🧠 AI Infrastructure & Power

• AI compute race still looks Earth-first, not orbit-first. The key pushback on $SPCX orbital data centers is simple: launch, maintenance, and communication costs eat the power-savings edge. Speed matters more than elegance — “He who strikes first wins.” 2

• Nuclear power is becoming a corporate compute-and-energy hedge. $WMT signing its first nuclear power agreement with $CEG for ~176MW from Dresden is another proof point: large enterprises want reliable, long-duration, carbon-free baseload, not just intermittent clean power. 5

🧩 Enterprise AI & Fintech Platforms

• $ZETA x $PLTR is more than a headline partnership. $ZETA popped over 8% premarket and reclaimed $20 because the setup plugs Zeta Data Cloud into Palantir Foundry, putting Athena on top of enterprise-grade ontology and governance. Management’s guide of over $100M in annual revenue over the next few years matters, but the deeper angle is distribution into higher-trust enterprise AI workflows. 7

• $SOFI Composer is not an agentic “AI trades for you” product. The sharper read: it turns natural-language prompts into rules-based strategies, with preset weights, conditions, and filters. The real utility is pre-trade backtesting across different market regimes, so users can see the logic before execution. 13

🧱 Semiconductors & Memory Cycle

• $TSM pricing power is getting stronger. TSMC pushing a 5–10% price hike across advanced nodes, including 7nm, signals management wants to capture the same pricing momentum memory peers are enjoying. This is not just cost pass-through; it is a margin-reset attempt in leading-edge capacity. 1

• DRAM looks structurally tight through 2030. The key driver is HBM, which burns far more wafer capacity per bit than standard DRAM. As HBM rises from 18% to 35% of total capacity, it tightens both HBM and commodity DRAM at the same time. That is the bull case for a longer memory upcycle. 18

• $MU earnings could be a volatility event, not a routine print. After the last couple of days of massive chop, positioning looks fragile into the report. 15

📉 Market Stress & Dip-Buying Setup

• Korean equities are not just selling off on noise. The market dropped nearly 10%, its biggest one-day fall since March 4, and triggered a halt after falling more than 8% intraday. The key catalyst flagged is tax-reform discussion around taxing unrealized gains — a policy shock, not just macro beta. 6

• KOSPI -10% looks ugly, but context matters. The index only round-tripped to where it traded ten days ago. That makes the move feel more like a violent reset than a full trend break. Healthy flush, not necessarily panic regime. 12

• Dip-buying is still the market’s default playbook, but the open looks dangerous. $QQQ was down nearly 3%, with $MU earnings on deck. Bounces may still show up, but key levels matter more than chasing the first green candle. 19