I've been wanting to write a small article about the investment prospects of the AI industry based on $Microsoft(MSFT.US), but haven't started yet. $Meta Platforms(META.US) has already revealed some of the concerns I wanted to address. Currently, the pricing of the AI industry is not based on the present but on sustained performance prospects. This is the premise for understanding pricing: the market is rushing to buy AI, not for current performance, but because its growth is so rapid that it requires scrambling. Once the logic of scrambling weakens, what follows is a correction or even a decline.

Meta's performance can be summarized as follows: current (Q1) performance is good, but AI expenditures are higher and faster than expected, while revenue growth is lower and slower. As a company in the mid-to-lower stream of the AI industry chain, Meta's AI business will face "reduced efficiency and increased costs," inevitably casting doubt on the future of cloud and other AI businesses of more upstream companies like $Amazon(AMZN.US), $Alphabet(GOOGL.US), and $Microsoft(MSFT.US). If the midstream is unsustainable, questions arise about whether data center businesses can continue to grow and profit, and whether institutions still need to scramble to buy. It's not yet time to call AI a bubble, but perhaps the rush is no longer necessary.

Let’s see how the post-market performance outlook of Google and Microsoft, the two major cloud service providers, sets the tone for Q2—not making grand conclusions, just laying a small coffin.