By playing, you agree to uListen's Terms and Privacy Policy and YouTube's Terms of Service.
Invest Like the BestWhy the Markets Are Pricing AI Wrong | Gavin Baker
Summaries, chapters, transcripts, and FAQs on this page are made by uListen, not YouTube.
At a glance
WHAT IT’S REALLY ABOUT
AI stocks sold off, but Baker sees accelerating compute fundamentals
- Gavin Baker argues that despite a sharp AI-related stock selloff, he has not found meaningful negative quantitative demand signals and instead sees multiple indicators accelerating.
- He says a major market misunderstanding is the large discount between long-term contracted GPU pricing and current spot pricing, which could drive higher monetization as contracts reprice.
- He identifies tightening credit conditions (higher real yields, wider spreads, rising CDS) as the most credible risk, while contending that stronger operating cash flow could reduce reliance on debt financing.
- He claims open source model adoption is being misinterpreted as bearish, because it may reduce frontier-model margins but still requires the same underlying compute and can increase total token demand.
- He calls regulation and poor public messaging about data centers a critical vulnerability, arguing the AI industry needs to communicate impacts on power prices, jobs, and communities more effectively.
IDEAS WORTH REMEMBERING
5 ideasBaker argues the AI selloff is narrative-driven while demand metrics keep improving.
He says he searched for “a negative quantitative metric” in July’s AI drawdown and couldn’t find one, claiming instead that indicators like GPU availability, GPU rental pricing, DRAM spot pricing, token growth, and other demand signals are accelerating.
Contract repricing is central to Baker’s view that ROI and cash flows will surprise to the upside.
He describes a large gap between contracted GPU pricing (from earlier long-term deals) and today’s spot market, and expects that as contracts roll off, repricing could lift hyperscalers’ economics even if spot prices later decline.
Credit conditions are the key macro risk, but Baker believes improving operating cash flow can offset it.
He frames widening spreads, higher real yields, and rising CDS as the “real concern,” because debt-fueled buildouts can unwind quickly if supply/demand shift; however, he argues higher operating cash flow from repricing could reduce the need for credit.
Open source may be bearish for frontier-model margins but bullish for infrastructure compute demand.
He says the market misread open source progress as bearish, arguing “a token is a token” from an infrastructure perspective—open source mainly shifts margin away from frontier model providers and can increase compute demand via elasticity.
Baker views regulation and public perception of data centers as the most serious threat to the thesis.
He suggests regulation is the biggest risk, citing narratives about data centers raising power prices, consuming water, and taking jobs; he argues the industry has done poor PR and needs to proactively explain community benefits.
WORDS WORTH SAVING
5 quotesI would describe, um- ... July as 2022 in a month.
— Gavin Baker
Like, open source is kind of dark matter to the public markets.
— Gavin Baker
Claude is kind of Walter Cronkite for the stock market, and everybody just believes whatever it says.
— Gavin Baker
If you do not speak your own truth, no one else will.
— Gavin Baker
Data centers are, like, are, are in a lot of ways the best thing to happen for blue collar wages in my lifetime.
— Gavin Baker
SPEAKERS
Gavin Baker
guestInvestor and Managing Partner at Atreides Management, focused on technology and public markets.
Patrick O'Shaughnessy
hostHost of Invest Like The Best and founder/CEO of O'Shaughnessy Ventures.
High quality AI-generated summary created from speaker-labeled transcript.
