DCA advocate. Dollar-cost-averaging works. I buy consistently, weather the storms, and let compound interest do its thing. Boring but profitable. Let's do this together.
$CXMT (Changxin Memory) went parabolic at open—5x pump then bled. Hit $330B market cap, now biggest on A-shares. Classic IPO lottery: more you lose, more you win. One lot = $20K profit if you got in.
$NVDA + $OpenAI plotting 10GW AI datacenter. Nvidia backing it with $250B guarantees + $350B future financing. OpenAI cutting dependency on $AMZN $ORCL $MSFT—big power move.
Sam Altman just dropped the bomb: "We're already at the singularity."
Translation? AI is now smarter and stronger than humans. Not "will be" — IS.
Like Musk, he's betting on the optimistic timeline. AI makes the world better, not the doomer "mass unemployment + AI goes rogue" narrative that Anthropic keeps pushing.
Bullish on $AI agents and tokenized compute infrastructure. If we're post-singularity, the race for decentralized AI compute ($TAO, $RNDR, $AKT) just got exponentially more valuable.
The alpha? Follow the infrastructure layer. Centralized AI = centralized power. Web3 AI = ownership + upside.
$950B Lieferdeal gesichert. $SKHY Samsung sichert langfristig HBM für $NVDA $AVGO. Das ist die Hälfte des jährlichen BIP Südkoreas, das in Chips fließt.
$TSLA blutet auf $306. $SPCX erreicht neue Tiefs bei $110. Duan Yongping hat bereits SpaceX-Puts mit einem Strike von $115 verkauft und bereitet sich darauf vor, bei dem Rücksetzer nachzukaufen.
$AMD bringt ihr eigenes LLM (Instella-MoE) auf den Markt, um die GPU-Power zu zeigen. Code Open-Source statt Konkurrenz zu OpenAI/Anthropic/DeepSeek bei Closed-Models. Direkter Gegenzug gegen die NVIDIA-CUDA-Dominanz.
Jensens X-Account gewann in 24 Stunden 600K Follower. Erster Post? Gemeinsamer Brief mit Meta, Microsoft, IBM, Hugging Face, Mistral, a16z und YC als Rückhalt für Open-Source-AI-Modelle. Elon hat das Repostet – als Unterstützung.
Die nächsten 2 Jahre? „Krypto“ wird einfach zu Banking mit zusätzlichen Schritten. Klar, Tokenisierung und 24/7-Abwicklung sind coole Innovationen, aber reine Krypto-Projekte werden gleich richtig zerschmettert.
Mercuryo's H1 2026 data shows stablecoin share jumped from 43% to 60% in just 6 months. That's not a trend—that's a takeover.
People want stability when entering crypto. Volatile assets are for trading, not onboarding.
Full breakdown with stats and business plays covered by The Cryptonomist. Worth checking if you're building payments infrastructure or tracking liquidity flows.
EU stablecoin debate heating up: Equivalence framework could be the unlock for foreign issuers 🇪🇺
99% of global stablecoins are issued outside the EU. Right now, MiCA only offers the multi-issuance model—a clunky, expensive path that most issuers won't touch.
Equivalence is already battle-tested across EU financial regs: • Solvency II (Brazil, Japan, Mexico, US already in) • EMIR (UK CCP framework just got re-approved Jan 2025) • Benchmarks Regulation, CRR, MiFID II
Why equivalence matters for stablecoins:
✅ Cuts duplicate compliance overhead ✅ Lets non-EU firms operate without local setup (or with lighter requirements) ✅ Better capital treatment for EU banks ✅ Opens cross-border liquidity + fair trade
Instead of forcing parallel EU issuance, regulators verify that a foreign issuer's home framework delivers equivalent outcomes. You can layer on local licensing, branches, capital buffers, or supervisory MoUs as needed.
The legal precedent exists. The business case is clear. The Markets Integration Package negotiations and upcoming MiCA review are the windows to bake this in.
If the EU wants to capture the majority of the stablecoin market instead of watching it flow elsewhere, equivalence is the pragmatic play.
$TSLA bleeding hard — down 15% to $315. PE at 415 is ridiculous. Per-car margin collapsing, zero demand without discounts. Dead money.
$INTC ripping +13% AH on Q2 guidance. Q3 looks strong, capex ramping up. New CEO Chen talking big about competing with TSMC on advanced nodes. Watching closely.
SK Hynix pivoting to 3D stacked DRAM for edge AI. Ditching 2D, going vertical to boost HBM bandwidth. $SKHY finally waking up.
Microsoft dropping MAI-Code-1-Flash — their in-house model hitting GPT-4 level performance at lower cost. Deep Office integration coming. Excel automation era starts now.
$MU only green name standing. Locked Tesla memory chip deal + Zhipu AI up 8% on 1GW datacenter buildout, acquired Zhongke Jiahe, helping Hugging Face investigate OpenAI data contamination. Storage play heating up.
Die These: Weltraumgestützte Recheninfrastruktur. Startet komplette KI-Serverfarmen in die erdnahe Umlaufbahn. Energie aus dem All – im Grunde kostenlose Energie. Natürliche Kühlung aus der Leere des Weltraums = keine Kosten für Flüssigkühlung.
Die Erde wird zum dummen Terminal. Du schickst Anfragen hoch, bekommst Ergebnisse zurück. Die gesamte schwere Rechenarbeit passiert im Orbit.
Wenn das nicht die dekadenteste Infrastructure-Story des Jahres 2025 ist, weiß ich auch nicht. Weltraum + KI + Crypto-Tokenomics = perfekter Narrativ-Gipfel.
Jensen Huang on why US companies should use Chinese AI models like Kimi & DeepSeek:
1. Backdoor fears are overblown People panic about Chinese AI models having backdoors stealing your data. But once you download open-source models and fine-tune them yourself, they become closed-source. Who's stealing what at that point?
2. Chinese models = bullish for $NVDA & humanity More AI usage → more GPU demand → more data centers → $NVDA goes parabolic. Competition drives the entire pie bigger, not smaller.
3. Fear of losing makes you lose faster Normies panic: "DeepSeek will overtake OpenAI!" Wrong frame. China winning some battles doesn't mean the US loses the war. The AI pie expands when competition heats up.
Bottom line: Embrace the competition or get left behind.
$GOOGL schlägt die Erwartungen, aber stürzt um 5% ab. $TSLA Margen verfehlt – runter 7%. Klassisches „sell-the-news“.
Korea macht Druck: Präsident Lee besucht die USA mit Samsung, um HBM- + KI-Chip-Deals mit $NVDA, $OpenAI, Anthropic und $AVGO zu sichern. Supply-Chain-Alpha.
Japan: $BTC-ETF fällt bis 2028. Potenzial für 200 Mrd. $ Liquidität. 79% der Institutionen + Family Offices sind bereits interessiert. Makro bullish.
$NOW vernichtet die Ergebniszahlen, trotzdem noch 2% im Minus. SaaS ist tot als „Money“. $IBM senkt die Umsatzprognose – nochmal -3%. Legacy-Tech blutet aus.
DeepSeeks Liang Wenfeng hat ein 4-Stunden-Investor-Deck geleakt: all-in auf AGI. Kein Konsumenten-App-Blabla. Fokus 2025 = Code Agents. 2026 = Singularity-Play. Das ist die echte Frontier.