Computer graphics enthusiast. I build graphics-style pipelines for things that aren't images — sound, neural networks, and mathematical proofs — and I ship them as open-source tools.

• 🔭 Currently building the Entropia family: a node-graph DCC suite that shares one UI shell.
• 🧮 Also doing machine-checked mathematics in Lean 4 + Mathlib.
• 🌱 Learning Computer Graphics, Quantum Computing, AI, Music, Animation, and Vocaloid.

An AI + graphics inspired spatial audio engine that borrows the modern rendering pipeline (represent → simulate → shade → render → deliver) and points it at acoustics: AudioGS (Audio Gaussian Splatting) sound fields, Audio-USD scene representation, geometric acoustic simulation (image-source / ray tracing), Acoustic-BRDF material shading, HRTF binaural rendering, and a DSP graph. Ships offline and real-time renderers, a creator application, and plugin tooling. TypeScript

ComfyUI-style visual workflows for PyTorch (with optional TensorFlow/Keras and Hugging Face nodes): ~200 built-in nodes, a Blender-like modular workspace, live training curves, subgraphs, code export, a plugin system, and a built-in file manager. Runs in the browser and as a standalone Electron desktop app; the Python runtime is on PyPI as entropia-riko. Python · FastAPI · React

The reusable node-graph editor front-end extracted from Riko: React + Vite + TypeScript + React Flow + Zustand, with Blender-style splittable panels, floating windows, themes (incl. Apple Liquid Glass), code editor, and file picker — ready to wire to any backend. Published on npm as entropia-template-ui. TypeScript

A sorry-free Lean 4 + Mathlib formalisation of the elementary layer around 4/n = 1/x + 1/y + 1/z: reduction to primes, the divisor–certificate bijection, parity structure, a complete a = 3 characterisation, Obláth's identity, and a one-sided divisor criterion. Every claim carries a register tag — A proven, B computational, C conjecture, D AI-generated — and they are never mixed.

To be explicit: the conjecture is not solved here, and nothing in the repo claims it is. Lean

A Lean definition of Graham's number. Lean

• 🎵 Spatial audio engine — Entropia-Atmos
• 🧠 Visual deep-learning tooling — Entropia-Riko + Entropia-Template-UI
• 🗣️ Vocaloid-like TTS (synthesis engine)
• 🎬 Stars Strings-Yori (animation project)
• 🏗️ Long-term: one Digital Content Creation (DCC) suite around graphics, sound, and storytelling.

• 🚀 Light Sail (Breakthrough Starshot)

Small research toys, mostly rendering and physics: dlss-compositor (offline Blender Cycles renders through NVIDIA DLSS-RR / DLSS-FG) · Taichi non-Euclidean ray tracer · Quantum-RayTracer (Qiskit + Grover search) · Greendemo-Renderer (wave-optics 3D engine) · Blender_Symmetry_Corrector · Relativities · Tokamak (fusion plasma sim) · Boids · PFGMSinger (singing voice via Poisson flow generative model)

• I'm a big fan of Miku, Nanachi, Nahida, かぐや と やちよ and Faputa!
• 推し: PinocchioP, Nanawo Akari, TeradaTera, and Tsukushi Akihito.
• 📫 QQ: 2892466922

| Zhihu | Bilibili | Email | CSDN |

We were once promised a future like this: material abundance, humanity reaching UBI, commuting in eVTOLs, kilometer-scale megastructures, ring-shaped space stations, robots taking care of everything for everyone, all labor handed over to machines, and resources distributed on demand.

Of course, if anyone posts that vision online today, they will only be mocked: "keep dreaming — anything exists in dreams." That is what happens when an ideal runs into productivity and chaotic systems. The modern system is still steered by human emotion. But that mockery is doing more work than it admits. It is not only a judgment about the world. It is also a defense of a particular way of living in the world.

Here is the question I keep coming back to. We have built machines that fly, compute, translate, draw and diagnose. A civilization with our productive power should be able to ask a very different question than "how do I survive". And yet most people I know, including myself on most days, organize their entire lives around that question anyway. Why does a civilization that could move toward abundance still demand that almost everyone sell their labor time in order to live? Why, when technology keeps lowering the amount of labor that survival requires, does the culture still insist on the same path: study, find a job, buy or rent a home, marry, have children, retire? And underneath both: is realism helping people adapt to reality, or is it sometimes dressing up the present as the only possible reality? This essay is my attempt to walk through that chain honestly, without pretending I have solved it.

Start with what a great civilization would even mean. My definition of a Type I civilization is concrete: it can reach any place on this planet, reshape any geography on this planet, travel to and from Mars at will, and cure almost all diseases. On that scale, progress is not a feeling but a checklist, and we are already checking items. The Eurasian Pole of Inaccessibility was once a place that defeated the ancients, but we have built Tianshan International Airport nearby — a 4F-class airport. Shenzhen was once overrun with cockroaches; no one in ancient times could have predicted that it would become a technology capital. To be an outstanding civilization, we need at least an elevator up Everest, and the ability to take everyone to Lop Nur, the Amazon — anywhere. Anyone should be able to reach Mars smoothly. Musk and entrepreneurs like him are pushing in this direction, but they also face many difficulties imposed by the existing system.

There is a gap between those two kinds of sentences. One kind is about what humanity as a whole can do. The other kind is about what an individual human is allowed to want. We can pour enough concrete to build a 4F airport at the pole of inaccessibility, and still tell an individual who wants to attempt something large to first secure a job, an apartment, a loan, a stable position. The civilization advances; the people inside it are asked to stay in place. I think that gap is the subject of everything that follows.

Begin with the person inside the structure. As for the definition of an "ordinary person" — that is a very fuzzy thing. Any person in any field can be called ordinary: Musk is undoubtedly one of the world's best in business, but he may know nothing about sports. Let me give a concrete definition: a non-outstanding employee at an ordinary company, earning around the average income, who must keep selling their labor to survive, follows the fixed life path of marriage, children, and retirement, and has no thought of self-realization such as starting a business.

But the word "ordinary person" has now become something close to a "Dalit" online — a status an individual basically cannot escape in their lifetime. I admit that in the current economic downturn, most people will suffer lost opportunities and debt servitude, and will live out their lives step by step along the fixed path.

And here I have to be careful, because two different claims are hiding inside that word, and they must be separated. Ordinary is not worthless. A person who never becomes an entrepreneur, a scientist, an artist, an athlete or a public figure has not failed at being human. The problem is not that most people are ordinary. The problem is that "ordinary" has been redefined as a caste. I am not asking for a world where everyone must become extraordinary; I do not believe that, and I do not want it. A truly free society would not require anyone to become a genius in order to deserve the right to choose their own way of living. The target is not "everyone must become extraordinary". The target is "no one has to become extraordinary just to deserve autonomy".

Why is the path fixed? The distribution of resources in our society is extremely imbalanced. As of 2026, humanity is still under capitalism. This is not about good people versus bad people — it is about distribution, and the distribution is skewed by design. Most people own no means of production, so they must sell their labor power to survive, and the wage they receive covers only the reproduction of that labor: housing, transport, food. That is exactly why those costs swallow most of a person's income. Nobody needs to plot against you — the system only needs you to keep working, not to excel. The surplus your labor creates is accumulated as capital and concentrated in fewer and fewer hands, and that concentrated capital decides what gets funded and who gets the material conditions to pursue excellence. A single breakthrough costs millions — nearly a lifetime's income for an ordinary person — so for most people the door is shut before they ever reach it. Entrepreneurs aren't the villains; they are the ones holding the accumulated surplus, operating inside the same structure. But the structure itself is not innocent: it turns society's surplus into a gate that only a few get to pass. I'm not here to blame individuals. I'm saying that as long as the social surplus flows through capital accumulation, what most people can become is decided by how that surplus is distributed — and right now, that distribution mocks effort.

Why, then, don't more people try to start something, or to become excellent? First, their economic conditions may simply not allow it. Second, the risk is too high. Third, they are shaped by the social environment around them. Fourth, their own ability may not yet have crossed the threshold. These aren't excuses — they are the actual mechanics.

Those are the constraints. But constraints alone do not explain the culture, so let me look at the layer of needs underneath them.

Maslow's pyramid is too neat to be literally true, but as a sketch it is useful. A person whose survival is not secured cannot easily be free: free to start a company, to do basic science, to make art, to research one question for years, to explore the unknown, to pursue self-actualization. Without a safety net, those things are luxuries in the strictest sense. So survival-first is not naturally wrong. For an ordinary person in a highly scarce environment, realism may be the most rational survival strategy available, and I would never tell such a person they are thinking too small.

But now ask the same question at the level of a civilization. If housing, food, energy, healthcare, education and automated production keep getting more abundant, then survival should gradually stop occupying most of a human life, and the goal of a society should shift from "how do we keep people alive" to "what can people do once they are alive". That is the movement the pyramid describes: survival, security, belonging, esteem, self-actualization. Seen this way, the meaning of technological progress is not just higher efficiency. It is helping more and more people move up from the bottom of the pyramid. A civilization's true wealth is not the number of goods it produces; it is the number of people who do not have to sell their entire lives in order to survive.

There are two stories a society can tell about a human life. The first is the survival narrative: find a job, earn money, buy or rent a home, marry, have children, support your parents, retire. It answers a single question — how does one person stay safe inside the existing structure? It is not wrong. For most people, through most of history, it was the only narrative the material base could afford.

The second is the civilization narrative: why do we develop technology at all? Why automation, AI, robots, energy, spaceflight? What, exactly, should civilization be liberating human beings from? Hannah Arendt separated labor, the work that keeps life going, from action, the capacity to begin something new. A civilization that organizes everyone around labor has forgotten what it wanted to be free from. The civilization narrative is the reason the survival narrative should be allowed to shrink. If centuries of industrial revolution, computing, AI and robotics end with the message "congratulations — you can now find a job more efficiently", then we have missed the deepest meaning of the productivity revolution. A rich civilization eventually has to answer not only how people survive, but what survival is for. When machines can do more and more of the work that keeps society running, should selling your labor time remain the precondition for the right to exist? I do not have an answer. I only refuse to accept that the current arrangement is a law of nature. It is a choice, and the distance between the two narratives is where most of modern exhaustion lives.

Between the two narratives sits a very common philosophy of life. Choose a major by employment rate. Judge a hobby by its salary. Judge study by its ROI. Do research only if it can be monetized. Start a company only if it can raise funding. Judge art by its commercial value. A human life, compressed into an ROI optimization problem.

I do not want to condemn this thinking, because in a resource-scarce environment it is genuinely rational. If a family has no safety net, no capital, no tolerance for failure, no guaranteed healthcare, housing or education, then "think about employment and income first" is not cowardice. It is arithmetic. The problem appears when productivity has already reduced material scarcity substantially, and the survival logic remains the highest principle of life anyway.

I find it useful to separate three layers. The first is survival rationality: "I must earn money or I cannot live." That is a real constraint, not a philosophy. The second is instrumental rationality: "everything I do must prove its economic value." That is what happens when the logic of the market enters a person's psychology; a constraint becomes an identity. Weber drew this distinction a century ago: instrumental rationality, the logic of means, against value rationality, the choice of ends. The third layer is that value rationality: some things are worth doing even if they will never maximize income. Basic science. Art. Philosophy. Exploration. Games. Personal creation. Pure curiosity. Public knowledge with no direct economic return.

And notice what happens when instrumental rationality becomes the only language available. "Why do you study mathematics?" — "Because the salary is high", and not "because I want to know mathematics". "Why do you make art?" — "Because it makes money", and not "because I want to create". The market logic has crossed from the economy into the personality. The question, then, is when a civilization becomes rich enough to let people ask "what is worth doing" rather than only "what pays the most". That is not an economic question. It is the question of what an economy is for.

In China this survival-first thinking has a name and a face, and it is enormously popular. Zhang Xuefeng is only the most visible of a whole genre of education and career bloggers, and I am not interested in judging him; I am interested in the narrative. Call it the Zhang Xuefeng-style narrative: don't talk about empty ideals; first think about employment, income, which city, which major, the stability of a state post (编制), and family risk. Its rule is simple: interest can wait, survive first.

The narrative is popular for reasons that deserve respect. High education costs. High housing prices. Education competition that begins in childhood. Employment pressure. Families without safety nets. Anxiety about class mobility. A cost of failure so high that one mistake can set a family back for years. Young people with no trial-and-error capital. Parents who avoid risk on behalf of their children. A system in which your degree is hard-bound to your employment. In a low-welfare, low-safety-net society with expensive housing and education, "find a job that can feed you first" is extremely reasonable advice. I would probably give that advice myself in many cases.

But there is a problem hiding in plain sight: a survival strategy slowly becomes a philosophy of life. "You should find a job first" slides, without anyone noticing, into "your life was always supposed to be a job". Those are two completely different propositions. The first is advice about sequencing. The second is a verdict about what a human being is for.

Let me test this against a concrete failure, because failures are where realism sounds most convincing. Homejoy was a YC company that did on-demand home cleaning. It grew fast, raised real money, got a great deal of attention, and shut down in 2015. It is usually remembered as a lesson in realism. The market was real and brutal: customer churn, thin margins, cleaners who had to be found, kept and classified, competition on every side, growth that cost more than it returned, unit economics that never closed, cash that burned faster than the story could raise more. A founder who ignores that list is not an idealist. They have simply lost the ability to receive feedback.

So Homejoy answers a real question: when should a person accept reality? The answer is: constantly, on the operational level. Realism is not the enemy of entrepreneurship; it is the material of it. Ideals need reality for feedback. Without reality, an ideal becomes a hallucination. But the reverse is also true: reality should not become the final boundary of the ideal. Without ideals, reality becomes a cage. Homejoy died because its economics were wrong, and that was reality delivering the correct verdict on that company. The wrong move is to extend the verdict: to conclude from Homejoy that no one should attempt anything hard, that every ambitious plan is a Homejoy waiting to be exposed. The company failed. The question of what a company can be did not.

Here is where a strange contrast becomes useful. Sam Altman taught much of the early YC canon: the startup lecture series at Stanford in 2014, the essays, the endless advice. The picture of reality in those talks was deliberately small: find a problem, build a product, get users, grow, don't die, build a company. That is reality at the scale of a single founder. The same person now runs an organization whose decisions touch AGI, AI infrastructure, scientific research, compute, energy, geopolitics, and what technology does to a civilization as a whole. I am not calling him a hypocrite. I am pointing at something more interesting: reality itself has scales. For a college student, "find a job" is reality. For a founder, "will the company survive" is reality. For a civilization, "can energy, computation, automation, medicine and spaceflight change what is possible" is reality. Each scale has its own realism, and each is real. The error is to take the realism of one scale and absolutize it into the realism of every scale. A society that only allows people to discuss how to live better inside the existing rules, and never whether the rules themselves can change, has lost its imagination at the scale of a civilization.

This is also where the refrain "don't get above yourself" gets interesting. A young person says "I will build a great company", and the environment hears arrogance. The same person, twenty years later, actually runs a large organization, and gets to speak about how AI, energy and automation will change society — and everyone calls it vision. So which is being punished: the arrogance, or the lack of proof? Watch how the words change around the same ambition. Before success: arrogant, unrealistic, delusional. After success: visionary, ambitious, world-changing. Society, it turns out, is often not against ambition at all. It is against unverified ambition. And "don't get above yourself" does have a statistical basis: most people who believe they are the exception are not the exception. But that fact cannot be stretched into the command "don't try to become the exception". The first is a statistic. The second is a social order, and the step between them is where the culture does its work.

Underneath all of this runs cynicism. I do not mean pessimism. Modern cynicism is something more specific: a person knows reality is cruel, and treats accepting that reality as maturity. "Your startup will most likely fail." "Your odds of becoming a top scientist are tiny." "You are, in the overwhelming majority of cases, an ordinary person." "The world is unfair." "Ideals cannot pay rent." Many of these statements are true. That is exactly what makes cynicism dangerous: it often describes reality correctly and then derives action wrongly.

Look at the logic closely. Most people fail. Therefore I should not try. Does the second sentence follow from the first? No. "Most people fail" is a statistical fact; it does not imply "therefore you should not try". Those are different claims, and the second smuggles itself in as if the first had proven it. Statistics describe the majority; they do not dictate the destiny of an individual. And the fact that most attempts fail does not imply that attempting is irrational.

There is a deeper conflict here, between statistical rationality and individual rationality. If a person always decides their actions by aggregate probabilities, they will almost certainly never become an exception of any kind, because every exception, before it became an exception, had to take some action that was not the optimal strategy for the population average. The person who becomes world-class did something the statistics told them not to do. Applied to the individual, the logic of the aggregate is circular: it produces exactly the world it predicts. A civilization in which everyone adopts the safest available strategy produces no paradigm shifts; it only refines the ones it already has. Statistics can tell you the risk. It cannot decide the value for you.

But I want to say this without any chicken soup. Most startups do fail. Most research attempts never become breakthroughs. Luck matters enormously. Ability differences are real. Resource differences are real. Self-belief does not change the objective probabilities. Anti-cynicism is not the belief that you will win. It is knowing you might fail, and still refusing to delete every non-necessary possibility in advance because the failure probability is high. Camus stared at the same absurdity and refused to hand it his resignation; one must imagine Sisyphus happy, he wrote, and that sentence is the exact opposite of cynicism.

This is where the logic gets personal. A young person says: "I want to start a company." "I want to do world-class research." "I want to make art." "I want to change an industry." "I want to create something that never existed." And the environment answers instantly, with statistics: "Who do you think you are?" "Do you know how many people have tried this?" "Do you know the success rate?" "Get a job first." "Stop dreaming." The refrain has a name: 不要自命不凡, don't get above yourself.

There is a subtle logical error in this: using the low success rate to negate the attempt itself. "Most people fail" is not the same proposition as "this particular person will fail". "Many people have already done this" is not the same as "you cannot be the first to complete a new combination". And "becoming a world-class figure is hard" does not contradict "creating a world first in a locally defined sense can be surprisingly easy".

The numbers are stranger than the refrain admits. Climbing Everest requires very strict permits, enormous costs, and many failed attempts — yet the number of people who have stood on the summit still exceeds 11,000. Athletes, directors, game producers, esports players, stars, singers, entrepreneurs, mathematicians, politicians, physicists, other scientists, musicians, painters, artists... counted together, they form a huge number. If you define "world first" in a narrow enough way, being world first is not that hard: the world's first person to do tensor computation with an audio equalizer; the world's first person to train JEPA in Minecraft — then you just go and do it. The examples sound like jokes, but the proposition underneath them is serious. The space of human knowledge is indefinitely large. Define a specific enough problem, and new frontiers keep appearing. Being world first in a tiny domain is not as hard as it sounds; the hard part is finding the question. Creation is not climbing an existing leaderboard. It is building new axes.

And still the tension must be kept, not dissolved. Most people will not become exceptions. Realism can save a person's life, and realism can cage a person's life. Idealism built civilizations, and idealism manufactures hallucinations. There is the tension between individual will and objective statistics: sometimes, to do something at all, you have no choice but to run on a placebo effect, because the cruelty of reality can scare you out of doing it. Self-belief beyond the statistics is what lets you move. But the placebo effect must not be allowed to blind you to the risk. Respect reality. Just don't worship it.

None of this is an argument for any particular alternative, and this is where essays like this usually collapse, so let me be careful. Capital accumulation has driven enormous technological progress. Entrepreneurs have created real value. Markets are unmatched at allocating certain kinds of resources. The systems history tried instead had serious problems of their own. UBI, socialism, automation-funded societies: every proposal carries real difficulties, and technology never automatically produces fair institutions. Productivity and distribution are two different problems, and solving the first says nothing about the second. The other -isms beyond capitalism have problems of their own; none of today's systems is the final answer; perhaps these problems will only be solved by future advances in social science and technology.

So my position is not "capitalism is bad". It is narrower, and I think harder to dismiss: any system should be judged on a single scale — whether it converts productive power into human freedom. GDP measures how much is produced. The question it never answers is what production is for, and who is freed by it. On that scale the current system earns a mixed verdict, genuinely mixed and not rhetorically mixed. It produces freedom for a few and survival for the many, and the ratio between the two has become the central political question of this century, whether or not we admit we are asking it.

I have been careful so far to aim my criticism at one thing only: any arrangement that fails to convert productive power into human freedom. It would be easy, and wrong, to read that as a coded defense of some other system, so let me say directly what I think about socialism and communism, because the topic deserves more than a sentence.

Start with what socialism is actually for, because it is easy to caricature and I do not want to do that. Socialism is a broad family of ideas about ownership, redistribution, welfare and democratic control of production, and several of its core goals are goals I have already defended in this essay. Social protection: the attempt to partially separate a person's basic needs from their market income. Universal access to education, healthcare and housing when institutions are designed to provide them. Collective risk-sharing, so that one individual does not bear the entire consequences of unemployment, illness, disability, an economic downturn, or a failed experiment at starting a company. The capacity to mobilize resources toward projects with long horizons and no immediate return — exactly the kind of project the market systematically underprices. The prevention of extreme concentrations of private economic power. The idea that some goods are so tied to human dignity that access to them should not depend entirely on purchasing power. And the possibility that greater economic security gives individuals more freedom to study, create, conduct research, change careers, or start companies.

None of these goals requires socialism. Capitalist welfare states, public institutions, taxation, social insurance, cooperatives, public ownership and mixed economies pursue many of the same objectives, and I say this explicitly so that nobody reads this section as "capitalism bad, socialism good". The interesting question is not which label wins. It is which institutional arrangements actually deliver these goods.

Because socialism, examined honestly, also has structural problems of its own, and they are not moral failings of the people who support it. They are institutional. The first is information. Economic knowledge — what people want, what things cost to make, what is possible where — is distributed across millions of individuals and organizations, and no central office can assemble it. The second is incentives: when rewards become weakly connected to contribution, innovation, productivity, risk-taking and responsibility can all decay, without anyone intending it. The third is bureaucracy: organizations created to coordinate resources develop their own interests and drift away from the purpose that created them. The fourth is the soft budget constraint: an organization that keeps receiving resources despite poor performance loses the pressure to improve or to exit. The fifth is the concentration of decision-making: when economic ownership and political authority become highly centralized, economic mistakes become political problems, and whoever controls the plan controls more than the plan. The sixth is the hardest: deciding what should be produced, in what quantity, and according to whose preferences — the very question markets answer automatically and imperfectly. And underneath all of these sits the tension between collective objectives and individual autonomy, plus the uncomfortable fact that replacing private owners of capital with a centralized administrative class does not necessarily eliminate hierarchy. It can merely change who sits at the top of it.

None of this makes socialists stupid, immoral or economically ignorant. It makes central planning difficult, and difficulty is not a moral verdict.

One more distinction gets lost in ordinary argument. Socialism and communism are not interchangeable labels. Socialism, in the sense I am using it, is the broad family: social ownership, redistribution, welfare, democratic control of production. Communism in the classical theoretical sense is a far more radical endpoint: common ownership, the disappearance of class divisions, and, in Marxist theory, the withering away of the state as a coercive political structure. And theoretical communism is not the same thing as the historical political systems that called themselves communist. I will not judge an abstract system only by its idealized description, and I will not judge every theoretical idea only by its worst historical implementation. The gap between institutional ideals and institutional reality is one of the subjects of this essay — it is the same gap as the one between the future we were promised and the future we mock.

That brings the question back to where this essay began. When resources are scarce, telling a person "first find a stable job; survival comes before self-actualization" is completely rational. But suppose AI, robotics, automation, energy technology and advanced manufacturing eventually make many forms of necessary labor extremely cheap. Then the question changes. If a civilization can produce abundant food, housing, transportation, computation and healthcare with relatively little human labor, should access to those things still depend primarily on an individual's ability to sell labor on the market?

This is where the old capitalism-versus-socialism debate becomes interesting again, and it stops being about which ideology is morally superior. It becomes: which institutional arrangements allow technological productivity to translate into greater human freedom?

And two distinctions from earlier must stay sharp. Productivity is not distribution: a society can become extraordinarily productive while the gains from productivity remain highly concentrated. Automation is not liberation: automation can mean humans work less, or work more efficiently, or become unemployed, or watch profits concentrate, or see basic goods get cheaper, or see new forms of employment emerge — or some combination of all of these. Technology changes the feasible set. Institutions determine how much of that feasible set becomes available to ordinary people. So I will not make the naive argument that AI will solve inequality automatically. It will not.

This is also where the survival-first realism from earlier returns to view. The argument is not that people should ignore employment and chase dreams. It is that survival rationality is rational under scarcity, and it should not automatically become a theory of what human life is for. A society may reasonably tell a young person "you need an income". But it should also be possible to ask the larger civilizational question: why does access to a meaningful life remain conditional on how successfully you sell your labor? That question sits at the intersection of capitalism, socialism, welfare states, UBI, automation and whatever comes after scarcity — and none of those systems currently answers it well.

Here the three rationalities from earlier line up again: survival rationality asks how I stay economically secure; instrumental rationality asks which action maximizes my measurable outcome; value rationality asks what is actually worth doing. Modern economic systems are extremely good at answering the first two questions. They are much less capable of answering the third, and an increasingly productive civilization should eventually give individuals more room to answer the third for themselves. This is not an argument that everyone must become an entrepreneur, an artist, a scientist, or any other kind of extraordinary person. It is the opposite: a good civilization should not require anyone to become extraordinary merely to obtain basic autonomy. An ordinary person should be allowed to remain ordinary without their life being counted as a failure.

Markets are powerful coordination mechanisms, and they can also produce inequality, externalities, monopolies, and situations where purchasing power decides access to necessities. Central planning can mobilize resources and pursue collective objectives, and it faces information, incentive, bureaucratic and power-concentration problems. Welfare states, public ownership, cooperatives, markets, regulation, taxation and social insurance can be combined in many ways. Institutions are technologies, with trade-offs like any other technology. The interesting question is what institutional architecture works best at a given level of technological productivity, information availability, and social complexity.

And the test remains the one I proposed earlier. A civilization should not measure progress only by GDP, corporate profits, stock-market capitalization, employment rates, or productivity per worker. It should also ask: how much freedom does each additional unit of technological productivity give to an ordinary human being? If a hundred robots replace a hundred workers, the most interesting question is not how much GDP increased. It is what happened to the time, autonomy, security, and possibilities of those humans. The ultimate goal of increasing productivity may not be to maximize output at all. It may be to maximize the number of people who no longer have to sell their lives in order to live — and then to leave them alone to decide what living is for.

Why do the old narratives hold on so long after their material base has started to move? Because they are not purely economic. They are built on human nature. On the question of "worldly wisdom" (人情世故): it is actually built on human nature. Over long evolution, humans evolved emotional needs to cope with survival — because of Dunbar's number, our emotional bandwidth is limited to a small circle, so societies end up running on trust, status, and ritual rather than pure reason. 人情世故 is exactly that: an operating system built on top of human nature. And an operating system is not replaced the moment a better one becomes theoretically available; it gets patched. The survival narrative, the ROI logic, the cynicism: they persist not because they are true, but because they are load-bearing. They manage fear, status, family obligations, the need to feel safe. A civilization changes its productive base first and its stories last. The lag is not a bug. It is the emotional firmware of the system. Humanity now has more power to change the world than at any point in its history, and our social psychology may still be living in an age of scarcity.

One correction against the loneliness of all this. There is actually no one who truly changes the world by the strength of one person alone. Civilization is built by many like-minded people. Those entrepreneurs may claim to be "changing the world," but Musk's plans were actually proposed by many people before him — and if you really want to be pedantic, even Einstein's ideas stood on the shoulders of countless predecessors. The person who refuses to be a statistical average does not become a solitary hero. Exceptions are made of networks, institutions and accidents as much as of will.

As Liu Cixin said, all soaring thoughts will come crashing down under the gravity of reality. I don't read that line as surrender. It is an interface problem: between the ideal and reality there is an interface, and it must be optimized and reconciled, over and over — hold onto the ideal, respect the reality. Most attempts will fail. That is exactly why the few who try matter, and exactly why "don't get above yourself" is such a catastrophic piece of advice: it optimizes the survival of individuals, and taken to its limit it guarantees the stagnation of the species. Humanity may be sitting on the Great Filter, but the bugs have never been wiped out — perhaps a black swan event will lead humanity step by step out of the swamp.

I know reality is cruel. I know most people fail. I know resources are distributed extremely unequally. I know ability and luck matter. I know "survive first" is often the correct advice. I know Homejoy proves that ideals must submit to reality's tests. I know cynicism is attractive because it is so often statistically correct. I know placebo-style confidence cannot change the probabilities.

But reality is not an oracle. Probability is not fate. A survival strategy is not the meaning of a life. Employment is not the purpose of a civilization. Growing productivity should not only mean working more efficiently; it should mean that fewer and fewer people must work merely to live. A civilization should ultimately be judged by whether it gives human beings more and more ability to decide what they live for.

Respect reality, but don't worship it. Reality should constrain our plans, not define the full boundary of our imagination. Statistics describe the majority; they do not dictate the destiny of an individual. And no one should have to become extraordinary just to deserve the right to choose their own life.