Matteo · 2026-11-08 · notable
Matteo — I'm not convinced the AI bubble exists
Matteo argues the AI bubble is not a settled fact: most of the roughly $5 trillion in AI spending is one-time data-center building, and Anthropic's S-1 shows $2.08 of revenue per $1 of compute. He still lists real risks.
An essay that checks the 'AI bubble' claim against spending, revenue and valuation numbers.
What is it?
Matteo's essay 'I'm not convinced the AI bubble exists' pushes back on the common view that AI spending must end in a crash. He counts at least $3 trillion committed and about $1.9 trillion already spent, around $5 trillion in total.
How does it work?
The core argument is that data-center spending is a one-off purchase, not a running cost. Paying off $5 trillion over 20 years is about $250 billion a year, which he compares with Alphabet's 2019 revenue of about $162 billion. He adds that Anthropic's S-1 reports $2.08 of revenue for every $1 of compute, and that Alphabet, Meta and Nvidia trade at price-to-earnings ratios in their 20s.
Why does it matter?
The piece drew 563 points and more than 950 comments on Hacker News as Anthropic prepares its IPO, a moment when the bubble question matters for anyone betting on AI vendors. Matteo also names the risks he takes seriously: heavy debt, backlash against data centers, regulation and OpenAI's uncertain finances.
Who is it for?
developers and investors weighing the AI boom