Reality Check

Source: Where’s Your Ed At Date: 2025-04-28 URL: https://www.wheresyoured.at/reality-check/

Summary

Ed Zitron argues that the generative AI boom is financially unsustainable: OpenAI spent $9 billion in 2024 to lose $5 billion, yet analysts project $125 billion in revenue by 2029, with $3 billion attributed to agents that don’t yet exist at commercial scale. Hyperscaler signals reinforce the skepticism — Amazon’s generative AI revenue is a fraction of its data-center investment, and multiple reports indicate pauses in capacity expansion. Zitron’s thesis is that the bubble is a narrative product, held aloft by failed incumbents (deteriorating Google Search) and uncritical financial media, not by demonstrated enterprise value.

Implications

• AI demand curve scrutiny: The gap between infrastructure spend and recognized revenue is becoming harder to paper over; expect tighter capital allocation from hyperscalers and slower enterprise adoption curves than 2024 projections assumed.
• Agent product pressure: The “agents as revenue” thesis depends on products that remain unreliable at basic tasks; this signals near-term risk for teams that have priced in autonomous-agent capability in their roadmaps.
• Market correction readiness: If even one major revision to OpenAI’s revenue guidance lands, it re-prices the entire sector’s valuation multiples — a tail risk worth tracking in the short term.
• Feeds thread: AI investment sustainability / enterprise adoption reality-check.