They eventually sold the rope that would hang them.

The collapse of the rentier economy. It used to be said that "capital is cannibalistic," which I initially thought was a metaphor, but after being imprisoned, I realized they were serious. It used to be said that "capital will sell the rope that strangles it," which I also initially thought was a metaphor, but after the "AI revolution," I realized this is also true. Recently, a little-known research firm in Silicon Valley, Citrini, released a 7,000-word research report, "The Global Intelligence Crisis," triggering a bloodbath in the US stock market. IBM suffered its biggest single-day drop in 25 years, plummeting 13.1%. Software stocks such as Microsoft, Oracle, and Accenture all plummeted, and financial stocks Visa, Mastercard, and American Express were not spared either. Delivery platform DoorDash, private equity firm Blackstone Group, and KKR all suffered losses exceeding 8%. This report didn't say anything particularly new, but it did point out one thing: the AI ​​bubble will burst before it bursts its own, triggering the US GDP bubble first. I previously wrote an article arguing that we may never catch up with the US GDP. Because in the US's numerical games, according to their rules, we will never win. The reason is simple: all economic activities require electricity. In terms of electricity consumption, China's economy is already twice the size of the United States'. In terms of newly installed power generation capacity, China's economic development is several times faster than that of the United States. It can only be said that the US Census Bureau has a clear advantage. Of the US's $25 trillion GDP, consumption contributes $21 trillion, accounting for 82% of total GDP. Of this $21 trillion, only $5.9 trillion (23% of GDP) is actually spent on physical goods, while non-physical consumption reaches a staggering $11.4 trillion (44.8% of GDP). In other words, half of the US's massive GDP is comprised of "services." This is what is known as the US's "ghost GDP." These are the outputs reflected in national accounts but never actually entering the real economy. Almost all of these are so-called high-value service industries. What does this include? Partly, it includes knowledge workers, such as programmers; but a larger part consists of America's unique "brokerage industry": such as lawyers (actually legal brokers), accountants/tax advisors (financial brokers), insurance brokers, financial brokers, real estate brokers, consulting (business brokers), and so on. These knowledge workers and brokerage industry practitioners, although they are a "rent-seeking class" who profit through licensed monopolies or de facto monopolies, their work is difficult to standardize, forcing the capital they rely on to employ a large number of people to complete the related work. According to US statistics, these white-collar workers account for 50% of the US workforce but contribute 75% of disposable income, making them the main consumers in the US, and consumption is the main driver of brokerage in the US. However, due to the widespread adoption of AI, large capital can use AI to replace these human workers in these brokerage industries. As profit-driven capital, it naturally no longer needs these "rent-seeking" peripherals to share profits. The replacement of software industries and SaaS by AI only impacts some programmers; but the replacement of these "rent-seeking" peripherals has a much greater impact. For example, in insurance brokerage, AI's quotes are accurate and there's no need to worry about "information asymmetry" profiteering, so why would there be a need for human brokers? It can calculate the bottom line for dozens of insurance companies in a second, so Willis Towers Watson fell 15%, impacting the entire US insurance industry. For example, tax advisors and accountants: AI provides tax avoidance rules quickly and reasonably, without any emotion, and charges only one percent of human fees, so Charles Schwab and LPL plummeted 7%. Real estate agents: US real estate agencies charge a 6% commission, plus various miscellaneous fees, making a fortune. With the arrival of AI, AI agents equipped with real estate databases and decades of transaction data can instantly replicate professional knowledge. With white-collar workers gone, commercial real estate naturally collapsed, with Jones Lang LaSalle and CBRE plummeting 20%. Previously, companies that relied on their competitive advantages and first-mover advantage to reap monopolistic profits have all been wiped out by AI. The most typical example is IBM, which, with its COBOL language developed sixty years ago, became the universal language of the global banking industry. Because it was difficult to replace, its mainframes could passively generate revenue from the banking sector. Now, Claude has released an AI tool that can directly convert this outdated language into a commonly used programming language, instantly eroding IBM's competitive advantage. Even Indian outsourcing is suffering, because AI is now more user-friendly than Indians and doesn't have an accent… Another example is Nvidia, which relied on its CUDA ecosystem as a competitive advantage, almost exclusively enjoying the benefits of GPUs. However, because AI tools can automatically convert the CUDA ecosystem into a non-exclusive language, Nvidia's monopolistic profits may be fatally impacted. These industries have begun large-scale workforce reductions, which will not only affect consumption but also financial stability. These individuals are the backbone of the US mortgage market, bearing over 80% of all mortgages. If their jobs are replaced by AI, mortgage default rates will rise rapidly, directly triggering a subprime crisis 2.0. The income of high-quality borrowers will be permanently damaged, potentially causing the collapse of the $13 trillion mortgage market. Home prices in tech hubs like San Francisco and Seattle have begun to fall. The pressure has spread from the housing market to private lending and life insurance companies, triggering a systemic pricing disaster. Unlike the 2008 subprime mortgage crisis, the problem this time is not poor loan quality, but rather a structural error in the assumptions about future income made by white-collar borrowers due to the impact of AI. This is more difficult to solve than subprime 1.0. The private lending market also faces the risk of collapse. For example, Zendesk defaulted on $5 billion in loans because AI agents replaced customer service, becoming the largest private lending software default in history. Blackstone and Blue Owl both suffered huge losses due to lending large amounts of capital to software companies and commercial real estate, freezing investor withdrawals. Therefore, the AI ​​revolution hasn't impacted producing countries yet, but rather triggered the collapse of America's unique "rent-seeking economy." The American consumer economy and credit market, built upon these rent-seeking classes and their surrounding communities, will also suffer. Technological revolutions can bring about social revolutions, which is precisely the fundamental reason why various states in the US are actively opposing AI, even sparking anti-AI sentiment. Will capital, in its pursuit of final profits, abandon AI, this golden goose? Developing AI will cause a bubble to burst, and not developing AI will also cause a bubble to burst; everyone is only calculating their own gains, ultimately leading to the rapid collapse of the entire system. Using technological progress to eliminate the rentier class has happened many times in history, but unexpectedly, this time it has happened to the United States itself. The disappearance of the rentier class will directly breach the "consumption base" and "credit base" of the American economy, triggering a chain reaction. Last time, the steam engine replaced slavery, and the Civil War killed 10% of the American population. Will the same story unfold this time, with AI replacing white-collar workers?