Meta, Microsoft, Amazon, Google earnings diverge on AI infrastructure spending

During this earnings quarter, Meta, Microsoft, Amazon, and Google all reported financial results while disclosing substantial capital expenditure on AI infrastructure. The coverage analyzed in Stratechery's earnings coverage reveals that despite each company's commitment to AI, Wall Street's confidence in these investments varied markedly. Meta's earnings were characterized as "a bit disappointing," with the company's future promises about AI products proving "more disconcerting" than reassuring to investors. The concern appears centered on the company's ability to monetize its AI infrastructure spending in the near term. In contrast, Microsoft presented a more compelling narrative. Its earnings showed what was described as "clarity of strategy, lower costs, and a tangibility of application." The company appears to have successfully conveyed both the necessity and the ROI logic behind its capex. Google's earnings seemed to confirm what analysts call "the Anthropic hedge"—a reference to the company's investment in the AI safety company—while Amazon's capex case was anchored to the justifiability of spending given the company's existing cloud business and the potential to serve demand within that context. The underlying pattern suggests the market is no longer simply rewarding raw AI spending; instead, investors are discriminating based on three key factors: whether frontier AI models have high or lower costs, whether immediate monetization pathways exist, and whether the company's strategic vision is clear and coherent. Companies that can articulate how infrastructure spending translates into measurable business returns are winning investor confidence, while those still articulating future potential face skepticism.