qing dynasties, Shaoxing had too many people and too little land, and a great many educated men with no room in the exceedingly narrow imperial-examination path; huge numbers of intellectuals took the route of the “clerk”, becoming the advisors of officials at every level — the “shiyé”.

In the Qing dynasty, Shaoxing shiyé were found all over the country.

They were expert in penal law (justice) and money and grain (finance), and through their legal knowledge, administrative ability and insight into human affairs they shaped the empire’s grassroots governance to a large degree.

As a city it is called the Venice of the East, a water city.

Unlike many “ancient towns” that are all surface, the water system of the old city of Shaoxing is still working today. It’s an organic living thing made of rivers, bridges, streets and taimen (the deep courtyard mansions peculiar to Shaoxing).

Shaoxing’s urban layout embodies the wisdom of high-density settlement.

The brewing of Shaoxing yellow rice wine is an ultimate understanding of water, rice and time. It is not just a drink but a cultural sign, tightly bound to sacrifice, festivals and the rites of life. Its complex brewing process and mellow taste are a portrait of the Shaoxing character: patient, meticulous, devoted to tradition.

Centered on the China Textile City in Keqiao, Shaoxing has become a world-class textile distribution hub.

Understanding Shaoxing yellow rice wine

The brewing of Shaoxing yellow rice wine is an ultimate understanding of water, rice and time.

Water is the source, the objectively existing law; water is the blood of the wine, the source of all things and the starting point of life.

Only good water brews good wine, just as a person’s talent, foundation, or the environment they were born into.

The highest good is like water; water benefits all things and does not contend.

Time is even more magical: time is the bone of the wine, the most mysterious and most crucial catalyst in the brewing process, a respect for time and law, following nature, governing by non-action.

Rice is the flesh of the wine, the material base and energy source of the brewing process. Shaoxing yellow rice wine uses top-quality glutinous rice. Rice represents solid, visible, shapeable real matter.

The steadiness of a small town

It feels like there were a lot more interesting things.

For example, the owner of the homestay I booked was very hard to reach.

I even felt there were a lot of sloppy steps in the process.

Neither the room nor the platform told me the wifi, and asking got no reply; only much later when I asked on the platform did I get in touch …

Couldn’t reach them after booking, and didn’t connect until three in the afternoon the next day.

Went to a cafe; Amap’s location was way off.

Went to another cafe; the opening hours shown on Amap weren’t in effect, I thought it had closed down, and was about to take a cab to the next one, then the owner showed up for work, so I cancelled the ride …

On checkout day I forgot to pick up a package.

The owner found it

and personally delivered it to the front desk of my new hotel.

Maybe it also takes putting yourself in their shoes — warmth, professionalism and attentiveness.

How important is verticality in architectural spatial aesthetics

I’ve realized — the presence of space in between gives a completely different felt experience.

A church’s dome → connects the human heart with the sky.

A Japanese dry landscape garden → creates layers of breathing through the height of the stones.

External space is the projection of a mental structure.

Light + verticality + negative space = spiritual space.

Verticality naturally makes people settle down.

You suddenly “go quiet” — that’s actually a kind of inner-outer alignment.

When Indians talk about China’s robots

When a robot enters an Indian hotel,

it’s very hard for it to come back out.

Jasmine, light roast beans — generally Yirgacheffe

Jasmine, light roast beans — generally Yirgacheffe, floral Geisha from Panama, Nicaragua floral lots, etc.

The aroma tends toward jasmine, white flowers, citrus and honey.

The acidity is generally bright, tending toward citric or orange-blossom acidity.

The mouthfeel tends to be light, tea-like and transparent.

Compared with dark roast, there’s no caramelized bitter-sweet balance.

Because I drank it as a hot latte, I felt the acidity got softened to some degree.

Light roast relatively preserves more acids, sugars and ester aromatics during extraction.

It allows more freedom of expression.

Extraction, generally speaking, is the process during brewing where soluble substances in the ground powder dissolve into the water. Two metrics:

TDS — the mass percentage of soluble solids in the finished liquid.

Extraction yield — the total mass of solids dissolved into the drink divided by the original mass of coffee grounds.

As for 冰博克

冰博克 is generally good for making dirty.

冰博克 is usually obtained by freezing it in the fridge the night before, then letting it melt the next day.

The milk suitable for latte art generally works like this: essentially latte art is whipping air evenly into milk to form a fine, stable foam, then combining it with coffee.

Protein content needs to be high, because when heated, casein denatures and forms a network structure.

Too low a protein content (like some skimmed milk) makes the foam loose and quick to collapse; too high a content (like evaporated milk, 冰博克) makes the foam too thick, hard to flow, and actually hard to pour a clear pattern with.

Generally the foam doesn’t easily collapse, because if the protein content is low it’s loose, and if high it’s too thick and doesn’t flow easily.

Fat adds “creaminess” and “gloss” to the foam; if the fat content is too low (skimmed milk); and if there’s a lot of fat, it also flows poorly and easily scalds.

The reform of the “three assets”

After “land finance” dried up, a necessary means to “keep local government finances on life support”. It tries to revitalize dormant state-owned assets, turning them from “dead money” into “live money”, to cope with imminent debt pressure and the spending needs of “guaranteeing livelihoods, guaranteeing salaries, guaranteeing operations”

If successful, it will build a virtuous cycle of “effective assets → effective debt → effective investment → new effective assets”

“Everything can be securitized” may lead to inflated asset valuations and create a new financial bubble. If the future returns on these assets (like highways, data centers) fall short of expectations, it will trigger a new round of debt defaults

If mishandled, in the end it may just move debt from the left hand to the right hand, further overdrafting government credibility

The essential cause is the complete bankruptcy of the fiscal model; the old development model centered on “land transfer fees” has collapsed, and local governments are stuck in a structural fiscal crisis

In the short term, the revitalized funds can steady local finances so they don’t cut off supplies

But in 3-10 years some negative effects may appear: a systemic rise in the cost of living, rising fees for some infrastructure

Rising costs for enterprises to use data will in the end definitely be passed on to consumers; for ordinary people the long-term price may be paying for today’s “government fiscal relief” with “higher prices and less convenient public services”

But it’s also possible it will succeed in the future; if it does, the money from selling assets and raising prices gets efficiently invested in new productive sectors, and the returns from those new sectors (like the data industry, high-end manufacturing) start feeding back into the finances

But… the actors (local governments), under huge short-term debt repayment pressure, will involuntarily choose the “zero-sum” shortcuts (raising prices, dumping assets) and ignore the difficult “positive-sum” path (innovation, efficiency gains)

Taking it slow and experiencing the local lantern festival

Looking back, it was also a year ago in Thailand

The lantern festival I saw

The folk lantern festival has a lot of atmosphere

Villages, small towns, temples and other places also hold rituals of releasing sky lanterns or floating water lamps.

Thousands of drifting sky lanterns in the night sky

Releasing a sky lantern symbolizes “letting go of bad luck, welcoming good luck / a new beginning”

Maybe only once or a few times in a lifetime; it’s a visual experience

Releasing lanterns together with others in the dark, making wishes, sharing the same moment

Suddenly remembered the lantern festival and the cremation temple

Both of these gave me a huge emotional and memory impact

A year later I still feel like I’m there

Thousands of sky lanterns rising, the night sky lit up with warm orange light. This is an experience of “collective beauty”: soft, flowing, romantic

Music, prayers, laughter, night wind, firelight interwoven - a “ceremony of light”, symbolizing people letting go of worries and welcoming a new beginning

The cremation temple is more like a quiet, direct facing of death

Flames, ashes, the river, weeping, the sound of monks chanting. The air carries the smell of a life ending

Heavy, quiet, real. It makes you face the most fundamental question of humanity: the brevity and the cycle of life

Trying some coffee lately, I suddenly felt coffee’s philosophy is time and space

A very important point in distinguishing people’s relationship with coffee is digestion/absorption versus sensation

The whole journey of coffee from seed to cup is essentially the product of a specific space at a specific time

How it’s constructed, what the spatial growing environment of the place is like - the soil on the high-altitude slopes of Ethiopia, the volcanic ash of Colombia, the humid climate of the Vietnamese highlands…, what kind of flavor space it has - and the environment where you taste it matters enormously; it completely changes your perception of the same cup of coffee

Time naturally accumulates; the processing method of coffee beans (natural, washed, honey) is essentially a delicate control of “time” - the roasting time, the extraction time and the drinking time - like a time capsule, finally presented in front of you

Digestion/Absorption - a physical, biological relationship, a “me-first” relationship. You don’t care about the coffee’s “time and space”, you only care about what it can bring your body - energy, alertness, wakefulness. Coffee here is “fuel” or a “tool”. It’s a one-way “taking”

Sensation/Feeling - an experiential, spiritual relationship, a full mobilization of the senses. Your smell catches the floral, fruity notes; your taste distinguishes sour, sweet, bitter, salty; your touch (the mouth) feels its body and texture

Pinterest

I’ve always been puzzled about what exactly is so appealing about Pinterest

Especially in trend reports I saw before

More and more young people (including Gen Z) are joining; for example, one report mentioned that Gen Z is the fastest-growing user group

By gender, the male-to-female ratio is 3:7

By age distribution, the largest age bracket is 18-34. For example, the 25-34 group has a relatively high share. Another report said women aged 18-24 and 25-34 each account for about 19-20%

The US is the biggest market; Pinterest is essentially a “visual inspiration + saving + discovery” platform: users often use it to browse inspiration boards for home decor, fashion, beauty, travel, DIY, recipes and so on

Pinterest’s core value is letting users discover the inspiration of “I want to do this / I want to buy that / I want to go there”

Pinterest’s model is also very worth learning: it lets users “discover first -> save -> act”

One thing it emphasizes strongly is “inspiration discovery + saving + planning + action”

In the US, high-income households make up a relatively large share of Pinterest users; for example, users from “40% of US households with annual income over $150K” are on Pinterest

Shanghai’s sea

The Huangpu River has always been crucial to Shanghai; it is Shanghai’s mother river

Shanghai’s urban formation, port development and commercial center construction all depended on the Huangpu River

Since ancient times the Huangpu has been a main artery of inland shipping, connecting the inland Yangtze with the East China Sea, an unavoidable route for early goods and passenger transport

It was also a center of foreign trade, helping Shanghai become an international metropolitan area

Along both banks formed financial and cultural core districts like Lujiazui, the Bund and the Xuhui waterfront

By comparison, Shanghai’s sea: the Port of Shanghai (including Waigaoqiao and the Yangshan deep-water port) is the world’s largest container port, crucial to global trade

The foreign trade, import-export business and free trade zone economic activity brought by maritime shipping are the main drivers of Shanghai’s modern economic scale

Looking at the modern economy and future strategy, the sea matters more, because Shanghai’s internationalization and global trade depend heavily on the East China Sea and its deep-water ports

Thoughts on China’s stock market

Short-term speculators lean toward a zero-sum game

In short-term price swings the money you make comes from what others lose. Total market wealth doesn’t increase, it just transfers — speculating on themes, chasing hot spots, harvesting retail investors — a classic “chip-transfer game”

Over the long term it’s more about investing, which is a non-zero-sum game

If the investment target is a business that truly creates value (profit growth, rising dividends, successful innovation), then society’s whole cake gets bigger and shareholder returns come from new value the company creates, not from others’ losses

But what the stock market presents is human greed and fear themselves

When greed wins → the crowd chases highs, price detaches from value

When fear spreads → the crowd sells off, price falls far below value

The minority who really make money are anti-human-nature: buying when others fear, selling when others are greedy

The market is always pricing human nature

Benjamin Franklin

A practitioner, America’s Renaissance man

Deeply involved in America’s founding, he’s the only Founding Father to have signed the country’s three most important documents (the Declaration of Independence, the 1783 Treaty of Paris, the US Constitution)

What draws me most to Franklin is his curiosity about the world and his drive as an explorer who reshapes the world with his own hands

Through a hands-on approach, experiments, and practice

A model of knowledge coming from practice

He also wrote two books: one is Poor Richard’s Almanack, which influenced many Americans

The other is The Autobiography of Benjamin Franklin, telling the story of how he grew from a poor boy through self-striving and moral cultivation, becoming an early model of the American Dream

Compounding in The Almanack of Naval Ravikant

Compounding is essentially causality too

Underneath it is a series of micro-decisions in a value system

It’s actually a person, within their rich information network, making fast intuitive decisions to select and eliminate

Filter out what interests you, then dig deep into the underlying principles

All the returns in life — wealth, relationships, knowledge — come from compounding

One causal chain in life is: idea → resolve → action → habit → result

High probability: if the long-term direction is right, the win rate is high

Small actions: reduce inner resistance, act immediately

Don’t break the chain: you can rest occasionally, but never give up

A rather interesting line about tool-driven approaches: hand the decisions to people with higher cognition than you, and let them be your tools hahaha

Lei Jun’s current reputation is the exact opposite of before

Narrative and product, service experience mismatched, trust overdrawn

What sincere marketing fears most is the facts being exposed

Public relations crisis response — and instead it’s not transparent

The macro social environment is declining, anti-chicken-soup, refusing to let anyone stand out

Quantified language is itself a tool

Purely quantified language is also a formal language.

Math, logic, programming — these are all formal languages.

The question is what problem you use it on.

Goal-directed mechanism.

Everything in the world can be abstracted into a goal-directed mechanism.

Cybernetics is the core paradigm of modern technology.

Quantify -> goal -> sensing/input -> compare the gap -> action/output maybe -> compare the gap.

Compare the gap -> action/output maybe -> compare the gap, a looping mechanism.

This can explain a lot of mechanisms: air conditioners, computers, artificial intelligence.

Quantifying meaning.

When we talk about meaning, we also need to relate a word or concept to the words around it for meaning to arise.

We can’t explain a word with that word itself; we can only explain it with something that isn’t that word.

This is also the negative theology in theology: what cannot be described is God, God is not this, God is not that.

Approaching the essence of a thing through negation; the core of meaning is a void.

So how does the word “meaning” flow and get described — there has to be a word that cannot be described, we can’t say meaning, and then meaning naturally starts flowing.

So the operating mechanism of human language is the primacy of negation.

So because of positivity, you get homogenized competition, you get scarcity, and it produces a sense of lack in you.

The Hertz heuristic framework — time, space, action

Time (When)

Not just clock time, but timing, rhythm, sequence.

In economic activity: investment timing, product life cycles, market cycles.

On the philosophical level: Kairos (the opportune moment) vs Chronos (clock time).

Space (Where)

Not just physical location, but field, structure, relational networks.

In economic activity: market positioning, supply-chain layout, information asymmetry.

On the philosophical level: field theory (Bourdieu), topological thinking.

Action (How)

Strategies, paths and methods for solving problems.

In economic activity: business models, technology roadmaps, organizational processes.

On the philosophical level: practical wisdom (phronesis), means-end rationality.

These three can cover every dimension.

Time: the dimension of dynamic evolution.

Space: the dimension of static structure.

Action: the dimension of agentive transformation.

Any economic phenomenon can be decomposed into: at what time (time), in what position (space), in what manner (action).

It is itself part of systemic thinking.

Time = system evolution.

Space = system structure.

Action = system intervention.

The emptiness of not knowing what you want to do

I always feel it’s because of the absence, the vacancy of long-term goals; life or career has no direction

From time to time there’s also a lot of anxiety and pressure, crushed by study, work or family so you can’t think clearly

It could also be choice burden, cognitive overload caused by too many choices

Or maybe it’s a lack of self-knowledge

Understanding Pang Dong Lai

Actually it comes down to two core questions

Why are the salaries so high

Why does a small city have the ability to make money

Pang Dong Lai isn’t an ordinary low-price retail supermarket; it takes the “service + experience + quality” route. For example, reports point out that although prices aren’t necessarily low, consumers are willing to pay a premium for “peace of mind”, “quality” and “service”

And employees get a lot of vacation, especially middle managers whose annual leave approaches 180 days

The higher salary is also for high investment, high return

Making high-quality operations in a cost environment, marginal profit provides space

Pang Dong Lai is located in third- and fourth-tier cities within Henan province, and has built up brand reputation, customer loyalty and community relationships there

In a small city there are few competitors; an excellent brand with differentiated service can get a relatively big share

Pang Dong Lai’s stores “become tourist attractions” - meaning customers from out of town will drive there specially to shop

Although the scale is small, efficiency is very high, and on top of that land, labor and rent costs in third- and fourth-tier cities are lower than first-tier cities, so there’s a very low base

Pang Dong Lai’s customers have strong spending power

In 2024, Pang Dong Lai’s supermarkets in the Xuchang area had annual sales of 5.945 billion yuan

This approaches Sam’s Club’s 2024 annual revenue in Wuhan

Including Xinxiang, 2024 revenue was also over two hundred thousand

This almost exceeds the annual revenue of a single Sam’s Club in Wuhan

Looking at consumption characteristics, Pang Dong Lai doesn’t rely only on tourists’ spending; more importantly locals are willing to pay

Several taxi drivers shared that people around them also go to Pang Dong Lai often, especially young people

Sugar supplementation during exercise

High sugar intake is linked to knee health deterioration in multiple ways

Especially for athletic performance — but excessive intake sets off a series of physiological processes that can cause knee pain, joint damage and aggravate conditions like osteoarthritis

So for everyday simple exercise, focus on proper hydration and a balanced diet, and try not to supplement sugar extra

For sustained high-intensity exercise, supplement carbohydrates appropriately

The evolution of short video

It’s more like society ✖️ psychology ✖️ technology ✖️ economics.

Mobile internet penetration → everyone has a camera and a playback device;

algorithmic recommendation → precise feeding of personalized content;

compression coding and bandwidth optimization → high-definition video played with low latency;

AI editing and generation → lowering the barrier to creation;

interaction design and interface optimization → seamless swiping, instant feedback.

Technology is irreversible, but essentially it can be redefined.

For example, you can give up certain technologies, or redesign the ethical boundaries of certain technologies.

The existence of technology is irreversible, but the rule of technology is reversible.

If a film or TV series takes a very long time to reach the basic three-step cycle

If a film or TV series takes a very long time to reach the basic three-step cycle — setup, confrontation, resolution —

that’s a very strong feedback, very easily an addictive loop, and the rhythm is extremely fast, giving the brain no chance to settle. It lowers the barrier to entry: a “psychological algorithm” exploitatively optimizing human nature.

The current monetization mechanism of short video and its ecological role

The current monetization mechanism of short video and its ecological role: the basic mechanisms are gradually becoming standardized.

At the same time, social and cultural standards haven’t matured.

For example, today’s short videos lack depth, no aesthetic or cultural sediment has formed, and the content is over-algorithmized.

Data is verification and reasoning, and also copying

A very good way to understand this world, to vividly reason out some of the problems in the world today,

and then fill in the holes in that part.

When something can be seen with the naked eye as capable of completing a commercial loop, there’s no need to validate something that might complete a commercial loop.

Today’s screenwriters have no room to play. They’re controlled by the platform or the director, with no ability to negotiate, and rarely insist on artistic creation.

The reason is the emergence of big IP in 2015: only big IPs could control every link, or internet capital took over the whole process, turning it into links and an assembly line.

A kind of data thinking; it loses the joy of exploration, and it’s also very hard to surpass truly disruptive innovation.

But all film and TV works are played out through a process of exploration and negotiation, not an accounting mindset — negotiating out a path of their own.

Artistic achievement and commercial achievement are both like this. Negotiation is an inevitable process; a good product and model come out of negotiating.

The decibel level inside a cafe

Generally the most recommended range is between 55-65 dB.

This range is considered the golden zone for combining creativity and focus.

Too little is generally too quiet, anything below 50 dB.

55-65 dB is light noise, which easily stimulates divergent thinking and promotes creativity.

Handan’s food is so cheap

Though it’s cheap,

…

the portions are also huge.

I feel embarrassed eating it.

AI keeps getting more advanced

How exactly will photography change?

In the past the value of photography was recording reality

AI can generate realistic images and video without any real shooting

The line between real and fictional is blurring

Shooting “the real world” is no longer the only way to obtain an image

Photography may no longer be mainly about generation, but may turn into an important tool for narrative, for ways of perceiving, and for personal expression

In the past the biggest meaning of photography was “what I shot is real”

Authenticity is a scarce commodity in photography

But when photography combines with AI, how do you build your own personal perspective and expressive perception, witnessing and trust, proof of authenticity

The relationship between people and time

The relationship between people and the world

The relationship between people and themselves

Developed countries have cleaned their surface via institutions, capital and globalization

Developed countries have cleaned up their surface reality through institutions, capital and globalization mechanisms

Cities are cleaner, the air is better, society is more equal, and it looks like a “model of human progress”

The energy, raw materials, cheap goods, even waste processing that let them maintain a high quality of life

Global ecological whitening; it stratifies “cleanliness and happiness” geographically

Every country, every consumer, every company is unwittingly participating in this structure

We are always displacing costs, not bearing the consequences

We are always playing the inert weakling, protecting the environment, when in fact between countries we are still living in a selfishness disguised as civilization

Instead, let all countries share cleanliness and happiness with dignity

The EU implements “Extended Producer Responsibility” (EPR), requiring companies to be responsible for a product’s full life cycle

Shooting - black and white scenes

First there’s the problem of light; some light can pollute the view, and black and white can draw attention to the rhythm of shadow

Then sometimes too many colors also distract; a complex street signboard, a crowd in bright clothes - removing color can highlight the people themselves

Then there are scenes that need to carry a nostalgic, abstract or surreal quality

Photographer Ansel Adams once said black and white isn’t a deprivation of reality but a “reinterpretation of reality”. Color is part of reality; black and white is a translation of feeling

When is black and white appropriate? Maybe in the moment of shooting, ask yourself why you want to capture it