Episode summary
The Reflexive Risk in Financing AI Buildout Is Real, Says Investor Who Nailed AI Hardware Trade
Summary
The guest argues that the AI trade has shifted from owning semiconductors and other hardware to owning the businesses that enable and monetize actual usage. He says the easy thesis was that hyperscalers would spend cash flow on compute, but now more of the buildout depends on equity and debt financing, which makes the trade more reflexive and more vulnerable to a sentiment break. He remains constructive on selected AI infrastructure, but says the more asymmetric opportunities are in “agentic utilities” and in businesses that help convert model capability into revenue.
He also says the bottleneck in biology is no longer idea generation or dry-lab design. Using Claude to design a scorpion antivenom, he says AI produced hundreds of candidates in about 64 hours and 17 bound in the lab, but the slow part was still synthesis, testing, and waiting on biological processes that cannot be compressed. His conclusion is that AI is commoditizing drug design, while the durable trade is in experiments, lab automation, CROs, and other preclinical infrastructure.
On that basis, he is skeptical of broad biotech hype and thinks the value of early drug assets may fall as candidate generation gets easier. He is more constructive on companies with proprietary data, strong execution, or direct exposure to the preclinical bottleneck, and he also sounds more positive on homebuilders if rates stop rising aggressively.
Key takeaways
- The guest says the AI trade has rotated away from semis and toward businesses that capture real AI usage and downstream revenue.
- He argues that financing the AI buildout is becoming more reflexive because incremental spend is increasingly funded by equity and debt rather than operating cash flow.
- He says Claude helped generate 211 scorpion-antivenom designs, 17 of which bound in lab tests, but the slow bottleneck remained synthesis and biological testing.
- He is bullish on experiments, lab automation, and CROs such as Charles River because AI increases candidate flow and preclinical demand.
- He thinks broad biotech is less attractive because easier drug design may commoditize early-stage assets and reduce the value of simple “one-asset” stories.
- He sounds relatively constructive on homebuilders if rates stabilize and the market has already priced in a lot of hikes.
Notable quotes
“the AI produced these molecules and and 17 of them bound in a lab and it's pretty mind-blowing”
“If you're concluding that the demand from AI over the next year or two is going to come from how much compute it's going to take to cure cancer, you're probably going to be disappointed”
“what we're seeing now really strongly which is the building aspect of it the it's it's I don't want to say priced in, but everyone is aware of it”
“the biggest fear is zoomed out at least what we've seen is model capability improves and then lab ARR goes up”
“we really want to be long experiments right and and and anything that can make experiments quicker”
How the episode opens
The AI produced these molecules and and 17 of them bound in a lab and it's pretty mind-blowing. But at the same time, there is an overwhelming amount of hype here. If you're concluding that the demand from AI over the next year or two is going to come from how much compute it's going to take to cure cancer, you're probably going to be disappointed. You know
Why? Back in 2023, Catrini was one of the earliest voices telling folks to own the AI hardware trade. He was talking about SKH on my podcast before most people had heard of it, including me.
But today he's light on AI hardware. Why did he cut it and what would bring him back? Since then, it's just an objective fact that Saturn has become undoubtedly one of the hottest investment minds on Earth.
It's just a fact. I was just watching an interview with Blackstone Steven Schwarzman and Steve unprompted brought Saturni up because of his 2028 global intelligence crisis piece. Now he's gone further.
He had Claw design a drug, a scorpion antivenenom actually, and it worked better than anyone expected. Which raises the real question. If AI is commoditizing drug design, what is the new bottleneck?
His answer will surprise you as it surprised me and so will his trade. This episode is brought to you by Seraya Partners who manages the return to tangibles ETF lens ticker less the Seraya thematics ETF built for a return to tangibles. Learn more at surmaya ETF.com and see the show notes for disclosures.
Let's get into it.
So we're we're going to talk about what the AI trade and then the scorpion stuff. Yeah, because you know I'm just looking through the trendex and kind of a light bag when you're looking at uh the the AI hardware almost nothing.
From the episode’s captions: roughly the first two minutes, lightly cleaned. Keep watching on YouTube →
Topics
- ai hardware capex
- drug discovery automation
- preclinical bottlenecks
- ai margins and profitability
- semi investment thesis
Companies and tickers mentioned
- NVDA
- META
- GOOGL
- GOOG
- MU
- WDC
- Sarmaya Partners
- Anthropic
- CoreWeave
- DeepSeek
- Nvidia
- Meta
- Alphabet
- Micron
- Sandisk
Watch the episode
Summary, takeaways and quote selection by Octagon. Quotes are short excerpts; the full conversation is in the episode.
More from Monetary Matters
-
AI is Heading for a Massive Oversupply: Top Hedge Fund Manager’s Prediction | David Orr -
Why Market Fragility is Building into a Crisis in 2027 | Cem Karsan & Noel Smith -
Inside a Short Seller’s Private Credit Investigation | Laks Ganipathi -
The End of Cheap Capital & What it Means for Markets | Russell Clark -
Breaking Down the Markets Massive Post-Midterm Moves with Andy Constan -
Michael Howell: Pain Now, Money Printing Later
Monetary Matters, this episode and its artwork belong to Jack Farley. Octagon is not affiliated with the show; this page is an independent summary.
