For me, the next step isn't just about whether Cap

For me, the next step isn't just about whether Capex will increase further,

but about three things:

1. Whether enterprise customers continue to expand;

2. Whether AWS's growth and profit margins can be maintained;

3. When cloud providers' operating cash flow will start to support this round of investment.

After reading the earnings reports of Microsoft and Amazon,

I actually feel: AI infrastructure is even more certain than last quarter.

Not because the numbers look good, but because the people paying have changed.

Microsoft said,

about 90% of Microsoft Cloud revenue comes from customers outside frontier model companies,

and all new commercial RPOs this quarter also came from these customers.

It's not just frontier model companies like OpenAI burning money,

but a broader range of enterprise customers are also paying.

On Amazon's side,

AWS accelerated from 28% to 37%,

AI and chip businesses each generated over $25 billion in annualized revenue,

with order backlog nearing $500 billion.

Amazon's free cash flow is negative,

and it expects to invest approximately $220 billion in cash Capex for expansion this year.

But this is exactly why I am more confident:

As long as enterprise payments continue to spread,

today's construction spending has the potential to

become thicker cash flows after the peak passes.

Thinking one step further:

Enterprise payment → Cloud provider revenue growth → Continued investment in hardware and power → More orders returning to the supply chain.

We still need to continue observing how much efficiency gains AI can bring to each industry,

but the step of "enterprises willing to pay for it" is already much clearer than last quarter.

What do you think? 🤔