Meta is trying to sell two stories at once: a grand consumer-AI future and a brutal near-term spending bill. Mark Zuckerberg says personal AI agents will become a daily utility for billions; investors, judging by the market reaction, are still asking who pays for the dream.

On Meta’s latest earnings call, Zuckerberg sketched an expansive vision in which AI agents handle everything from health and finances to relationships and household tasks, working continuously on a user’s behalf. He argued that within five years it is “extremely unlikely” that billions of people won’t have such an agent, and said messaging platforms like WhatsApp will become central to that shift.¹ Meta is also framing those agents not as a side project but as “the foundation for our next wave of products and revenue lines.”²

That optimism runs into a harder number set. Meta’s free cash flow plunged 91% to $784 million in the quarter as capital spending on servers, networking, and data centers surged, with one report noting the company’s AI push is “swallowing almost all its free cash flow.”³ The Financial Times captured the market mood bluntly: “Meta shares tumble as Mark Zuckerberg tries to sell his vision for AI ‘agents’.”⁴

Still, Meta insists the spending is already doing real work. Zuckerberg told investors he is “excited about how AI is helping our teams speed up product development,” as the company rolls out a string of standalone apps and experiments faster than before.⁵ He also pitched a broader business case beyond consumer assistants, saying Meta sees “a large enterprise opportunity” in APIs, business agents, compute, and software services for companies.⁶

That leaves the real tension unchanged. Meta’s bull case is that AI will improve ads, spawn new apps, unlock enterprise revenue, and eventually make personal agents mainstream. The bear case is simpler: this is another massive Zuckerberg moonshot, only this time the infrastructure bill lands long before the payoff does.