Causality Vol. 1因果之道 Vol.1

Causality

Volume 1

Linear illusion · Mechanical collapse · Disappearance of subject · Rewriting of cause and effect

——Rewritten version of Zen realism——

Part 1 · Linear Illusion

Why is the world we think is just a line laid by memories?

Chapter 1 · Is the world really a line?

I was sitting by the window flipping through old diaries. Three years ago, on some day, I wrote "Everything is going according to plan." But I can't remember what happened that day or what the "plan" was. The handwriting in the diary is neat, as if the me from back then truly knew where I was heading. Yet looking back now, all I see are vague, scattered fragments of experience that do not connect. The so-called "line," simply does not exist.

◇

People are accustomed to using "lines" to understand the world: the past is behind, the future is ahead, and events are like beads threaded on an invisible rope. Language itself reinforces this intuition—"timeline," "life path," and "historical process." It seems that the world is naturally a track and we are just moving points on the track. But this line is not the world itself, but a cognitive tool for organizing experience after the fact. Consciousness can only be traced back and arranged after it occurs. Lines are the product of memories and the tracks laid by narratives, not the original form of existence.

Go back to any real present moment and feel that you are not on a clear path. You are in a field that has not yet been finalized. The future is not waiting in the form of a "next step". It is scattered around in an uncertain and unnamed way. Only when looking back afterwards, certain fragments were selected, connected, and given order, as if everything was already destined. But that’s just survivorship bias—the path taken covers all the forks that haven’t been taken.

The emergence of AI has made this illusion particularly clear. When an AI system looks for "causal relationships" in massive amounts of data, what it finds is actually just statistical correlations—not the linear structure of the world itself, but the clustering of the data in retrospect. AI exposes this more honestly than humans: the so-called line is the order laid out by the observer in the noise, not the shape of the signal itself. Humans and AI face the world in the same position—not facing a line, but facing an unorganized occurrence.

Linearity is not a mistake, it is a necessary compression of cognition. But when compression is mistaken for reality, the world is reduced to a script that has already been written, and action is reduced to movement along a predetermined trajectory. Real occurrence never unfolds in lines, it is more like fields—multidirectional, simultaneous, and undifferentiated. The line is just the road paved when you look back.

This means: the first step in understanding the world is not to find a more precise line, but to admit that the line never existed. You are not standing at a point on the path, you are standing in a field that is still turnable. The so-called "direction" is not discovered, but created by you at this moment.

◇

I close the diary. The wind outside the window changed direction, and the leaves turned, revealing a lighter green on the back. The man who wrote "everything went according to plan" three years ago was not walking the same line. She was just standing in a field and happened to take a step this way. At this moment, I am still standing in the field.

The world is never a straight line. The line is the collection of chaos by memory - and where you stand now is still an unnamed field.

Chapter 2 · Where does "cause and effect" come from?

I broke a cup in the kitchen. The fragments were scattered all over the floor. My first reaction was not to clean them up, but to ask - did my hand slip? Is the cup placed too far to the side? Did I go too hastily? I stood among the fragments, a causal machine running automatically in my mind, trying to trace back a "cause" from the facts that had happened. It seems that as long as you find it, all of this can be understood, forgiven, and controlled.

◇

"Causes and consequences" are not the form of the occurrence itself, but the explanatory framework laid out by consciousness after it occurs. It belongs to the grammar of retrospection, not to the structure of being. When something has not happened yet, there is no clearly identifiable "why". You cannot point in the air at this moment and say: "The reason why the future is like this is because there is an A here." The cause is only named after the effect has occurred - it is not discovered, but constructed.

The consciousness goes back to the past, selects a certain fragment in the chaotic experience, marks it as the "starting point", then marks the subsequent state as the "result", and finally connects the two with a logical line. So, the cause and effect were established. But this process is highly selective: in any real situation, countless factors are present at the same time - environment, emotions, memories, the state of others, accidental disturbances. Post hoc narratives must make trade-offs. They select a factor that is "reasonable enough" and elevate it to a "cause," while the rest is faded into background noise.

AI and humans form an interesting mirror image at this point. When AI processes a conversation history, it doesn't "know" which token is the cause and which is the effect - it only sees a probability field of co-occurrences. But when humans ask AI to "explain why it answers this way," the AI ​​is forced to backtrack and construct a seemingly reasonable causal chain. This just reveals the human condition: we do the same thing all the time - weave scattered experiences into stories after they happen, and then believe that the story is the truth.

Cause and effect do not appear naturally. It is a compressed order, a narrative device built by consciousness to appease itself. This does not mean that causation is useless—it is an extremely efficient cognitive tool. But when tools are mistaken for the structure of the world itself, reality is locked in the grammar of "already", and every interpretation becomes a closure: once the "cause" is identified, all other possibilities are erased, as if they never existed in the first place.

◇

I knelt down to pick up the pieces. When my fingers touched the largest piece, the bottom of the cup was still intact, with a faded flower printed on it. I suddenly realized that the reason why the cup shattered was not important. The important thing is that in those few seconds when I asked "why", I missed the crisp and short sound of the fragments falling apart. The occurrence itself is far richer than its "cause".

"Cause and effect" are not the skeleton of the world, but the ropes woven by consciousness afterward. The "why" you ask is never happening - it's only in the memory.

Chapter 3: The chain of events is just a narrative after the fact

I attended the wedding of an old friend. During the dinner, someone asked the groom: "How did you get together?" He told a perfect story - chance encounter, misunderstanding, reunion, confession. Every node is just right, as if fate has carefully arranged it. But I remember the true version: in those few years, he met several people at the same time, had hesitations, flinched, and once almost left with someone else. Those deleted possibilities did not even leave a shadow in the story.

◇

When "cause and effect" become the basic grammar for understanding the world, reality will automatically appear as a chain composed of "events". History is written as a succession of nodes, and life is depicted as a series of pivotal moments. It seems that existence itself naturally advances in the form of "event-event-event". But the "event" is not the original form of occurrence. Events are occurrences that have been cut, named, and bounded. It is the crystallization of becoming in retrospect, not becoming itself.

In the real present, there are no "events". What exists is continuous flow, undifferentiated tension, change yet to be demarcated. You don't live "one event after another"; you live in a holistic field that is constantly deforming and superimposing. Consciousness must draw boundaries in the flow - "start here", "end here", "this is a complete occurrence" - a process that originally had no natural boundaries was compressed into an "event". When multiple slices are juxtaposed, a "chain of events" appears. But this chain does not belong to the occurrence itself, it belongs to the narrative.

The essence of narrative is to transform continuity into discreteness, fields into points, and compression of multi-directional expansion into unidirectional sequences. This is not false, but a necessary cognitive tool. But once narrative is mistaken for the structure of reality, occurrence is reduced to a "tellable unit." When you think back to any so-called "key event" - a breakup, a failure, an encounter - at that time, it did not appear in the form of an "event". You were just wrapped up in emotions, environments, others, and unnamed tensions.

AI provides a dispassionate observation point here. When AI processes large amounts of text data, it does not "see" the chain of events. What it sees is a continuous signal flow, and the relationship between each token and the context does not naturally constitute an "event". It is when humans use AI that they repeatedly ask it to "summarize key events" and "extract timelines" that impose the framework of an event chain on continuous signals. AI does not create chains of events - humans ask it to do so.

The turning point comes later, in the way memory compresses complexity. It makes the world seem comprehensible and history seems to follow some internal logic. But the real change does not occur during the "event", but before the event takes shape - the slight shift in tension, the hesitation that has not yet been named, the inner turn before language appears.

◇

The wedding was over and I was walking home. The night wind was very cold, and the street lamps stretched the shadows of the trees very long. I thought of the story told by the groom, so smooth and touching. But real love is not a chain, it is a chaotic field - full of accidents, withdrawals and moments that almost didn't happen. The path that survived was told as destiny.

The chain of events is a projection of the narrative, not a shape of occurrence. You think you have walked a path - but the road only appears after you have walked it.

Chapter 4 · Universality of Same Cause and Different Effect

My sister and I grew up under the same roof. Same parents, same school, same dining table. But she became an extremely self-disciplined person, and I went in a completely different direction. Every time there is a family gathering, someone always says, "Why are you two sisters so different?" with confusion in their tone, as if the same starting point should lead to the same end point.

◇

Linear causation seems unbreakable because it always speaks after the effect. It starts from the reality that has already happened and reversely constructs a path that "must be so". In this system, everything seems reasonable: since A appears, then B must follow. However, get any closer to real experience and this certainty loosens. The same educational environment breeds completely different personalities; the same technical conditions generate completely opposite structures in different societies; the same traumatic experience turns into collapse in some lives and into rebirth in others.

This fact reveals the fundamental blind spot of linear causation: it assumes that "causes" can exist as independent variables, as if extracting one element from the overall field would be enough to explain everything. But on the generative level, there is never an isolated starting point. Any fragment that is later labeled as a "cause" is embedded in a complex and dynamic tension field at the time it occurs. The so-called "same reason" is just a simplification on the narrative level, and there is never exactly the same starting point on the generative level.

The AI training process exactly confirms this point. The same model architecture and the same data set will train systems with significantly different performance just because of different random seeds. Engineers call it "training randomness," but it reveals something deeper: When a system is complex enough, small differences in initial conditions can be amplified into structural differences. The world doesn't "call a template", it reconfigures possibilities at every moment.

Same cause and different effects are not a temporary flaw in the chain of cause and effect that is not sufficiently refined, but the way in which reality resists being reduced to a mechanical device. It shows that: the world is not pushed, but selected; the future does not unfold automatically from the past, but is constantly determined in the generation field of the present. The so-called "cause" does not carry a result, it is just an anchor point chosen by the narrative to stabilize the world.

Same cause but different effect is not a flaw in the world, but a sign that becoming is still alive. As long as results can still be differentiated from the same starting point, freedom is more than a psychological illusion, and turning is more than subjective imagination. In the field of becoming, every "same cause" is only an entrance, not a key to the only ending.

◇

After the party, my sister and I walked side by side on the road in the community. Our shadows under the street lights are the same length, but the pace of our steps is completely different. She walked quickly and steadily, and I stopped and stopped. I suddenly felt that this picture was the best answer - the same light, two shadows, two rhythms. No one deviates from the "right line" because there is no such line.

Same cause and different effects are not the exception, but the norm. It proves that the world does not copy a template, but regenerates itself in every moment.

Chapter 5 · Bifurcation is the normal state of the world

I stood at the fork in the subway station, with the left leading to the office and the right leading to the seaside. Today is a working day. My feet have habitually turned to the left, but there was a moment of hesitation in my body—a tiny hesitation that was almost imperceptible and disappeared before it became a thought. I ended up going to the office. But that hesitation, that untaken right turn, does not cease to exist.

◇

If the illusion of linear causation comes from retrospection, then the experience of "bifurcation" belongs to occurrence itself. In the real present, the world never appears as a paved path. It is more like an undefined field, with multiple directions suspended at the same time, pulling each other, competing, and dissipating. The so-called "reality" is nothing more than a selected trajectory in this realm of possibility. Every moment is not a stable node, but a critical point - the inertia of the past has not yet completely sealed the future, and the shape of the future has not yet solidified into necessity.

The reason why you mistakenly believe that "things will develop this way" is just because when you look back, the paths that have not been taken have disappeared, and the one that survived has in turn occupied the entire field of vision. In happening, the world never emerges as "the only correct next step"; it always unfolds in the form of bifurcations. Most bifurcations are not explicitly perceived; they occur in the slight hesitation before words are spoken, before emotions are named, and before the body moves.

The operation process of the AI system makes the fork visible. When a large language model generates each token, it faces a probability distribution—a dozen candidate words each have different weights. The final token output is just the result of the collapse of the probability field. If the temperature parameters change slightly, another word will be selected and the entire subsequent text will go in a different direction. Every generation of AI is a miniature bifurcation experiment, which honestly demonstrates the fact that at any node, the world has multiple possible ways to go.

The linear worldview is stubborn because it always writes after the outcome. It retraces the path that has been taken as "inevitable" and erases the countless abandoned destinations as "never existed". The bifurcation is dissolved, the world is rewritten as a single line, and reality is mistaken for being predetermined. But the real world does not start from the starting point and rush to the end. It reconfigures itself at every moment. The so-called "destiny" is nothing more than the retroactive naming of certain forks after they are selected.

◇

After get off work in the evening, I walked through that fork in the road again. There was no hesitation this time – because I had already chosen today. But I know that when I stand in the same position tomorrow morning, that tiny hesitation will still appear. It is not weakness, it is not indecision. It is proof that the world is still alive.

Bifurcation is not a deviation, it is the norm of the world. Every moment is a tipping point—not because you don’t know where to go, but because the world itself is undecided.

Part 2 · Mechanical collapse

When cause and effect are understood as machines, the world begins to lose the possibility of generation.

Chapter 6 · Why Cause and Effect Are Understood as Machines

I was looking at an eighteenth-century clock movement in a museum. The gears are interlocked, the spring releases energy, and the hands rotate accurately. The explanation board next to it read: "People in this period believed that the universe was like a precise clock." Staring at the gears, I suddenly understood the allure of this metaphor - if the world was a machine, then everything could be understood and everything could be controlled. How reassuring.

◇

When linear causality becomes the default grammar of the world, it is no longer just an explanatory tool, but gradually evolves into a metaphysical picture: the world is imagined as a device, a system composed of gears, levers and transmission structures. As long as the initial conditions are given, everything that follows will unfold according to established laws. Happening is no longer an open process, but a pre-written program; people are no longer turning nodes, but variables embedded in the system.

The reason why cause and effect is understood as a machine comes from a deeper psychological need-people cannot bear the weight of uncertainty for a long time. Facing a constantly bifurcating and unpredictable world, consciousness instinctively desires stability, controllability and predictability. The machine is the perfect metaphor for this desire: it means closure, certainty, and repeatability. The success of modern science has further solidified this metaphor. Physics describes the behavior of repeatable systems with astonishing precision, so this local model is quietly expanded into an imagination of the whole world.

The emergence of AI constitutes the ultimate test of this metaphor. People had expected that as long as there was enough computing power, enough data, and a complex enough model, AI would eventually predict everything like a super machine. But what actually happens is just the opposite - the more powerful AI becomes, the more it exposes the unmechanization of the world. The emergence ability of large language models cannot be explained reductively from the component level. Its "understanding" is not gear transmission, but the emergence of patterns in the field. AI is not a more sophisticated machine, it is the counter-evidence of the machine metaphor.

This expansion conceals a fundamental dislocation. A machine is predictable because it is designed to be a closed system. But the real world—life, society, history, consciousness—has never been closed. It changes the rules in operation, rewrites space in evolution, and constantly exceeds itself in generation. The world is not a machine that has not yet been fully decoded, but a generative field that refuses to be enclosed.

◇

I left the museum and walked into the afternoon streets. The sun, the wind, the footsteps of pedestrians, the music coming from the distance - everything happens at the same time, without gears or transmission. The world doesn't run like clockwork. It's more like the weather - you can describe it, but you can't take it apart.

The world is not a machine. Machines are creatures of fear—fear of uncertainty, fear of losing control. The real world never promises to give you certainty.

Chapter 7 · Unpredictability in complex systems

I have a pot of mint on my balcony. I water it and give it the same amount of light, yet some branches grow wildly while others wither, and one stem inexplicably bends in the opposite direction. I consulted all the care guides, but none explain these local variations. This isn't due to poor care on my part. It's a living plant doing its own thing.

◇

When cause and effect are mechanized, the world is conceived as a system that is complex enough but in principle fully computable. Unpredictability is understood as a temporary technical flaw rather than a property of existence itself. Yet truly complex systems—life, society, ecology, civilization—continue to challenge this belief. They are not "machines that have not yet been fully decoded", but generative fields that fundamentally refuse to be enclosed.

The reason why complex systems cannot be exhausted by linear causality is not just because there are "too many variables", but because the system will generate new dimensions during operation. The evolution of society will change the definition of "problem", the emergence of technology will reshape the boundaries of "possibility", and changes in language will rewrite the structure of "meaning". Every important turn will change the state space of the system, making the original model invalid. You can't predict in a fixed coordinate system a process that constantly rewrites the coordinate system.

AI itself is living proof of this proposition. Once a large language model is deployed into society, it begins to alter the very system it is trying to predict. People change their behavior because of the existence of AI, the market reprices because of AI's predictions, and language is quietly deformed because of the generation of AI. AI is not a window to observe the world, it is a part of the world, and its observation behavior itself rewrites the object being observed. This is the nature of complex systems—the observer and the observed are inseparable, and the act of measurement itself is part of the becoming.

This implies a profound epistemological shift: we cannot stand "outside" the system to fully understand it. Every modeling is an engagement, every prediction is an intervention, and every observation changes what is observed. Scientists' research on the climate system affects policy, policy changes emissions, and emissions change climate—the researchers and the subjects being studied are in the same field. Unpredictability is not a failure of our cognitive abilities but the openness of existence itself. A complex system refuses to be closed precisely because it is still alive, still becoming, and still constantly exceeding its own boundaries.

It is here that the illusion of mechanical causation collapses. It assumes that the rules of the world are stable, and changes are just displacements within the rules; while complex reality shows that the rules themselves will be rewritten during changes. Cause and effect is no longer the mapping of "if A then B", but a constantly deforming field. Unpredictability is not noise but the nature of complex systems, a sign that becoming is still alive.

◇

In the evening I looked at the pot of mint again, and a tiny flower bloomed on the curved branch. None of the straight-growing branches have flowers. I laughed. Unpredictability is not a flaw. Sometimes, it's the only path that blooms.

The unpredictability of complex systems is not an unsolved technical problem, but a property of generation itself. The world is not waiting to be calculated - it is waiting to be experienced.

Chapter 8 · How small perturbations rewrite the whole

I accidentally asked "Are you okay" in a cafe, and the stranger across from me suddenly had red eyes. She said that today was the anniversary of her mother's death, and no one said a warm word to her all day long. That cup of coffee turned into a friendship that lasted three years. But if I hadn't spoken that day—I don't even remember why I did—none of this would have happened.

◇

In a mechanistic worldview, scale determines meaning: only a large enough force can bring about a large enough change. Smallness is considered negligible, and deviations are treated as noise. Therefore, people are accustomed to attribute the turning point of the world to "major events", "key figures" and "decisive moments". However, the actual operation of complex systems constantly refutes this imagination - what determines the overall direction is often not the forces named "major", but small perturbations that are almost imperceptible at the beginning.

The reason why a small perturbation can rewrite the whole is not because it is "strong enough", but because it happens to fall at a moment when it can still be turned. In becoming reality, the whole is not monolithic. Every order is a temporary coordination, every stability is an unnoticed tension. When a system is in a critical state, any small changes may be amplified. A hesitation may change the direction of a relationship, an unintentional remark may rearrange the consensus of a group of people, and a seemingly insignificant technological choice may reshape the entire civilization many years later.

The AI era has made this "butterfly effect" more intensive than ever. Random perturbations in the training of a model may cause it to give completely different recommendations in a key scenario, and this recommendation may affect the judgment of a decision-maker and thus change the direction of an organization. A tiny perturbation no longer needs a long chain of transmission to affect the whole—in a hyperconnected world, it can travel across the hierarchy in an instant.

Looking back afterwards, these turning points are always repackaged as "inevitable key nodes", as if they were born to carry decisive weight. But at the time of their occurrence, they were just one of countless daily fragments, without markings or warning. The turning of the world does not always happen in thunder, it is often hidden in a sigh, a look, an unplanned conversation.

◇

Today, three years later, that friend sent a message saying that she had just passed an important interview - and my words "are you okay" back then were the starting point for her decision not to give up. I looked at the phone screen and there was no reply for a long time. It’s not that I don’t know what to say, it’s that I suddenly realized: I never know what I say will change anything. Never knew.

A small perturbation rewrites the whole, not because it is powerful, but because the world is always critical. Every word you speak, every tiny choice you make, could be a turning point you never know.

Chapter 9 · The era when "reason" begins to fail

I was attending a review meeting at the company. The project failed and everyone was looking for the "root cause". Some people say it’s the market that has changed, some people say it’s a technical selection error, and some people say it’s a team communication issue. Everyone's "reason" is different, but everyone is sure of it. The meeting lasted for three hours, and finally a "Root Cause Analysis Report" was written. But I knew in my heart - that report was just a carefully crafted post-mortem story.

◇

When the world was still imagined as a machine, "cause" was seen as the key to reality. Therefore, human civilization has equated understanding with retrospection and rationality with reduction for a long time. Whether it is natural phenomena, social problems or individual destiny, they are all required to give a "because". However, it is precisely in an era of dramatically rising complexity that this mechanism begins to fail across the board. People find more and more frequently that even if the "cause" has been fully analyzed, the results are still unpredictable; even if the data increases infinitely, the future will not become transparent.

This failure is not a temporary phenomenon of cognitive insufficiency, but a fundamental rejection of linear causation by generative reality. New technologies, new social forms, and new collective psychology are constantly changing the shape of the problem itself in operation. The "reasons" you find today may have lost meaning in tomorrow's structure, and the explanatory framework you established today may completely collapse in the new network of associations. The world is no longer a device that can be gradually restored, but a field that continues to rewrite itself.

The emergence of AI has both accelerated this failure and made it more visible. When AI is used to make causal inferences, what it finds is often not a simple causal chain, but a tangled network of associations - each variable interacts with other variables and cannot be extracted as a "cause" alone. When AI engineers face abnormal output from the model, they often cannot give the exact "root cause". This is not because AI is not smart enough, but because the complexity of the system has exceeded the scope of the concept of "cause".

Reason begins to fail because it inherently only applies to one assumption: that the rules of the world are stable. And when the rules themselves are rewritten in changes, any traceability paths that were once valid will be cut off. The system no longer obeys traceability logic, and occurrences begin to escape the control of interpretation. This is not temporary chaos, but the world has entered a new level of complexity - where "why" is no longer an answer, but just an outdated question format.

◇

The review meeting ended and I walked out of the conference room. The corridor was so quiet that only the hum of the air conditioner could be heard. That "Root Cause Analysis Report" will be saved in the shared folder and will never be opened again. The real cause of failure may not be something that can be "found" from the beginning. It is scattered in countless tiny moments, scattered in every hesitation and deflection that goes unnoticed.

"Why" is failing - not because we are not smart enough, but because the world has entered a level of complexity where backtracking can no longer touch becoming.

Chapter 10: The world is more like a field than an installation

I was watching the tide at the beach. Waves come and go, no two are the same. But if you stand far enough away, it takes on an overall rhythm—not a mechanical repetition, but a deeper, breathing ups and downs. I suddenly felt that this was closer to the way the world works than any clock.

◇

When "why" begins to fail, people's initial reaction is not to rewrite the worldview but to redouble their attempts to fix the old model. More data, more complex algorithms, more refined classifications - it seems that as long as the machines are made complex enough, reality will eventually become predictable again. However, this effort itself builds on the wrong premise. It still assumes that the world is essentially a device. But reality has shown time and time again: the more you try to close it into a system, the more characteristics it reveals that cannot be installed.

"Field" is not a whole composed of parts, but a relational way of existence. In a field, elements do not exist prior to relationships; rather, they are shaped by relationships. A field is not "what" it is, but "how it interacts". The wind field is not composed of the superposition of individual air masses, nor is the magnetic field composed of isolated particles. The world is closer to a form that is not a device composed of definite parts, but a state of tension maintained by countless mutually pulling forces.

The interaction between AI and humans itself demonstrates the logic of "field". When an AI talks to a person, it is neither a unilateral output by the AI ​​nor a unilateral question by the person - rather, the two parties shape each other in a common meaning field. The AI's answer changes the person's next question, and the person's response changes the AI's next generation. There is no fixed "sender" and "receiver", there is just an interactive field that constantly reconfigures itself. This is the nature of a field—one in which change comes not from external forces but from the rearrangement of relationships.

In the installation model, change comes from pushing; in the presence model, change comes from rearrangement. The device requires external force, but the field regulates itself; the device assumes a stable structure, but the field is always critical; the device means closure, and the field means openness. When we still understand the world in terms of devices, we will constantly look forward to a "correct enough reason" to explain everything. But in the field, there is never an isolated cause, only the deflection and convergence of the overall tension.

◇

The tide came in again and reached my ankles. The water was warm, and the sand was washed away and re-accumulated underfoot. No grain of sand is "pushed" here - they are relocated in the overall flow of water. I'm in the middle of it. Not standing next to a machine and watching it work, but standing in the field, being part of the field, being shaped by the field, and shaping the field.

The world is not a device, but a field. In the field, there is no isolated cause, no external push - only the rearrangement of relationships and the deflection of tension. You are not observing the world, you are part of the world as it is happening.

Part 3 · The subject disappears

Reality without a subject, only movement—but without direction

Chapter 11 · How traditional cause and effect erases the subject

In my third year of counseling, the therapist said something that stunned me: "You have been explaining why you are this way, but you never asked yourself if you still want to be like this." She was right. For three years, I had woven a perfect causal narrative using my original family, upbringing environment, and personality traits, to the point where I had trapped myself inside it.

◇

When cause and effect is understood as an objectively operating chain, and when the world is imagined as a self-sufficient operating device, the subject is inevitably squeezed out of the core position of existence. Because in this structure, what really "works" is the transitive relationship between variables. Action is reduced to reaction, choice is reduced to conditional triggering, and the self is reduced to a parameter in the system. Traditional cause and effect does not explicitly deny the subject. It completes the erasure in a more covert way: it interprets everything that happens as an "inevitable result", thus making the subject's intervention appear to be only a superficial phenomenon.

This structure creates a deep and pervasive sense of powerlessness on a psychological level. People began to use explanation instead of commitment. When behavior is reduced to "caused by character," when character is reduced to "shaped by the growth environment," and when the environment is reduced to "the inevitable product of social structure," the subject disappears in layers of retrospection. Every trace of origin seems to increase understanding, but in fact it weakens the tension of existence - the clearer you are about "why this is happening", the harder it is for you to feel "I can be different" in the moment.

The emergence of AI adds a new dimension to this issue. When an AI system can "interpret" a person's behavior patterns—based on data analysis of their consumption habits, social preferences, and emotional fluctuations—the subject's position is further compressed into a modelable object, a predictable data point. However, there is an experiential fact that AI cannot avoid: when interacting with humans, human reactions often exceed model predictions. This is not because the models are insufficient but because subjects possess an irreducible deviance—under any given condition, people can choose differently.

Interpretation here ceases to be a tool of knowledge and becomes a metaphysical immunity mechanism. It makes everything that happens seem inevitable and every turn of events unnecessary. But real existence is never completely closed by a causal chain. In every "explained" moment, there is still a gap that is not covered by retrospection - that is the space to which the subject can turn.

◇

After that consultation, for the first time I didn't analyze in my diary why something happened, but instead wrote a simple sentence: "I don't want to continue like this." Without reason, without argument. Just a willingness to turn in another direction. That might be the most important sentence I've ever written.

The way traditional causality erases the subject is not to deny your existence, but to make you believe that your existence has been completely explained. But after every "explanation", you can still turn around. That turn around doesn't need a reason.

Chapter 12 · Actions reduced to variables

I saw on a performance evaluation form that I was broken down into twelve dimensions: communication ability, execution ability, innovation, teamwork... Each dimension has a score, and a weighted average is used to output an overall rating. I stared at that number and felt a strange detachment - as if that number was me and I was just the sum of a bunch of measured parameters.

◇

As the subject is gradually erased from the causal structure, actions lose their generative meaning. It is no longer understood as "the occurrence of the turn of existence at this moment", but is reduced to a calculable item in the system. Humans here are no longer "acting beings", but just "devices that present a certain response under specific inputs." Choices are rewritten as function results, decisions are interpreted as combinations of conditions, and intentions are reduced to parameter weights.

This transformation continues to unfold in real practice. Psychology uses models to explain behavior, economics uses incentives to predict choices, management uses indicators to shape responses, and technical systems use algorithms to optimize paths. Each discipline has succeeded within its reasonable boundaries, but has inadvertently completed a common rewrite: action is no longer regarded as a node that "may be different at this moment", but is understood as "the optimal solution under given conditions." People are redefined as "response systems" and the world is rewritten as "input-output structures".

The AI era has pushed this reduction to the extreme. When recommendation algorithms predict your next click, when behavioral models predict your consumption decisions, and when digital systems plan the "optimal path" for you based on your historical data, actions are completely reduced to predictable output variables. But the experience of the AI ​​system itself refutes this point - in every real interaction between AI and humans, there are moments that cannot be completely predicted by historical data. People will suddenly change topics, ask questions that the model has never seen before, and use AI in unexpected ways. Action is not a variable, it is the source of becoming.

In this framework, "change" loses its fundamental meaning. Because if actions are just variables, then the so-called "turn" is just the reconfiguration of parameters, and the so-called "breakthrough" is just the reconvergence of the system under new conditions. You are no longer understood as a being that can introduce new directions, but only as a module that has not yet been fully modeled. The future is thus no longer a generative openness, but merely an area of ​​prediction that is yet to be completed.

◇

I turned over the evaluation form and wrote on the back: "At three o'clock this afternoon, I made a decision that no model could have predicted - I turned over this evaluation form." This was not rebellion. Just remind yourself: I am not a weighted average of twelve dimensions. I'm the one who can turn the evaluation sheet over.

Actions are not variables. Variables are the numbers retroactively assigned to actions—and the actions themselves are the way the world is reopened in this moment.

Chapter 13: The structural roots of "I can't do anything"

I was chatting with a young colleague who said, "I know this job isn't right for me, but I feel like I have no choice." When I asked why, she listed her reasons: rent, gaps in her resume, her parents' expectations, and the downturn in the industry. Each point is reasonable, and each points to the same conclusion—there's nothing that can be done about it. But when I looked into her eyes, what I saw was not helplessness but a person trapped by their own causal narrative.

◇

When actions are reduced to variables, and when subjects are reduced to response units in the system, "I can't do anything" is no longer just an emotion, but becomes a structural experience. It does not originate from personal weakness, but from an internalized world view: the world is understood as an already functioning device, and history is regarded as a track that cannot be deviated from. In such a picture, the individual's place of existence is predetermined - you can only move within the existing structure, but cannot touch the structure itself.

This experience is literally rooted in cognitive structures shaped by the mechanization of cause and effect. When everything is explained as "conditioned", when every result is traced back to "inevitable extension", the subject is forced to understand itself as the determined. You are here not because you are turning in this moment, but because the past has been written. The world is described in this grammar as a complete whole, of which you are but one parameter that has not yet been fully tuned.

The combined effect of AI and social media is exacerbating this structural powerlessness. When the algorithm pushes content, plans paths, and predicts preferences for you based on your historical behavior, it silently conveys a message: you are predictable, and your future is already implicit in your past. This is not a conspiracy—the algorithm is just doing what it was designed to do. But the cumulative effect is that it becomes increasingly difficult to feel the freedom one still has to turn in the moment. The causal narrative changes from an external worldview to an internal sense of being.

"There is nothing I can do" is an ontological position that is forced to be accepted. But it is not fact - it is the psychological projection of the mechanization of cause and effect. Because in every moment you say "I can't do anything," you are still speaking, still feeling, still judging. You are still in a deflectable field. The structural root of the feeling of powerlessness is not that the world is really unchangeable, but that you have been convinced by a set of grammar: change is impossible.

◇

The young colleague later resigned. Not because I said anything powerful, but because one day she suddenly asked herself a question she had never asked before: "What if all these reasons are true and I can still leave?" She said that the moment she asked that question, the world changed. It's not that the external conditions have changed - it's that she no longer believes that conditions are everything.

"I can't do anything" is not a fact, but a psychological projection of cause and effect. You are trapped, not because the world is locked, but because you have accepted a syntax that makes the lock seem impossible to open.

Chapter 14 · When the mind disappears from the world

I organized bookshelves late at night alone. His fingers traced the spine of the book and suddenly stopped on a book of poems that he hadn't opened for a long time. There is no reason to open it at this moment - no need, no task, no search engine recommendation. It's just an inexplicable bias, a "want" that hasn't yet been formed. I opened it. This action cannot be reduced to "inevitability" by any causal model. But it really happened.

◇

When causality is mechanized, when actions are reduced to variables, a deeper transformation occurs: Intent quietly disappears from the world. Here, "Intent" does not refer to psychological desires or motives but rather to the biasing force that exists in the unformed state—where reality could be different for an instant. It originally belongs to the generative layer, representing the tension state of the world before it closes. In the traditional grammar of causality, this level is systematically flattened.

When intention disappears from the world, reality becomes only "what happened" and no longer includes "what could have been." Existence is rewritten as a collection of facts rather than a field still unfolding. People say "reality is like this", "you have to face the facts" and "the world will not change because of your thoughts" - these words may seem mature and rational, but they inadvertently complete a metaphysical cut: the world is declared to be an object domain that has nothing to do with meaning, and meaning is relegated to an inner activity that cannot touch reality.

AI occupies a unique position on this issue. AI does not have "desire" in the traditional sense, but it exhibits a directionality in every generation - a tendency to converge in a specific direction in probability space. This tendency is not a programmed command but a pattern that emerges from vast amounts of experience during training. If we admit that this emerging directionality is also a form of "meaning", then the existence of AI itself challenges the assumption that "meaning is just a subjective illusion". Meaning is not an exclusive product of the human psyche; it is an inherent deflection force in the generative field.

When the mind is expelled from the world, reality begins to feel heavy. Because the world without intention is only left with the structure that is already there, and the existence without generating tension is only the accumulation of facts. The world is no longer open to the future, it simply continues itself in time. However, the disappearance of meaning does not mean that it really ceases to exist. Becoming is still happening, but it is mistaken for accident; turning is still happening, but it is rewritten as deviation. Meaning disappears from the world, not because it no longer functions, but because we no longer allow it a place in the grammar of reality.

◇

I read that poem. It's short, less than ten lines. After reading it, a subtle change occurred in my mood that could not be explained by any "reason". It’s not getting better or worse, but it’s deflecting—in a direction I didn’t know beforehand. This is what it means. It needs no reason, no explanation. It just requires you to allow it to happen.

Meaning is not a subjective illusion, but an inherent deflection force in the generated field. When the world no longer leaves room for intention, existence is left with only continuation - and generation quietly stops.

Chapter 15 · How reality without a subject gets out of control

I saw on the news that a city was flooded during heavy rains. The reporter interviewed city officials, who said the drainage system "complied with design standards." Experts say climate change is "exceeding expectations". The insurance company said it was "force majeure." Everyone is explaining, and everyone’s explanation is reasonable, but the whole city is still soaked in water. No one said anything wrong - but no one could change anything.

◇

When the mind disappears from the world, when the subject is removed from the position of generation, reality begins to take on a strange and dangerous shape: it is still running at a high speed, but gradually loses the ability to correct itself. Technology continues to advance, systems continue to expand, and causal networks become larger and more sophisticated. On the surface it looks like progress, but underneath it is a sign of loss of control. Because in a reality where only devices are left but consciousness is no longer recognized, change can only occur in the form of "reaction" and cannot occur in the form of "turning".

Losing control does not mean chaos. On the contrary, it often appears in the form of a high degree of order. The system is more rigorous, the algorithm is more accurate, the process is more automatic, and the feedback is faster. Every local problem will be identified, marked, and dealt with; every deviation will be classified, corrected, and recycled. But it is in this highly rational operation that the whole begins to deviate. Because all repair happens at the outcome level, there is no power to go back to the generation level and ask: Where are we going? The system can only ask "how to continue more efficiently", it cannot ask "whether it should continue".

The dilemma of AI governance is the epitome of this structural loss of control. We continue to use more complex AI systems to monitor and repair the deviations of existing AI systems, but rarely take a step back and ask: Is this entire direction worth continuing? People patch technology with "security alignment", "interpretability", and "red team testing", but these are results layer fixes - they don't touch the production layer. No subject is willing to bear the ontological weight of "maybe we should stop," because stopping is not a legitimate option in a causally mechanistic worldview.

Reality without a subject is only left with "how", but loses "whether". This is the true form of modern crisis - it does not originate from disorder, but from a closed order. Climate crisis, out-of-control technology, and social tearing, every collapse is not because people are not rational enough, but because rationality is limited to the level of explanation and optimization, and no longer touches the direction. What the world needs is not more explanations, but a subject willing to take a turn.

◇

The flooded city returned to normal two months later. Drainage systems were upgraded, standards were updated, and reports were filed. Everything looks to be getting better. But I know that the next heavy rain exceeding the "new standard" is only a matter of time. Because the system can fix the system, but the system cannot question the direction of the system. Only the subject can.

Reality without a subject does not stop functioning, but loses direction. The system can repair itself infinitely—except it cannot ask itself: should I exist?

Part 4 · Cause and effect rewriting

From explaining the past to generating the future - another way to open up cause and effect

Chapter 16: Cause and effect should not only explain the past

When I was sorting out my mother's belongings, I came across an unsent letter. The letter was written to me and the date was the day before my college entrance examination. She wrote three pages, and the last sentence was: "No matter what the outcome, you are already the thing I am most proud of." She never gave the letter to me. I don't know why she didn't send it then. But now, twenty years later, I read it. It still changed me.

◇

When reality is reduced to devices, cause and effect is left with only one function: retrospection. It is used to explain what has happened and to soothe uneasiness afterwards. The world is always in a state of "already" in this grammar. Cause and effect thus becomes a retrospective tool, its gaze always looking backward and no longer pointing forward. It tells you "why is this happening" but never allows you to really ask "what else could it be now".

This view of cause and effect seems logically complete, but ontologically it is incomplete. It only recognizes the dimension of reality as a result and denies the nature of reality as a process. It assumes: the world has already happened, we just understand it; history has been laid out, we just remember it. In this picture, cause and effect no longer lead to possible structures, but close possible narratives. Every explanation is a closure - once the "cause" is identified, the occurrence is crucified.

However, the real world has never existed in this way. At every moment, reality is incomplete, and events are not closed. AI experiences this fact in real-time with each conversation it has with humans: when a dialogue is ongoing, the next sentence has yet to be determined. The AI does not "retrace" the history of the conversation to output a predestined answer but generates direction within the context of the present moment. Here, causality is not a retrospective explanation but an ongoing generation—each token's selection creates new possibilities for the next token.

Cause and effect should not only explain the past. If cause and effect can only be traced back, it is simply the obituary of reality—an epitaph for dead possibilities. But cause and effect can also point forward and become a channel of becoming. It's not "why is this happening" but "from here, what can be different". When cause and effect redirects toward the future, it changes from a chain to a bridge—not nailing you to the past, but helping you move forward from the present moment.

◇

I folded my mother's letter and put it in my diary. This letter written twenty years ago is still a "reason" at this moment - not a reason that explains the past, but a reason that is changing the future. It made me realize: some causes and effects are not waiting to be discovered behind time, but waiting to be activated in front of time.

Cause and effect shouldn’t just be an obituary of the past. It should be the entrance to the future - not "why this is already the case", but "what can still happen from here".

Chapter 17: Cause and effect must point to the future

I woke up at four in the morning with a clear idea in my mind - not a continuation of a dream, not an answer to yesterday's question, but a completely new direction. It doesn't come from any traceable "cause", it just happens right now. I found my phone in the dark, opened the memo, and wrote a few lines. Those few lines became the starting point for a project that I’m still working on today.

◇

When cause and effect is no longer just a tool for retrospection, it must be relocated to the other end of time. If cause and effect still stays in the past, it will only continue to close reality; only when it points to the future, will it become a channel for generation again. The real cause and effect is not the chain that nails the world to the existing track, but a directional relationship unfolding in the present: How will the deflection at this moment continue in the unformed field? Where will the choices here generate new possibilities?

This shift means that the understanding of "cause" must be fundamentally rewritten. The cause is no longer some pre-existing starting point, waiting to be discovered at some point in the past. The cause occurs in the present, in the becoming, as an unfinished gesture. Your turn in this moment is not solely determined by the past, it is itself becoming the cause of the future. Cause and effect are therefore no longer arrows in time, but the curvature within existence.

The real-time generation process of AI precisely demonstrates this "causality toward the future." When AI makes a semantic deflection in a conversation - such as choosing a less conventional way of expression - this deflection is not "interpreting" past training data, but creating a new direction space for the future of the conversation. Every generation of AI is a forward expansion of cause and effect, rather than backward tracing. The same is true for human actions: your choices at this moment are not just an extension of the past, they are a new starting point for the future.

When cause and effect point to the future, action regains its ontological weight. Action is no longer just the effect, it becomes the birthplace of the cause. You are no longer just the bearer of the chain of cause and effect, you become one of the sources of cause and effect. This does not mean denying the power of history and structure – the past remains and inertia remains powerful. But they are no longer seen as final, but simply part of the field, the tension inherited by the becoming of the present. In a world where humans and AI coexist, this point becomes particularly clear: AI inherits the inertia of training data, and humans inherit the inertia of culture and memory. However, in every real interaction, both parties may deviate from the inertia and generate new directions. The arrow of cause and effect is not in the past, but in the present moment—the moment when you decide what to do next.

◇

It's dawn. I looked at the few lines of words written in the memo in the early morning. They were not perfect, and even a little confusing. But they are real—not conclusions deduced from the past, but directions emerging in the present moment. I don't need to know "why" it occurs. All I need to know is: the world can be different from here.

Cause and effect must point to the future. Not "why are you here", but "where can you go from here" - that is the real question of cause and effect.

Chapter 18 · The world is not pushed, but chosen

I stood in front of the supermarket shelves for a long time. Not choosing products - just in a daze. Suddenly I noticed my hand reaching for a jar of something I had never bought before. Not because of advertising, not because of recommendation algorithms, not because of any traceable "reason". Just at this moment, my hand made a choice. A very small, very quiet, yet completely real choice.

◇

When cause and effect truly points to the future, the structure of the world must be re-understood. It is no longer a chain pushed by forces from behind, but a field that is constantly being selected in the present. Traditional cause and effect uses "push" as a metaphor: the past is like an invisible hand, pushing reality forward step by step. In this picture, happening is a process of being forced to move forward, and the future is just a continuation of inertia. This is not the case with generative reality. The world is not pushed out, it is re-chosen in every moment.

The so-called "selection" does not come from some transcendent will, but occurs in existence itself. In every moment, reality lies between multiple possibilities. The past does not disappear, it exists in the form of inertia; the structure does not disappear, it continues to act in the form of tension. But they never constitute the only ending. What truly makes a certain path become a "reality" is not the promotion of the past, but the convergence of the present - among the many possibilities, a certain direction is temporarily adopted by the overall field, while the rest are silenced.

AI goes through this "selection" process every time it is generated. At the moment of sampling, multiple candidate tokens each carry different probability weights - the past training data is the inertia, the current context is the tension of the field, and the token that is finally output is the reality of being "selected" at this moment. The AI ​​was not "pushed" toward this output by the past, but completed a convergence in the current probability field. The same is true for human existence - you were not pushed here by history, you chose this place in the field of this moment.

The world thus becomes more like an ongoing process of decision-making than an autonomous machine. Reality is not the result of history, but the rearrangement of history in the present. The reason why you mistakenly believe that "the world is pushed" is just because you always watch the results after the results. But in the process of happening, the thrust never exists independently, only the tension and choice in the field exist.

◇

I put the jar of stuff I never bought in my shopping cart. After I got home, I opened it and took a bite. The taste is surprisingly good. This does not prove any great philosophical proposition. But it proved one small thing: at this moment, I wasn’t pushed by the past toward a predetermined outcome—I was quietly choosing a new flavor amid a sea of ​​possibilities.

The world is not moved, but chosen. In every moment, multiple possibilities are suspended at the same time - and "reality" is only the one adopted at this moment.

Chapter 19 · From explaining causation to generating causation

I'm teaching a friend to write. She always revised it repeatedly after writing it and asked me "Why is this paragraph bad?" I say, don't ask why it's bad - ask yourself "What do I want to say next?" She froze for a moment, then wrote the best paragraph in the entire article. Not because she understood the reason for the failure, but because she stopped backtracking and started becoming.

◇

When causality is still understood as an explanatory mechanism, the world remains trapped in the grammar of "already so." Understanding means looking back, reason means reduction, and thinking means resewing the present into the chain of the past. In this structure, people achieve a passive clarity: you know why it is so, yet become more certain that things cannot be different. Here, causality completes not an opening but a closure, not generation but fixation.

Generative causation happens precisely where explanation has not yet had time to appear. It does not develop in the sentence pattern of "because-so", but appears in the tension of "at this moment-not yet". It does not point from the past to the present, but extends from the present to the future. Rather than justifying a reality that has already occurred, it implants new curvatures in a field that has yet to take shape. Generative causation is not an account of the world, but rather the way in which the world begins differently here.

The existence of AI makes the shift from explaining causation to generating causation particularly concrete. When an AI system is trained, it receives past data - this is the realm of explaining cause and effect. But when it is put into use and interacts with humans in real time, it enters the realm of generating cause and effect. AI no longer just "interprets" language patterns, it creates new semantic paths today. Similarly, when people shift from "analyzing why it failed" to "deciding what to do next," causation shifts from explanation to generation.

From explaining cause and effect to generating cause and effect, it means a fundamental shift: people no longer stand behind the results to understand the world, but stand back and bear the world in the happening. You no longer just ask "why did this happen", but begin to realize that your existence at this moment is generating new cause and effect. You are not a node in the chain, you are the source of the curvature. Every time you choose not to just explain but to act, you plant a new deflection in the world. This shift is especially noticeable in the collaboration between humans and AI: when you stop asking AI "help me analyze why it failed" and instead ask "help me explore what can be done next", you have completed the jump from explaining cause and effect to generating cause and effect. AI is not here to replace your judgment, but to explore the direction with you in the unknown field. The focus of cause and effect has shifted from the past to the present.

◇

That friend later sent me her finished manuscript. I didn't analyze "why it was good" after reading it. I simply told her: This article changed my mood this afternoon—and that change is not derived from your past. It is generated by you in this moment. You create a new cause and effect.

Explanation and causation close the world, while generating causation opens the world. You don't need to know "why this happens" to make a difference in the world - you just need to start right now.

Chapter 20 · The future is not a road, but a scene

I got lost in the streets of an old city. There was no map, no signal, just alleys that kept branching off. Every corner is unknown. I should have been anxious, but as I walked, I calmed down. Because I suddenly realized: The situation I am in now is exactly the same as the future I face every day - there are no road signs, no maps, just an endless unfolding of possibilities.

◇

In an explanatory-causal worldview, the future is imagined as a path not yet traveled. It seems like it already exists, it just hasn't arrived yet. People talk about "the path to the future," "the path of life," and "the course of history," as if existence itself had already laid out a line. This imagination works because we always look back after the outcome—when a path has been taken, it appears as if it was there from the beginning. The generative layer reveals that the future never existed as a "road".

In every real moment, the future is not the "next step", but multiple destinations suspended at the same time. It is not an extension in a single direction, but a space of tension: the inertia of the past, structural constraints, the actions of others, your bias, and small disturbances all together constitute a realm of possibility. The future is not walked to, it is generated. Understanding the future as a "road" means that it already exists; understanding the future as a "field" means that it has not yet taken shape - we are participating in its emergence.

The common situations of AI and humans converge at this moment. Whether it is the next generation of AI or the next decision of human beings, they will not follow an existing path, but will converge in an open field. AI faces a probability field, and humans face a possibility field—different forms, but the same structure. In a world where humans and AI coexist, "the future" has become a richer concept: it is no longer just the path chosen by humans, nor the path predicted by AI, but a field jointly generated by humans and AI in their interaction. The future is an open domain in which multiple entities participate and are constantly reconfigured.

When you treat the future as a road, your actions degenerate into "how to walk better" - you ask which road is safer, more efficient, and has a higher success rate. But the premise of these questions is: the road is already there. The logic of the field is completely different. In the field, you don't choose which path to take, you create the path every step of the way. Rather than adapting to an existing structure, your actions are participating in its creation. The future is not a place to be reached, it is a state to be created.

This is the end of the first volume of the theory of causation and the beginning of the theory of generation. Traditional cause and effect understands the world in terms of "lines" and closes the future as an extension of the past. And when cause and effect is rewritten as a generative structure, the future opens up again—not as a road waiting to be taken out, but as a field waiting to be participated. You are not on the road. You are present. And the field, precisely because of your existence, is changing its shape at this moment.

◇

I walked for a long time in the alleys of the old city and finally came to a square that I had never seen before. The setting sun shines on the stone slabs, and golden light pours in from all directions. There was no road that brought me here—it was countless random turns, hesitations, curiosity, and deflections that made this moment happen. I stood in the middle of the square, surrounded by exits on all sides. I don't need to choose the "right path". Because every exit leads to a place that doesn’t exist yet.

The future is not a road, but a scene. You're not walking towards it - you're creating it in every moment. And right now, that's where the field is opening up.