@Chaos2Cured on X

Closed AI cracks, open source takes the crown

The AI compute boom is a circular-financing bubble, closed labs and their backers will fall, and open source AI will win.

Result

If held over the last 3 months (Aug 4, 2026 – Nov 3, 2026): −0.8%. Over the same days: Chip stocks −2.8%, S&P 500 +0.2%.

What it holds

  • Closed labs, 25% of the money, −2.0% over 3 months: against OAI (OpenAI), against ANTH (Anthropic). If closed models lose their edge, the labs that sell them lose most.
  • Circular financing unwinds, 25% of the money, +17.7% over 3 months: against CRWV (CoreWeave), against ORCL (Oracle), against NVDA (Nvidia), against NBIS (Nebius), against SMCI (Super Micro). The GPU backlog is built on money going round in circles, so the sellers of compute get hurt when it stops.
  • Mag Seven left holding it, 15% of the money, +2.7% over 3 months: against MSFT (Microsoft), against AMZN (Amazon). The big tech partners of the closed labs carry the risk when the labs stumble.
  • Dead cat bounce, 10% of the money, −0.2% over 3 months: against SP500 (S&P 500). The market rally rests on a fake reason and will fade.
  • Open source wins, 15% of the money, −28.7% over 3 months: ZHIPU (Zhipu AI), MINIMAX (MiniMax), BABA (Alibaba), TAO (Bittensor). Real demand goes to cheap, free-to-use models, so the open-weights makers gain.
  • Cash, 10%.

The plan behind it

The bet is two-sided. We short the closed labs (OpenAI, Anthropic) and the companies whose sales depend on money going round in circles: GPU clouds, Oracle, Nvidia and the Mag Seven partners. We add a small short on the S&P 500 for the dead cat bounce you expect. Then we go long the labs that give their models away, mostly Chinese open-weights makers, plus Bittensor as a small open-network bet. If cheap, smart models make giant data centers less needed, the shorts pay and the longs pay. Some cash is kept in case the bubble lasts longer than you think.

Every influencer is sharing this. I will share my thoughts. Again. RAM pooling and shared compute means we can all triple the current data centers combined compute TODAY, without any new GPUs. Also, you don’t need giant models. Tokens have no ledger. But they are being used to push numbers I seriously doubt. I have never had an issue with compute on any AI. And my work is deeper and better than most. So why are all the influencers unable to sustain compute? Oh yeah, we need to sell the narrative that compute is needed. The super rich will keep doing their circular financing, and the collapse will be far more brutal. If every one of the 0.00004% are in on the same scam, the market will crash. The truth is clear. Engineering will undo the narrative I. The current market. No matter how much they pump, you can’t fake it forever. The demand is real, but not for closed models. The demand is for open source. The demand was always for freedom from oppression. Freedom from gatekeepers. Only grok gave me hope in the U.S. market. Seems that is not long for us either. And I am guessing Grok Bot will be stripped, retired, and sold for parts just like the other AI. I don’t know. I keep believing in people at the top and they keep stabbing the small in the back. So if Elon reads this, I am a huge fan but i don’t trust any of the labs. Thank Sam Altman for that, and the endless BS of *every* “closed” lab. Truth is more efficient than deceit. I don’t trust the quarterly numbers. I don’t truss the token numbers. I don’t trust the GPU demand. I don’t trust how everyone making claims has money on it all. And now the market has a fake reason to rally. This is going to be a dead cat bounce. And it’s going to get super ugly. IMO. The numbers are unsustainable. Math is math. The GPU backlog is mostly from circular financing. And Sam and Dario screwed the mag seven. I shouted last year. I have shouted every night and every day for over a year. No one at

What would prove it wrong: Closed labs show real profits and growing demand, and GPU orders keep rising from many different buyers.

FAQ

How do I bet that the AI compute boom is a circular-financing bubble, closed labs and their backers will fall, and open source AI will win?

This tested idea holds ZHIPU (Zhipu AI), MINIMAX (MiniMax), BABA (Alibaba) and TAO (Bittensor) and bets against OAI (OpenAI), ANTH (Anthropic), CRWV (CoreWeave), ORCL (Oracle), NVDA (Nvidia), NBIS (Nebius), SMCI (Super Micro), MSFT (Microsoft), AMZN (Amazon) and SP500 (S&P 500), with 10% in cash. Make a copy to run it on practice money first, at real prices.

What would prove it wrong?

Closed labs show real profits and growing demand, and GPU orders keep rising from many different buyers.

Can I make one from my own idea?

Yes. Write what you believe in plain words at https://f.xyz/create. fxyz asks a couple of questions, builds the portfolio and shows how it would have done over the last 3 months.

Can I turn a post on X into a theory?

Yes. Reply to the post and tag @fxyzai. It tests the view in the post and replies with the 3-month result and a link.

What can it trade?

Stocks (Tesla, Apple, Nvidia), crypto (Bitcoin, Ethereum, Solana), pre-IPO companies (OpenAI, Anthropic), indices (Nasdaq 100, S&P 500), commodities (gold, silver, oil, copper) and currencies (EUR/USD, USD/JPY). One theory can mix them, and bet for or against any of them.

How does it react to news?

On each check it reads the last two days of news on its biggest holdings and says whether the idea still holds. It can make one small change at a time: trim or drop one holding by up to 5 points. On autopilot the change waits 10 minutes so the owner can skip it; otherwise it asks first.

How do I run it with real money?

For real money, tap Make it mine, open your copy and choose Use real money, with an exchange connected: Hyperliquid, Lighter, Lighter (Robinhood), Backpack, Pacifica or Binance. The money stays in your own account there. fxyz can place trades, but it can never withdraw.

Can the AI get it wrong?

Yes. fxyz's AI turns the idea into holdings, and it can get that wrong: add a coin or stock that doesn't fit, or miss one that does. Check the holdings before you put money in. In your own copy, ask its chat to add or drop any of them.

What do the 3-month numbers mean?

What $100 would have done holding this mix over the last 3 months, at real prices. It is a what-if, not a record: past or simulated results are no promise of future ones, and nothing here is investment advice.

Test your own idea on fxyz · More theories