To The Moon

Quick facts check:

Is Alibaba Chairman in an executive position? No, Eddie Wu Is the CEO.

Is Alibaba Chairman from a technical or AI background to know better? No, he used to handle Alibaba’s financial and law structure, and doesn’t have experience in technology.

Did Alibaba increase AI spending in 2025? Yes

And did Alibaba buy Nvidia chips, or make their own chips? They make their own chips so obviously they want to claim US is in a bubble.

- Chinese firms are also increasing their AI spending, then are they also in bubble too? Yes if we apply’s Joe Tsai’s warning then China is also in a bubble. Better to sell Alibaba stocks.

- Is Alibaba Chairman a credible source on AI spending? No, US’s downfall will help Alibaba to sell their chips. They wouldn’t be making those chips themselves if it is a bubble.

- Has there ever been a decline or bubble in infrastructure spending?

No, hyper-scalers tend to spend a lot on servers and data centers historically. The same is true for all big tech companies back in 2010. Spending on IT and infrastructure has been growing for the past 20 years and hasn’t stopped. From the early Dotcom days of servers to the cloud computing boom—and now to the AI boom—infrastructure spending has never shrunk but has continually increased from ($30-50billion per year to $212-220 billion per year since 2010).

There hasn’t been an infrastructure spending bubble (not even during the Dotcom bubble) because infrastructure usage isn’t restricted to a specific industry. It powers the needs of all businesses across the spectrum.

- Is the case of chips inventory renewal a big issue?

Not only for providing AI services but also for their core services, hyper-scalers hold the world’s data. Nvidia chips power accelerated computing, and older chips can still handle current CPU workloads. Developers are also using older chips to run and build older and smaller AI models, while new chips allow for bigger and more powerful model to be created. Revenue and profits offset long-term inventory renewal. Spending on AI chips is not substantial when considering the size of these companies, and investment in servers and server equipment has been essential since the CPU era.

Just taking a look at Meta:

- Meta spent over $100 billion on the Metaverse (VR), yet it generated only $2.3 billion in revenue.

- Their current investment in AI is roughly equivalent to the Metaverse expenditure, but the returns are far superior—a 21% year-over-year increase reached $48.39 billion in the fourth quarter of 2024. The revenue is already offsetting part of their spending.

- Is there even a case for OpenAI and Meta (those working on AGI) to stop buying new chips?

No way, the development of newer models are more and more compute-power demanding. Never ever in the history of technology that we have stopped at developing CPU capabilities, it has always been the case of not powerful enough. It has never been the case of we have enough or too much compute power. If that is the case, Nvidia would never exist.

Fake news?

None of the current news make any sense or have any rationale. Instead of celebrating the growth of Nvidia’s competitive advantage, the fake news are selling the prospects that we have enough or too much compute power. This has never been the case in technology and never would be.

Obviously like DeepSeek, this is another Chinese comment aimed at downplaying and hitting back at US technology to boost their own narrative. Market reaction is too much as it doesn’t consider political biases. Likewise, they did make a similar claim to their own semiconductor industry? Or just taking aim at US specifically?

Here they are selling BYD for having very advanced technology, but none of the Chinese investor claim that BYD is too advanced and will be a bubble when BYD is leading. Whereas when Nvidia is too advanced, they claim it is too advanced and there is a bubble because too many people are buying?

Obviously this is Market Manipulation 101.