AI is probably the same.

I recently finished reading a book related to the Internet, *The Trap of Algorithms*. Its authors, Ariel Ezrachi and Maurice E. Stucke, are law professors at the University of Oxford and the University of Tennessee respectively. This is a critical book in the traditional sense, with a familiar left-leaning style that carries the aloof sense of being the only sober one in a crowd of drunks.

The book itself is not highly controversial. Its core argument is that algorithms and data are changing the nature of market competition, seemingly benefiting consumers but potentially causing actual harm. Judging from its content, it is just novel enough, and for readers like us who do not know much about the major Internet enterprises in the US stock market, it serves as a mirror.

What Has the Internet Changed?

After the emergence of the Internet, the changes to daily life are subversive, especially in the United States.

The first change is online sales.

"In the life insurance industry, the prevalence of Internet price comparison sites led to an 8%-15% drop in the price of life insurance policies in the 1990s, while consumer surplus increased year by year at a level of 115 million to 215 million US dollars."

"In the era of Walmart's frantic expansion, the harm it brought to regional businesses was no less than an earthquake. Once Walmart chose to open a store in a certain place, local small shops often had to surrender and give up their sales, and the prosperity of the main street faded away from then on."

"When Walmart began to lose its dominant position, who took its place? The answer is undoubtedly Amazon." "The vast majority of products sold on Amazon can adjust their prices every 15 minutes."

The second change is the comprehensive monitoring of consumers and the transformation of advertising delivery methods.

"In 2013, 28% of advertisements in the global digital display advertising market were delivered through programmatic methods. By 2017, this proportion accounted for 60%-70%."

The New York Times once recorded such a story: A middle-aged father angrily went to a supermarket manager to demand an explanation: his daughter was still in high school, but the supermarket had sent her coupons for baby clothes and cribs. The manager called the man a few days later to apologize, only to find the man rather embarrassed, admitting that his daughter was indeed due to give birth in August that year.

There is another more interesting story: "At an important board meeting, a senior executive looked flustered and eager to leave the meeting. Others hurriedly asked him why. The executive pointed to his mobile phone and said: 'I just received a series of advertisements for alarm devices and security systems. I think my house might have been broken into.'"

Price Discrimination That Targets Different Consumers Differently

Theoretically, price discrimination has its economic foundation.

"For most industries, 20% of consumers contribute 80% of the profits."

"Economic theoretical research shows that when commodity suppliers can effectively distinguish different consumers (that is, carry out price discrimination), they will maximize their profits."

Charging different prices to different consumers can extract more consumer surplus along the demand curve. The core of this theory is to set higher prices for customers with relatively rigid consumption demand, and offer more competitive discounts to consumers who are less willing to spend.

For example, "This insurance company (Allstate Insurance Company) is issuing bills with different prices to customers by 'observing' whether customers will go elsewhere to compare prices when buying insurance". Another example: "According to a 2015 social survey, the online tutoring service of Princeton Review's college preparation courses in Asian-American communities is 1.8 times more expensive than that in other regions. The survey also found that even in Asian-American communities with lower income levels, the price of the courses is still significantly higher than in other regions."

The High-Tech You Imagine Is Never That Elegant When Implemented

The Magnificent Seven of the US stock market: Apple, Microsoft, Google, Amazon, Meta, Tesla, and NVIDIA, have contributed more than half of the gains of the US stock market for a long period of time.

There is also a saying in the investment circle that technology stocks in the US stock market, high dividend stocks in the Hong Kong stock market, and micro-cap stocks in the A-share market are the three long-term sources of excess stock returns. The Magnificent Seven of the US stock market have always been the representatives of technology growth stocks.

But different from everyone's impression, among the seven giants of the US stock market, except for NVIDIA, the remaining six are essentially consumer stocks.

"In the final analysis, Google is a super platform that generates revenue relying on advertising." "Google pays Apple nearly 1 billion US dollars every year to ensure that the latter sets its search service as the default search program on Apple devices."

The search engine industry has to make profits to survive after all.

"Google's data statistics and analysis tool Google Analytics will provide its partners with various analysis reports based on these user data, helping the latter to test the effectiveness of their advertising marketing and user experience."

"However, it is undeniable that search engines will inevitably reserve a prominent advertising position for their 'sponsors'. In practice, search results are divided into two categories: sponsored results and organic results. For paid sponsored results, advertisers pay search engines based on the number of clicks on the advertising links."

"Robert Epstein also pointed out that Google not only manipulates information dissemination for commercial motives, but also actively manipulates public opinion to interfere in political affairs."

"Like Google, the vast majority of Facebook's revenue comes from the advertising business, and this proportion is increasing year by year. In 2012, 2013 and 2014, this revenue accounted for 84%, 89% and 92% of the total revenue respectively."

As for Amazon, as we all know, online retail is its main business.

Therefore, the long-term bull market of the US stock market is essentially a long-term bull market driven by consumption, not a victory of technology. After rising for so many years, the latest price-earnings ratio of the seven tech giants of the US stock market is only 25.7 times (Wind US Seven Tech Giants Index, October 30, 2026).

What Has AI Brought?

The US stock market's third-quarter earnings season also presented a familiar scenario similar to that once seen in the A-share market: blue-chip stocks plunge sharply, bringing huge market volatility.

Although Google's financial report is quite good, its free cash flow turned negative, and its stock price fell by more than 10%. As for Intel, due to the sluggish PC market caused by the skyrocketing prices of memory and chips, its stock price plummeted by 28% directly.

All this illustrates a fact: the market is getting less and less patient with AI capital expenditure.

The logic that "the US stock market will keep rising forever" has not been broken so far, because the US stock market is essentially a consumption-driven market. With monetary easing comes consumption, and consumption brings profits.

However, excessive AI capital expenditure has undermined the foundation of this logical chain. The trillions of US dollars in annual AI capital expenditure will require subsequent revenue at the level of tens of trillions of US dollars per year to cover. But where exactly this revenue will come from is still unclear for now.

There is a blind spot that many people do not understand: why several US tech giants, especially Google, have been persisting in investing in AI.

Because search is Google's traffic entrance and the foundation of its existence, and AI is subverting all this. Originally, consumers turned to Google to query everything, but now they are starting to turn to AI. Although the sign is not obvious, once the inflection point comes, Google will only go downhill from then on.

This is the problem Google is facing now: it may lose all its investment if it pours money into AI, but the risk of not investing is even greater. Apart from Google, the other several giants are more or less facing the same dilemma.

These things have happened over and over again in history: from the British canals in the 18th century, the railways in the 19th century to the Internet bubble in 2000, almost all major technology investments led by private capital have gone through the "boom-bust" cycle.

Canals, railways and optical fibers are left behind, but most of the people who invested in them lost their money.

AI is probably the same.

This article is written based on public materials, for information exchange only, and does not constitute any investment advice.

This article is from the WeChat Official Account "Jinduan" (ID: jinduan006), author: Yu Laotou, published with authorization from 36Kr.