If you're feeling an inexplicable sense of security, or that you can rest easy after "learning a few AI tools," you may be in a dangerous physical state: gliding.

Gliding feels very similar to flying. You have speed, the sound of wind in your ears, and even a brief ascent. But physics tells us that gliding is essentially a process of energy dissipation - you're relying on potential energy accumulated in the past and gravity is irreversibly pulling you toward the ground.

This gravitational force is called "zero marginal cost of intelligence".

Look at what just happened to Swedish financial giant Klarna in 2024: they announced 700 layoffs and deployed an OpenAI-powered customer service system. The ending was grim: the AI handled 2/3 of all customer service conversations, the equivalent of 700 full-time employees, and customer satisfaction was flat, with resolution times reduced from 11 minutes to 2 minutes. This is expected to add $40 million to the company's bottom line.

The scary thing about this case is not "how powerful AI is", but "how worthless mediocre intelligence is".

Are those 700 employees not working hard? They may be well-educated, have good communication skills, and even work hard every day. But on a capital statement, their skills (communicating, understanding, responding) become a generic bulk commodity.

Sam Altman, CEO of OpenAI, has formulated Moore's Law of Everything: as AI evolves, the price of intelligence will approach zero.

In the past, writing a logical article, drawing a beautiful illustration, or writing a piece of running code were scarce "intellectual assets" for which you received a high salary. Today, these are becoming "infrastructure" like tap water.

When you stay up all night in the company doing PPT, writing weekly reports, organizing Excel, please wake up and realize that the "professional ability" you are so proud of is becoming a relic of the old days. All of us, at the moment, are just relying on the old professional inertia "gliding".

The labor trap: why "learning AI" can't save the workers?

In the face of this crisis, the market is full of a cheap antidote: "hurry to learn AI, become a person who can use AI, you will not be eliminated."

This is a huge logical trap.

In an economic perspective, this advice is not only ineffective, it's even harmful. Because it allows you to continue to indulge in the old game of "improving labor efficiency" while ignoring that the rules have changed.

First of all, technology was never meant to "serve the experts", but to "eliminate the experts".

The biggest role of AI is to lower the threshold. In the past, it took 5 years to learn art in order to draw the commercial draft, now a white person can use Midjourney to generate two days. In the past, it took three years to learn Python to write a crawler, and now a liberal arts student can use Cursor to run through.

If you are just an "operator", your moat is filled. For the boss, there are more alternatives, and your bargaining power is reduced; for the market, there will be an extreme surplus of content and code, and the unit price will be extremely depreciated.

Second, AI is multiplication, not addition.

This is the core algorithmic logic:

Result = Capital*AI (Result = Capital ✖️AI)

The "capital" here is not just money, but ownership, means of production and core data.

• If you are a laborer: your capital is almost 0. AI amplifies your "output", not your "value". You use AI one hour to do the work of the previous ten hours, the boss will not give you ten times the salary, but only give you ten times the amount of work, or lay off nine colleagues around you. In a business model that simply sells time, the high efficiency of AI ultimately translates into labor market involution and low prices.
• If you're the owner: you own the business flow, you own the customers, you own the rights, AI is your super-leverage to not pay social security, to not call in sick, and to work 24 hours a day.

This is the cruel "labor trap" .As long as you are still trying to make money by "selling units of time", no matter how skilled you are at using AI tools, you are still competing downward.

In the AI era, "execution" is the least valuable. The only way to survive the flood is to "own".

Redefining "Primary Capital": The Ticket to Noah's Ark

Since the road of "selling intelligence" does not work, where should we escape?

The answer is: to do what AI can't do and capitalize on it.

In the old days, capital usually meant cash, land, or buildings. But in the AI era, we need to redefine "raw capital". The real capital is those scarce resources that cannot be generated by algorithms trained on massive amounts of data.

If you want to survive in the next ten years, you must immediately start accumulating the following three kinds of "new capital:

1. Trust capital and distribution rights:

AI can generate 10,000 perfect articles a second, but it can't generate a "Luo Xiang", nor can it generate a "Li Jiaqi".

The cost of content generation is zero, which means "content flood" .In the ocean of information, the scarcity is no longer "good answers", but "trusted sources" .When everyone is using AI to produce content, "who's talking" is 10,000 times more important than "what's being said".

Having 1,000 hardcore followers (distribution rights) is more of an asset than having 10 AI skills. Don't just be a behind-the-scenes producer, build a personal brand, build private traffic, and build a real, live user relationship.

2. exclusive data and dark knowledge:

ChatGPT is fed with the growth of public data on the Internet. Anything that Baidu and Google can search is cheap for AI.

But what can't it learn? It's the non-standard data cluttering up your company's intranet, the intuition you've accumulated over 10 years in the industry, and your insights into the deeper needs of specific customers.

Stop doing generic work and go deeper. To solve those non-standardized, complex, and even need to "run legs" to solve the dirty work. Only with private data can you train your own vertical model. This is your private mineral, do not easily contribute it to the public big model. 3.

3. Ownership of the system:

This is the most critical point. You must own a "system", even if it is small.

If you are a programmer, don't just be proud of it, you need to own a SaaS product that "solves problems" or a piece of automation code that consistently produces results; if you are a salesperson, don't just rely on the company's platform, you need to own your customer network.

The laborer is the part of the system, the capitalist is the owner of the system. AI is the perfect part, it will replace all the expensive parts (you). But AI can't be the owner. Go be the one who pushes the button, not the button itself.

Extreme Greed: A Guide to Arming Yourself for Action

The window of time left to us is not much. We are in the "iPhone moment" of AI technology. It's not perfect, but that's the opportunity. Before it evolves into its full form, you need to arm yourself with an "extremely utilitarian" approach.

First, from "work" to "work".

Whether you are now in a large factory or a small company, please immediately develop a "mini CEO" thinking.

Instead of focusing on the "tasks assigned to me by the boss", focus on the "closed loop of the business". Try to build a minimal business project in your spare time - even if it's a paid community, writing an e-book, or running a Xiaohongshu account.

The purpose is to practice the feeling of "mastering the whole picture". When AI takes over the executive level, only those who know how to define the problem, organize resources, and distribute value can master AI.

Second, the crazy hoarding of "semi-finished assets".

Take advantage of the stage where AI still needs human guidance, and frantically make your tacit knowledge explicit.

Write the experience in your head into SOPs (standard operating procedures); finish your half-conceived novel; write a prototype of the tool you want to develop.

You are now a super-individual with a plug-in. Don't waste your time brushing up on short videos; spend it creating assets. These assets in the future, is your bargaining chip with AI.

Third, establish "non-standard" real connections.

Go meet real people. Go for a drink, go for a chat, go for a handshake.

The more digitized you get, the more expensive biological connections become. The high-end services and core opportunities of the future must circulate in private, offline networks of people that AI can't reach. This is the last moat of being a carbon-based creature.

The Endgame

In the world of the future, there may really be only two kinds of people:

One kind of people are those who are bred by algorithms. They live on cheap entertainment and a basic income, their time is worthless, they are batteries in a vast digital system.

One kind of people are the ones who feed the algorithms. They own the capital (data, IP, systems), they just need to issue instructions, thousands of AI will work tirelessly for them, creating amazing wealth.

From "batteries" to "manipulators", the doorway in between is slowly closing.

On the eve of the full-scale advent of the AI era, please stop the simple anxiety and blind efforts of self-motivation.

Go build capital, fight for ownership, and build trust.

Arm yourself, not to defeat AI in this cruel era, but for the future, when you stand in the center of the storm, clutching the ticket to Noah's Ark - the name of that ticket is "irreplaceable assets".