To The Moon

If there’s no moat, who wins in AI? In a healthy supply chain, very few companies are making an outsized profit because high profits get competed away. Without some kind of moat or competitive advantage, there’s no pricing power or differentiation. There’s usually one power player, and everyone else is competing around the margins to gain a foothold as a commodity supplier, a niche modular supplier, a distributor, or play some other important, but often less profitable role. The iPhone is the perfect example of this. $Apple(AAPL)$ makes a gross margin of nearly 40% on its hardware, and the business overall has a 32.6% operating margin. $Samsung Electronics Co., Ltd.(SSNLF)$ is far less profitable in smar

Here’s the part of the AI trade I think investors are underestimating. If you're long $NVIDIA(NVDA)$ $Advanced Micro Devices(AMD)$ $Taiwan Semiconductor Manufacturing(TSM)$ $NEBIUS(NBIS)$ $Bloom Energy Corp(BE)$ and other AI infrastructure names, you're ultimately long the spending decisions of the biggest AI model companies. OpenAI and Anthropic matter enormously. They are among the companies driving the demand for compute, chips, power and data-center capacity. If their growth expectations or funding plans change, the impact can travel through the entire AI infrastructure chain. R

Investing is about risk and reward. And I think the current state of the AI buildout has more risk built in than many investors want to think. Those risks came to light over the weekend. In essence, Dario’s argument is that AI development should be “paced” in three ways: Embedded evaluators to verify the safety of models. This is similar to how big banks are regulated today. AI labs coordinate to create safety frameworks and limit “unchecked AI progress”. Coordination globally. I have a lot of thoughts, and I don’t think there are 100% certain answers to the safety, risk, or future development of AI. But I do think the argument and the fact that all of the major lab CEOs seemed to agree bring forward some major risks for investors. And it seems to be pulling the rug out from under some of

$Microsoft(MSFT)$ was formed in 1975, and the PC revolution started only a few years later. But it would take 24 years before half of the homes in the U.S. had a PC. In 1994, the first online payment was made, but it was 12 years before $Shopify(SHOP)$ was founded. $Amazon.com(AMZN)$ launched in 1994, and even today, only about 17% of purchases are made online. $Apple(AAPL)$ iPhone launched in 2007 when nearly everyone already had a phone in their pocket, and it still took six years for half of Americans to adopt the smartphone. Consumer adoption of products often takes longer than we think or remember, which is both an

When it comes to autonomous vehicle economics, I think we may be focusing on the wrong number. Vehicle cost is not the biggest variable. Utilization is. A robotaxi spends far more of its life generating revenue than sitting in a driveway, so how often that vehicle is actually carrying passengers can completely change the economics. Here’s the simple example from my model 👇 🚗 $70K vehicle → 30 rides per day vs. 🚙 $30K vehicle → 25 rides per day The more expensive vehicle can still generate better economics because it is being utilized more heavily. That’s why trying to win the market simply by making the vehicle cheaper can backfire. If lower pricing reduces the number of rides or revenue generated per vehicle, the cost advantage starts getting overwhelmed by utilization. And this is where

$Tesla Motors(TSLA)$ is coming at robotaxis with a simple thesis: Efficiency wins. Take the biggest part of the demand curve — one or two riders, short trips, dense metro areas — and drive the cost per ride as low as possible. Tesla’s Cybercab is clearly designed around that philosophy, with a small two-seat configuration and a focus on low operating costs. But there’s one problem. People don’t always choose the cheapest option. If cost were the only thing that mattered, everyone would take the bus. People pay for convenience. Comfort. Safety. Cleanliness. Predictability. Privacy. And sometimes, simply a better experience. That’s where $Uber(UBER)$ has an interesting advantage. Uber doesn’t need to provid

Over the past year, historic spending on the AI buildout has arguably kept the economy afloat. Yet, despite that historic spending, real GDP growth (growth on top of the rate of inflation) was just 0.5%, 2.1%, and 1.5%, respectively, in the past three quarters, well below what experts thought it would be coming into the year. That’s not a great rate of growth given the level of capital investment and there are plenty of signs consumers are being squeezed by high oil prices and interest rates. The problem is, I don’t think either can or will get any better anytime soon. More on that in a moment. Why Gasoline and Interest Rates Will Stay High One of the big misses a lot of prognosticators and CEOs made coming into 2026 was expecting interest rates to be lowered, boosting economic growth. For

$Joby Aviation, Inc.(JOBY)$ is making the eVTOL story a little easier to actually model. A new unit economics tool lets you play with the numbers yourself — how many rides per day does an eVTOL need, what price per ride makes the business profitable, and how quickly can the aircraft pay for itself? That last part is especially important. The site now has two models: ✈️ eVTOL Economics Test ride volume, pricing, profitability and payback period. 🚗 Autonomous Vehicle Economics Look at the potential ROI of an autonomous vehicle. The idea is simple: instead of just talking about the future of air taxis and autonomy, put the assumptions into a model and see what actually has to happen for the economics to work. And payback period may be one of the most

A large percentage of the Asymmetric Portfolio is invested in companies that could have major tailwinds from autonomous driving. My thesis is that many companies will make autonomous vehicles, leading to the modularization of components and technology, and aggregators like $Uber(UBER)$ ( ▼ 0.26% ) and $Lyft, Inc.(LYFT)$ ( ▼ 3.24% ) being huge winners as supply is commoditized. The view of a more autonomous future is consistent with many investors, but how I envision that future is very different. The market still thinks that $Tesla Motors(TSLA)$ ( ▼ 5.92% ) — who first promised Teslas could soon drive across the country fully autonomously in January 2016 — will d

The argument around $Tesla Motors(TSLA)$ ’s FSD has become pretty simple. Cost vs. safety. Tesla’s approach has always leaned heavily toward making autonomy work with a simpler hardware stack. But real-world driving isn’t predictable. When something goes wrong, safety often comes down to redundancy — having another system available when the first one fails. That’s why the FSD debate isn’t really just about whether the system can drive. It’s about whether the system has enough backup when the real world throws something unexpected at it. Tesla says FSD is still supervised and does not make the vehicle fully autonomous. That distinction matters. The bigger question is whether a lower-cost approach can eventually deliver the level of redundancy peopl