China’s tech giants — Alibaba, Tencent, and Baidu — were all founded between 1998 and 2000. By the end of 2000, the number of internet users in China had jumped from 3000 in early 1995 to 22.5 million. In 2001, China joined the WTO. Urbanization accelerated, and the growth of the middle class fueled demand for luxury goods, tourism, and better nutrition. The number of private cars in China went up from 1 million in 1992 to almost 10 million by 2002. Many people envisioned a hopeful future in which they could acquire new clothes, new luxuries, and new technology in the new millennium.
But the “many” did not include the 100 million people residing in the Northeast — roughly 8.5% of China’s total population as of 2000. By the 1990s, urban shrinkage, which is measured by sustained population loss, had already taken hold across 52 cities in the Northeast. And of the 68 cities across China whose populations diminished continuously into the 2010s, half were in this region. The regional birth rate has been trending lower than the national average for more than three decades, and net outmigration has become an increasing problem since 2000. In 1990, the Northeast represented 8.66% of the country’s population; by 2016, that proportion had dropped to 7.9%. The one-time cradle of China's industrial development has become a place that many would rather not raise kids or live in, given the choice.
In the span of a decade, Chinese society simultaneously experienced rapid economic growth and extreme economic precarity. Individuals were offered transformative opportunities and faced catastrophic crises, all due to the same factors put in place by a select elite who generated the incredible promise and acute challenges modern China still faces. To many Americans watching AI reshape their economy, this narrative may sound familiar, though calls to regulate, pause, or stop the technology reflect a belief that the transformation can still be steered or stopped. That option did not exist for Chinese workers in the 1990s.
[...]
For China’s policymakers, slowing development was never an option. A 1931 quote from Joseph Stalin — “落后就要挨打 (luohou jiu yao aida) or “those who fall behind get beaten” — that adapted by Mao Zedong in 1956 permeated society, serving as a cornerstone of high-level policy narratives. In China’s mnemonic practices, this phrase, linked to the idea that only development can sustain a nation’s independence, is the most significant lesson from the past, necessary to remember from China’s 20th-century history of war and colonization. “The reform is painful but rewarding,” wrote the state in 2012 in reference to the previous century.
At the turn of the century, then, the policy question was therefore not whether to reform; instead, it was how to make the transformation less painful. The government attempted to address the pain. In 1998, the state established re-employment Service Centers, which provided laid-off workers with living allowances, basic social security, and job training. The state taxation administration introduced tax incentives for businesses that hired displaced workers. Xiagang workers were entitled to tax exemptions, fee waivers, and preferential access to microloans when starting small businesses or seeking new employment. The Minimum Living Security System was established in 1999 to guarantee basic income for urban residents and expanded to rural areas in the 2000s. Higher education grew in 1999 and university attendance increased 600% in less than 10 years. This expansion was partially aimed at delaying China’s youth from entering the job market, thus leaving spaces for the re-employment of laid-off workers.
For some workers, these policies provided a bridge. But the scale of the problem overwhelmed the response. Funds were too small or simply did not arrive. When funds did arrive, they rarely reached the people they were meant for. In one case, one former deputy director of the city-level Development and Reform Commission — an institution responsible for implementing national economic policies — embezzled the subsidies of 556 xiagang workers.
Even as market reform and industrial upgrades brought new job opportunities, there were simply not enough: In 2004-2005, 24 million people entered the workforce, but only 9 million new roles were created. Even within these new jobs, there was a mismatch between supply and demand. The workers who had been laid off were predominantly in their forties and fifties with industrial skills, while the foreign companies entering China wanted fresh university graduates or young rural migrants who were willing to work for less. And though the expansion of higher education benefited many, it eventually produced young workers who were overqualified for many jobs, resulting in high youth unemployment that persists in China today. And much of the suffering was silently buried under cold numbers and grand policies.
[...]
If economic restructuring was an unstoppable force of nature, then the only possible response was to move with it before it moved without you. Xiang Biao diagnosed this as a "last bus" mentality: a collective fear that missing the opportunity to seize a piece of post-socialist accumulation meant missing everything. You either catch this bus towards success or be left out forever. It was a frenzy born not of greed or enthusiasm, but of the desperate realization that the old world was gone and the new one had no reserved seats. What began as a northeastern industrial experience has, amid decades of social change and competition, became a prevalent psychological structure spanning different socioeconomic classes and regions.
[...]
In the late 1990s and early 2000s, learning English was the last bus. Globalization was the irreversible trend; only by learning English could Chinese people interact with the greater world. The state mandated English education as a core Gaokao subject and pushed it into primary schools in 2001, giving rise to cultural phenomena like "Crazy English" (疯狂英语; fengkuang yingyu), wherein tens of thousands of people gathered in public stadiums to scream English phrases at the top of their lungs in a desperate collective bid for fluency. In the late 2010s, the mobile internet boom was the last bus. As tech giants like Alibaba and Tencent offered unmatched salaries in other industries, millions rushed to learn coding and enroll in computer science degrees in universities that were aggressively expanding computer science programs, only to find themselves facing a constantly decreasing employment rate.
In 2023, understanding AI was the last bus, and over 250 thousand people paid for rudimentary AI crash courses, terrified of being rendered obsolete overnight. In 2026, OpenClaw was the last bus, with thousands of people — retirees, white-collar workers, housewives — lining up outside tech company offices for engineers to install the agent directly onto their phones.
[...]
When polled only three decades later, perhaps every respondent genuinely believes that AI is good for both society and for themselves. Or perhaps they see AI as another surgery necessary to survival, knowing full well that flesh will be cut away and discarded, but convinced that the pain borne by individuals — however devastating to them — is small against the benefits at large. The polls, as they are written, cannot distinguish between these narratives. .
From the article:
[...]
[...]
[...]
[...]