09/21/2026
Meta stock is up 33% in one month. Because Mark Zuckerberg is putting agentic AI directly into the hands of its 3 billion monthly users with Muse, and developing ways to monetize that on Facebook and IG.
For a year I have been talking to my clients about how to buy Hypercalers, as those stocks failed to catch the stock market rally: Meta, Google, Amazon & Microsoft. Essentially Big Tech players investing trillions of dollars in AI capex, and often derided for it.
The current discussion about 'slowing' AI is triggered by the breakneck speed at which it's now learning. The debate revolves around control of frontier models, not a notion that the physical buildout of capacity needs to decelerate. If anything, AI's increased potency is likely to force greater corporate adoption, via fears of price competitiveness.
Importantly, scale matters.
In general, the paradox is that as AI models become more capable, they lose differentiation. It may therefore be that future value accrues to what surrounds them: proprietary data, computing power, distribution and the enormous amounts of capital required to keep investing.
Google, Meta and Amazon already have vast user bases, global distribution, proprietary data and formidable balance sheets. AI could represent a monumental acceleration in the economies of scale each already enjoys.
I could be wrong. The leaders of these companies could be wrong. They stake their careers and shareholder value on spending hundreds of billions of dollars building for an AI future yet to fully materialize.
Former Google CEO Eric Schmidt argues, if you believe AI is a once-in-a-generation technology capable of transforming almost every aspect of life, it's incumbent on Big Tech to embrace it and invest aggressively. Savita Subramanian at Bank of America says Big Tech is simply moving away from an era dominated by financial engineering and share buybacks to investment in productive capacity.
But AI skeptic Ed Zitron fears corporate America and developed nations won't generate sufficient revenue and cash flow to ultimately justify that capex. His case is worth hearing in The Compound and Friends episode, “The Four Horsemen of the AI Apocalypse.”
I don't dismiss Ed Zitron because he's almost a lone wolf; no big bank would tolerate any analyst publicly doubting the most powerful corporations in the world. I just don't think he fully appreciates the immensely profitable nature of selling vast tech stacks of services to clients 'caught' in your AI cloud, in perpetuity.
This is part of what I discuss with my clients every day.
You are welcome to join that conversation.