Am I Meant To Be Impressed?

2026-05-06 · Source: Ed Zitron's Where's Your Ed At · Field: Finance & Economics — Capital Markets & Investment Management, Corporate Finance & Treasury, Economic Analysis & Policy · Depth: Intermediate, extended

Summary

Big Tech's reported AI revenue growth is largely an illusion, driven by circular financing and massive capital expenditures with little actual return. Microsoft's \$37 billion AI revenue run rate is over 71% dependent on OpenAI, which is projected to lose over \$25 billion in 2026. Similarly, Amazon's \$15 billion AI revenue run rate relies over 80% on Anthropic, a company losing billions annually. Google, while not disclosing AI revenues, sees significant cloud growth from Anthropic's commitment to spend \$200 billion on its cloud and TPUs over five years, comprising over 40% of Google's revenue backlog. Overall, OpenAI and Anthropic constitute over 70% of the AI industry's revenue and GPU compute capacity, despite both being deeply unprofitable and sustained by continuous capital infusions from the very hyperscalers they pay. Meta has also spent over \$150 billion on AI capex since 2023 with questionable returns.

Key takeaway

For investors evaluating Big Tech's AI growth, you should critically examine reported AI revenues, as they are largely driven by circular financing with unprofitable LLM companies like OpenAI and Anthropic. Your investment decisions should account for the significant risk posed by this artificial demand and the massive, unproven capital expenditures. Demand transparent, itemized AI revenue disclosures, rather than vague "annualized run rates," to assess genuine market adoption and long-term profitability.

Key insights

Big Tech's AI revenue and compute demand are artificially inflated by circular financing with unprofitable LLM developers.

Principles

• Hyperscaler AI revenue is heavily reliant on a few unprofitable LLM customers.
• Circular financing relationships create an illusion of AI market demand.
• Massive AI capex has not yielded sustainable, diversified revenue streams.

Method

Hyperscalers invest in LLM companies (e.g., Anthropic, OpenAI), which then use that capital to pay the hyperscalers for compute and services, artificially inflating reported AI revenues and compute demand.

In practice

• Demand transparent, itemized AI revenue reporting from Big Tech.
• Question "annualized run rate" figures without clear definitions.
• Assess AI investments for genuine, diversified customer demand.

Topics

• AI Finance
• Circular Financing
• Hyperscaler Revenue
• OpenAI
• Anthropic
• GPU Compute Capacity

Best for: VP of Engineering/Data, Director of AI/ML, Executive, Investor, CTO, Tech Journalist

Related on AIssential

• Premium: The AI Compute Demand Story Is A Lie — The "AI compute demand" narrative is largely an illusion created by hyperscaler subsidies to a few key AI companies.
• Massive AI Has a Terminator. It’s a Tiny Thinking Engine — Massive AI fakes intelligence with scale; true AI needs a better mathematical foundation.
• Premium: What If...We're In An AI Bubble? (Part 3) — The AI industry's financial models, particularly for compute demand and revenue growth, are built on highly speculative and potentially unsustainable projections.
• Ai is pricy — AI hardware demand may be inflated by a temporary training phase among concentrated hyperscaler buyers.
• Amazon Looks to Raise At Least $37 Billion with Bond Sale — The AI boom is driving massive capital expenditures and corporate debt offerings across major tech companies.
• Why Anthropic’s $30 billion milestone is shaking OpenAI’s valuation — OpenAI's high valuation faces investor skepticism amid strong competition and Anthropic's rapid revenue growth.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Ed Zitron's Where's Your Ed At.