The Corporate-Government Parallel

How Corporations Work

In a stock market, public companies’ stocks and bonds are traded. Investors seek the best parking lot for their money. And they do it by picking good companies. How to pick one? By evaluating the company by certain criteria: revenue, profit, net margin, YoY growth, P/E (profit/earnings multiplier in terms of the valuation).

The main brain of the company is the CEO. Behind the CEO, there is a board of directors. The board chooses him, so he serves for the good of shareholders. If the CEO is doing badly, the board replaces him with another one. Quite efficient approach. Look at CEOs of big corporations, Microsoft, Google, Apple. They run by efficient managers. And if they’re not, they will be replaced by others.

So, how do they measure efficiency? How do they know that a CEO is either efficient or not?

Efficiency

If the CEO doesn’t bring enough value to shareholders, he’s getting fired. End of story. But long before that, shareholders ask a simple question:

“Did the CEO make us richer?” This question can be deconstructed into several:

1) Did market cap grow?

2) Did revenue & profit grow?

3) How efficiently is capital used?

And other questions, like market share, hiring, culture, etc. We won’t go deeper.

Money is always at the core of any business. It’s about moving money to the right place at the right time.

I thought... It’s similar to how the government works. You pay salaries from taxes to the board and the CEO, and fund projects. Why not apply these principles to them and see the flow? Hm, let’s apply these corporate principles to government, that flaws become evident:

The Government Contrast

Now, let’s replace general shareholders with citizens, CEO with president, and the board of directors with a congress or parliament, and reevaluate everything.

In theory, citizens are the primary shareholders of their government, with Congress/Parliament serving as the management team behind the president (CEO). Ministers can be called a board of directors. Compartment starts and ends right now. Problems begin.

Evaluating government efficiency will lead to disappointment and confusion. But that’s how it is.

1. Measuring Government Efficiency

Public companies have clear performance metrics, and governments have theirs: GDP, poverty rates, income inequality, and unemployment figures. However, unlike corporate metrics that drive immediate action, government metrics rarely lead to changes in leadership or policy.

Why is that? Money drives politics, like business. Citizens as taxpayers pay politicians' salaries, public services, major projects. All this money comes from citizens to improve their lives, not politicians.

The fundamental difference emerges in how "returns on investment" are measured and distributed. In a corporation, shareholders invest capital and expect dividends or stock appreciation in return. The performance metrics are clear, standardised, and regularly reported.

In government, citizens invest tax dollars but receive no direct "dividend statement" showing their return. Instead of quantifiable returns, citizens receive abstract promises of "public goods" that are difficult to measure. How do you quantify the value of a new highway or defense spending to each individual taxpayer? The "ROI" becomes impossibly vague.

Even worse, while corporations must account for every dollar spent, governments maintain "classified" budgets and vague spending categories. Imagine telling corporate shareholders that billions of their investment dollars went to "miscellaneous expenses" that cannot be disclosed! Yet citizens accept this level of opacity as normal.

In conclusion, you can’t influence money politics (business efficiency). It’s run by whoever men in suits.

Back to corps for a sec. Try to overspend some funds in the company. You’ll go to jail. Do something stupid in the government = you’ll live peacefully. So it’s a game with low personal stake. We shouldn’t react like it’s normal setup. It’s wrong.

To make you feel even weirder, I’ll dive into the accountability issue, when the difference will become even clearer.

The Accountability Gap

Imagine you're a shareholder in two organizations. In the first one - a corporation - you notice the CEO making poor decisions that hurt your investment. What happens? The board steps in, and if things don't improve, that CEO is gone within months. Your power as a shareholder matters.

Now consider the second organization - your government. You're still technically a shareholder (a citizen), and you're still providing the money (through taxes). The politicians, like CEOs, receive salaries and manage budgets with your money. But here's where the parallel breaks down spectacularly: when these 'political CEOs' fail, can you fire them? In 90% of cases, you can't.

Think about that disconnect. A corporate CEO mismanages funds, they're out. A politician mismanages an entire country's budget? It becomes an abstract "government issue" - nobody's personal failure. They continue in their position, spending your money, making decisions that affect your life, all while facing minimal personal risk.

The most frustrating part? You chose this president, this political 'CEO', just as shareholders approve their corporate leaders. But unlike shareholders, when your chosen leader loses credibility or demonstrates incompetence, you're stuck. There's no board meeting to call, no vote to organize. Your power as a citizen-shareholder vanishes the moment after the election.

Okay, no power, no accountability, maybe they’re telling you what they do at their work? What about transparency? I want to know where my tax $ went. Lmao, I’m so naive.

Transparency Theater

As public companies are obligated to issue public reports, your government has the right to hide anything they want from you and not stand for spending numbers. Every cent of corporate spending is tracked in quarterly reports, while billions in government spending can vanish into "classified" categories.

This lack of transparency extends beyond just financial reporting. While corporate shareholders have clear mechanisms for exercising their power, citizens face a different reality. Think of public protests and strikes - the ultimate expression of citizen disapproval. Yet strikers have no real power, because it's up to the people in charge to listen or not. Democracy functions until people need to actually exercise their power - then suddenly, that power vanishes.

But perhaps the most alarming contrast between corporate and government structures isn't about accountability or transparency, but about the fundamental quality of leadership itself. At least with corporations, there's a baseline expectation of competence. The government? Not so much.

The Competency Crisis

While technology enhancement drives progress in the private sector, government technology seems designed to suppress rather than serve. The root of this problem? Competency and continuous learning lives, gov is tech which suppresses life.

Management in corporations is constantly studying something. They are sent to expensive workshops, courses, mentorship, consultancy etc. Because it’s important that managers “live in the future” and know what to do next.

While politicians have the power of passing laws on topics, where they have absolutely no clue of what’s going on and what effect it will have. And nobody pushes them or punishes for such behavior. This accepted incompetence at such powerful levels is staggering.

Okay, we’re on the same foot now, but I think it wouldn’t be fair to show only the beneficial side of such an approach. It has its own limits and issues.

The Limits of Corporate Comparison

Decisions take long to apply and show results.

In a private or public company, when business decision is made, you can see the results in the next financial report right away. Either your revenue drops, profit goes to losses, net margin efficiency changes, or the opposite positive direction. Money flows extremely fast.

When you talk about decision-making in the government, you mean reforms of any kind. Economical, political, etc. The reforms sometimes take years to get any traction, and only then their effect could be evaluated by efficiency. So the cycle of judgment is much longer and it would be bad to judge reform/law right away.

People can’t choose efficient managers.

For normal people, it’s pretty hard to choose an efficient manager, because they’re not taught to do so. What they actually do is electing good politicians, which is a separate thing from good effective manager.

Politicians talk, managers execute. I remember a famous quote from ex PM of UK Tony Blair: best policy is the best politics. It means if you apply popular reform/law it will make you a brand, and you don’t need to talk much with audience to “sell the idea”. It’s hard to acknowledge, much easier to believe useless speeches of populistic figures.

We all live in big countries, which aren’t effective.

Democracy, different social groups, different beliefs. Too many species under one roof makes it inefficient as hell, ‘cause you need to acknowledge every opinion to make a decision. As a result, you can’t make all societal groups happy.

Big countries run like Google Corporation. Bureaucracy, lawyers, overspendings, random shutdowns, all of it. To make it efficient, you need to separate it into different companies. It becomes obvious in politics too, because certain areas across the globe are trying to separate from the mainland. Call it a trend, Catalonia and Spain, Quebec and Canada, etc.

We should live in city-states like the Greeks did. It wasn’t an empire or the strongest country, but an agglomeration of many peaceful and efficient governments.

Society teaches us that life is sacred and creating new life is humanity's greatest gift. While I acknowledge the importance of human life, I question whether reproduction should be our highest purpose.

This isn't an argument against the value of human life. Rather, it's a question of whether reproduction should be our highest calling. To answer this, we need to examine a wider perspective.

Nature of reproduction

Reproduction is embedded in our evolutionary history. If our ancestors hadn’t reproduced, we wouldn’t exist. This biological imperative is hardwired into us as a survival mechanism—the drive to pass on our genes. So let’s say for 50,000 years it was a growing stage. We’re progressing in all life areas, increased population, and so are other Earth species.

Can you pursue the same agenda now? I don’t think so. The world has changed, and resources are limited. If the human population grows at the same rate + we learn to prolong our lives to 100-150 years, it would be a disaster. There’s no free land and drinkable water. There is a shortage in certain regions already. It’s already happening.

From a resources perspective, it’s a bad decision. But it’s not enough for some people to say no, the world won’t be okay. Why? Because it’s not always their agenda, but the state. Our economies are organized for constant growth without limits. But we could change it. Let’s explore how we ended up here.

Need for optimization

Nobody thinks of optimization, because no one wants to live in the moment where everything stops and systems fail. That’s why we have long obligations, contracts, corporations aren’t temporary, and bonds issued on 20-30 years and medicine progresses = to prolong compliance and grow. We suggest that everything will be fine.

Every state puts its citizens in a position of prolonging the system, requiring people to give birth and influence in all ways. It’s called paternalism. You think of China and their birth control, right? It’s not only them. Mostly all states organized around reproduction as the main action in the economy cycle. Give birth, so your kids could pay taxes, which is retirement check for someone else.

For a long time, civilizations thrived on this, but given the planet’s state, I don’t think it’ll work in the future. We’ll need new mechanisms to organize the state, probably selling other agenda to social groups, which can sound as cyberpunk, but we’ll be pushed to do so anyway. I call to organize states around knowledge creators.

Understanding value

If the state has no agenda and no interest in you giving birth, you grow a kid for yourself. You grow a kid for fun, power over them, ego, or to save your genes.

I ask myself: Did the planet ask me to share my genes? Who did overall? Is there a world population deficit? The answer is NO.

Plus, I thought about my situation with ADHD, ASD and vulnerability to depression. Is it a good idea for me to reproduce? NO.

People should have kids to maintain the system, but it’s not the purpose of each human being. The reason isn’t eugenics, but that many people can express themselves and create knowledge instead, and be more useful for society in those roles.

Knowledge to outlive successors

Everything about the world and human lives is knowledge on how this world works. Think about the Beatles. They existed 65 years ago, but people still love their music. They reproduced such valuable knowledge about the world, that 65 years later, we still love it.

Einstein with his theory of relativity, Columbus mistaking the US for India, Gagarin as first man in space, Chaplin as prominent actor. All creators of knowledge. Some lived centuries ago, and we still praise them.

The conclusion is that the only thing that can pass generations and outlive is knowledge produced during life.

Knowledge is beneficial for society. More people should express themselves and reproduce knowledge. Create anything you want, for everything you want, interested, passionate.

This newsletter is my call to create something. I’m not equating myself with Einstein, he’s ahead. I hope it serves as a source of knowledge, provoking thoughts and pushes thinking. That’s the goal.

Invisible tendency

In any system, business or government, two stages are natural: growth and optimization. Companies invest in new technologies, then optimize spending. Government grow in people, then fire for optimization. The same goes for our time on Earth.

Green, ESG, and environmental organizations were founded 50-60 years ago to optimize life for the Earth. That’s why it’s happening.

Produce knowledge. It’s the most valuable thing you can do.

This is the final piece in a three-part series exploring how humans and GPT models, handle information: consuming, processing, and now, generating.

Check other parts:
Part 1
Part 2

Intro

Every human generates information 24/7.

All generating processes divide into external and internal. Internal is what we have in our brain: goals, thoughts, fears, everything inside.

External generation refers to what we produce for society and the world: books, articles, videos, events - our jobs, businesses, everything we create.

Paul Graham generates essays for his blog, Sam Altman rules OpenAI and that’s how he produces information. Giving birth is also a process of generating information, because you transferred one gene to create another.

This post is my way to generate information. I had a white list, nothing. And then here’s the piece. I generated it from my head.

Cool insight

The information we generate depends on what we've previously consumed and processed.

As we discussed earlier, generating goes well if you worked hard on the previous parts. You can actually see what people consumed based on what they create. The YouTuber, for example, watches other YouTubers to catch the styles, starry talent, vlogging, he consumes that and processes to generate information like GPT does.

If you feed your brain with cheap dopamine, like Instagram shorts or LinkedIn posts, then generation won't work. You won't produce anything; you'll just get a dosage of dopamine, and that's how it ends.

The Core Theory: Information as Legacy

I theorize that our life’s purpose is to create information for future generations. It’s the most valuable task—creating something out of nothing. From 0 to 1.

The information’s value is measured by its existence and truth. Look at Einstein. He created a theory of relativity in 1915, received Nobel prize, and 110 years after, it’s still true. That’s the information’s value.

To assess the value of information, look at it after years. Do people know this? Do they remember? Does this knowledge transfer globally?

It explains why you shouldn't argue about bad movies, music, books, any art. Time will tell and value them. If it's nothing, everyone will forget about it. You don't know will those singers become legends. Wait through time, it will be clear.

Despite global changes, we consume major information till 40, then our brain activity declines, so we consume less valuable knowledge. Your brain is active, till you understand the knowledge and create something. Learn until your brain is flexible and adaptable, you won’t be in the stage. Later, you will become conservative and unlikely generate something useful.

Check what you consume, as it will affect the generation process.

The Cycle of Information

Information generation isn't isolated. It's deeply connected to what we consume and process. Like a GPT, we can only generate info based on our training data—our experiences, knowledge, and understanding.

Creators and Sustainers: A Necessary Balance

Society operates through two essential groups: creators and sustainers. Neither is superior; both are crucial for progress.

Not everyone generates information through books, blogs, or movies, and that's fine.

For example, gig workers like taxi drivers don't produce information, but they're valuable because they sustain the generation cycle. Doctors heal people, psychiatrists improve mental health, and fitness trainers keep bodies fit, allowing longer lives and more time to create.

I'm not saying generators are better than sustainers. Both parts are needed; otherwise, creators couldn't survive. Sustainers produce food, water, clothes, and build hospitals. It's crucially important.

The opposite is also true. If the system just sustainable, it doesn’t grow or develop. There aren’t any innovators. Sustainers need creators to produce knowledge, science, and improve life. That stuff is crucial too.

The most valuable thing you could do in your life is to produce information and hope it will be valuable and used. Nothing could kill it, as information doesn’t age or die. It can only be forgotten or remembered. Only 1 option to pick.

Example of Lasting Generated Information

Henry Ford created the first car, and 100 years passed since its invention. Though he wasn’t the creator of the “car” concept, he made it accessible and affordable for everyone.

He made “cars for people, not just the rich.” He created something new that outlived him - Ford company still exists, and engines work the same way as 100 years ago. It doesn’t matter if Ford generates innovations now, but we won’t forget Henry Ford’s impact. Remember and praise the hero.

Control of Information Generation

It’s clear why governments control education systems and narratives: they want to shape how people generate information. They aim to produce predictable outputs from citizens, guiding them toward government-approved paths.

That’s how governments recruit people for war or dangerous jobs. Humans can’t have those desires from childhood, right? Someone puts them there, right?

The title suggests they feed your information at school at university, then they control labels we discussed in the previous article. Labels are shortcuts to meaning, so they control labels and your thoughts on the consumed information, following the government agenda. They control the generating process, but it still doesn't work for everyone because people like Henry Ford didn't produce what was expected.

You can call them rebels because they go the opposite way; they decide. To be a creator you need to be a rebel, otherwise you'd join the sustainers group which is a different but important job. You need to be a real rebel to believe in yourself, create something valuable, and take action.

Fine-Tuning Your Information Generation

Just as AI models can be fine-tuned, humans can optimize their information generation capabilities. This requires careful attention to:

Input Quality (Information Diet)

If you care about your information generation, be intentional about your sources, quality, and quantity. Your outputs will depend on your inputs.

Processing Time

Leave empty slot in your calendar for doing nothing. Don’t pick up the phone. Process what you have. Don’t rush into consuming non-stop.

Generation Methods

Express yourself fully. Start a blog, write a diary, build a business. This is our footprint on human history. Don’t think of yourself as Caesar, but important element in the universe.

3 steps to fine-tuning

1. Ask yourself

1) What datasets do I consume? Which are high quality? Low quality?

2) Who do I talk to? What do I read? How do I spend my free time?

Our ability to critically analyze data and decide. GPT can’t do that.

2. Filter sources

Create an info deficit, not surplus. Avoid cheap dopamine sources like social media, TikToks, reels, binge-watching.

3. Replace low-quality datasets.

Reflecting on bad datasets will help you find high-quality data to replace them. You probably don’t know which is which, so just try and you’ll find one. Datasets must boost your creativity.

My high-quality data sources:

• Every.to — Daily tech essays

• Big Think — Scientific and philosophical interviews

• Andrew Chen newsletter — Startup and Tech insights.

• NASA+ streaming — Space exploration content

Conclusion

The information is eternal: we consume, process, and generate. Unlike AI, we can choose our inputs and shape our outputs. Whether you're a creator or sustainer, your contribution matters. Be intentional about the information you consume and conscious about what you generate.

Remember: the value of information isn't measured in likes or immediate feedback, but in its longevity and impact across generations.

The Crucial Role of Meaning in Information Processing

Following our exploration of how information shapes our minds, let's dive deeper into perhaps the most important aspect: meaning itself.

Every piece of information needs meaning. It's not enough for your brain to just take in raw data - you need to understand what that information means for you. For GPT its the same. It can't skip this phase.

That raises the question: "What meaning is? How does meaning look for human? And for GPT?"

Core Differences Between Human and GPT Meaning

For humans, meaning is action-oriented. It helped our species survive. We process every piece in the matrix of survival and thriving in the future. Ex: we learn math for so many reasons, from calculating money, splitting bills, planning, avoiding scams. Every piece of information becomes a potential future action. That's the meaning for us.

For GPT, meaning is simple: what pattern will come next? What's next, the most likeable token in the sequence? It's statistical prediction without subject's understanding.

On this concept we differ from AI a lot. We aren't on various levels, but different dimensions. Thinking statistically is hard for us, and for GPT information doesn't necessary mean action.

But despite those differences, we live and thrive in same data types.

Types of Data Processing

To understand how both humans and AI process information, we need to examine two fundamental types of data:

• Unlabeled data: Raw information without built-in meaning, like plain facts or observations

• Pre-labeled data: Information with pre-assigned meaning (which I call labels) like a teacher's explanation.

Understanding Labels: Our Mental Filing System

Every piece of information needs meaning. It's not enough for your brain to just take in raw data - you need to understand what that information means for you. This is where labeling comes in.

Labeling is your brain's survival mechanism. Think about encountering a snake - there are 450 dangerous species out of 3,000. Would you stand there calculating probabilities, or would you accept the pre-existing label "snake = potential danger" and react? This labeling system is how we survived as a species.

Labels work as cognitive shortcuts:

• If you don't know wolves are dangerous, you might try to pet one

• If your ancestors didn't label "strange berries = potential poison," they wouldn't survive

• Most people accept the labels they're given because that's how our species survived

Picture your brain's file system:

• Eyes label visual information

• Ears label audio signals

• Smell labels environmental data

• Each sense creates immediate meaning through labels

A label either exists or it doesn't. When you encounter new information, your brain:

1. Searches for existing labels

2. Applies them if found

3. Creates new labels if needed

4. Updates old labels when necessary

This brings us to the crucial battle - who gets to create these labels in your brain?

GPT: The Master of Unlabeled Data

While we've explored how humans process information, let's contrast this with how AI, specifically GPT, handles data.

GPT doesn't try to understand meaning - it just calculates probabilities between patterns in text. This makes it the master of unlabeled data for two key reasons:

1. Scale and Cost: Labeled data would make model development too expensive

2. Statistical Accuracy: Numbers don't lie, but human labels always carry bias

For example, when it sees "painting houses," it doesn't understand the concept of decoration or murder. It just knows that 99.9% of the time this phrase appears in home improvement contexts... and 0.1% in Italian mob stories.

That's why GPT models sometimes hallucinate - they follow patterns, not truth. If enough examples said 2+2=6, GPT would happily agree.

Human on the other side

Take this example: "Elon Musk bought Twitter. He implemented X, Y, Z changes."

Sounds like perfect unlabeled data, right? Just facts, no interpretation. But humans can't help but label things. Try reading that headline without asking: Was this good? What were his motives? What does this mean for users?

Humans are kings of labeled data because:

• We need meaning to survive

• We process information through action potential

• We can't handle raw statistics like GPT

• We rely on pre-existing labels to function

Most of our knowledge (about 80%) comes pre-labeled, unlike GPT's 20%. We need these labels because:

• We can't process everything from scratch

• We need mental shortcuts to survive

• We learn slowly (18+ years of education)

• We build on previous human knowledge

In fact, processing raw data would be fatal for humans. Imagine having to calculate probabilities for every decision - we'd be extinct before finishing the math on whether a predator is dangerous. That's why we evolved to use labels as survival shortcuts.

The Modern Battle for Labels: Who Controls Your Mind?

Several major forces compete to control what things mean in your brain:

State vs Individual

The government wants predictable citizens and controls meaning through:

• Educational system enforcing official narratives

• Laws defining acceptable behavior

• Media regulations shaping information flow

• National symbols and values

Example: China's AI models are trained to never criticize the government or mark Taiwan as independent land. The state literally controls what meaning their citizens can access.

Education vs Experience

Schools aren't just teaching - they're programming meanings:

• History is written by winners

• Scientific theories become "truth" until disproven

• Cultural values are embedded in curriculum

• Authority figures define "right" and "wrong"

But personal experience often challenges these taught meanings. Think of Newton's physics - taught as absolute truth for 200+ years until Einstein's experience showed a different meaning.

Media vs Critical Thinking

Modern media fights for control over meaning through:

• News framing ("good" vs "bad" actors)

• Entertainment shaping values

• Social media echo chambers

• Algorithmic content curation

The battle here is between accepting pre-labeled meanings or thinking critically about information.

Family vs Society

The oldest meaning-makers in conflict with modern labels:

• Traditional values vs new social norms

• Inherited beliefs vs personal discovery

• Family roles vs individual identity

• Cultural preservation vs adaptation

Winning the Fight: Tools for Maintaining Control

Now that we understand the forces trying to shape our understanding, let's explore practical tools to maintain control over our own meaning-making process:

To maintain control over your own meaning-making:

Question existing labels

Take the "painting houses" example. When someone says they "paint houses," 99.9% of people think home improvement. But in mob stories, it means something entirely different. This shows how the same phrase can carry completely different meanings depending on context. Never assume your initial interpretation is the only valid one.

Seek multiple perspectives

Think of Ground.news - they attempt to show "pure facts" by comparing different media perspectives on the same story. Even their attempt to be "unlabeled" becomes its own kind of label. By examining multiple viewpoints, we can better understand how different groups assign meaning to the same information.

Balance Speed and Reflection

While quick reactions can save us from danger (like avoiding a snake), not everything requires immediate labeling. Some situations demand rapid responses based on pre-existing labels, while others benefit from careful consideration. The key is knowing when to act quickly and when to pause for deeper analysis.

Curate Your Information Diet

Just as China's AI models are trained to avoid certain topics, our minds can be shaped by the information we consume. Choose your sources carefully and deliberately seek out diverse perspectives. Remember that controlled information leads to controlled thinking.

Challenge Authority Labels

Even established "truths" can be wrong - just look at how Einstein's theories superseded Newton's "absolute" laws of physics. Question not just what you're told, but who's doing the telling and why. Watch for competing labels, like when traditional institutions (such as the Catholic Church) clash with evolving social values.

6. Practice Active Meaning-Making

Don't be a passive recipient of others' labels. Actively engage with information, question assumptions, and be conscious of who's trying to influence your understanding. The battle for meaning never ends, but active awareness is your strongest defense.

The Ultimate Challenge

Understanding how we process information and assign meaning is only half the battle. As we've seen, both humans and AI have distinct ways of handling data and creating meaning. But there's an even more fascinating frontier ahead: how do we move from processing information to generating it?

While GPT excels at pattern recognition and humans master meaning-making, the real power lies in creation. How do we take all this processed information and transform it into something new? How do different labels and meanings influence what we create? And most importantly, what makes human generation of information unique in an age of AI?

These questions lead us to our next exploration: the fascinating world of information generation, where we'll discover how humans and AI approach the creative process in fundamentally different ways.

These questions lead us to our next exploration: the fascinating world of information generation, where we'll discover how humans and AI approach the creative process in fundamentally different ways. Armed with an understanding of how meaning shapes our minds, we're ready to explore the ultimate expression of human cognition: creation itself.

GPT models consume human knowledge, process and generate something. All of it from human knowledge. I got the idea, of reversing this idea, and learn something from how GPT models learn about us. This piece is collection of thoughts, things I figured out. I just play with my creativity. Enjoy.

Check other parts of the cycle:
Part 2

Part 3

The GPT-Brain Connection

GPT stands for generative pre-trained transformer. It means input of information causes output of information’s generation. We call its answer to our prompt, and give its description as good, correct, or wrong, hallucinated, delusional, etc.

To better understand analogy with the human brain, let’s revise the life cycle of GPT and then switch to human.

Life Cycle of GPT Models

Stage 1. Consuming information — The model studies large datasets, which contain plain text, articles, websites, pdf, music, videos, etc.

Stage 2. Processing information & fine-tuning — The model recognizes patterns, determines probabilities, and generally connects dots. If wrong, then feeds additional datasets (fine-tuning) to better understand the context.

Stage 3 — The model awaits an order from the human, which we call a prompt. The prompt serves as a trigger activating the ability to process and match patterns across its internalized knowledge. The response should “make sense” for us.

Stage 4. Generating — The model analyzes the prompt provided by the human, and generates the most probable and contextually appropriate response based on the data it was trained on. The cycle repeats over and over.

Abstract

Imagine your brain as a GPT model being constantly trained by the world around you. Every experience, every interaction, every piece of information becomes training data. But here's the fascinating part - different systems are actively competing to be your primary trainer, each trying to shape how you think and what you believe.

This piece explores what happens when we reverse-engineer our own minds using the GPT model as a framework. What can we learn about ourselves by understanding how these systems compete to train us?

Why does this matter? Because understanding which datasets influence your decisions is crucial for living a conscious, self-directed life. Some datasets you're born into (like your country and family), others you choose (like your education and social circles), but all of them shape how you think and act.

The key difference between your brain and GPT is that while AI passively accepts its training data, you can choose which datasets influence you - if you know they exist. So, let's break down these competing systems by their power and control over your mind.

1. State Dataset: You Can't Uninstall It, Only Replace It .

We were born in the country. It’s top-level dataset, 1st in hierarchy. This means that the state creates rules for other datasets to operate within. Its always there, since no place on Earth out of jurisdiction. Some state will train you, the way they want to. Let’s break down their influence.

Personal Life & Values

• When to marry and have children

• What constitutes a "normal" family?

• Views on privacy, individual rights, and social norms

Economic Framework

• Career paths and industries to pursue

• Financial freedom through tax policies

• Retirement and long-term planning

Social & Political Structure

• Political ideologies and alignments

• Healthcare decisions (vaccines, insurance)

• Legal boundaries and freedoms

Reality Distortion

Government isn’t interested in you understanding the truth. It fights for its own truth matrix. It wants you to think that state’s dataset is the crucial, we know what’s right and wrong, you need to believe us. State uses other datasets within to influence ur matrix. And they do it successfully.

Like GPT models that can't recognize their own hallucinations, citizens under heavy state influence often can't see the programming. Just as GPT confidently generates false information, citizens can passionately defend state programming they've never questioned. Take COVID-19 reporting: same event, completely different "truths" depending on the state dataset - from China's "victory over virus" to Sweden's "balanced approach."They put labels on information, replacing critical thinking, which undermines their authority. So you don’t think bout the meaning, but just label and live with it. Think about it: state can feed you a patriotism, making more kids, send u to die for the country, what media, movies, videos to consume. That's the scale of State's dataset influence.

Dataset Flexibility

Even the most controlling states can't maintain perfect dataset control. There are always cracks in the matrix: underground movements in USSR, VPNs breaking through China's firewall, banned books and foreign radio in North Korea.

When you train yourself a lot, you know there are plenty labels on the same thing. A market is states’ narratives. Corruption, freedom of speech, military policy, immigrants, etc. So you can replace the bad one with a good one. Just pick whatever you think is right. Move to the country. Anyway, you can’t totally exclude the state, but replacing Germany with Italy, or US with Canada. Why not?

Hidden Dangers

Labels and datasets aren't a function of scientific research. It's opinion of those who create laws, and not necessarily proved by experience, experiments, etc. The scariest part? Entire populations can run on corrupted data without realizing it. Nazi Germany didn't happen overnight - it was a systematic corruption of an entire nation's dataset, making the unthinkable seem normal.

Real Example

Just compare China and Western civilization or look at Iran. In Iran, the same geography and the same people had completely different datasets pre/post 1979 revolution. In one generation, the state completely rewrote how people think and behave.

Or two Koreas. Those are the same people. Now they’re totally different because states feed them different datasets.

What You Can Do About It

1. Cross-reference with other state datasets.

2. Study how your state's "truth" changed over time

3. Travel to experience different reality matrices.

4. Build a network of independent information sources.

5. Learn to spot state programming patterns.

2. Family Dataset

Your first training data will be your second if not. Like a local AI model running on top of the state's larger system. Moral, good and bad varies from family to family. We can see it by comparing ourselves to others. Some families can push u into family life, others to construct a successful career, or picking prestige instead of true desires. So its influence also high. It certainly influenced by the state, but can both fully support the agenda or not.

Think of parents as prompt engineers for your developing brain-GPT. Your parents give feedback on your actions during childhood and through your life. Find a job, find a girlfriend (or boyfriend), buy a house, take a mortgage, have kids, etc.

It's first dataset if your folks oppositional to the politics, because they will filter what gov says and extract best out if it. Why it could be second? Cuz ur folks aren't protected from being influenced and truth followers of state agenda. They fill feed with the same thing as gov: die for the country, go to army, whatever else. It's like having admins who either protect you from malicious code or help install it. My condolences here, because no government could be trusted.

Areas of Deep Training

Just like GPT has different training domains, family programming covers multiple areas:

Core Values & Beliefs

• Moral compass (good vs. bad)

• Religious or spiritual beliefs

• Attitude toward money and success

Relationship Patterns

• Conflict resolution

• Emotional expression

Emotional Programming

• How to handle conflict

• When to trust or doubt

• What love looks like

• Which feelings are "acceptable"?

The quirky detail about family datasets? They can act like antivirus software or malware for state programming:

Dataset Characteristics

Antivirus Mode:

• Some family datasets create resistance, even in strict autocracies.

• Parents install critical thinking filters against state propaganda.

• Family values can override state programming.

• Opposition mindset passes through generations.

• Think Soviet kitchen talks, or underground reading clubs, discussing democracy.

Malware Mode:

• Some families amplify state programming.

• Parents become like state agents, reinforcing the official narrative.

• The family dataset completely merges with the state agenda.

• Critical thinking gets blocked at home level.

• Example: Families reporting neighbors during the Cultural Revolution

The Freedom Patch

Unlike state datasets, family programming isn't final. You can:

• Install your own critical thinking modules.

• Run different values in parallel.

• Choose which family code to keep.

• Build firewalls against toxic programming.

• Create your own upgraded version.

3. Education Dataset

Think of education as your brain's mandatory training program. Just as GPT needs pre-training to function, humans need systematic dataset installation. That's why every state creates educational systems. It's the most efficient way to install standardized programming in young GPTs.

Think of schools as miniature versions of society where your brain-GPT learns how the real world operates. During these years, your neural networks are most vulnerable to programming. That’s why the state has a lot of control here.

Core Installation

1. Society in Miniature

• First experience with hierarchy (teachers, principals, popular kids)

• Learning to compete and cooperate

• Understanding rules and consequences

• How grading systems install performance metrics

• How schedules program time management

2. Knowledge Installation

• Official version of history

• Approved cultural values

• Basic tools for survival (math, language, science)

• Pre-labeled datasets in form of textbooks

• State-approved processing patterns

3. Future Citizen Training

• Accepting authority

• Working in groups

• Installing standardized thinking patterns

• Programming social behaviors

Training Environment Analysis

Just like AI systems need specific training environments, educational datasets shape how we:

• Process information

• Interact with authority

• Handle structured tasks.

• Develop processing patterns.

Dataset Characteristics

Key Features:

• High update frequency (daily reinforcement)

• Standardized content delivery

• Peer validation systems

• Tracking performance metrics

• Built-in reward/punishment loops

Hidden Influences

• Social hierarchy programming

• Competitive mindset installation

• Conformity training modules

• Time management conditioning

• Authority response patterns

Just like AI systems need specific training environments, educational datasets shape how we:

• Process information

• Interact with authority

• Handle structured tasks.

• Mass-scale knowledge installation

• Core science and math modules

• Universal access protocols

• Standardized processing patterns

System Vulnerabilities:

• State bias in history interpretation

• Self-preservation loops (never criticizing its code)

• Outdated processing methods

• State requirements filter truth.

Alternative Education Systems

The quirky detail about education? In democracies, private institutions can run modified versions of the training program. Think of them as custom-built GPTs with different training parameters:

Alternative Dataset Installation:

• Challenge state narratives.

• Install critical thinking modules.

• Run experimental learning protocols.

• Process information from multiple cultures.

Real-world Examples:

• Stanford's "question everything" protocol

• MIT's "mind and hand" processing model

• Alternative schools' pattern-breaking systems

• Multicultural datasets

The Disruption Paradox:

These institutions create valuable anomalies in the system (think Gates, Musk, Zuckerberg). But here's the quirky part - while individual disruptors drive progress, too many could crash the entire state operating system. That's why access to these alternative datasets remains limited.

4. Business & Enterprise Dataset

Have you ever thought how businesses & enterprises train us? Not literally, but with datasets. Decomposing this to foundation will lead u to growing your own food, making our clothes, building your electricity line, phone, car. Impossible, right? That's how businesses became the most subtle programmers of your brain-GPT. Their training methods are everywhere:

1. Social Behavior Programming

• How we communicate & connect (Messengers & Social media) -

• How we date (Tinder and dating apps)

• What we share and why

2. Consumption Patterns

• What and when we buy (Amazon's "you might also like")

• How we think about money (credit cards, buy now pay later)

• When to upgrade our lives

4. Information Flow

• News consumption (paywalls, recommended articles) -

• Search results (Google's algorithm deciding what's relevant)

• Social media feeds

Reality Distortion

The mind-bending question: when you tap "buy now," is it really your choice, or just successful programming? Even scarier - can you tell the difference? When u choose to watch a movie, or listen a song, was it really your choice? How did you do it?

Dataset Characteristics

Unique Features:

• The only dataset that openly admits it's training you is called marketing.

• Uses direct feedback loops (buying patterns)

• Constant A/B testing on human responses

• Votes counted in dollars, not ballots.

Power Dynamics:

• Operates within state framework but has its own agenda.

• Sometimes more powerful than state itself.

• Creates progress state couldn't be achieved alone.

• Fights state for control over your GPT

Hidden Dangers

The US case study:

• Health industry programming expensive treatment as normal

• Tech companies installing addiction patterns

• Insurance companies rewriting risk perception

• Marketing replacing needs with wants

So what do you make out of consuming phase?

1) There is no such thing as a new idea. You don’t have your own thoughts, values, principles. All of that internal luggage is a mix of consumed datasets from around. Movies u watched, books u read, people u listened to, state u lived in. Anything that your brain produces is a result of data they trained u on.

2) Datasets constantly compete for training you, whatever and wherever you live. They lose control, take it back, and cycle proceeds. Your state, family, corporations, religion, friends, media. You have only one sense of attention. And they compete for it. At some period u stopped talking with ur parents, dropped faith in government, found it, forgiven by parents and again. Cycle repeats..

3) Some GPTs (people) are more independent than others. For unknown reasons, some GPTs don’t buy datasets sold by other systems; instead, they start repurposing and creating their own. “Rule or be ruled.” Heard that? Datasets lose control over you, as u become independent, and it’s your call to reject their narrative. as u gain power, and datasets lose power over u, question your judgment. The systems constantly fight for ur attention. Attention is ur prompt. ask and tell appropriate questions. You listen to your parents less and less with years, don’t u?

4) Life always be unfair in some sense, because some GPTs are born in USA and Europe, and others in Bangladesh and Somalia. Their environment, government, and education will all function according to the state and local conditions of life, whether it’s a poor or rich place. You can see it a curse in some sense, which u won’t be able to purify or change, when it already happened. The same could be said about not only state, but city, family, neighbours, and district.

5) Some datasets cannot be changed, but replaced. Nobody obligates you to live in a homeland or city. You don’t have to be friends with former classmates. Move away, change a school, find new job, new friends. Of course, cannot be replaced, but still, nobody obligates u to have relationships with. You can choose.

6) Any object in the universe produces information and could be considered a dataset. If the object could be seen, then just by its existence it produces information about its existence. Clothes, book, laptop, wardrobe with t-shirts. What u can do about it? Variety of options. Stare at it, think of it, interact, write about it. Its black, its wide, warm, cool, whatever. Literally any object produces information, acting as a micro-dataset which trains you to act, deal with it.

7) GPT stops consuming u not. it just reads what we have. u live in the world and constantly consume information. The kettle is whistling, neighbor is shouting. Just be existing, sitting in the room and consuming information. That’s empty, that’s the pen I lost, my favorite Spotify playlist. You never stop. As u never stop consuming, you can’t make something out of it. Make pauses in consuming.

Questions I ask myself: What type of datasets I currently consume? Do I believe in it? Is it wrong? How can I fact-check my beliefs? What is considered a quality dataset? Is social media aligned with it?

It is important to realize what you consume. While Part 1 might make it seem like consuming datasets is all we need to rule the world, that's just the beginning. Data without meaning are just useless symbols. In the next part, we'll explore how to make sense of it all.