Never really got to explain my reasoning for my $SIVB short - but suffice it to say now if you marked their holdings to market their book value would be negative

Analog semiconductor lead times are now up to six months!!

Unfortunately, Analyst #3 will be joining Cloudflare.

I remember in Q3 2023 we had like a 3 month long drawdown in every important AI name, all the people who thought AI was BS felt vindicated while the ones of us who believed continued buying. I think that same thing would kill half of you right now.

If you’re a bit more open to some ideas that aren’t AI bottlenecks, today the average of our small themes picks is 60bps.

I think the natural reaction of those familiar with fins is to assume that parabolic deposit growth sets up for disaster, which is basically a trauma response from SVB. However, I’ve spoken with Erebor - they know what they’re doing and are deeply thoughtful and considered operators. It’s not hard to imagine how “SVB but more attentive to the risks” results in an amazing business model.

Must be way better to be a VC now vs 10 years ago. SaaS era pitches probably packed with nonsense. Now it’s in vogue to invest in defense where the pitch can be: “We made a drone.” “Okay is it good?” “It is faster than the other drones” “Cool, here’s a check”

Palantir, the open source AI company?

Perhaps the best evidence of how fast AI has moved is that just three years ago this was impressive to us.

Even the moon over manhattan is long memory??

Fwiw Tae Kim is the Tae Kim of technology newsletters.

Hi, it’s me. I’m the guy at Citrini whose job it is to go to the software store every once in a while. It’s actually amazing I still have a job. What a freaking joke, right?

Before the ‘29 crash forced consolidations there were hundreds of car companies in the US. Robotics will follow a similar trend. Still, in the interim, I’m pretty stoked to robot mog. “Oh you went with the <perceived lower tier robot>? How quaint.”

More robots without legs

If your stock sells off into America’s 250th birthday you have to pay double corporate taxes this year

Regardless of whether this prediction comes true in the specifics, it’s almost certain that it comes true in the broad sense. DRAM prices are not going to keep going up the way they have without a significant incentive to use it more efficiently. You need to be an extremely firm believer in Jevon’s Paradox as it applies to memory to remain long the OEMs here, otherwise it is essentially a bet that’s short innovation out of necessity.

“I am buying a bunch of neoclouds because countries will need to secure compute to secure their future, it’s an existential threat” “Seems like they’d probably just nationalize them then?” “Lol, yeah. You got me, I’m just trying to maximize my AI beta”

We only write 1-2 primers a year because we want them to be durable. It’s simple to publish something market relevant for 6 months, but our goal is a piece you can return to years later and still derive value. Our Robotics primer meets that standard. citriniresearch.com/p/themat…

We're coming out of stealth. We've built our first racks after a successful A0 tapeout, $1B in customer contracts, and $800m raised. Early customer tests show us achieving SOTA throughput, latency, and power efficiency on inference workloads. Our first racks ship this summer.