Welcome to Nofinity

Summary follows...

The Nivona Cube is a fully automatic coffee machine that borrows key techniques from manual espresso machines (portafilters) to overcome common flaws: uneven extraction and channeling.

• Biggest plus: You see the ground coffee before brewing and can distribute it evenly (tamping).
• Traditional super-automatics create uneven coffee cakes → channeling → bitter and sour drinks.
• Manual leveling ensures uniform extraction → much better flavor.

• Removable brew basket (hand-wash only).
• Conical burr grinder (grinds 12 g in ~8 seconds).
• Grind settings: Fine (espresso) to coarse (lungo), but overall too coarse for true portafilter espresso.
• Auto-tamping adjusts to coffee amount → clean puck.
• Dead space (~12.5 g): Old grounds affect first brew; tip: preheat the basket for better results.

• 12 g coffee → 28–32 g brew → TDS ~6% ✅ qualifies as espresso.
• Taste: Solid, but limited by crema valve and grinder.
  • Crema valve (fancy term for dual-wall basket) creates artificial pressure → fake crema (not flavor-relevant).
  • Grinder limit: Particle size in coffee-crema range (400 μm peak) vs. espresso (200-300 μm).
• Temperature: Highest setting recommended (measured ~85°C without resistance).

• 12 g coffee → 70 g brew (1:5.8) → best coffee-crema from a super-automatic in a long time.
• Clean, full-bodied, sweet, cocoa-like – no stale machine taste.
• Brew profiles:
  • Dynamic (1 bar): reduced flow → bitter-metallic; better for lighter roasts.
  • Constant (3 bars): recommended → slow pressure build → balanced, similar to vibe-pump espresso machines.
  • Intense (5 bars): extends extraction time → pointless for lungo; only for experimenting with light roasts.
• Pre-infusion: Wets only ⅓ of the coffee bed → minimal effect.

• Best for:
  • Households wanting high-quality longer drinks (80–100 ml) rather than pure espresso.
  • Beginners who want more control than a standard super-automatic.
  • Those valuing easy cleaning (no internal mess).
• Not ideal for:
  • Espresso purists → grinder too coarse, crema valve → dedicated espresso machine + grinder (e.g., Dedica + DF54 for ~€ 450) is superior.
• Price: ~€ 430 (DE) / CHF 500 (CH).

1. Preheat brew basket → massive flavor improvement.
2. Use regularly (avoid long pauses) to prevent stale dead-space coffee.
3. Quality coffee: Freshly roasted (check roast date), e.g., Compadre (darker roast) – balanced, chocolaty.
4. Grind settings:
   • Espresso: finest grind + minimum water (~12 g / 30 ml).
   • Lungo: medium grind (level 3) + 70 ml.

Conclusion: The Nivona Cube is not a portafilter replacement, but the best compromise in the super-automatic market for long, clean coffee drinks. Manual distribution is the game-changer – simple, effective, and fun.

This video demonstrates how to connect Claude with the data platform Quiver Quant to gain a legal edge in financial markets. Instead of using Claude just as a chatbot, it gets access to real, proprietary data – including:

• Insider trades (Corporate Insider Filings)
• Congressional trades (Congressional Trades)
• Institutional holdings (Institutional Holdings)
• Lobbying data and federal contracts

1. Claude subscription (or GPT with a similar prompt)
2. Quiver account (free basic version, premium from about $30/month for API access)
3. Generate API key and paste it into Claude Code – either via MCP connector settings or directly in Claude Code
4. Setup time: approx. 10 minutes

1. Spotting Insider Buys

Prompt: "Find every company where three or more different insiders made open market purchases in the last 60 days (exclude option exercises). Rank by number of distinct buyers and total dollar value."

• Advantage: Buys are often a stronger signal than sales (sales may be for tax or diversification reasons).
• Warning: Raw data can be misleading – e.g., 31 buyers at TSM looks impressive but could be a one-time compensation event. Focus on outliers and cross-check with fundamentals.

2. Analyzing Congressional Trades

Prompt: "Show me the most recent congressional trades filed. Explain the difference between trade date and filing date."

• Key nuance: Insiders file within 2 days, Congress has 30–45 days. So congressional data is better for research, insider data for live alerts.
• Example (Palantir): Combine congressional trades, federal contracts, and lobbying spend → reveals the company's political positioning.

3. Finding Early Signals

Prompt: "Show me tickers where fewer than 5 distinct people traded but activity is increasing – versus crowded tickers. Explain how you define 'increasing'."

• Outcome: You spot trends before they become saturated – the "cleanest early signals."

To get instant notifications, build a bot:

1. Create a Telegram bot via BotFather
2. Send the token to Claude Code
3. Define criteria (e.g., "Alert me on any new insider buy over $1M")
4. Ensure 24/7 uptime (e.g., Mac Mini for $10/month VPS or your own computer)

Tip: With the bot, you can also use voice commands to query live data anytime – e.g., during a beach walk.

• Pros: Real-time data, no outdated web scraping, easy setup, high customizability.
• Pitfalls: Raw data needs interpretation (e.g., batch buys vs. genuine signals). Insider sales aren't always negative. Congressional data is often delayed.
• Overall: The system is legal (all data comes from public filings) and provides a real information edge that most investors don't use.

All prompts and bot setup guide are available in the video description (Free School Community).

• An attacker gained access to an Italian government email account (PEC system) and sent fraudulent but formally valid European Investigation Orders to Revolut.
• Revolut’s compliance desk responded for six months, leaking data of about 680 customers: passport scans, selfies, home addresses, bank statements, and full Bitcoin transaction histories.
• Victims were selected in advance based on their visible crypto wealth in public blockchain data.

• A KYC dossier is a complete profile: name, photo, address, and assets. Unlike passwords, this package cannot be changed and puts individuals at physical risk.
• Victims like Mark Karpelès (former Mt. Gox CEO) and Felix Romer (Gamdom founder) received blackmail attempts before Revolut’s official disclosure.
• The attacker demanded 6000 Monero (~$3M) in ransom; Revolut refused to pay.

• EU law forces financial institutions to release customer data to authorities quickly – hesitation is treated as obstruction.
• At the same time, AML regulations require storage for 5–10 years, while GDPR demands data minimization – a direct contradiction.
• Technologies like ZK-Passport (zero-knowledge proofs) could verify documents without storing them, but are politically ignored.

• Similar breaches at Ledger (272k customers in 2020) and Coinbase (69k customers via bribed support contractors) show a systemic issue.
• Physical attacks on crypto holders are rising sharply: 52 documented wrench attacks in H1 2026, including 20 home invasions.
• Conclusion: Self-custody and data minimization offer better protection than supposedly regulated platforms.

This video kicks off a series on how to strategically invest $1,000,000. Part 1 focuses on stocks and their role in a portfolio, tailored to individual risk tolerance.

• The S&P 500 has multiplied 11 times in 16 years, the NASDAQ nearly tripled in 4 years.
• Sentiment is euphoric: 'Make as much money as possible while the party lasts.'
• However, a 50% crash is a real possibility.

• 40% of the S&P 500 consists of just 10 companies – almost all in AI.
• Nvidia is the key driver: 70% of revenue comes from only 5 clients (Google, Microsoft, Amazon, Meta, Oracle).
• These clients' AI revenue is 70–75% from OpenAI and Anthropic, both burning cash.
• Question: What if this financing collapses? The real depreciation of AI chips is 2–3 years, not the 6 years on the books.

• GMO expects negative real returns for US stocks (–7% per year over 7 years).
• Value stocks (e.g., emerging markets, UK homebuilders, Xiaomi) offer 5–6% real returns with lower risk.
• Advantage: Higher dividends (up to 5%), low P/E (~10), and protection during downturns (dividend reinvestment compensates losses).

• The key isn't 'Which asset class performs best?' but 'What fits your risk profile?'
• Speculation (AI, S&P 500) vs. value investing with substance and dividends.
• Hedging (e.g., put options) will be discussed in the next video – viewers are invited to share their thoughts.

The series continues every Saturday with bonds and other strategies – always aiming to find the best plan for your portfolio.

US 10-year Treasury yields have surged above 5% – the fastest pace since the 1980s. This creates a stress test for equities.

• Oil & Diesel: Rising diesel prices fuel inflation; Trump threatens an export stop that could raise fuel costs in Germany by up to 25%.
• Earnings expectations downgraded: For the first time in months, more analysts cut earnings forecasts than raise them.
• Oracle crisis: AI mega data center “Jupiter” invokes force majeure – financial strain threatens the sector.
• OECD risk scenario: Warns of potential 15% stock market decline by 2027.
• Weak market breadth: 60% of S&P 500 stocks trade below their 100-day moving average.

• US economy booming: GDP Nowcast above 5%, PMIs at multi-year highs, 47% of companies plan to hire.
• Inflation expectations stable: Yield rise reflects strong growth, not runaway inflation.
• Yields in context: At 5.1% vs. nominal growth >5%, the level is reasonable, not alarming.
• AI investment intact: Nvidia B200 chip lease rates +30%, hyperscaler cloud revenues rising, South Korean exports strong.
• Bears outnumber bulls: 48% bearish sentiment – a contrarian signal that often precedes rallies.

• Weekly World Income ETF (HANetf): Weekly dividends via covered call strategy, 0.5% expense ratio, target ~8% p.a. (not guaranteed). Holds big tech, >81% US exposure.
• Memory stocks (Micron, SK Hynix) at historically low valuations – potential rebound.
• Low volatility & Dividend Aristocrats more resilient during yield spikes.
• Latin America: Brazil (93% of global niobium – key for AI/power grids), Argentina (Milei reforms). Political upside if Flávio Bolsonaro wins.

The rising yields primarily reflect strong growth, not an inflation explosion. Short-term volatility is possible, but historically rapid yield increases have had negligible impact on 1-year stock returns. Overwhelming bearish sentiment may actually be a contrarian buy signal.