The $6 Trillion Question. Our first video is live.

This week, the AI trade met its first real resistance.

New York became the first state to say no more data centers. IBM became the first megacap to lose a quarter of its value in a day this cycle. And the inflation report handed the Fed exactly the cover it needs to keep money cheap — which is to say, to keep the whole machine running.

We picked this week to publish our first video: a 10-minute anatomy of the six trillion dollar bet holding up the market. The timing was luck. The thesis wasn’t.

Ten companies hold 41% of the S&P 500. AI infrastructure drove 92% of U.S. economic growth. And more than $800 billion now moves in a circle between the same few companies — Nvidia funds OpenAI, OpenAI pays Oracle, Oracle buys Nvidia chips.

▶  WATCH THE FULL VIDEO — 10 MIN

The buildout, the circle, the productivity paradox, both crash scenarios, and the surprisingly strong bull case — every number sourced. If it changes how you see the AI boom, subscribing to the channel is the best way to support this work. The next episode takes on the strangest mystery in economics: the productivity paradox.

New York became the first U.S. state to impose a ban on AI data centers — a direct constraint on the buildout that currently accounts for nearly all U.S. growth.

Data centers need three things: capital, chips, and permission. The market has priced the first two obsessively and the third at zero. If other states follow New York — and grid strain plus water politics say some will — the $6 trillion machine has a bottleneck nobody modeled.

Watch the follow-on trade: states that welcome data centers just became more valuable. Texas, Virginia, and the desert West are the other side of this story.

IBM stock collapsed roughly 25% in a single session after preannouncing its Q2 results — one of the sharpest megacap wipeouts of this cycle.

The lesson isn’t about IBM. It’s about what this market does when an AI-era story wobbles: no partial credit, no soft landing, straight to repricing. Cybersecurity names rallied the same day on AI-spending commentary — the money didn’t leave, it rotated. That’s what a crowded trade looks like when it gets nervous: violent moves within the theme before anyone questions the theme itself.

June consumer prices fell 0.4% month-over-month — the biggest monthly decline in six years — with core CPI flat on the month and running 2.6% annually.

Cooling inflation means cheaper money for longer, and cheaper money is oxygen for a debt-financed buildout. The new Fed chair testified before Congress this week with senators asking specifically about data centers and AI — read that again: AI infrastructure is now a congressional monetary-policy topic.

19 new billionaires: The AI boom keeps minting winners at the top of the stack — even as bank earnings season surfaced AI-driven job cuts at JPMorgan. The two-speed economy our video calls the quiet part of the story.

Chips and China: A U.S. trade official said “very few” Nvidia H200 chips have actually shipped to China — the export spigot everyone argues about remains barely open.

Apple goes small: Apple is reportedly in talks with a startup that shrinks AI models to run on an iPhone. The counter-thesis to the data center buildout, quietly, from the biggest company of them all.

Banks print: Big-bank earnings came in strong, from Goldman’s SpaceX IPO fees to blowout trading desks. The financing layer of the boom is getting paid either way.

Week 6 of 52: Learn to spot circular revenue.

When Company A invests in Company B, and B spends it with C, and C buys from A — revenue is real on every income statement and fragile everywhere at once. This pattern decided the dot-com crash, and it’s running again at 10x scale.

This week’s action: Watch our video’s money-circle chapter (minute 3 to 5), then pick any AI company you follow and ask one question of its latest quarter: who is the customer, and where did the customer’s money come from? The person who can answer that question sees the cycle before the headline does.

Connect this week’s dots. New York just proved the buildout has political limits. IBM proved the market has no patience for wobbles. The CPI print proved the Fed will keep the money cheap. Constraint, fragility, and fuel — all in the same seven days.

That is not a market at equilibrium. That is a market waiting to find out which force wins. Our video ends with the three lines that will answer it: hyperscaler free cash flow, credit spreads on AI-linked debt, and whether customers ever start reporting real gains. This week added a fourth: watch the statehouses.

Earnings season accelerates — the AI capex line in every megacap report is the only line that matters this quarter.

Fed chair testimony fallout — markets parse every word on data centers, rates, and the inflation glide path.

States after New York — watch which legislatures float data center bills next. The map is the trade.

Video 002 enters production — the productivity paradox: why every technology revolution takes decades to show up in the numbers, and what that predicts about AI’s payoff.

We’ll be watching all of it.