Value of $10,000 invested 15 years ago: Nvidia: $4,043,000 Tesla: $2,490,000 Netflix: $310,900 Eli Lilly: $265,300 Amazon: $256,300 Alphabet: $216,800 Apple: $211,200 Mastercard: $203,200 Visa: $167,600 Microsoft: $157,000 Costco: $138,700 Home Depot: $90,800 Starbucks: $52,900

DUNNING-KRUGER EFFECT 🆚 SOCRATES After seven months on @X, one thing has become increasingly clear to me: this platform is a perfect illustration of the Dunning-Kruger effect. Most people drown in a stream of shallow facts that don’t add up to knowledge – they only create the feeling of being well-informed. Professor Faber captured this brilliantly in Ray Bradbury’s Fahrenheit 451. Explaining to Montag why people abandoned books, he said: “Cram people full of facts, fill them up to the brim with data they never digest, and they’ll feel brilliantly informed – without noticing how fast they’re heading toward the abyss.” Many here write just to write. Many read just to read. And both share the same trait – a deep conviction that they already know enough. Most minds are stuffed with so much scattered “noise” that there’s no time or energy left for actual thinking. And yet, as Faber also said: “We have plenty of time. The question is whether we have time to think.” When something goes wrong in life, career, or markets, the cause is rarely conspiracies, governments, geopolitics, or specific leaders. The real reason is much closer: the content we choose to consume. We shape it ourselves – with our attention, our likes, our replies, our comments, or by ignoring it. We are what we eat. We are what we drink. We are what we read. We are what we watch. What we give our time to – people or things – is our choice, and it shapes us either way. If you want a strong body – go to the gym every day. If you want good health – stop drinking, smoking, overeating and cut down on sugar. If you want to develop your mind – read. Here are my favorite authors: Asimov, Balzac, Bradbury, Belyaev, Boussenard, Bulgakov, Clarke, Dick, Dostoevsky, Dreiser, Dumas, Fitzgerald, Goethe, Huxley, Poe, Ritchie, Sacher-Masoch, Stevenson, Tolstoy, Twain, Verne, Wells, Nabokov, Orwell, Heinlein. They will gladly open their pages to you. And one last thing. Every watch I own carries an engraving of a line attributed to Socrates: Οὐδὲν οἶδα ὅτι οὐδὲν οἶδα “I know that I know nothing.” I see these words every morning when I put my watch on, and every evening when I take it off. They keep me grounded. If each of us tries – even a little – to be closer to Socrates, and not to the person who “knows everything” after reading two random posts, the world around us will start changing for the better. But not by magic. It starts with us.

Education Saturday! 📓 For all my Photonics followers, you'll want to learn this $AXTI $LITE $COHR $GLW $POET etc. The Photonics Supply Chain: Start to Finish. Step 0: Mining and Refining It starts with a metal called indium. Indium is what makes high-performance data center lasers possible today. Without it, you can't build the components that move data at the speeds AI demands. Indium has no dedicated mines. It doesn't get extracted on its own. It's a byproduct of zinc refining, meaning it only gets recovered when zinc smelters have the equipment and economic incentive to capture it from their waste. Step 1: The Substrate Refined indium gets combined with phosphorus to create a material called Indium Phosphide, or InP. InP has a unique property in that it can generate light directly from electricity. Silicon, the material that runs everything else in computing, cannot do this. That's why InP is the go to for the lasers inside data center optical components. The first thing you make with InP is a wafer which is a thin, flat disc that serves as the foundation for everything built on top of it. InP wafers are expensive, brittle, and difficult to produce at large sizes. The industry is only now moving to 6-inch wafers. For context, standard silicon chip fabs run on 12-inch wafers. That size gap is a big part of why photonics capacity is so hard to scale quickly. Step 2: Epitaxial Growth A bare InP wafer still can't do anything useful. To create a laser, you have to grow extremely thin additional layers on top of it, each just a few nanometers thick (a human hair is roughly 80,000 nanometers wide). This process is called epitaxy. The exact chemical composition of each layer determines the laser's wavelength, power, and efficiency. Get it slightly wrong and the entire wafer is scrapped. This step requires specialized equipment found in very few places in the world. It's rarely talked about, but it's one of the most critical bottlenecks in the entire supply chain. Step 3: Wafer Fabrication Now the actual circuit gets built. Using techniques similar to semiconductor chip manufacturing (patterning, etching, depositing materials etc.) engineers carve microscopic structures into the wafer: the channels that guide light (called waveguides), the cavities where light gets amplified, and the components that switch it on and off. Unlike standard chip manufacturing, at this time this cannot be done in a regular semiconductor fab. It requires a dedicated photonics facility. These take years to build, qualify, and ramp. There are very few of them in the world. Step 4: Dicing and Yield The finished wafer gets cut into individual chips. Each chip is then tested to see if it actually works to spec. The percentage that pass is called yield and it's one of the most important numbers in this business. Low yield means high cost per working chip. Improving yield is one of the biggest levers on profitability, and it's hard-won through years of process refinement. You probably won't see it reported directly, but it's hiding inside gross margins. Step 5: Component Assembly A working laser chip still can't be used on its own. It has to be physically aligned to an optical fiber with tolerances finer than a fraction of a micron and combined with other components like light detectors and signal modulators to create a functional optical sub-assembly. Automating it reliably at high volume remains one of the hardest manufacturing problems in the industry. The assembled component also has to be hermetically sealed inside a protective ceramic and metal enclosure. Data centers run hot, and moisture or dust will degrade a laser chip quickly. These hermetic packages are specialty components with few suppliers and long lead times and they have shown up as a bottleneck alongside InP when demand spikes. Step 6: The Transceiver Module The optical sub-assembly goes into a housing along with a DSP chip (a processor that cleans up and interprets the light-based signals) plus a circuit board and casing. The result is a pluggable transceiver: the finished module that slots into a switch or server in a data center. These individually get tested before it ships. That testing process is slow and expensive, and it's a hidden constraint on how fast output can actually scale. Note, in the world of CPO this will change. Step 7: Into the Data Center It plugs into a port on a network switch inside the data center, the hardware that routes data between thousands of servers. And none of it moves an inch without the fiber it runs through. Ultra-pure glass strands, thinner than a human hair, carrying light signals between every switch, server, and building. Follow stocks in this space? Drop a ticker below and I'll tell you exactly where they sit in the stack 👇

Amazon is difficult to own because it has diminished free cash flow from debt... I say stay in it but i know it went from cheap to expensive for a lot of people after that last q...

INVESTING IN BLUE-CHIP STOCKS AND LONG-TERM HOLDING – WHAT’S THE CATCH To my surprise, X is full of popular accounts making, in my view, dangerously misleading claims – like “10 Dividend Stocks to Buy and Hold Forever” @dividendology. I clearly see how this could cause harm in the stock market, but it’s hard to imagine how it could help – it’s like going into a fight with Conor McGregor after watching a 5-minute YouTube video on the most effective UFC chokeholds. There’s a persistent myth: to make money in the stock market, you need to buy the best stocks of the best companies and hold them as long as possible. Some even suggest holding them… indefinitely. The main flaw in this theory is that the list of blue-chip stocks constantly changes. For illustration, here’s an infographic of the top 10 US stocks since 1960. Investors who bought blue-chip stocks 60, 30, or 15 years ago often missed the mark. Among the past top companies, five went bankrupt, some were acquired, and many of the remaining ones are now in poor shape. NOTABLE EXAMPLES 🔹 General Motors $GM – a global auto industry leader of the 20th century, producing up to 40% of all cars worldwide. In 2009, the company went bankrupt and was restructured with US government support. Today its market cap is around $53B, less than Maruti Suzuki India Ltd at about $58B. Symbolically, the American giant now ranks below the Indian subsidiary of a Japanese brand. 🔹 Xerox $XRX – the icon of the 1970s office era, whose name became a verb. Today its market cap is $399M. 🔹 AOL – the largest internet provider of the 1990s with 30M users. In 2000, it acquired Time Warner for $165B. Today AOL no longer exists. 🔹 Kodak $KODK – controlled 90% of the film market in the 1990s, missed the digital revolution, went bankrupt in 2012. Today it survives in niche markets like printing and retro products. Market cap: $513M. 🔹 Sears Roebuck – $AMZN of its time: catalogs delivered to every American village. It couldn’t survive the e-commerce shift and finally went bankrupt in 2018. Which of the “untouchables” – $MSFT, $V, $MA, $ASML, $AVGO, $COST, $CNI, $WM, $MPLX, $JPM – will still be thriving in 20 years? Which ones will disappear completely? What do you think?

Today, we reported our Q4 and full-year 2025 financial results. The highlights include: - ARR as of year-end was $1.25B, ahead of our most recent guidance of $900M–$1.1B - This paves the way for significant continued growth. We are on track to end 2026 with ARR of $7B–$9B

Yesterday I was asked about $SPX / #M2SL. While answering, it hit me – not everyone actually sees the essence of what I write. When was the first touch? January 12. Take high-risk assets like $NBIS, $APP, $HOOD, $SOFI – when did their corrections actually start? The second touch was January 27. Check the charts yourself for any asset you follow. We are now in a correction tied to a “two-touch” all-time high from 25 years ago. And I warned about this more than 3 months ago. Below are two pinned posts from January 12 and 27 – about the touch itself. Even further down are two November posts explaining the $SPX / #M2SL indicator. If you want to understand what’s happening, start there. Please be honest in the comments – did you read all four posts before? Is the content difficult? First – even when I explain things as simply as possible, the subject itself is comparable to open-heart surgery. By definition, it can’t be easy. Second – the problem isn’t how I explain it (I almost convinced myself it was). The problem is that many simply aren’t familiar with the underlying material. You don’t build a skyscraper starting from the 14th floor. You start with the foundation. Everything you need for that foundation is in the pinned post. Missed it? Time to catch up.

📣 SoFi delivers another record-breaking quarter! In Q4 we delivered a record $1 billion in adjusted net revenue and had the biggest increase in new members and new products in SoFi history. This is durable growth, powered by our one-stop shop. Full results here: businesswire.com/news/home/2…