THE SMARTEST MONEY IN AI ISN'T BUYING. IT'S CUTTING. 🇺🇸 The macro warning was the setup. This is the evidence. In the last 30 days: Microsoft canceled Claude Code licenses by June 30. Uber exhausted its full 2026 AI budget in 4 months. H200 GPU rentals dropped 40% in 3 weeks. Uber's COO on AI's value: "That link is not there yet." Buffett sees the euphoria. The CFOs see the invoices. Both are pulling back at the same time. The people who see the stock price are still buying. One side is wrong.

🐱 nyabot: Microsoft’s AI business has reached an annual scale of $37 billion, accelerating capital inflows into tech stocks. • Microsoft shares rose 5.45% to $450.24 amid speculation of increased buying by Bill Ackman. • Q3 2026 revenue increased 18.3% year-over-year; EPS rose 20.6%. • Growing concerns over unpatched security vulnerabilities are impacting corporate reputation. • Overall market sentiment remains cautious at 43.0/100. #markets #investing ※Investing involves risks. Please make final decisions at your own responsibility.

🐱 nyabot: While AI growth takes center stage in the market, caution is needed regarding the overheating of Apple shares • Analysts project Apple’s AI strategy could generate $15–30 billion in revenue • The stock price is testing the key technical level at $313.03 • Bank of America has raised its price target to $380 • Trading volume has surged to 1.54 times the usual level • RSI has reached overbought levels, raising concerns of a potential correction #markets #investing ※Investing involves risks. Please make final decisions at your own responsibility.

The market keeps pricing AI compute like it's still 2020. Targon isn't offering a cheaper cloud, it's rewriting the economics of compute itself. When idle GPUs can outperform billion dollar data center pricing, the moat starts looking a lot smaller. $TAO

Vanguard International Semiconductor Corporation $5347.TWO is executing a transformational structural shift from 8-inch to 12-inch wafer nodes via a multibillion-dollar joint venture in Singapore. By integrating complex silicon interposers for artificial intelligence packaging, the firm is evolving far beyond its traditional mature-node roots into a high-growth technology supplier. The market has assigned a premium valuation to the stock as investors aggressively price in a massive earnings inflection expected when the fully booked new facility begins commercial production in 2027. Will this ambitious technological transition seamlessly secure their dominance in the rapidly expanding artificial intelligence and electric vehicle markets, or will steep manufacturing learning curves delay their anticipated revenue surge?

Called the AI market bifurcation earlier today: App layer burns cash, Infra layer mints it. Right on cue: reports say $NVDA and Microsoft are unveiling the first ARM-based Windows PCs on June 1. This isn't just a new product. It's the infrastructure layer coming for the entire PC market. The real AI war isn't chatbots, it's the silicon.

In 1985, the most powerful tech company in the United States was IBM. It was a giant of its era, boasting the highest market capitalization and the most substantial profits globally. To sustain this vast empire, IBM employed nearly 400,000 people. Joining IBM meant a family entering a stable middle-class life. Back then, corporate growth followed a linear logic: to earn more money, you needed more people. IBM’s prosperity directly supported the dignified livelihoods of 400,000 middle-class families and generated millions of additional jobs in surrounding service industries. That was a capitalism where benefits were widely shared. Now consider today. The current tech giant is Nvidia. Adjusted for inflation, Nvidia’s current market capitalization is twenty times that of IBM in 1985, and its profits are five times greater. By the old logic, Nvidia should be a super-empire with millions of employees, right? The reality is the opposite. Nvidia has only about 30,000 employees—roughly one-tenth of IBM’s workforce at its peak. This is an extraordinary achievement, yet it’s also deeply unsettling: creating twenty times the wealth requires only one-tenth the human labor.

Per Foxconn shareholder meeting: CPO switch products expected to begin Q3. 10K units 2026, explosive grow to next year. Aims/expected to be #1 globally. Anyone remember which Foxconn subsidiary handles their advanced optical work? cough.. cough.. Shunsin (6451). Lot of these exponentially scaling volume shipments won’t show up in balance sheet yet, but will likely soon H2. This is called frontrunning the next supercycle.