// capital expenditure

Source: Morning Brew

Alphabet, Amazon, Meta, and Microsoft are treating AI capex as table stakes for market dominance, not discretionary spending. Capex growth is outpacing revenue gains. Some companies report double-digit increases. The bet is explicit: whoever builds the largest, most capable compute clusters controls the next computing paradigm. This is about securing asymmetric advantages in foundation models and inference capacity, not quarterly earnings. The spending pattern creates a dependency trap. All four are locked into a capital arms race that punishes restraint. Any one pulling back on AI spending would be read as capitulation and trigger immediate market repricing. Margins are under pressure in the near term, but the companies are absorbing those costs as the price of entry.

Source: Forward To Friend

Microsoft, Google, Meta, and Amazon are collectively committing roughly $700 billion to AI infrastructure by 2026—a sevenfold increase from current spending. These companies treat computational dominance as essential competitive advantage. This scale of capital deployment will reshape supply chains for semiconductors and data center real estate, create hard constraints on competitors without equivalent balance sheets, and lock in winner-take-most dynamics before AI's actual commercial ROI becomes clear. The bet also reveals management's confidence (or desperation) that current generative AI capabilities justify spending equivalent to the entire annual R&D budgets of most Fortune 500 companies.