I though the big news of the day would be NVDA earnings, I was wrong. Intraday the TACO (Trump Always Chickens Out) trade started making the rounds of Wall Street. I would argue, as I said yesterday, that spikes in yields have caused him to walk back some statements.

I would also argue that poking the bear is not a good thing. Then after hours came news that the courts ruled he doesn’t have the power to impose tariffs. I’m not a lawyer but I do include a tweet above and below indicating that Goldman thinks he can do it anyway,

but markets are nicely green on this, and NVDA earnings. Now we have uncertainty on what Trump will do from here, which also creates uncertainty about what the Fed will do. I wouldn’t do anything major here until we hear more from Trump and perhaps gain a bit more clarity.

Yields are up so far this morning, keep an eye on that

We take a deep dive on NVDA earnings below, one key highlight that we hit a few days ago, and will hit again because it’s so big, is robotics……

Although, this could be a threat to autonomous vehicles…..

Wall Street analysts always behind the curve….

I went on Schwab Network yesterday to preview NVDA earnings….

The stock has already rallied so much off the liberation day lows I worried about whether they could do enough to test the year highs, especially since a lot has happened since those highs. Looks like I was wrong, but I still would be looking for dips and wouldn’t buy the rip. What I do think this does is clear the path some more, after the other Mag 7 names, to keep buying back into AI infrastructure. I ran the earnings through GPT and we seem to be on the same page…..

Forget the China charges. Forget the political noise. NVIDIA (NVDA) just proved—again—that it’s not only the AI king, it’s the entire throne room.

On Wednesday, Nvidia reported a jaw-dropping $44.1 billion in quarterly revenue, up 69% year-over-year, and smashed Wall Street’s already aggressive expectations.

Even with a $4.5 billion hit from China export curbs, the company printed $18.8 billion in profit, and signaled more strength to come. You’d think a company losing access to an $8 billion market would blink.

NVIDIA didn’t flinch.

NVIDIA’s edge has never been just about hardware. It’s about control of the AI developer stack. Every chip runs on CUDA—its proprietary software ecosystem that locks in developers like Apple did with iOS.

That’s the crux. While Wall Street obsesses over China and $5.5 billion write-downs, Huang is out in the Middle East, Asia, and now Europe, signing infrastructure deals with sovereign AI buyers.

Think of it this way: Microsoft sells cloud. NVIDIA sells civilization.

Despite a $2.5B sales loss from H20 chip restrictions and an ongoing $8B China headwind, Nvidia's forecast came within 1% of consensus—and gross margins are set to expand to 72% next quarter.

But here’s the catch:

You do not sell the infrastructure winner of a global arms race—especially not after it just shrugged off a geopolitical gut punch and still printed 69% YoY growth.

If you own it: hold it. If you missed it: wait for a pullback and buy it.

Don’t overthink it. This is Act II of the AI Supercycle, and NVIDIA still owns the script, the stage, and the box office.

Bottom Line:
Nvidia just proved it can lose China—and still outgrow the rest of tech. There are few companies in the history of markets that have created demand for $45 billion worth of silicon in one quarter.

This isn’t a chipmaker. This is the foundation of modern AI civilization.

And it’s still a buy.

Nancy’s been early—and right—on plenty of next-gen names. Tempus AI (TEM) might be the biggest swing yet. But this morning, Spruce Point Capital came out swinging with a short report, arguing that Tempus is just another overhyped diagnostics firm riding an AI label.

Shay Boloor’s rebuttal is worth your attention. Because he nails what most public-market investors still don’t get:

Tempus isn’t trying to sell tests. It’s trying to own the interface between human health and machine learning.

Spruce’s core argument boils down to this:

Here’s why that’s short-term thinking in a long-duration game.

Tempus runs a data pipeline business. The tests are the toll booth—every sample that goes through its system feeds the ML flywheel. Think Palantir’s Foundry, but for genomics, histology, and clinical trials. The AI revenue today is small—but the data asset is massive.

Acquiring Ambry is vertical integration 101. It ensures control of sample quality, logistics, and throughput. Same playbook Amazon used with planes and Tesla with batteries.

Spruce tries to paint Tempus as a billing-code roll-up. But the real bet is that Tempus becomes the “ChatGPT” for your health records—the thing your doctor asks before deciding what to do next. If that interface works, the company becomes indispensable.

Yes, valuation is rich. Yes, revenue is skewed toward diagnostics. But that’s always the case in the early stages of a data-platform company. Just like Palantir in 2020 or Axon before SaaS flipped the business model.

If Tempus turns its clinical pipeline into a sticky AI interface—used by oncologists, trial sponsors, and health systems—the monetization paths are exponential.

Spruce is measuring the wrong thing. This is not a company selling tests. It’s a company building the pipes through which future healthcare flows.

Shay Boloor said it best: Tempus is not about reimbursement codes—it’s about becoming the medical interface layer.

Nancy was early. The shorts are early—for the wrong reasons.

‍