AI arms race, discipline premium
The AI capex boom is accelerating across Big Tech, but investors are rewarding the company that can spend while still throwing off cash. Alphabet’s strong free cash flow and guidance for disciplined capital spending placed it ahead of peers in market reaction, even as Microsoft’s cloud growth outperformed. Meta, Alphabet and Microsoft are all signaling larger 2026–27 data center footprints, long-term supply deals for advanced GPUs, and rising power procurement costs. The near-term trade-off is visible: higher depreciation and lower margins today for dominance in model training, inference and AI-native services tomorrow. Funds with profitability screens gravitated to Alphabet’s cash machine; traders stayed cautious on names where returns remain more narrative than numbers.

What the numbers imply for 2026
Two tensions define the next year: compute scarcity and power. Even with more Nvidia capacity online, buildouts hinge on grid hookups and substation timetables. That shifts advantage to firms with energy offtakes and on-site generation plans. On product, the leaders are racing to stitch AI into consumer search, productivity suites and ads without breaking UX or trust. Regulators are watching model safety, child protection and content provenance—particularly after hurricane deepfakes again flooded social feeds this week. Bottom line: Wall Street will indulge AI capex, but only if the cash engine keeps humming and the roadmap shows measurable unit economics at inference scale.