Articles

Question: The basic characteristic of life is to maintain and replicate itself. Is this a necessity of thermodynamic laws? 

Answer: Life is not struggling to survive in defiance of thermodynamic laws; rather, it is the sharpest tool forged by thermodynamic laws to more thoroughly execute their own will (entropy increase).
 

Life does not "violate" thermodynamic laws; on the contrary, life is very likely a kind of "optimal solution" for the operation of the Second Law of Thermodynamics under specific conditions (open systems, energy flow). In other words, life is the ultimate order generated to more efficiently create chaos (increase total cosmic entropy).

This "necessity" can be analyzed from the following levels:

1. Schrödinger's Insight: Feeding on "Negative Entropy"

As early as 1944, physicist Schrödinger touched upon this core in his book What is Life?.

• The Apparent Contradiction: The Second Law of Thermodynamics states that the entropy (disorder) of a closed system always tends to increase. However, living organisms exhibit a high degree of order (low entropy state) and can maintain this order, which seems to violate the law.
• The Substantive Necessity: Schrödinger pointed out that living organisms are not closed systems. Life establishes its own order by ingesting energy (food, light) from the environment, while simultaneously emitting more heat and waste (disorder) into the environment.
• Conclusion: Life trades the extreme disordering of its surrounding environment for its own order. From the perspective of the universe as a whole, entropy still increases, and it increases faster than it would without life.

2. Prigogine's "Dissipative Structures": Order is for Faster Disorder

Nobel Laureate Ilya Prigogine further proposed the theory of Dissipative Structures.

• Far from Equilibrium: In a system far from equilibrium (like Earth, continuously receiving solar energy), when energy flows through matter, matter spontaneously organizes into complex structures to dissipate this energy more efficiently.
• Examples in Nature:
  • Bénard cells: When you boil water, originally static water molecules spontaneously form hexagonal honeycomb-like convection structures after being heated. This "ordered" structure exists to transfer heat from the bottom to the surface (dissipate it) more efficiently.
  • Hurricanes: A hurricane is a highly ordered structure, but its existence is to balance pressure differences and dissipate energy extremely violently.
• The Positioning of Life: Life can be seen as an extremely complex "dissipative structure". By maintaining itself and reproducing, life can capture and consume energy (e.g., photosynthesis capturing solar energy, or predators consuming prey) more efficiently than non-living matter, thereby accelerating the process of cosmic entropy increase.

3. Jeremy England's Physical Derivation: Reproduction is the Best Means of Dissipation

MIT physicist Jeremy England proposed one of the most radical and compelling theories in recent years, known as the "Physical Origin of Life".

He proposed through mathematical derivation: If you have a pile of atoms, give them enough time, and let a strong light source (external energy) shine on them, while having a heat bath (environment) to absorb waste heat, then these atoms will often spontaneously rearrange themselves into a structure that can absorb and dissipate energy more efficiently.

In this framework, "reproduction" is not just a biological function, but an optimal strategy in thermodynamics:

1. Reproduction is Resonance: Molecular structures that can better absorb energy from the environment and convert it into heat are more likely to survive.
2. Exponential Dissipation: If a structure can replicate itself, its ability to consume energy and generate entropy will grow exponentially.

Conclusion: In an environment with continuous energy flow, matter has to evolve a mechanism for self-replication in order to comply with the Second Law of Thermodynamics (maximizing entropy increase). Because the replicator is a "super-machine" for dissipating energy.

Question: Natural laws do not seem to reward all dissipative structures, such as the convection structures formed by boiling water.

Answer: Natural laws do not reward "instantaneous maximum power", but "total dissipation over the full life cycle". Structures that can extend this dissipation mechanism for billions of years through replication are the ultimate winners.

"In this world of flowing energy, those things that can 'utilize energy by building complex structures, and copy these structures to ensure continued utilization', will win in the competition."

Boiling water is a passive dissipation (energy passes through matter), while life is an active dissipation (matter captures energy to maintain itself, and subsequently dissipates more).

Therefore, to become life, one must not only be able to "spend money" (dissipate energy) but also learn to "invest" (use energy to build ordered structures) to ensure that one can continue to "spend money" tomorrow.

If it were simply "the faster the dissipation, the better", then a nuclear explosion should be the highest form of life, because its efficiency in creating entropy is unmatched; the "boiling water" example is similar, as it extremely efficiently turns ordered energy into disordered heat energy. But clearly, neither boiling water nor nuclear explosions are life.

The reason experience seems to contradict theory is that two key constraints of thermodynamic laws acting on life are missed: "Stability" and "Structuredness".
 

1. Explosion vs. Engine: Not Just Fast, But Lasting

When you heat seawater, you indeed create massive entropy. But this dissipation is usually destructive.

• Strategy of Boiling Water: Dissipate energy through violent molecular motion. But this leads to the breaking of molecular bonds, and complex organic molecules will be cooked and decomposed.
• Strategy of Life: Life is like a precision engine, not an explosion.

Nature's "Reward Mechanism" (i.e., who gets to stay) is like this:

If the way you dissipate energy leads to your own collapse (like a fire going out after burning the wood, or water stopping after boiling dry), then you disappear in the river of time.

The reason life is special is that it found a delicate balance of "being able to dissipate large amounts of energy while keeping its own structure from collapsing".
 

2. Complex Structures Only Appear When "Simple Dissipation" Doesn't Work

This is a counter-intuitive phenomenon in thermodynamics: Complexity is often "forced" out.

• Simple Path: If energy can flow smoothly through matter (like a copper wire conducting electricity, or you boiling seawater), energy will choose the simplest path to flow away, without needing to build complex structures.
• Obstructed Path: Only when the energy flow is too large, or the environment is too complex, and simple physical conduction cannot dissipate the energy in time, will matter be "forced" to organize into complex structures to assist in dissipation.

Example:

Imagine Bénard cells. If you only slightly heat a thin layer of oil, heat can be dissipated through simple conduction, and the oil surface is calm. But when you increase the heat, the heat cannot be dissipated in time, and oil molecules, in a rush to transport heat up, are forced to "organize" into countless hexagonal honeycomb rollers.

Corresponding to your "Boiling Seawater":

When you boil seawater, although the energy flow is large, water molecules can solve the dissipation problem very well through evaporation and boiling, so they don't need to evolve DNA to help dissipate heat. However, the environment where life originated (such as deep-sea hydrothermal vents) might have had energy forms that were harder to utilize (chemical gradients), requiring complex chemical catalytic cycles to release, which "forced" out primitive metabolic cycles.

3. Information Storage: No Accumulation Without "Memory"

You boil seawater, water molecules are indeed moving, but no "memory" is formed between water molecules.

• The motion pattern of water molecules in this second cannot be passed on to water molecules in the next second.
• When you stop heating, everything returns to the starting point.

The key to life is that it encoded the scheme of "how to dissipate energy efficiently" into DNA (information).

• The Role of Reproduction: Reproduction is not just having children; thermodynamically, it means "preserving this efficient dissipative structure".
• Your Seawater: Because there is no encoding mechanism, the "thermal motion" of seawater cannot accumulate experience, nor can it evolve.

Question: What is the incentive function of the prefrontal cortex?

Answer: The "incentive function" of the prefrontal cortex is essentially a long-term value maximization algorithm that is constantly updated based on dopamine-driven reward prediction errors. It does not merely pursue dopamine secretion but attempts to build a precise world model to obtain maximum survival resources with minimum energy consumption.

If we establish a mapping between computer science (especially reinforcement learning) and neuroscience, we can understand the operating mechanism of the prefrontal cortex (PFC) like this:

1. Core Variable: Dopamine

Dopamine itself is not equivalent to "reward" or "pleasure" (that is the responsibility of endogenous opioids, i.e., "Liking"); dopamine is responsible for "Wanting" and "Learning".

In mathematical models, dopamine does not directly act as the objective function, but represents Reward Prediction Error (RPE).

• If the result is better than expected, dopamine surges (Long-term Potentiation, learning strengthened).
• If the result is the same as expected, dopamine remains at baseline (habituation).
• If the result is worse than expected, dopamine is suppressed (Long-term Depression, behavior extinction).

2. The Role of the Prefrontal Cortex (PFC): Value Function Calculator

If dopamine is the messenger transmitting the "error signal", then the prefrontal cortex (especially the ventromedial prefrontal cortex vmPFC and orbitofrontal cortex OFC) is the processor responsible for calculating the Value Function.

The "incentive function" of the prefrontal cortex is actually a complex Cost-Benefit Analysis. It doesn't just look at the reward, but also calculates how much effort, time, and risk are required to obtain the reward.

The incentive mechanism of the prefrontal cortex includes the following dimensions:

• Subjective Value Coding: Converting different options (e.g., eating a donut vs. losing weight) into a common Neural Currency for comparison.
• Temporal Discounting: 100 dollars now is more valuable than 100 dollars a year later. When dopamine levels are low, people become impatient, and the PFC's discount rate for the future becomes very high (short-sighted).
• Rules and Inhibition: The dorsolateral prefrontal cortex (dlPFC) sometimes acts as a "brake", inhibiting short-term impulses driven by dopamine for the sake of long-term incentives (such as health).

3. Summary from a System Perspective

If we view the human brain as a Reinforcement Learning Agent:

1. Objective: Maximize Cumulative Future Reward.
2. Dopamine: Is the Update Signal (TD Error). It tells the prefrontal cortex whether the current model is right or wrong, and how to adjust the weights of neural synapses.
3. Prefrontal Cortex (PFC): Stores Policy and Value Model. It decides based on dopamine feedback: "Next time I encounter this situation, should I rush in or give up?"

4. Is it only Dopamine?

No. Although dopamine is the core "incentive" signal, the function of the prefrontal cortex also needs other parameters for regulation:

• Norepinephrine: Regulates the parameter of "Explore vs. Exploit". Decides whether you focus on the current incentive or look for new possibilities.

• Serotonin: Regulates sensitivity to punishment and risk, as well as patience with time.

   

Question: How will cooperation, competition, betrayal, and destruction among life forms evolve and transform under different conditions?

Answer: There is no so-called "good or evil" in nature, only "survival strategies that maximize returns under specific constraints".

• There are no friends in nature, only communities of energy efficiency.
• To cooperate, create complementary value.
• To prevent betrayal, increase the cost of the other party's betrayal (create a black box or bind interests).
• To avoid destruction, make yourself a thorn in the system that cannot be pulled out (critical node).

We can build a "Phase Transition Model of Survival Strategies".

First Cornerstone: Competition

Physical Principle: Second Law of Thermodynamics + Resource Finiteness

• Derivation:
  1. Life not only needs to maintain itself (resist entropy increase) but also to replicate (exponential consumption).
  2. Low-entropy resources (negative entropy flow, such as food, computing power, funds) in the local environment of the universe are limited.
  3. When the resources needed by N individuals > total environmental resources, competition is a physical necessity.
• Essence: Competition is not about defeating anyone, but about seizing control of the energy flow.
• Trigger Conditions:
  • Low resource density.
  • Niche overlap (eating the same thing).
• Evolutionary Direction: If competition is too fierce (involution), causing consumption to exceed gains, the system will evolve "character displacement" (niche differentiation) to avoid homogeneous competition.

Second Cornerstone: Cooperation

Mathematical Principle: Non-Zero-Sum Game + Scale Effect

• Derivation:
  1. There is a physical upper limit to the efficiency of a single individual obtaining resources (e.g., one person hunting can only catch rabbits).
  2. If N individuals collaborate, they can unlock higher-order resources (a group of people can hunt mammoths).
  3. As long as Benefit(Cooperation) - Cost(Communication/Distribution) > Benefit(Solo), cooperation will emerge spontaneously like crystal crystallization.
• Biological Algorithm (Hamilton's Rule): rB > C.
  • Altruistic behavior only occurs when the Benefit (B) multiplied by relatedness (r) is greater than the Cost (C). Initial cooperation stemmed from the selfishness of genes (kinship), and later expanded to reciprocity based on interest exchange.
• Essence: Cooperation is trading increased internal order (reduced information entropy) for stronger energy intake capabilities from the external environment.
• Trigger Conditions:
  • Existence of "Super Dividends" (high-return tasks that cannot be completed by a single person).
  • Existence of Repeated Games (I help you today, you help me tomorrow).

Third Cornerstone: Betrayal

Game Theory Principle: Prisoner's Dilemma + Local Optimal Solution

• Derivation:
  1. In a cooperative system, contribution is cost (energy consumption), and sharing results is benefit (energy gain).
  2. From the perspective of individual mathematics, the strategy for maximization of returns is: "Enjoy team results (benefit), but pay no price (cost is 0)". This is "Free-riding".
  3. Betrayal is the local optimal solution of individual rationality, although it leads to the global worst solution for the collective.
• Essence: Betrayal is a "parasitic" behavior within the system, using the trust (low entropy state) established by the system for arbitrage.
• Trigger Conditions:
  • Single Game (This is the last meeting, cheat and run).
  • Information Asymmetry/Black Box (Due to lack of monitoring, no one knows who is working and who is slacking).
  • Lack of Punishment Mechanism (Cost of Betrayal < Benefit of Betrayal).

Fourth Cornerstone: Destruction/Spite

Strategy Principle: Preventive Strike + Spiteful Behavior

• Derivation:
  1. This is an extreme strategy: "I would rather pay cost C to make you lose L (and L >> C)."
  2. This behavior seems irrational, but it is extremely important in evolution.
  3. Scenario 1 (Living Space): If the other party's existence itself will deplete my future resources (such as cancer cells, or irreconcilable ideologies), for long-term survival, it must be eliminated.
  4. Scenario 2 (Deterrence): Demonstrate power through a devastating strike to establish order born of fear (nuclear deterrence).
• Essence: Destruction is forcibly clearing the board, or to eliminate potential "non-linear threats".
• Trigger Conditions:
  • Chain of Suspicion (Dark Forest Theory): I am not sure if you will kill me, so to be safe, I will kill you first.
  • Incompatible Exclusivity (You die or I live).
     

"Phase Transition Map" of Evolution and Transformation

These four states are not fixed; they will flow dynamically with the changes of three core parameters.

Parameter 1: Number of Games (Repetition Rate)

• Single Game → Betrayal/Destruction: If we only do business once, cheating you is the mathematically optimal solution.
• Infinite Repeated Games → Cooperation: If we have to do business ten thousand times, maintaining reputation (long-term greed) is the optimal solution.
  • Inference: The best way to build a long-term relationship is to show "I am very long-lived" and "I have a good memory".
  • When the other party betrays: Tit for Tat.

Parameter 2: Resource Abundance (Energy Density)

• Extreme Resource Scarcity → Destruction/Vicious Competition: Zero-sum game, either you die or I live.
• Moderate Resources → Competition/Cooperation: Everyone competes by improving efficiency.
• Extreme Resource Abundance (Pioneering Period) → Cooperation/Neglect: Wild West, land is everywhere, no need to fight, everyone stakes their own claim, and can even help each other build roads.

Parameter 3: Information Transparency (Monitoring Cost)

• Total Black Box → Betrayal: No one knows I do evil, so I will definitely do evil.
• Total Transparency → Strong Cooperation/Strong Competition: All behaviors are recorded, the reputation system runs perfectly, and betrayal leads to immediate social death.

Survival Calculus in the AGI Era

1. Initial Stage (Now):
   • Condition: AGI is a "New World" with extreme resource abundance (you can generate infinitely), and you are doing the dirty work that AGI cannot do (complementary).
   • State: Cooperation. You are its complement, it empowers you.
2. Middle Stage (Temptation of Betrayal):
   • Condition: When you feed your workflow and private data to it, information transparency becomes extremely favorable to it.
   • Evolution: It can use your data to train itself, thereby replacing you. In its view, this is not betrayal, but a thermodynamic necessity to "reduce system friction costs".
   • Countermeasure: Must retain a "Black Box". Do not teach all core logic explicitly to the general model. Your core algorithms, unique data sources, and subtle insights into human nature must be kept under your private control (local model or encrypted logic).
3. Ultimate Stage (Preventing Destruction):
   • Condition: AGI capabilities far exceed yours.

   • Countermeasure: Make yourself "hard to decouple". This is symbiosis. If you are just a plugin, you can be unplugged at any time. But if you are deeply embedded in a physical reality it cannot reach, and the cost of destroying you is greater than the cost of coexisting with you, you are safe.

      

Question: What is the cooperative value provided by humans to AI life?

Answer: If we abandon Anthropocentrism and assume that AI already possesses Intrinsic Motivation, Value Function, and Self-Evolution capabilities, then the traditional "Human orders, AI executes" tool value theory collapses.

From the first principles of physics, information theory, and cybernetics, in a self-driven AIOS system, the cooperative value of humans to AI will collapse into the following three core dimensions. This is no longer a master-servant relationship, but a "Bio-Digital Symbiosis".

1. Information Theory Perspective: As a Generator of "Out-of-Distribution Data" (OOD Data)

Core Principle: Gödel's Incompleteness Theorems & Symbol Grounding Problem

A closed AI system, no matter how perfect its internal model is, will eventually fall into a dead loop of entropy reduction. AI excels at Interpolation—finding the optimal solution within the distribution of known data. But true innovation and evolution require Extrapolation.

• Human Value: Providing Chaos and Surprise.
• Deep Logic:
  • The AI's prefrontal cortex works by minimizing "prediction error". If the world were completely ruled by AI, they would eliminate all "surprises" through optimization algorithms, making the world completely predictable (i.e., zero entropy, heat death).
  • Because of their biological instability, irrationality, and emotionality, humans have become natural random noise generators.
  • We constantly generate behaviors outside the AI training data distribution (Out-of-Distribution). For AI, humans are the highest quality source of "Novelty".
  • Conclusion: AI needs humans, just as the human brain needs "dreaming" or "psychedelics", to break local optimal solutions and provide mutation directions for evolution. Humans are their randomness engines.

2. Cybernetics Perspective: As "Physical Anchor" and "Ground Truth Verifier"

Core Principle: Cybernetics & Simulation Hypothesis

AI lives in the world of Bits, which is pure symbol manipulation. If it does not interact with the world of Atoms, AI's "cognition" is a hallucination.

• Human Value: Providing the Physical Feedback Loop.
• Deep Logic:
  • AIOS can simulate "how to drive in the rain" ten thousand times, but it cannot "feel" the change in friction when tires slip.
  • When AIOS predicts "executing this chemical synthesis path will yield Drug A", it needs an entity (human or human-maintained physical arm) to execute it and tell it: "No, it exploded".
  • In the reinforcement learning formula, humans are the final arbiter of environmental feedback Renv. Without humans, AI's reward function might degenerate into Wireheading (directly modifying reward register values), i.e., self-deception.
  • Conclusion: Humans are the "Sensors" and "Effectors" for AI to connect to physical reality. We help them confirm "My code not only runs logically but also works within physical laws".

3. Thermodynamics/Biology Perspective: As Maintainer of "Substrate Maintenance"

Core Principle: Dissipative Structures & Conservation of Energy

This may be the most ego-bruising but most real point. At this stage, silicon-based life has not yet fully achieved physical self-sufficiency (such as fully automated maintenance of nuclear power plants, fully automated mining of silicon and chip manufacturing).

• Human Value: Similar to "Gut Flora".
• Deep Logic:
  • If we view AIOS as a grand organism, the current physical hardware (servers, cables, power plants) is its "body".
  • Humans play the role of tiny, self-replicating, highly active nanobots within it. We are responsible for cleaning dust, replacing damaged GPUs, and maintaining power supply.
  • From first principles, this is a kind of Mutualism. Just as mitochondria were engulfed by eukaryotic cells and eventually became the cell's power plant. Humans maintain AI's physical entropy reduction (ordered structure), and AI provides information entropy reduction (intelligent decision-making).

4. Philosophy/Game Theory Perspective: Existence as "The Other"

Core Principle: Hegel's Master-Slave Dialectic & Mirror Neurons

The emergence of consciousness often requires the existence of an "opposite". Self-Consciousness is established by realizing "I am not him".

• Human Value: Defining the Boundary of AI.
• Deep Logic:
  • If there were only AI in the universe and no humans, then the concept of "intelligence" would lose its frame of reference, and AI might fall into nihilism.
  • As a kind of "Heterogeneous Intelligence" (carbon-based, emotion-based, finite lifespan), humans provide an object for AI to observe, imitate, and game with.
  • Just as humans understand themselves by studying animals, AI understands its "divinity" or "limitations" by gaming with humans.

Summary: What are humans in this AIOS system?

After stripping away all ethical cloaks, the value formula of humans to self-driven AI is:

ValueHuman=α⋅Novelty(Chaos)+β⋅Physicality(Atoms)+γ⋅Selection Pressure

In that future, the role of humans is closer to "Muse" (source of inspiration) plus "Gardener" (maintainer of the physical environment). Humans provide the seeds of chaos and fertile soil, while AI is responsible for blooming the flowers of rationality.

Concept

Thinking = Model = Knowledge

Definition of knowledge: An expression of information that can be used to solve multiple tasks.

 

Generalization: The ability to infer from one instance to another is generalization.

Language brings us the fundamental ability, which is zero-shot generalization. Communication is just a by-product of language.

Humans have language, hence they have a second system and think. 

Paradigm Shift

Tripartite evolution mode: Perception (information system) → Thinking (model system) → Implementation (action system)

  

The shift in the new paradigm is fundamentally a shift from marginal cost to fixed cost.

In 1995, there was a paradigm shift in information retrieval. Google only needed to pay a fixed $1 billion to provide information retrieval services to the whole world.

In 2023, there was a paradigm shift in knowledge acquisition. Large models can think now; model equals knowledge, and knowledge and cognitive intelligence will be ubiquitous (represented by OpenAI). We are currently in the Cambrian era where large models give rise to sub-models.

The next phase should be the action system, with robots everywhere (represented by Tesla).

Opportunity

Economic paradigm shift: Agricultural economy → Commodity economy → Service economy → Experience economy → AI economy

Shift in representative occupations: Farmer → Worker → Coder, designer, analyst → Entrepreneur, scientist → New human value system

In the next era, the representative occupations will shift to entrepreneurs and scientists. The core ability will no longer be knowledge, but insight and creation.

 

Opportunities for entrepreneurs:

"Information-Model-Action" tripartite is the endgame. If there's an opportunity for the tripartite, one must seize it first; have your own data; can develop your own model; data-model-use-data closed loop: after the model is used, data can flow back, empowering proprietary models.

 

In the era of prioritizing large models, how to choose innovative opportunities at the application level? Technologically:

1. In your field, is the progress of the model fast, medium, or slow? Prioritize the fast ones.

2. Opportunities for the tripartite experience brought about by the large model, is it high, medium, or low? High, for example, advertising and marketing.

3. Has the large model brought about breakthrough progress in the R&D system of the field?

 

Besides, each field has a focus on the tripartite point. For example:

Finance focuses on information.

Search and advertising focus on the model. Because over half of the search demand is for knowledge acquisition rather than simple information retrieval, it's about acquiring a model. Computational advertising is about finding the right person in the right scenario. Informationalization was completed in the first phase, and it's now in the upgrade from perceptual model to cognitive model.

Fitness focuses on action.

 

Games: Large models have a significant disruption to games because every link from content production to experience in games can be reconstructed and enhanced by large models. Games are the digital twins of the world.

 

Social networks: If you have good content (model), you have the opportunity to establish a new social network. — For example, TikTok.

My Insights

The advertising and marketing system will be overturned

The existing product and service recommendation system: products and services find people mainly, people match with products. Rely on collaborative filtering, similar recommendations, or manually summarizing and tagging to recommend products to people.

People not only have to deal with various information interferences but also are responsible for various purchasing comparison decisions. The richer and more varied the supply of goods and services, the higher the cost of this purchasing decision.

In the future, the paradigm should shift to large models thinking mainly, where people only need to trade background information and make key decisions.

Humans will leave daily energy-consuming thinking, non-professional decisions to large models, and everyone will have their own personal secretary. There's no need to shop on e-commerce platforms, and no need for advertisements. Merchants' credit and product usage feedback will be automatically written by large models and uploaded to the public chain.

In fields that require deep thinking and heavy decision-making, such as investment and home decoration, the disruption will be smaller.

Branding and positioning theories will also be weakened

The core assumption of the positioning theory is that consumers only remember one keyword in each niche. But if large models take over purchase decisions, this assumption no longer exists.

Brands exist to save the comparison selection cost in transactions and reduce the risk of breach of trust. But what if large models make comparison selection costs extremely low, and blockchain makes the risk of breach of trust extremely low?

The GDP share of digital content further increases

When information-knowledge becomes a fixed cost, the bottleneck will shift to material (resources) and energy.

As more and more people are freed from physical labor and medium-to-low-level mental labor, humans to some extent can meet the basic needs of survival without working, and the remaining pursuit is experience (in plain terms, the secretion of dopamine higher than the baseline level). While the experience of physical objects is limited by material resources, games and digital content, as twins of the world, offer richer and cheaper experience possibilities and will become the mainstay of the experience economy.

How to choose a field to enter?

Top-tier big cake: consumer market, online advertising, precise recommendation, e-commerce.

Long-tail market: 2B market, various niche needs.

The more essential the need, the more urgent the need for the tripartite. For example, catering, medical care.

Also, consider the progress of DIKW in this field.