The selloff in semis is largely driven by short selling heading into the hyperscaler prints this week, per UBS TMT: “Our PB data over the last two sessions shows HFs beginning to put on tactical shorts in U.S. Semis ahead of hyperscaler capex estimates due tomorrow and Thursday night (see chart below). That’s a notable shift. For earlier momentum unwinds, the selling was almost entirely driven by longs getting cut. This is new shorting behavior in Semis. In fact, U.S. Semis saw -2.8 standard deviations of net selling on Monday, with more than 75% of that flow driven by short sales. Meanwhile, on the other side of the ledger, hyperscalers were the most heavily bought group in U.S. equities yesterday, fueled by some good old-fashioned short covering. For context, HFs have now largely cleaned up the hyperscaler shorts they put on ahead of GOOGL earnings last week. That said, while positioning remains net long, enthusiasm has clearly come down a few notches. Hyperscaler exposure is now sitting at its lowest level in the last two years, with the current long/short ratio at 2.4x versus 13.6x at the January 2025 peak”

Preserving this Greek carve-out sends precisely the wrong signal. It tells companies that sufficiently valuable commercial interests may eventually secure exemptions. It tells member states that delaying agreement can produce favourable outcomes. And it tells Moscow that Europe’s political unity remains negotiable when influential industries are affected.

There is no reciprocity between China and the rest of the world that can compel Chinese developers to adhere to foreign AI regulatory standards. Particularly if Chinese devs publish open-weight models, there's not a lot the U.S. can do to prevent Americans from using them.

Memory Liquidation The AI memory market is not the telecom boom, and it is not the housing bubble. What we are seeing now is a leverage event: too much leverage, too much crowding, and too much exposure piled into the same trade, all of which now need to be unwound. That matters because a real supply crunch in memory has been amplified by positioning, acute shortages and sharply higher prices tied to AI infrastructure demand are real. Yes the easy money in the AI trade has been made! Let’s be explicit about what that means. Parabolic charts are not proof of durable fundamentals; they are often evidence of momentum, leverage, and borrowed conviction feeding on themselves. When a trade gets this crowded, price stops reflecting only supply and demand and starts reflecting how much fast money is trapped in the move. The underlying AI demand story is still real, and the fundamental supply-demand imbalance still exists. AI demand has forced companies to fight for dwindling memory supplies, while chipmakers prioritized higher-margin data-center chips and memory prices spiked sharply over the past year. But that does not mean every price swing is fundamental. Narrative follows price: when memory names surge, investors discover scarcity; when they break, they suddenly discover China risk or efficiency gains. That is why the analogies to the 1990s telecom boom and the housing bubble are only partly useful. In those episodes, supply ran ahead of demand, too much fiber, too many houses. Here, demand has outrun supply, but the stock market layered excessive leverage on top of a real bottleneck. As Graham observed, the market is a voting machine in the short term and a weighing machine in the long run. Right now, the vote is being driven by crowding, leverage, and forced selling. Over time, the market will weigh the underlying AI demand and the still-tight supply picture on their merits. What is being liquidated is not the existence of demand. It is the leverage wrapped around the story. Yes Parabolic charts that amplify crowded leverage one way bets should be avoided.

FT: "Prices for credit default swaps, popular tools to bet against corporate debt, tied to Oracle, SpaceX, Alphabet, Amazon, Meta, Broadcom and Nvidia have risen to record highs in recent days...'Credit markets don’t deal well with uncertainty, and the sheer unpredictability of the pace and cost of AI financing is triggering a serious crisis of confidence right now.'"

@servzero I've been using GitLab over them before acquisition talks were even rumors. GH just always rubbed me the wrong way. They have market share I'll grant but I've never liked their product. Been thinking of even self hosting my own personal git server cuz I'm tired of the cloud.

NVIDIA IS THE LENDER OF LAST RESORT Nvidia Guaranteeing Open AI $250 Bn Financing This is bullish for $ORCL $MSFT and $GOOGL - all of whom have counter-party risk on OpenAI. I added to our Cloud datacenter exposures last week. The big issue was never demand -- it was about credit risk...

2021: Meta deleted over a billion faceprints after a class action lawsuit alleged it collected biometric data without consent. They called it "an important step" in privacy. 2024: Meta settled Texas's facial recognition lawsuit for $1.4 billion. The largest biometric privacy settlement in US history. 2026: Meta launches a facial recognition system for every Facebook user. They deleted a billion faceprints. paid $1.4 billion in fines. then built it again. Meta buried the announcement inside a post about Facebook Marketplace's 10th anniversary. five paragraphs about buying and selling furniture. then: "oh and starting Monday, video selfie verification for everyone." The product is described as "free" and "optional" to get the badge. But the feature that makes it work, facial recognition that matches your selfie against every photo you've ever posted on Facebook, is not optional. It runs when you submit. Meta says the video selfie is deleted after verification. They also said, "We do not use your photos to build or improve facial recognition systems." This is the company that paid $1.4 billion to do exactly that without telling anyone.

“Russia’s economic deterioration will continue and will worsen as the clock ticks. There is currently no path for recovery. Putin has lost his place as a world power. The Damocles sword is there and it is getting heavier and heavier. The question is ultimately whether he is ultimately a realist or will enter the history books as self destructive fool. At this point no commentator knows the answer.” Hersh closed his analysis with a stark historical reminder: “Putin knows the ghost in the Kremlin closet: Revolution.”

Narendra Modi, India’s prime minister, announced a panel to reform the country’s exam system after days of mass protests over a series of exam-related scandals. On Saturday demonstrators, who were led by a group called the Cockroach Janta Party (after a judge called unemployed youths “cockroaches”), danced and chanted in the streets of Delhi after the education minister resigned. Few had expected the government to yield.

Bitchat migrates to @radicle - a decentralised git version control system and release hosting alternative, after India cracks down on GitHub hosting, where student protesters were using this Bluetooth-based offline messenger app.