Microsoft, Meta, and Alphabet disclosed that they had spent $32B+ combined on data centers and other capital expenses in Q1, as they accelerate AI spending
The spending that the industry's giants expect artificial intelligence to require is starting to come into focus — and it is jarringly large.
New York TimesKaren Weise
Context & Ripple Effects
This disclosure is an early marker in a spending arc that soon broadened: the four largest platforms raised first-half capex to $106 billion while building AI infrastructure. It matters because it shows Microsoft, Meta and Alphabet were already treating data-center capacity as a near-term strategic requirement rather than a marginal product expense.
Later coverage frames the outlay as the beginning of a sustained capital cycle: the same group’s 2024 spending was projected to exceed $200 billion, and subsequent reports questioned whether large infrastructure budgets would translate into income.
First-order effects
Microsoft, Meta and Alphabet commit more capital to data centers and related assets, increasing the near-term financial weight of their AI buildouts.
The disclosures give investors a clearer indication that AI plans at these companies require infrastructure spending at a scale measured in tens of billions of dollars.
Second-order effects
The spending raises the competitive cost of keeping pace in AI infrastructure, pressuring other large platforms to expand their own capacity commitments; that pattern was visible in the broader first-half capex increase that included Amazon.
As more budget shifts toward data centers, the companies face greater pressure to show that the resulting capacity supports revenue or other returns—a tension later surfaced when markets split over whether infrastructure spending could produce income.
Third-order effects
If sustained, the buildout makes access to capital and data-center capacity a more important determinant of who can compete at the AI platform layer, favoring companies able to fund multiyear infrastructure programs.
The pattern points to an AI infrastructure capital cycle in which investment decisions increasingly shape both product competition and investor scrutiny of returns, rather than AI development being funded primarily through ordinary operating budgets.
The trend: AI competition is becoming an infrastructure-led capital cycle, with leading platforms using escalating data-center investment to secure capacity and strategic position.
For a sense of how bonkers the money is that Big Tech is spending on AI (and many other big ticket projects): Exxon spent $26.3 billion last year digging oil and gas out of the ground. Google and Microsoft combined spent $26 billion IN THREE MONTHS on computer chips, servers, da…
Investment like this makes sense for Microsoft and Google, as they have cloud platforms heavily used by basically everyone. Meta? Not so much. This isn't to say many of Google's ML initiatives will pan out, because they won't. …
@EricJhonsa Between META, GOOG and MSFT capex is expected to go up almost 50B over last year. Seems like AAPL will need to step up their capex spend significantly as well [image]
Zuck on why he ramped CapEx with Meta down 70%: “It came from being behind [what TikTok was doing]....most of the times where we make some decision that seems good is because we messed up before and don't want to make the same mistake” [image]
The absolute dollars are astonishing. Capex is going to increase from $48B to $184B. But Gross Profit (ex fulfillment expense) is going to go from $208B to $714B.
$MSFT spent $14B on capex in FQ3, up from $11.5B in FQ2 and $7.8B a year ago. $GOOG spent $12B on capex in Q1, up from $11B in Q4 and $6.3B a year ago.
Google's market cap is at an all time high ($2T) and Microsoft saw a positive response on announcing huge AI spending while Meta stock tanked. Key difference is investors are no longer patient about returns. Microsoft and Google are making money from AI today via cloud & search. …
Looking at Capex to Gross Profit (to normalize for biz models; I deduct fulfillment expense from AMZN) for GOOG/AMZN/META/MSFT. From 2017 to 2024, Capex/GP is surprisingly only up from 23.2% to 25.8%. MSFT: 13% -> 26% GOOG: 20% -> 22% AMZN: 48% -> 26.7% META: 19% -> 29.7% [image]
Google's market cap is at an all time high ($2T) and Microsoft saw a positive response on announcing huge AI spending while Meta stock tanked. Key difference is investors are no longer patient about returns. Microsoft and Google are making money from AI today via cloud & search…
@davemcclure Cross-selling is one way to reduce customer acquisition cost. But creating a new business from nothing is a hard problem. Le Cun suggested on Lex that Meta could sell productivity software to restaurants. Payments is possible. Neither is easy.
oddly diff market reaction to Meta AI $pend vs. MSFT / AAPL — guessing this is due to historical Meta (over-)$pend on VR, hwvr this time seems like it's the right decision... wouldn't be surprised to see market revise their opinion in future.