Dear Member,

The Fed got a new chair on Friday. Kevin Warsh walked into a 30-year yield above 5%, inflation at 3.8%, and the most divided FOMC since 1992.

He also walked in with an explicit view that AI is a “significant disinflationary force” — which is his code for permission to cut rates into the largest industrial buildout in human history.

So while Wall Street debates ripples on the surface, here’s what the people actually building the thing are saying.

Nvidia’s founder, Jensen Huang, famously called a day in November of 2022 the “ChatGPT moment” — a day that changed everything. From that one breakthrough, he said, it was clear the world would go through a trillion-dollar transformation in how computers work.

We talked about it in my first AI-Innovation Portfolio note back in June of 2023 — and how it would likely usher in an era of multi-trillion dollar companies. We’ve seen it. Still, Jensen underestimated. It’s going to be far bigger than a trillion dollars. And it’s happening in years, not a decade.

With that, Jensen thinks we’ve recently had another one of those “moments.” It was February 5th. On that single day, two of the most important AI companies in the world didn’t just release a better version of a generative AI model. They released something fundamentally different. “Always-on” AI. AI that does your work. On its own. All day. All night. Without you in the room.

Jensen said what happened that day “did for AI, what ChatGPT did.” It changed the game the same way ChatGPT did three years ago. It was another world-changing moment.

Here’s why: Before February 5th, the demand for AI was driven by people.

People ask questions. People sleep. People go on vacation. The whole thing ran at human speed. After February 5th, it stopped running at human speed and started running at machine speed.

We’ve talked about the “always-on” inferencing, where every time someone hits enter, a machine pings a model, and the model goes into reasoning mode and creates output. When inference is running, the meter is running. Revenue is being produced. New data is being created.

Now it’s a machine hitting enter.

They don’t stop on Friday afternoon. They run around the clock, consuming electricity and computing power continuously. The demand doesn’t stop. The revenue doesn’t stop. And that’s why the datacenter builders have been building capacity to accommodate that demand as fast as they can. They knew this was coming — the demand curve going from normal to vertical.

This is referred to as the “singularity curve.” In singularity, AI capabilities broadly exceed human intelligence, and can rapidly self-improve.

Elon Musk says we are there.  When the world changes fast enough that the old model breaks.

With that, let’s talk about some commentary from these two guys (Jensen and Elon). In a tech revolution that’s accelerating by the day, it’s a huge advantage to hear what the most important CEOs in the world think about what’s happening, where it’s going, and what constrains it.

Jensen sat down earlier this year with podcaster Lex Fridman for a nearly three-hour interview. It’s the most expansive interview I’ve heard him give. Jensen calls the investment in AI computing capacity the biggest industrial buildout in human history.

In October, he said there was half a trillion dollars of demand to fulfill. At Nvidia’s developer conference, he upgraded that to at least $1 trillion through 2027 — and then he said he’s confident it will be higher. Nvidia is fulfilling on the demand.

In the Lex interview, Jensen explained why the demand outlook keeps going up. This is the “compute equals revenue” logic. Every dollar of compute capacity added is being monetized the moment it comes online.

The “tokens” are profitable.

When he says tokens, he means units of the output AI systems generate every time they answer a question, write a line of code, or execute a task. It’s not just measuring usage, it’s now clearly resulting in revenue. Not just revenue for the datacenter, revenue for the end user.

The latter is why token usage (output) is exponentially exploding. So much so, that customers are capacity-constrained. They can’t get enough. And they can’t get enough, because the compute generates output, the output generates revenue, and that’s why the revenue is insatiably funding more compute.

That’s the boom loop.

So, when Jensen was asked whether Nvidia could become a $3 trillion revenue company, he said yes.

For perspective, that’s more than 10 times the size of the recent annualized quarterly revenue. When asked about a $10 trillion market cap, he said Nvidia’s growth path is “extremely likely” and, in his mind, “inevitable.” That’s 2.5x the current market cap.

So, the CEO of the most valuable company in the world thinks 2.5x’ing from here is “inevitable.”

What is Elon saying?

Elon did an interview earlier this spring with tech futurist podcaster Peter Diamandis. Elon and Jensen are on the same page. He said he believes the economy could be 10 times larger in 10 years. And here’s what he said about the current environment…

“We’re in hard takeoff. Right now.“

He said, “I go to sleep, there’s some massive AI breakthrough. When I wake up, there’s another one.”

What does “hard takeoff” mean?

It means the human bottleneck on model progress gets taken out of the process, the machine improves itself.

The exponential part of the growth curve is here — where AI understands its own code and hardware. It begins rewriting its own software more efficiently. The progress accelerates because it can rewrite itself faster and more effectively than before. The cycle repeats exponentially, leading to a massive leap in capability, almost instantly.

And the timing of this becomes very interesting. Because Elon confirmed Optimus 3 production starts this summer — to be in high-volume production by summer 2027.

This is Tesla’s humanoid robot. And it’s important because the capabilities of this version (v.3) are what Elon has said many times, at scale — create effectively unlimited labor, and therefore, ultimately, a limitless-sized economy.

And with the latest model capabilities, Optimus 3 will likely improve itself.

So, this is very important perspective. Jensen runs the company that builds the hardware the entire AI industry runs on. He sees every order, every backlog, every bottleneck in real time. He thinks his company will be worth a third of the size of the current U.S. economy.

Elon is the richest, arguably most consequential person of our lifetime, with a direct line to the most powerful person in the world (Trump), and is building the AI, the robots, the energy infrastructure, and the rockets that may someday transport datacenters to space.

He thinks the idea that the economy could be ten times its current size within ten years is a “comfortable prediction.”

And he thinks the machines will eventually produce so much (so many goods, so many services) that they’ll run out of things humans can even think to ask for.

If they are even half right, we are living through the single largest economic expansion in human history.

Meanwhile, the daily financial news — the Warsh handoff, the tariff headlines, the interest rate guesses, the market going up or down on any given day — is a conversation about ripples on the surface of the ocean.

As Elon says, would you clean up the beach if you knew a Tsunami was coming (a limitless-sized economy)?

So, the people closest to building this thing are telling us plainly how large it’s going to be. It’s the abundance outlook. And in a world of abundance, you want to invest in scarcity.

That’s all for now. If you’re not yet a member, I recommend both portfolios for what’s coming.

Our Billionaire’s Portfolio owns the still-undervalued producers of the hard assets that fuel the AI revolution — copper, gold, oil and gas, and the legacy tech being revalued as AI-critical infrastructure. Our AI-Innovation Portfolio owns the scarce physical inputs the AI buildout cannot exist without — the power, the fiber, the connectors, the chip equipment, the network edge.

Two portfolios. One thesis.

See the track records here: Billionaire’s Portfolio → and AI-Innovation Portfolio →.