According to Asia Times on October 22, the development of the U.S. AI industry has moved away from technological progress and turned into a game of capital, creating a phenomenon of circular transactions.

The most typical example is the closed loop of funds formed between tech giants such as NVIDIA, Microsoft, OpenAI, AMD, and Oracle: NVIDIA invests billions of dollars in OpenAI, and OpenAI then uses this money to purchase NVIDIA chips; Oracle provides cloud services to OpenAI and also spends huge sums to buy NVIDIA equipment; AMD and OpenAI hold shares in each other, and Microsoft, in exchange for cloud computing power, invests in OpenAI.

This series of operations may seem like building an interconnected AI empire, but in reality, it is just a phenomenon of capital circulating within a circle.

What is called circular transactions, in simple terms, is when upstream companies invest in downstream ones, and the downstream ones use that money to purchase products from the upstream, thereby inflating both sides' revenue and valuation, creating an illusion of prosperity.

It is not substantive value creation, but a financial self-cycling mechanism, similar to the concept of a snake eating its own tail in finance.

In the AI field, this means that transactions between chip manufacturers, cloud service providers, and model companies are increasingly detached from market logic, becoming bubbles inflated by capital, appearing profitable on the surface, but actually relying on mutual blood transfusion to keep each other afloat.

OpenAI

This "circular transaction" in the U.S. AI industry is a sign of bubble formation.

On the surface, it demonstrates the vitality and collaboration of the AI ecosystem, but in reality, it is capital engaging in a fight with itself.

On one hand, these transactions inflate revenue and valuation, making investors believe that the AI industry has entered a high-profit phase, thus attracting more capital into the market;

On the other hand, this binding makes the fate of the entire industry highly interrelated, so that if one company faces problems, the whole chain could collapse.

These transactions are often packaged as strategic partnerships or ecological construction, hiding real financial risks.

It gives people the feeling of a Ponzi cycle in the AI world, and once the capital chain breaks, demand drops, and investors withdraw, the AI bubble will burst like the Internet bubble of the past.

Although NVIDIA's stock price is still high, nearly 90% of its revenue comes from AI data centers, and if clients such as OpenAI or Microsoft cut spending, the risk will be immediately transmitted.

Meanwhile, OpenAI itself is still in long-term losses, and its procurement of computing power and construction of data centers rely on financing support. This model of mortgaging future profits for current expenses is a typical high-leverage bubble operation.

It is a gamble, and it also indicates that the U.S. AI industry has already shown signs of being unsustainable. If the business could naturally grow, why would they blow bubbles?

NVIDIA

By contrast, China's AI industry has hardly seen such乱象 of circular transactions, and its overall development is clearly more stable and solid.

This is because the funding structure of Chinese AI companies is more prudent, and they mostly rely on mature business ecosystems to generate their own income, rather than relying on capital runs.

Moreover, the development logic of China's AI is to pursue technology and industry together, rapidly implementing in fields such as intelligent manufacturing, government systems, education, finance, and healthcare, pursuing real benefits, rather than simply chasing valuations.

Every step of China's AI is a result of accumulated strength, without inflated market values, seeking solid technical accumulation, and relying on real collaboration across the industrial chain.

In short, China's AI path is more like a national project, emphasizing long-term strategy and industry ecology, rather than a game of capital.

Altman Blowing Bubbles

That is why China's AI industry is more likely to last longer, while the U.S. AI boom has returned to its usual logic, trying to make a quick profit through short-term speculation.

This model of supporting valuations with capital and stories will collapse once the capital leaves.

That is also why U.S. AI companies are mainly To C models, as their goal is to pursue user scale, quickly inflate bubbles, and then monetize them.

While Chinese AI companies are more To B-oriented, focusing on solving industrial pain points and improving production efficiency. Even if growth is slower, they have stronger risk resistance and healthier profit structures, which naturally allows them to come out on top in the end.

Original article: https://www.toutiao.com/article/7564303399689208359/

Statement: The article represents the views of the author. Please express your opinion below using the [up/down] buttons.