3 thoughts on “Alphabet’s 2026 capex forecast soars; Gemini 3 AI model is a huge success”

1. Alphabet’s 2026 capital expenditure (CapEx) forecast is between $175 billion and $185 billion.. This projection significantly exceeded analyst expectations of roughly $119.5 billion and is roughly double the $91.4 billion the company is expected to spend in 2025.

   Key details of the CapEx announcement:

   -Focus on AI Infrastructure: The increased spending is for AI computing infrastructure. This includes servers, data centers, and networking equipment to support Gemini and Google Cloud initiatives.
   -Response to Demand: The investment is needed to meet “significant cloud customer demand” and to develop frontier AI models. AI investments are already driving revenue across the business.
   -Market Reaction: The announcement initially caused Alphabet shares to fall. The stock later recovered some losses. Investors weighed the aggressive investment against a strong Q4 2025 performance, where Cloud revenue grew 48%.
   -Context within Big Tech: This aligns with a broader industry-wide push by “hyperscalers” to dominate the AI market. Major tech players are expected to spend around $650 billion on AI-related capital expenditures in 2026.
   -The 2026 guidance suggests Alphabet is doubling down on its AI strategy. It is betting that current infrastructure investments will yield high returns despite potential short-term pressure on profit margins.

   Source: Google Gemini

   Reply

2. Google is benefiting indirectly from its massive AI spending on data center build outs, Nvidia chips and investments in Anthropic and other AI startups.  However, it is impossible to quantify if there is any positive ROI or profits.

   1.  Revenue for Google Cloud grew 48% year-over-year in Q4 2025 to $17.7 billion, surpassing analyst estimates and outpacing growth in previous quarters, heavily driven by AI infrastructure and solutions. , Google Cloud reported a significant increase in profitability, with an operating income of $5.313 billion compared to the $2.093 billion in Q4 2024.

   2.  AI Overviews and “AI Mode” in Google Search led to a 17% growth in Search revenue, with users conducting more, and longer, queries presumably due to new AI search/query features.

   3.  Gemini for home (which I have on a Google speaker & my Samsung Android phone)  is designed to generate profit for Google on speakers and Android phones. It does so primarily through subscriptions and by driving higher-value hardware sales. In late 2025, Gemini replaced Google Assistant on smart home devices. It moved from a free, ad-supported model to one with both free and paid options.

   a] Google introduced a paid AI model, with “Home Premium” (previously Nest Aware) costing around $10/month. This subscription unlocks advanced features like Gemini Live. It also enhances home security monitoring and enables more complex, conversational automation for Nest cameras and speakers.

   b] Google is releasing new hardware explicitly designed for Gemini AI. For example, a $99 Gemini-powered smart speaker captures revenue from users upgrading for better performance. Gemini makes advanced, exclusive AI features available on Pixel devices. This increases the perceived value of Google’s own hardware, driving sales.

   Reply

3. Related: Amazon’s $200 Billion Spending Plan Raises Stakes in A.I. Race

   Amazon said on Thursday that it would spend $200 billion this year on data centers, satellites and other big-ticket items as it invests heavily in the race for artificial intelligence. That eye-popping dollar figure, which topped Wall Street’s predictions by $50 billion, was the latest in a series of announcements by big technology companies that they plan to dial up their A.I. investments this year.

   Alphabet, said it would spend as much as $185 billion this year, and Meta said last week that its capital expenses, in large part to support A.I., could reach $135 billion. Last week, Microsoft reported sales and profits that surpassed Wall Street’s expectations, but its share price also dipped 10 percent after its capital spending was higher than predicted and its cloud computing sales grew slightly less than investors had anticipated.

   Industry executives have made it clear that they believe spending too little on A.I. would be a much bigger mistake than spending too much, even as the combined annual capital spending plans of Amazon, Microsoft, Meta and Google race past half a trillion dollars.

   “I think this is an extraordinarily unusual opportunity to forever change the size” of Amazon, Andy Jassy, the company’s chief executive, said in a call with Wall Street analysts on Thursday.

   Amazon and its peers have said they still don’t have enough data centers up and running to meet the pent-up demand for A.I. from customers. And they argue that when companies adopt A.I., they also use more traditional cloud services, sending more money into other parts of their operations.

   “You don’t have to be Sherlock Holmes to figure out that A.I. has driven these changes,” John Dinsdale, chief analyst at Synergy Research Group, said in a note on Thursday.

   Mr. Jassy said that customers were snapping up Amazon’s A.I. computing capacity just as soon as it becomes available, and that customers were moving more data to the clouds so they could power A.I.

   “We just have a lot of growth, and a lot of demand,” he said. Mr. Jassy did not directly answer questions about whether there was a red line that would make the company slow its spending.

   Mr. Jassy said in a statement that “strong demand for our existing offerings and seminal opportunities like A.I., chips, robotics and low earth orbit satellites” were the reasons for the spending spree.

   The company cautioned that its costs would eat into profit in the current quarter, with operating profits potentially declining to as little as $16.5 billion, about $2 billion less than a year ago.

   Amazon’s cloud computing division saw sales grow 24 percent to $35.6 billion as the business gained momentum after initially lagging in the A.I. race. It was the fastest growth in about three years.

   For the past several quarters, Amazon has been racing to open more data centers to meet demand. In late October, Amazon opened its largest data center, in Indiana, to begin serving the A.I. company Anthropic, which Amazon has also invested in.

   Amazon spent more than $38 billion on capital expenditures in the quarter, bringing its total for the year to more than $128 billion. That has largely gone to building data centers, though it also includes warehouses and other facilities for its e-commerce business.

   In October and January, Amazon laid off a combined 30,000 corporate and technology employees, a move that Mr. Jassy has said was driven by a desire to reduce bureaucracy. Engineers accounted for roughly half of those who lost their jobs in Washington, Amazon’s home state, according to disclosures filed there.

   https://www.nytimes.com/2026/02/05/technology/amazon-200-billion-ai.html

   Reply