Deep Dive
The Schelling Point Framework
James opens with a core investing principle: the Schelling point. It's the obvious answer that doesn't feel stupid—like two strangers meeting in a city where every building is identical except one giant red clock. They gravitate there without agreeing beforehand. He applies this to Bitcoin, arguing AGI can code a better version in minutes, but Bitcoin's 15-year consensus on scarcity as insurance against fiat collapse is unrewritable. The Schelling point isn't technical elegance; it's trust as the product. James doubles down: look for obvious signs already staring you in the face. CAPEX spending, robot factories staging, Nvidia chip orders, Anthropic's ARR growth—these are all Schelling points. Stop chasing noise; gravitate toward the signal.
Meta's Valuation Trap and Hyperscaler CAPEX Race
Meta trades below its 200-day moving average, chopping between $640-660. The kill zone sits at $525-530. James flags three problems: earnings and revenue growth are flat to down, net margins compressed hard in Q2 2026, and massive CAPEX fears plague the stock. Meta plans $217B by 2028, Google $341B. The question is ROI—can they generate returns on these bets? Meta's track record terrifies him: they sank $100B into metaverse and got nothing. Now they're betting again on AI, renting out compute infrastructure they built but didn't use. Competition for frontier models is violent and racing to the bottom unless you own the hardware. Wall Street targets $753 (27% upside), with a high of $1,000 and low of $580. James would buy at $525-530 and sell at $680 in a tax-free account, but he doesn't think Meta wins the frontier model race like Google or OpenAI.
Bitcoin's $120K Target vs SpaceX's Faster Returns
Over the next 24 months before the next halving, Bitcoin averages $120K target—87% upside from current levels near $60K. Solid return, but the question is timing: should you sell at the bottom and chase at $120K? No. This is where you stack, not sell. However, SpaceX offers 65% upside in just one year and could easily hit $600 (4x return). SpaceX turns atoms into AI compute and orbital compute; it's in the center of multiple areas with extreme moats nobody can catch. In the age of AGI, things are happening fast and demand is insane. If forced to choose, James sees SpaceX as the faster horse. A viewer with 37% Bitcoin allocation should rotate some into SpaceX—way too concentrated. Bitcoin's purpose is insurance against fiat collapse and trustless cash, not get-rich-quick scheme. But diminishing returns are clear; last cycle was muted.
Solana vs Base: X4O2 Reality Check
A viewer worried Base's X4O2 activity was eclipsing Solana. James clarifies: Solana processed 184.7M daily transactions versus Base's 9.5M—nearly 20x difference. Base does 3x ETH (3.2M) but nowhere near SOL. On fees, Solana collected $708K daily while Base only $51K. X4O2 is Coinbase and Cloudflare's payment protocol, but Coinbase can play games moving money between their own agents to fake adoption. The real story: AI agents can rewrite APIs and create their own protocols, so X4O2 may not be the standard going forward. James doesn't buy tokens solely because AI agents need transactions. Look at real world asset tokenization (Solana dominates 92%), stablecoin volume, DEXs, trading. These are the actual use cases. SOL thesis hasn't changed; if a layer-one wins, it's Solana. Everything can change, but the data supports it today.
Tesla's Tipping Point and Emotional Volatility
Tesla had two very different earnings presentations. One call, Elon was sick and subdued. ZEV credits were down, energy margin dropped from 39.5% to 20.4%, free cash flow negative, and CapEx spending over $25B this year spooked markets despite record Q2 deliveries and robotaxi miles. The other call, Elon talked SpaceX hitting $100B ARR in September and $1T revenue pulled forward to 2030—stock ripped. James sees tribalism. When assets don't print money fast, retail throws mud. But look at the Schelling points: FSD is tipping (TikTokers doing makeup in Teslas), over 2,300 robotaxis staged nationwide, factories building at unprecedented scale creating extreme moats, Cybertruck coming, Optimus coming, Tesla Semi sold out for a year. Retail holdings crashed from 44% to 16% as weak hands panic-sold; Druckenmiller just loaded Tesla calls. Smart money buys dips. Potential acquisition in 2 years gives big premium. James's biggest position is Tesla and he's not sweating it.
Memory Demand Explosion and Micron's Runway
New AI systems now remember previous queries; six months ago they forgot yesterday's questions. Optimus robots and Tesla FSD cars will need memory for parking preferences, garage locations, all that data. Memory demand will scale 200% yearly while production grows only 20%—the math is brutal. Micron goes from $800 to $1,600, then nuts projections by 2030. No slowdown in memory demand exists. Bitcoin is half James's retirement model and he holds no Micron, but everyone needs some because it's an easy double. The growth trajectory is relentless; this is a Schelling point nobody's watching hard enough.