The 10 Reasons Michael Burry Is Rooting for a Market Crash in 2026 (So Skynet Can't IPO)
The man who shorted the housing market now wants the AI bubble dead before OpenAI and Anthropic sell it to you.
Photo: FastBull
Let's stop pretending the AI boom is a sure thing.
Here is the uncomfortable truth: the guy who predicted 2008 now wants the entire stock market to tank. Not to profit from it. Just so two IPOs never happen.
This article is not a list of AI doom predictions. It is about why the most famous bubble hunter alive thinks OpenAI and Anthropic going public would be a disaster for everyone except their investors.
1. He posted the quiet part out loud
On September 29, Michael Burry wrote on his "Cassandra Unchained" X account that for the benefit of humanity, markets should "tank hard" and prevent the OpenAI and Anthropic IPOs.
- He is not hedging anymore. This is the full thesis, posted publicly.
- He wound down Scion Asset Management in 2025, so he is speaking as an outsider now.
- The man who bet against subprime mortgages is betting against your ChatGPT subscription.
(Source: TechSpot)
2. "Trillions of dollars" will be destroyed, in his words
In a follow-up post, Burry said both companies will "suck up and then destroy TRILLIONS of dollars of capital," and that will be the least of the damage they do.
- He thinks the capital destruction is the floor, not the ceiling.
- The IPOs are how this capital gets vacuumed in, from pension funds and retail investors.
- Nobody raising money at a $2 trillion valuation admits they plan to burn it.
(Source: The Cryptocurrencies)
3. He agreed with the Skynet joke
When an X user suggested tanking the market so that "Skynet can't IPO," Burry agreed with him.
- Skynet, the machine intelligence from Terminator that launches a nuclear war on humanity, is now a financial punchline.
- The joke lands because the IPOs really are the moment AI labs sell you the future and the risk.
- Burry saying yes to this is him cosigning the darkest joke in tech finance.
The so what: when the world's most famous doomsayer laughs at your IPO joke, the IPO is the joke.
(Source: TechSpot)
4. He let Grok explain his thesis, then cheered it on
A user asked Grok, xAI's chatbot, to explain Burry's stance. It described a dot-com-style bubble built on unprofitable LLMs that fall short of AGI. Burry publicly agreed with the answer.
- Let that sink in: a rival lab's chatbot made his case better than he could.
- The argument is that the models are unprofitable and not a path to AGI.
- If both are true, the valuations are pure theater.
(Source: The Cryptocurrencies)
5. Anthropic's own paperwork admits the product might end the world
Anthropic's IPO prospectus, seen by Reuters, reportedly shows the company could seek a valuation above $2 trillion. The same prospectus acknowledges its AI could pose an existential risk to humanity.
- Read that twice: $2 trillion valuation, and the fine print says "might end the world."
- This is the honesty rule in action. The prospectus admits the risk, then prices it at $2 trillion anyway.
- Burry's point: an IPO lets them monetize the risk before it materializes.
(Source: The Cryptocurrencies and TechSpot)
6. OpenAI keeps moving its own IPO date
Sam Altman said OpenAI will not go public in 2026, citing safety concerns. Reports now point to 2027, while Anthropic is expected to list after November's midterm elections.
- The dates keep sliding. The hype does not.
- Delaying the IPO while raising private money at ever higher valuations is a neat trick.
- Burry's read: delays plus danger talk equals IPO prep disguised as caution.
(Source: Economic Times and TechSpot)
7. "LLMs are not AI and won't be AGI"
In September, Burry called the AI safety warnings from OpenAI and Anthropic executives "self-serving." His line: LLMs are not AI, will not become AGI, so there is nothing to slow down.
- The slowdown talk, he argues, is incumbents protecting their lead.
- Open-source and Chinese models are eroding the incumbency advantage fast.
- "Hype & puffery" is what IPOs need, in his words. The danger narrative is the marketing.
(Source: Economic Times)
8. He already accused the hyperscalers of cooking their books
Last November, Burry accused big tech hyperscalers of artificially boosting reported earnings by understating depreciation on their AI infrastructure.
- Data centers are the physical proof of the spending spree.
- If depreciation is understated, the AI boom looks more profitable than it is.
- This is not a new grudge. He has been building the case for a year.
The so what: first he questioned the profits, then the models, now the IPOs. The thesis keeps widening.
(Source: TechSpot)
9. The competition is coming from everywhere
Burry's argument is not just vibes. Open-source models and Chinese labs are closing the gap, and every month the incumbents' moat gets shallower.
- When the moat shrinks, the only defense is scale, which means spending more.
- More spending means more capital raised, which means the IPO matters more.
- A crash would cut off the funding pipe that keeps the whole loop alive.
10. He has been right about exactly one thing before: everything
Burry is a trained physician who became famous for shorting the 2008 housing bubble, a story told in "The Big Short" in 2015. People still confuse him with Christian Bale, who played him.
- One legendary call buys you a decade of people listening.
- He no longer manages outside money, so he has no clients to impress.
- The Cassandra account name is not subtle. He knows exactly what he is doing.
The so what: he called the last bubble when everyone called him crazy. Now he is doing it again, louder.
(Source: TechSpot)
How to start
Whether you are a builder, an investor, or just someone with a retirement account, here is the practical version.
- Builders: stop assuming the API you build on will have a moat. If Burry is right, the winners will be whoever survives the shakeout, not whoever raised the most. Build portability into everything.
- Investors: ask what the unit economics look like without the hype. Unprofitable LLMs plus $2 trillion valuations is math that needs a miracle.
- Writers and curious minds: follow the depreciation story, not the keynote demos. Data center accounting is where bubbles leave receipts.
- Everyone: give yourself 90 days to stress-test your AI assumptions. If a crash cut AI funding in half tomorrow, which of your tools and bets would still be standing?
The real pattern
An IPO is the bubble's exit door. Burry wants to weld it shut.
IPOs turn private hype into public money. Once OpenAI and Anthropic list, the risk gets socialized across index funds, pensions, and your 401(k). The upside stays with the early investors. That is the trade Burry is describing, and it is the oldest trade on Wall Street.
He is not rooting for a crash because he hates AI. He is rooting for one because he thinks the IPO is how the bill gets handed to you.
Final thoughts
The scariest line in this whole saga is in the prospectus itself: $2 trillion. Existential risk. Same document.
Maybe Burry is wrong this time. Maybe. But the last time everyone was sure he was wrong, he bought himself a movie.
Yes, AI helped write this. It did the typing. I gave the ideas and I did the editing and orchestrating.
If this saved you money or taught you something, buy me a coffee. It keeps the articles coming.
Follow along. Clap if it earned it.