Here it comes, here it comes, the AI takeover hero has arrived.

It is said that the mountains are heavy and the waters are deep, casting doubt on the path; yet, when the willows darken and the flowers bloom, a new village emerges. You guys keep shouting that the AI bubble is about to burst, but in the end, you still have to wait a little longer. After OpenAI couldn't make ends meet, and Ultraman cried to the White House a few times, suddenly the news reported that princes from the Middle East brought hundreds of billions of dollars in cash to invest. Goodness, UAE capital is indeed generous and exceptionally easy to talk to. The so-called bottomless pit of investment, OpenAI, searched the whole world but couldn't find anyone willing to invest real money. Even Nvidia, which played along to boost its market value to $5 trillion, publicly stated that it would promise to invest hundreds of billions in OpenAI, but that was just talk; they wouldn't really give the money. When this news broke, it triggered a collective plunge in American tech stocks.

Among them, Oracle, which was misled by Sam Ultraman's grand vision and went on a borrowing spree to build data centers, suffered the most. Their so-called benefactor ran away, and the promised $10 billion earmarked for building data centers was never delivered, causing Oracle's stock price to halve from its peak. Benefactors who can easily pull out billions are not foolish; the notion that 'fools have plenty of money' is something we poor souls conjure up at home. For example, Blue Owl Capital, which had promised to invest in Oracle, is a very savvy private equity fund. Oracle's building of data centers was just a side investment, gaining shares in return. Making AI infrastructure can cause Oracle's stock price to plummet; it doesn't need to profit from these data centers, just need to boast in the market, and when the stock price soars, they can cash out in the secondary market and make a fortune.

So you see, capital plays investment very practically; small amounts can be gambled like dice, but when it comes to large sums, every account is calculated clearly. You must first find a buyer; otherwise, there's nothing to discuss. Oracle wasn't abandoned because its debt ratio was too high; of course, Oracle's debt is very high. Over the years, in order to climb higher on the wealth list, the octogenarian Larry Ellison went all in, crazily investing in data centers. Of course, investment also brings returns. Originally dealing in databases, Oracle once boosted its market value to a trillion dollars through this effort. However, the American AI model has encountered bottlenecks now.

In the Sino-American tech war, if we focus on GPU chips, we are certainly suppressed. How can we compete in others' systemic pathways? All the rules are set by them; no matter how well we perform, we are merely feeding their ecosystem. However, Tsinghua University is taking a different path; we are developing open-source large models, welcoming the world to use them. Our AI closely aligns with productivity and can quickly generate returns from the industry, forming a positive cycle of profitability. This significantly harms the AI model led by OpenAI. Investors are starting to believe that models like Google's might be better, as it inherently has a large user base, making monetization easier and the story easier to understand.

Thus, the market began to abandon OpenAI's circle of friends, and companies like Oracle are now being scrutinized for their debt repayment capabilities, which suggests they are no longer viewed as tech companies. Not long ago, in 2018, the entire American tech community was examining Tesla's accounts, gambling on when Musk would go bankrupt. Is anyone still discussing this now? When a group of people uses traditional methods to judge a developing tech company, it only indicates one thing: everyone believes this industry logic is failing, has no future, and is about to collapse. Otherwise, who looks at balance sheets? Aren't they just dreaming of market cap? So at the most dangerous moment, the princes from the Middle East grouped together to pour in money. Are they really just fools with too much money?

When someone points a gun at your forehead, you might do the most foolish things. A few days ago, I praised the American CPI data on Weibo for being sensible, obediently dropping to 2.7, then magically the unemployment rate also came up, just enough for the Federal Reserve to find an excuse to lower interest rates and expand their balance sheet. With the printing presses rolling, they can counteract the impact of yen interest rate hikes, and the liquidity released can help save the American AI bubble. At this time, the princes came with hundreds of billions in cash, taking a stake in OpenAI at a ridiculously high valuation exceeding $800 billion, a miracle in the history of world financing. Interestingly, at this time, prices of non-ferrous metals surged continuously, while oil prices kept dropping, with the commodities market swinging wildly. What is America doing? Launching a heavy offensive to beat up poor Venezuela, which holds the world's largest oil reserves.

The money from the princes in the Middle East has always held multiple meanings. If giving money can raise oil prices, would you not spend it? What if it has additional effects? For example, does Israel have magical offensive and defensive functions? It's not that OpenAI is particularly powerful, but rather that it is currently too important. If it collapses, the previous AI stock price structure will also crumble, and who knows how many big investors like Blue Owl will hasten to pull out. The consequences are something the Americans dare not gamble on. Hence, at a critical moment, they have to invite the princes to sponsor. As for returns, at a certain level, money is more of a symbol. When the decree comes down, who would dare to say no?

Long ago, I mentioned that America does not want, nor dare, to let the AI bubble burst for now; temporarily it still has this capability. So this time it was UAE capital; next time it might well be a Japanese consortium. Who knows? The root of the problem still lies in a very simple question: can AI find a reliable, long-lasting, and feasible profit model? At least the Americans currently cannot answer this. Fortunately, with ample funds, they can hold on for a while longer; at worst, they can just print more when the time comes.