> ## Content Index > Fetch the complete content index at: https://millionaire-insiders.ghost.io/llms.txt > Use this file to discover other available public pages before exploring further. # Big Tech Is About to Spend $697 Billion on AI, and the Bill Just Got Bigger - URL: https://millionaire-insiders.ghost.io/big-tech-is-about-to-spend-697-billion-on-ai-and-the-bill-just-got-bigger/ - Published: 2026-08-07T10:47:10.000Z - Updated: 2026-08-07T10:47:10.000Z - Author: Daniel Whitmore Sell-side analysts now peg 2026 capital spending at roughly $697 billion for just five US technology companies, up $173 billion from where estimates stood at the start of the year. Amazon alone plans $200 billion, a 56% jump from last year. Alphabet expects $195 billion to $205 billion, mostly for data centers, networking, and custom chips. Meta is guiding to as much as $135 billion, an increase of at least 59%. Microsoft's capex is projected at $105 billion, up 19%. That is an extraordinary amount of money being committed by a handful of companies to a single technology buildout, and it's still climbing every time these companies report. The scale here matters more than any individual number, because it means the AI infrastructure cycle isn't slowing down even as questions about return on that spending get louder. ### The chip makers are the clearest beneficiaries Global semiconductor sales hit a record $120.6 billion in May, up 104.1% year over year, marking the fifteenth consecutive monthly record. Broadcom's own guidance calls for AI semiconductor revenue to grow more than 200% year over year to $16 billion in the current quarter. Micron is guiding toward a $50 billion revenue quarter. These aren't speculative growth projections from early-stage companies. They're guidance from established suppliers describing demand that's already contracted and largely locked in. That distinction is worth sitting with. A lot of past technology spending cycles were driven by expectations of future demand that sometimes failed to materialize. This one is different in that hyperscaler capex commitments and chip supplier guidance are now confirming each other in real time, each earnings season validating the other side of the same spending relationship. The scale of this buildout is also reshaping how these companies talk about their own businesses. Data center construction, power procurement, and custom silicon design have moved from footnotes in hyperscaler earnings calls to headline agenda items, on par with the core advertising, retail, or software businesses that built these companies in the first place. That shift in emphasis is itself a signal of how central AI infrastructure has become to each company's long-term strategy. ### The jitters underneath the spending Not everyone is comfortable with the size of these numbers. Big Tech's roughly $600 billion in combined AI capex plans could accelerate automation rollouts across the economy, but it's also sparking real investor unease about whether returns will show up fast enough to justify spending at this scale. That tension, genuine demand confirmed by chip guidance on one side, and a valid question about payback timelines on the other, is the central debate running through every hyperscaler earnings call right now. **Whitmore's Watchlist:** NVDA (NVIDIA Corporation): The dominant supplier of the AI accelerator chips underpinning most of this hyperscaler capex spending. AVGO (Broadcom Inc.): Guiding toward more than 200% year-over-year growth in AI semiconductor revenue this quarter alone. MU (Micron Technology): Memory chip demand tied directly to AI data center buildout, guiding toward a $50 billion revenue quarter. For investors, the useful distinction isn't whether AI capex is real, the numbers above settle that question. It's whether a given company sits closer to confirmed, contracted demand, like the chip suppliers currently reporting record guidance, or closer to the still-unproven return-on-investment side of the equation, where hyperscalers themselves are being asked harder questions about payback periods on their earnings calls. ### Watching for the next signal The clearest tell for whether this cycle is sustainable will keep coming from the same place it has all year: hyperscaler earnings calls and the capex guidance embedded in them. If capex commitments for 2027 start decelerating rather than climbing further, that would be the first real signal that the buildout is maturing rather than accelerating. Right now, every quarter has brought upward revisions, not downward ones, which is itself informative about how these companies view the demand picture staying intact. The spending concentration among just five companies is also worth remembering. A pullback or disappointment from any single one of them would ripple through the entire chip supply chain given how much of current demand traces back to this narrow group of buyers, which is a real concentration risk sitting underneath an otherwise strong growth story. **Whitmore's Take:** The AI capex numbers are large enough that they're no longer a story about a handful of tech giants alone, they're a broad demand signal for the entire chip supply chain. Worth distinguishing between exposure to confirmed spending and exposure to the payback question still being debated on those same earnings calls. ![](https://storage.ghost.io/c/e9/10/e9109ad9-55f9-4e96-a078-46af25115156/content/images/2026/08/ai-capex-boom-2.jpg) *Written by Daniel Whitmore* *Millionaire Insiders*