Meta's ad dominance positions it as top AI stock play, analyst argues

The article's core thesis rests on a structural difference in how consumers and enterprises consume AI. Anthropic's recent $65 billion annualized run-rate revenue comes from enterprise subscriptions—a model that works because companies view AI as a productivity cost that drives revenue. Consumers, however, have historically resisted paying monthly fees for software, creating a monetization problem for most consumer-facing AI products.

Meta's existing advantages—its 3.6 billion daily active users, established ad infrastructure, and engagement metrics (Instagram time spent up double digits in Q2)—mean it can offer AI services free to consumers while capturing value through advertising. The article argues this isn't a sidecar business but a strengthening feedback loop: AI improves recommendations and engagement, which attracts more user time; simultaneously, AI gives advertisers better targeting and creative tools, increasing their return on investment and spending. Each cycle reinforces the other.

The $137.5 billion annual capex bet reflects Meta's confidence in this model. Unlike hyperscalers that can sell excess compute to outside customers, Meta is building frontier AI models to directly improve its own advertising operations. The analyst contends that as chips, cloud platforms, and AI models become commoditized, the real profit opportunity lies not at the infrastructure layer but in companies leveraging AI to improve existing high-margin businesses—and Meta's advertising business is among the most defensible.