The Shit to Know Thursdays

This market doesn’t ring a bell before it moves—it tests conviction. Days like today are designed to shake tourists, bruise narratives, and see who actually understands what they own. Headlines scream, prices wobble, timelines panic… and quietly, the future keeps getting built.

When the tape feels uncomfortable, that’s usually the point. Leadership rotates. Illusions get taxed. Real themes tighten their grip. If you’re paying attention, the noise isn’t confusing—it’s clarifying. This is where separation happens.

Let’s get into it.

Mag7
MICROSOFT TANKS ON FLACCID EARNINGS

Microsoft dropped the quarter, beat numbers, and still got smoked. Why? Because Azure growth slowed by just enough to crack the illusion. In a market trained to believe cloud curves only go up and to the right, even a tiny bend feels like betrayal.

This wasn’t panic about customers leaving or AI dying. This was investors waking up to the cost side of the AI equation. Models are cool. Compute is expensive. Power isn’t free. And capex is finally big enough that you can’t just hand-wave it away with buzzwords.

Software isn’t dead—but the fantasy version of software is. The market is done paying infinite multiples for mature platforms and wants proof that all this AI spend eventually turns into real, defensible profit. Welcome back to gravity.

M&A
SPACEX MERGING WITH XAI?

SpaceX talking merger with xAI is peak Elon. This isn’t about financial engineering—it’s about control. Launch rockets. Own satellites. Generate data. Train models. Distribute information. Close the loop. Repeat.

This is Elon doing what he always does: ignoring categories and stitching together systems. Cars weren’t enough—so he built batteries. Batteries weren’t enough—so he built factories. Now AI isn’t enough without orbit, bandwidth, and real-world data. This is vertical integration at a planetary scale.

People joke about “mega-corps,” but this is what one actually looks like being assembled in real time. While public markets argue over quarterly Azure deltas, Elon is laying down infrastructure that doesn’t care about narratives—only physics, scale, and time.

Theme Picking
THEMATIC STOCKS STAY STRONK

The indices looked rough, sentiment got weird, and traders started reaching for the macro excuse drawer. But leadership told the truth: the right themes didn’t break. Memory and space stayed bid because they’re not trades anymore—they’re requirements.

SanDisk came out and absolutely torched expectations. Datacenter demand wasn’t “solid,” it was aggressive. Customers aren’t debating AI spend—they’re racing to secure supply before it disappears.

That’s the tell. When the market sells stories but buys bottlenecks, you’re late-cycle in hype but early-cycle in infrastructure. Storage is oxygen for AI. Space is the new high ground. Money knows that, even if Twitter doesn’t.

MRAM: Everspin
MEMORY AND SPACE IN ONE STOCK? YES PLEASE

Everspin is the opposite of sexy, which is exactly why it works. MRAM is memory that doesn’t care about radiation, power loss, or chaos. That makes it perfect for satellites, defense systems, and anything Elon might eventually shoot into orbit.

This isn’t fast adoption tech. It’s slow, painful qualification tech—the kind where once you’re designed in, you don’t get replaced. Engineers don’t gamble in space. They pick what survives and stick with it for a decade.

That’s the VHLA angle: small market, real moat, structural tailwinds. While everyone chases the loud AI names, this is the kind of company that just quietly ends up everywhere important.