AI stocks have been falling pretty intensely in recent days, but I’m not really in the mood to chase the tape and guess a rebound.

Instead of comparing who’s down more, I’d rather wait for Alphabet’s earnings report after the close this Wednesday. Because it will very likely shape how the market views this wave of major AI infrastructure expansion next: whether Alphabet keeps committing to high spending, or starts seriously calculating the payback period.
Alphabet’s data from the previous quarter was already strong.
Google Cloud revenue broke $20 billion, up 63% year over year. Cloud backlog jumped straight from $240 billion to more than $460 billion, indicating that enterprise demand for AI compute hasn’t cooled off.
But the pace of spending is just as astonishing.
Alphabet’s Q1 CapEx reached $35.7 billion, and operating cash flow was $45.8 billion; after subtracting capital expenditures, free cash flow was only $10.1 billion. The company also raised its 2026 CapEx guidance to $180 billion—$190 billion.
So the real question isn’t whether there’s demand for AI. With $460 billion worth of orders sitting there, demand is basically not in doubt.
What we need to watch instead is how long it takes for these orders to be delivered, and how quickly they can turn into revenue, profit, and cash flow. After all, backlog is orders not yet confirmed—not money already earned.
After spending years in crypto, I see this logic as similar to how miners expand their machines.
Adding machines and boosting compute is obviously a good thing, but if the incremental output can’t cover equipment costs, electricity bills, and depreciation, the bigger the scale, the more obvious the cash-flow pressure becomes.
Alphabet is the same.
If Cloud can still maintain a high growth rate this time, without a notable drop in profit margins, and backlog keeps rising, it suggests that AI spending is already generating returns while the expansion continues. Such CapEx is still high, but at least there’s steadily growing business to absorb it.
On the other hand, if Cloud starts to slow down while CapEx keeps rising and free cash flow is still being squeezed, then we need to recalculate how long these investments will actually take to pay back.
So this time I won’t get overly optimistic just because one Cloud number looks good, and I won’t automatically go bearish just because CapEx is rising.
Only if Cloud growth, profit margins, and cash flow can hold together would I view this AI infrastructure pullback more as an opportunity; if revenue starts to slow while spending accelerates further, then I’ll wait.
I still like the AI direction, but liking the direction doesn’t mean you have to rush to buy every position.
Let’s first see if Alphabet can get the numbers straight this round!
#Alphabet