What breaks first if the 10-year stays this high: AI valuations or AI capex?

I've been thinking about these as two separate questions that often get lumped together.

**1. Does AI spending continue?**
Probably a question of competitive necessity, demand and access to capital.

**2. What are those future cash flows worth to shareholders?**
That's where the discount rate matters regardless of how important AI becomes.

A company can be completely right about AI and the stock can still have been priced for returns that become much harder to justify at higher long-term rates.

So I'm curious how people here are actually underwriting this:

**At what 10-year yield would you materially change the valuation multiple you're willing to pay for NVDA/MSFT/GOOGL/AMZN/META—or have you already?**

I'm more interested in the valuation math people are using than whether AI is a bubble.