An article that thoroughly explains the 200-year cycle law

In the past two hundred years, humans have been extremely optimistic about technological progress, always believing that new technologies would eliminate old jobs but create more new ones. They thought that although cars replaced carriage drivers, the world didn't get worse. So, there's no need to make a fuss if AI replaces programmers.

However, this logic has a fatal underlying assumption: the speed of technological iteration is roughly in sync with the human life cycle. The prototype of the car emerged around 1880, and Ford achieved mass production in 1910, with a gap of about a generation in between. But the current iteration speed of AI is unprecedented in human history.

In this article by Mr. Zhang Xiaoyu, the AI sociology tutor of the PPE course of Notesman and a rising star in the field of science and technology history, starting from the difference in the speed at which AI replaces human labor, he deduces all the way to the first and second halves of the Industrial Revolution, the long - cycle of capital, the chain reaction of de - globalization, and finally to the "Intelligent - Industrial Complex" that Peter Thiel has been planning for two decades. Ultimately, he reaches a chilling conclusion: AI is taking us back to a "High - level Middle Ages."

I strongly recommend that you spend more time reading it carefully because it presents a complete analytical framework for the cycle of civilization.

I. Differences between the AI Revolution and Previous Industrial Revolutions

1. The speed of the AI revolution is unprecedented

In previous industrial revolutions, new technologies replaced some human labor. When cars replaced carriages, carriage drivers did lose their jobs, but it didn't cause a subversive impact on human society.

So, why is the impact of AI different? Why won't new occupations emerge naturally after programmers and other occupations are replaced by AI, as was the case in the past?

In terms of the speed of technological progress, the two are not on the same scale. Around 1880, Benz and Siemens built the first - generation car prototypes. It wasn't until around 1910 that Ford achieved industrial mass production of cars through the assembly line. There was a gap of two or three decades, almost an entire generation.

Before Ford's mass production, a car cost about one or two thousand dollars. In the United States at that time, this amount was almost enough to build a house. Even after Ford reduced the cost to $800 per car through assembly - line production, it was only barely within the reach of ordinary consumers.

The rhythm of technological iteration in that era was in sync with the natural human life cycle. A carriage driver might have long realized that cars would replace his occupation in the future, but it took two or three decades for this technology to truly spread and enter every household.

By the time cars were everywhere on the streets, he would have reached retirement age and saved enough money for his old age from driving carriages. His next generation would not choose to be carriage drivers from the start, thus naturally avoiding the fate of being eliminated.

The occupational impact brought about by technological progress in the past could be smoothly digested through the natural generational replacement of humans. One generation could complete their careers smoothly, and the next generation would naturally shift to new tracks, without causing severe social upheaval.

However, the current progress speed of AI has completely broken this rhythm. OpenAI's GPT 3.5 was officially launched at the end of 2022. Just two or three years later, in 2024 and 2025, its capabilities were already sufficient to handle many jobs that required a doctoral degree.

It takes at least four or five years for a person to complete a doctoral program. That is to say, before you even finish your doctorate, your professional ability has already been surpassed by AI. This iteration frequency is unprecedented in any previous technological revolution.

People who don't understand the actual situation always make analogies with past industrial revolutions, thinking that history will simply repeat itself. But those who truly see the reality clearly have long realized that the pre - conditions of the two are vastly different, and thus have begun to re - reflect on how to proceed in the second half of the "Industrial Revolution."

2. The first and second halves of the Industrial Revolution

The past 200 years have been the golden age of the fastest progress in human history. Even after two world wars, the overall income level, quality of life, and level of civilization of humanity have achieved an unprecedented leap. This iron - clad fact has also made "technological progressivism" the most deeply ingrained and persuasive ideology in the past 200 years.

The core logic of technological progressivism is based on two premises: history is continuously moving forward, and progress is linear.

This theory of linear progress means that all the problems existing in the current society can be solved through continuous development. If the problems haven't been solved yet, we just need to keep moving forward and maintain the development rhythm.

The development experience of the past 200 years has almost perfectly verified the effectiveness of this logic. However, even before the popularization of automation technology and the emergence of AI, some scholars had begun to reflect: this default premise may have fundamental flaws.

This doubt has been confirmed in the core data of economics. There is a core concept in economics called Total Factor Productivity (TFP). Simply put, it is the contribution of per - capita GDP growth purely brought about by technological progress after excluding tangible factors such as capital investment, labor input, and fixed - asset investment. It is the core indicator to measure the real driving effect of technology on economic growth.

In the 1990s, American economists proposed the famous "Solow Paradox." At that time, personal computers had become popular in society, and everyone was talking about the information revolution and the third technological revolution. However, statistical data showed that per - capita GDP did not increase significantly as expected, and the value of technological progress was not reflected in economic growth at all. Many economists asked, "Where are the computers?"

From 1920 to 1970, the average annual growth rate of TFP in the United States reached 1.89, the highest level in history. During the so - called golden period of the information revolution from 1994 to 2004, the average annual growth rate of TFP was only 1.03, less than half of the previous level, and it only lasted for 10 years.

This shows that since the 1970s, the positive driving effect of technological progress on the improvement of human living standards has been in a continuous decline.

Regarding this phenomenon, the economics community has reached a core conclusion from the perspective of development economics: The technological attributes of previous industrial revolutions are fundamentally different. The core of the first two industrial revolutions was to extend the industrial chain, while the core of the third industrial revolution was to shorten the industrial chain. This fundamental difference directly determines the vastly different impacts of technology on society.

What is "extending the industrial chain"? It means that all the core inventions of the first two industrial revolutions, from steam locomotives, cars to electric lights, telephones, refrigerators, and televisions, are specific products. The birth of each new product will extend hundreds or thousands of supply - chain links, give rise to a large number of related enterprises, and create a large number of jobs.

The most typical example is the auto workers in Detroit in the 1950s and 1960s in the United States. Most of them only received secondary - vocational or junior - college education. After working in the factory for 20 years, they could grow into experienced senior engineers.

The core feature of the manufacturing industry is that experience and skills will continuously increase in value with work experience. Therefore, these workers had high incomes, could afford single - family houses, raise two or three children, and take the whole family on overseas vacations every year. Their lives were very decent.

The third industrial revolution, however, brought about the opposite effect of "shortening the industrial chain." Of course, robots and information technology themselves will also give rise to upstream supply chains, but the scale of these new jobs is far from enough to catch up with the scale of jobs replaced by automation.

At the same time, the links that cannot be replaced by automation will be massively transferred to countries with lower labor costs. From the overall perspective of society, the industrial chain either shrinks continuously or is transferred overseas. The final result is a significant reduction in domestic employment opportunities.

The remaining high - quality employment opportunities are almost all concentrated in industries with extremely high educational thresholds, such as finance, the Internet, and automation.

In the past, one could get a high - paying job in Detroit with a junior - college degree. Now, one must have a master's or doctoral degree to be eligible. The higher educational requirements mean that young people will postpone entering society, and the age of marriage and child - bearing will also be postponed. Naturally, the whole society will enter an aging stage.

The consumption demand of the elderly is much lower than that of young people, and the consumption and finance of the whole society will enter a contraction cycle accordingly.

This is the core logic of the second half of the industrial society, which is completely opposite to the expansion logic of the first half. This difference is fundamentally determined by the underlying attributes of technology.

There are two main criteria to judge whether a technology increases labor demand: whether the new technology creates new task scenarios or releases previously unmet potential demands.

"Creating new tasks" is easy to understand. There was no automobile industry in the past. After the birth of cars, a whole set of new tasks and jobs emerged, from car manufacturing, repair to driving, naturally creating a large number of jobs.

The core of this logic is: Whether automation can create jobs essentially depends on whether it can reach previously unmet demands. Once the demands are fully met, automation will shift from "creating jobs" to "replacing jobs."

Ultimately, the relationship between AI or any other new technology and employment and social stability is never determined solely by the technology itself. It must be embedded in the institutional framework of human society, the existing economic structure, and real demand scenarios to make an accurate judgment.

II. Universal Basic Income: 99% of people can have income without working

1. In the AI era, the relationship between people is facing reconstruction

The most fundamental challenge brought about by AI is actually the complete reconstruction of the relationship between the 1% and the 99% in the entire social structure.

For the 1% of elites at the forefront of AI, they really have the ability and resources to achieve this isolation. For example, distributing UBI is one way.

So - called UBI, or Universal Basic Income, simply means providing all social members with a regular basic living allowance unconditionally, without any requirements or pre - conditions.

In the eyes of these elites, distributing UBI is essentially spending money to buy social stability and a stratified and isolated social order. As long as ordinary people don't cause trouble after receiving the money and don't affect their scientific research and production using AI, this deal is very cost - effective.

Currently, almost all the top - tier groups in Silicon Valley are seriously discussing UBI. The core reason is that the explosive progress of AI since 2024 has turned the large - scale replacement of human labor from a future speculation into an ongoing reality.

2. UBI is essentially a solution to avoid social fragmentation and turmoil

This is also a very rare scene in human history: the top - tier capitalists are collectively and actively discussing giving money to ordinary people. And the closer you are to the technological forefront of AI, the more frequently you will hear this topic.

This is actually a purely political issue: They are not doing this out of charity but from the perspective of vested interests. They must provide a basic guarantee for the bottom - tier people to ensure that the whole society will not fall into large - scale conflicts and turmoil.

Since 2017, there have been more than a dozen rounds of UBI social experiments globally, but most of them are on a small scale.

Two of the most representative large - scale experiments are one led by Sam Altman, the founder of OpenAI, in the United States, and the other is a long - term experiment launched in Kenya in 2020 for a period of 10 years. It has been running for 6 years and has accumulated a lot of valuable data.

The Kenyan experiment also had a group - control design: some people received the money monthly, some annually, and some received the UBI for several years in one lump sum.

The result was very counter - intuitive: the group that received the full subsidy in one lump sum performed much better than the group that received the money monthly. After receiving the money, local poor people didn't spend it randomly. Most of them saved it as a family risk - resistance fund and even subsidized other members of the extended family.

When everyone has a basic guarantee, complex economic activities will naturally emerge, and the vitality of the whole society will be activated. This is completely consistent with the "emergence effect" we discussed before: when the basic conditions are met, complex order will naturally grow.

III. Capital in the Twenty - First Century and the Cycle of Capitalism

Ten years ago, Capital in the Twenty - First Century by French economist Thomas Piketty caused a global sensation. He did something seemingly extremely simple but subverted the mainstream economic understanding: he used a basic accounting framework to dissect the underlying logic of the continuous expansion of inequality in human society over the past 20 years.

There are four core variables in this framework. I'll break them down for you in the most popular way:

K (Capital): The total capital stock of the whole society, that is, the total wealth belonging to the capital side;

L (Labor): The total wealth obtained by all workers through labor. K + L together constitute the total wealth of the whole society;

Y: The total newly - added wealth of the whole society each year, which is roughly equivalent to the GDP we often mention and can be directly estimated by GDP.

The ratio of K/Y is the core weight of capital in the whole - society wealth distribution. You can imagine the whole society as a giant company. K is the principal invested by all investors, and K/Y is the dividend ratio that shareholders can get from the company's annual revenue each year.

On this basis, we then introduce two core growth - rate indicators:

Small r: The average annual rate of return of capital wealth, that is, the speed of "money making money"; Small g: The average annual growth rate of the total wealth of the whole society, that is, the overall growth rate of GDP.

According to the most basic fairness logic, these two