Meta’s AI Spending: Genius Move or Zuck Losing It?

Meta is spending aggressively on AI infrastructure—and the big question is whether that’s visionary or reckless. Andrew frames Meta as primarily an advertising business today, while Stephen digs into what Meta’s leadership is signaling about 2026–2027 capex visibility (and how murky “2028 and beyond” still looks). From there, the conversation turns into what investors actually have to believe to own Meta: trust in Zuckerberg’s control and decision-making, plus a belief that AI becomes a durable, monetizable platform shift.

Then you get into the messy parts: trust (AI agents with access to your life), security risk, and the ethical/legal overhang from lawsuits and public perception. The episode closes with Andrew walking through why the numbers look undeniably strong—growth rates that don’t make sense for a company Meta’s size, margin discipline, and shareholder returns—while still admitting the qualitative risks are what keep him on the sidelines.

What You Will Learn

How to think about Meta as an ad business today vs. an AI platform bet tomorrow

What Meta’s capex visibility (2026–2027) implies—and why “2028+” is the scary part

The two big “belief requirements” to invest in Meta: Zuck + AI

Where Meta’s AI strategy could fail: trust, monetization, cost curve, regulation

Why Meta’s financials look so strong even while the narrative stays controversial

Timestamps

0:00 Welcome back + why Meta/AI is the focus today

1:15 “Are you feeling AI fatigue?” + Andrew admits he’s all-in (NVIDIA)

2:05 What is Meta right now? Framing it as an advertising business first