AI, energy, and population are the three biggest forces that will reshape the world in the next two to three decades. Why twenty or thirty years? Because three lines happen to reach critical points at the same time during this window period: AI is likely to complete the replacement of most cognitive labor, clean energy is likely to complete the main replacement of fossil energy, and the flipping of the population map is a confirmed fact. Three forces move from trend to reality at the same time. This is also a change that most people today will experience in their lifetime.

AI reshapes social structure and class distribution, energy reshapes the geopolitical order, and population reshapes the world's center of gravity and power. The directions of these three variables are clear and irreversible. Its cumulative effect will bring about structural changes in the world. We do not need to predict the specific picture of the future. We only need to see the direction of each variable clearly to realize that the social order, economic logic and international pattern we are accustomed to will undergo profound changes.

AI: Excess of intelligence and the disappearance of the middle layer

AI is not another technological upgrade. Steam engines replaced animal power, industrialization replaced handwork, and computers achieved automation—every technological revolution in the past mainly replaced manual labor and mechanical repetitive labor. AI replaces cognitive judgment on a large scale for the first time. A large number of ordinary mental tasks - analysis, writing, programming, customer service, translation, basic decision-making - are being supplied infinitely at near-zero marginal cost.

This means that the market value of mental labor will depreciate significantly. Just as the Industrial Revolution made the skills of hand weavers worthless, AI will make the professional accumulation of a large number of knowledge workers face the same fate. But this time it is faster and affects a wider area - because cognitive labor covers a much larger range of occupations than any kind of physical labor.

The most direct structural consequence of this is the intensification of social differentiation. The gap between those who can use AI to amplify their abilities and those who can't is not widening linearly, but diverging exponentially. Although the technological revolution in the past eliminated old jobs, it also created a large number of new supporting jobs. Industrialization requires workers, managers, engineers, accountants, and sales. The expansion of the entire industrial chain has created a huge middle class.

AI is different this round. Its characteristics are: very few people can complete the work that used to require a lot of manpower. One person plus AI may be better than an entire team in the past. Of course, AI will also create new jobs, but these new jobs are not evenly distributed - they are concentrated at both ends: creative work at the top and manual services at the bottom, leaving the middle layer empty.

The middle class is not just a concept of income. It is the structural basis for the stability of modern society. The middle class has properties to protect, children to educate, and pensions to preserve, so they naturally prefer order, gradualism, and compromise. They are the source of votes for moderates in democratic politics, the main force in the consumer market, and the carrier of social consensus. In fact, this group is already shrinking. The number of Americans who consider themselves middle class fell from 53% in 2008 to 44% in 2014. The proportion of the German middle class fell from 60% in 1991 to 54% in 2013, and the proportion of the middle class among young people fell faster. This was still a trend before the large-scale popularization of AI. The emptying out of the middle class means that all these functions have lost their bearers.

Education is one of the sectors hardest hit. The modern education system is built on an implicit assumption: individuals spend more than a dozen years accumulating knowledge and skills, and then use their decades-long careers to recoup that investment. It's a return on investment model, and it's worked well for the past hundred years. But if AI brings the market value of most knowledge-based skills to zero, this model will collapse. A person spends four years studying accounting, law, and programming. When he graduates, he finds that AI can complete the same work faster and at a lower cost.教育不是不重要了，而是教育作为社会筛选机制和阶层跃升通道的功能，被大幅弱化。 It is more difficult for poor families to produce rich children, not because educational resources are unfair, but because the rate of return on education itself is collapsing.

The political consequences of the shrinking middle class are certain: polarization. The middle class is the filler in the middle of the political spectrum. After this group shrinks, the middle ground collapses, leaving politics with only two ends - an angry bottom and an anxious elite, with no buffer between them and no basis for consensus. The political space of moderation, rationality, and compromise disappeared, replaced by emotional mobilization and identity confrontation. This is not a peculiar phenomenon of any particular country. From bipartisan polarization in the United States, to the rise of the far right in Europe, to the spread of populism around the world, the same structural forces bear similar fruits in different political soils.

Political polarization will inevitably spill over. Polarized domestic politics naturally tends to be inwardly retracted and outwardly tough—because this is the only posture acceptable to both ends. Mild free trade positions lost political ground. Each country will develop its own version of economic nationalism. The domestic political consensus on which globalization depends has collapsed.

At the same time, AI is also eroding the foundation of globalization from an economic perspective. The core driving force of globalization in the past few decades has been labor cost arbitrage: moving production to places where labor is cheap and selling products to places with strong purchasing power. This is a concise and powerful economic logic that drives the unprecedented global industrial division of labor. But if AI makes labor costs unimportant, this logic loses its support. The reshoring of manufacturing is no longer a political slogan, but an economic rationale.

The political and economic foundations of globalization are being eroded by AI at the same time. This has a particularly profound impact on developing countries: the catch-up path that has worked well in the past few decades—first undertaking the transfer of manufacturing and then gradually moving up the value chain—is being closed.

Energy: Retirement of fossil energy and reconstruction of geopolitical order

Energy is the physical base of civilization. How much energy a civilization can mobilize determines how much it can do. For the past two hundred years, this base has been fossil energy—coal, oil, and natural gas. They shape industrial systems, urban forms, transportation networks, and the basic pattern of international politics.

This base is being replaced.

The cost curve of photovoltaic modules is already very clear: it has dropped by nearly 90% in the past decade, and it is still accelerating. Wind power follows a similar path. Energy storage technology is iterating rapidly. Even without considering nuclear fusion, the strategic scarcity of fossil energy continues to be weakened by continued cost reductions in renewable energy and energy storage. If nuclear fusion is commercialized during this period, this process will be greatly accelerated. Clean energy is not a supplement to fossil energy, but a substitute. Twenty or thirty years from now, we are likely to live in a world where energy costs have dropped significantly.

Many people think that the last bastion of oil is transportation fuel-cars, ships, and airplanes. Not really. Electrification has put oil’s role in ground transportation at risk. Synthetic fuels have also begun to be commercialized. As the cost of clean energy continues to decline, synthetic fuels will significantly replace petroleum fuels, including aviation kerosene, which is the most difficult to replace. The real moat for oil is the chemical industry. Global plastics, fertilizers, synthetic fibers, synthetic rubber, and almost all modern chemical products start from hydrocarbon molecules produced by the cracking of petroleum. Chemical industry is the "irreplaceable" field of petroleum.

But this moat is also being breached. Green hydrogen plus CO₂ captured from the air can be used to synthesize methanol. Methanol is an extremely versatile chemical platform molecule, from which major chemical product lines such as olefins, aromatics, plastics, and synthetic fibers can be derived. The bio-based route offers another alternative. The key bottleneck is not technology, but cost - the current green synthesis route is far more expensive than petroleum cracking. But the cost depends on electricity prices. As the cost of clean energy continues to drop, synthetic routes will eventually compete head-on with petroleum economically.

At that point, oil will truly be downgraded from a strategic resource to an ordinary commodity. It’s not that no one uses it anymore, it’s that it’s no longer irreplaceable.

The geopolitical consequences of this incident are far-reaching. Fossil energy is not just an energy source, it is the underlying operating system of the current international order. The strategic position of the Middle East is based on oil. The U.S.'s involvement in the Middle East, the financial foundation of Middle Eastern countries, and the geopolitical importance of the Gulf region are all ultimately due to the fact that the world cannot live without what lies beneath that land. Saudi Arabia still derives more than 60% of its fiscal revenue from oil. Russia's status as a major power is also highly dependent on energy exports - oil and natural gas have long contributed 30% to 50% of federal budget revenue. They are the Kremlin's most important single source of revenue and its main tool for exerting influence on Europe. The U.S. dollar’s ​​reserve currency status is largely supported by the U.S. dollar pricing mechanism for oil trade.

When the strategic value of oil shrinks significantly, all these will have to be repriced. For a country that relies on oil for more than 60% of its fiscal revenue, this is not just an economic adjustment, but a shaking of the foundation of the country's survival. What the Middle East faces is not a problem of economic transformation but an existential crisis - the wealth and status accumulated over generations are built on a depreciating resource. The same goes for Russia. Without energy leverage, Russia’s bargaining chips in international games will be significantly reduced.

Global trade will also be disrupted from the supply side. Fossil energy and its derivatives account for a huge proportion of global maritime trade. The essential feature of clean energy is localization - sunlight and wind do not need to be transported across oceans, and power plants can generate electricity wherever they are built. Energy localization means that the interdependence between countries due to energy trade has dropped significantly. Trade is not just the flow of goods, but also the glue that binds relations between countries. When this bond weakens, international relations will become looser and less predictable.

At this point, two independent lines - AI and energy - converge at the same conclusion. AI eliminates the impetus for global division of labor from the demand side: labor arbitrage no longer makes economic sense. Energy eliminates the need for global trade on the supply side: the most important commodities no longer need to be transported across borders. Two completely different variables, independent logical chains, point in the same direction: the form of globalization in the past few decades is coming to an end.

Population: The Southward Shift of the World's Center of Gravity

Population is the most certain of the three variables. The development of AI may encounter technical bottlenecks, and the energy path may experience unexpected breakthroughs or delays, but there is almost no uncertainty about the future of the population. Because the people who will determine the labor force and demographic structure twenty or thirty years from now have already been born today—or have determined that they will not be born. Once the fertility rate falls below a certain threshold, it cannot come back within decades. This is not a policy issue, it is social inertia. Few countries have successfully reversed declining fertility rates.

The global population map in 20 to 30 years has been basically determined today: the labor force in Europe and East Asia will drop off a cliff. According to United Nations population projections, China's working-age population will drop from approximately 980 million in 2024 to approximately 750 million in 2050, a decrease of nearly a quarter. Japan and South Korea have entered deep aging. At the same time, youth populations in South Asia, Southeast Asia and Africa will continue to grow. India's working-age population will grow by about 140 million over the same period, making it the world's largest labor market. Africa's population will double, and many African countries will enter the demographic dividend window. The center of gravity of the global population will clearly shift from the north to the south.

Population growth itself is a driving force. More people means more demands, more transactions, more trial and error, and more possibilities. A society with a majority of young people, if there are basic economic opportunities to take over, will be a hotbed for entrepreneurship and risk-taking, and the rhythm of the entire society will be forward; if not, the same population energy may also turn into turmoil. The opposite is true in an aging society - no matter how advanced the system is, a defensive temperament and a preservation mentality prevail. Culture certainly plays a role, but generational structure is a lower level variable.

In the past two hundred years, this power has experienced several major shifts. In Europe in the nineteenth century, it turned to North America in the first half of the twentieth century, and to East Asia in the second half of the twentieth century. Every shift is accompanied by a reshaping of the world structure. In the next twenty to thirty years, this power will shift to South Asia and Africa. This is not a prediction, but an extension of the timeline of what has already happened.

In the AI era, the shape of the demographic dividend has undergone fundamental changes. In the past, the demographic dividend was "cheap labor" - a large number of young people were willing to enter factories at low wages, providing cost advantages for global manufacturing. This dividend is indeed depreciating in the face of AI. But the new form of demographic dividend has not disappeared, but has become more important. Population means the consumer market - as income levels increase, South Asia and Africa will contribute most of the world's increased consumption. Population means talent base - extreme talent is a probabilistic event. The larger the population base, the greater the absolute number of geniuses appearing on the right end of the distribution. However, whether this potential can be realized still depends on whether basic education and the institutional environment can allow geniuses to be identified and cultivated. Population also means the fiscal tax base - the total amount of resources a country can mobilize, ultimately depends on how many people are creating and exchanging value. Whoever has the largest market will have the strongest demand gravity; whoever has the largest talent base will be most likely to produce leading innovations - provided that the institutional environment allows talents to stand out.

Looking at the plight of aging countries from the other side, it is deeper than what appears on the surface. The shrinking population not only means a shortage of labor - which may be partially offset by AI - but also means a continued contraction in domestic demand. The growth assumptions of all economic models over the past few decades have been based on market expansion. After the population U-turn, it is not that the growth rate has slowed down, but that the growth paradigm itself has failed. What Japan has experienced in the past thirty years is that policy mistakes are only superficial reasons. The deeper reason is the confrontation with the irreversible physical reality of demographic structure. Pensions and medical expenses will eat up an increasing proportion of the fiscal budget, squeezing investment in education, infrastructure and national defense. National capabilities are locked up when transformation is most needed - not only financially, but also mentally: aging societies tend to be conservative and rigid, and lose the ability to take risks just when they need to embrace change most.

And a more basic fact should not be ignored: South Asia and Africa still have a lot of industrialization and modernization "debts" to pay. The urbanization of billions of people, infrastructure construction, the establishment of manufacturing industries from scratch, and the improvement of public service systems—these historical processes that have been completed in developed countries have only just begun or are far from completed in southern countries. Just catching up on these lessons is enough to unleash huge economic growth momentum. This growth does not rely on globalization channels, but is endogenous and structural. Even if changes in the political system are not taken into account, industrialization and modernization alone can bring profound changes to these regions that will last for decades.

But the endogenous power of industrialization has its upper limit. To become a global force, domestic demand alone is not enough. And precisely at this point, there is a huge tension. The population power is moving southward, but the channels that carry it are closing. In the past, East Asia successfully transformed demographic dividends into economic growth by relying on the channels provided by globalization - cheap labor was integrated into the world's division of labor system through export processing, and gradually accumulated capital and technology. But this channel is being blocked by multiple forces: AI has eliminated the economic logic of labor arbitrage, energy localization has weakened the cohesion of international trade, and political polarization has closed the door to immigration.

Southern countries have growth momentum, but they may not be able to find the exports to release it. There is a historic dislocation between power and channels. Looking back at every power shift in the past - from Europe to North America, and from North America to East Asia - it took decades for the recipient to truly rise, experiencing wars, revolutions, and institutional reconstruction. Not once was it a smooth transition. This round in South Asia and Africa will most likely be no exception. Whether we can find new ways to transform the huge population energy into national capabilities and international influence is the biggest uncertainty among the variables of population, and it is also a key issue in determining the future world power map.

Conclusion

Three variables are irreversible and evolve independently, but they act on the same world at the same time. AI eliminates the middle layer, energy eliminates resource dependence, and population reshapes the bottom distribution of national power. Every pillar of the current international order—the globalized trade system, the geopolitical landscape dominated by fossil energy, the structural advantages of northern countries, and the social stability supported by the middle class—will be redefined.

The three variables also have a common direction: they are all weakening the power of the "center". AI has weakened the advantages of large organizations over small teams. Energy localization has weakened the control of resource exporting countries over importing countries. The collapse of globalization has allowed power to be recycled from the global system to regions and countries. The world order of the past two hundred years was based on the logic of centralization—big factories, big companies, big countries, and a unified global market. What shakes these three variables simultaneously is the logic of centralization itself.

Don't try to predict the final state. But something is certain: Everything we take for granted today—globalization, middle-class society, the fossil energy order, the world structure dominated by the North—is not stable on the scale of twenty or thirty years. Seeing the direction of each of these three lines is enough to reexamine all of our certainties today.