Episode summary
The AI Selloff Doesn't Match the Data | Top AI Investor Explains
Invest Like the Best
· · Hosts: Patrick O'Shaughnessy · Guests: Gavin Baker
Summary
Gavin Baker argues that the July 2026 AI selloff was a disconnect between public-market fear and still-accelerating fundamentals. He says he spent the month trying to find a negative quantitative signal and could not: GPU availability, rental pricing, DRAM spot prices, token growth, and hyperscaler operating cash flow all looked stronger, not weaker. In his view, the market is still underestimating how much demand is coming from OpenAI, Anthropic, open-source models, and the inference clouds that sit between them and enterprise users.
A major part of the thesis is that AI compute is being repriced higher as old contracts roll off, while many buyers are still sitting on long-term agreements signed at lower rates. Baker says that dynamic should lift operating cash flow at Microsoft, Meta, and Amazon enough that most of the buildout can eventually be funded internally rather than with debt. He also argues that the memory market is shifting toward long-term supply agreements and that breaking those agreements would be strategically dangerous for buyers.
Baker is also constructive on NVIDIA’s evolving model, including equity stakes and a credit-wrapper style structure that lets it participate in the economics around its chips. The biggest risks he flags are regulatory intervention, a real slowdown in demand, or sustained declines in GPU prices. But his bottom line is that the fundamentals still look stronger than the stock action, and the market may be too pessimistic about AI infrastructure demand.
Key takeaways
- Baker says July 2026 felt like "2022 in a month," but the underlying AI demand metrics were accelerating rather than decelerating.
- He argues hyperscaler operating cash flow from Microsoft, Meta, and Amazon accelerated, supporting a future AI capex wave without needing heavy debt financing.
- Baker says open-source models and inference clouds are not bearish for compute demand because cheaper tokens can expand total token usage and shift margin dollars into infrastructure.
- He sees long-term agreements in memory as strategically binding; breaking them could cost buyers future allocation when supply tightens again.
- Baker says NVIDIA is unusually well positioned because its chips are the most financeable, and its new equity-plus-revenue-share approach could deepen its moat.
- He identifies regulation, especially data-center restrictions, as the biggest downside risk to the AI buildout.
Notable quotes
“I would describe um July as 2022 in a month”
“The underlying fundamentals are improving and stocks”
“every metric is accelerating”
“a token is a token, and you need the exact same amount of compute to make a token”
“I haven't been able to find one that is like a quantitative metric”
How the episode opens
I want to be scared, you know. I don't want to feel like a lunatic watching these stocks get more cheaper thinking the expected forward returns are going up. My main kind of mission out here this week — Is like pressure test.
Yeah. — Yeah. — Find tell me something negative. — But I haven't been able to find one that is like a quantitative metric. The underlying fundamentals are improving and stocks. Nvidia is actually, as we record this, at its lowest forward PE of the last 10 years. The market 100% thinks they are significantly overrated. Gavin, it's only been two months. Like the model release cycles, the gap between our podcast episodes are shortening.
We're basically uh you and I are basically on a model release cadence at this point. — Well, I was I was sensitive to criticism that um that I think somebody pointed out that um our podcasts were coincident with like local uh market peaks and nobody can say that after this.
What's on your mind? It's been a crazy crazy month. — Yeah. I would describe um July as 2022 in a month. — Yeah. There are some fundamental negatives which which we should talk but like on on the whole the balance of fundamentals I think is improving significantly. Loads of AI names are down 50 60% from their highs. We'll call it 40 to 60% in a month in a straight line. And I asked you before we started you've you've been out here for the summer. Have you heard a single negative quantitative metric about AI?
Yeah. — A single instance of deceleration? — Nothing. — Nothing. — In fact, every metric is accelerating. — And to your point, not just blind optimism from people excited about AI.
From the episode’s captions: roughly the first two minutes, lightly cleaned. Keep watching on YouTube →
Topics
- ai selloff vs fundamentals
- gpu pricing
- hyperscaler operating cash flows
- open source ai infrastructure demand
- memory market long-term agreements
Companies and tickers mentioned
- NVDA
- Nvidia
- SpaceX
- Anthropic
Watch the episode
Watch “The AI Selloff Doesn't Match the Data | Top AI Investor Explains” on YouTube
Summary, takeaways and quote selection by Octagon. Quotes are short excerpts; the full conversation is in the episode.
More from Invest Like the Best
Invest Like the Best, this episode and its artwork belong to Colossus | Investing & Business Podcasts. Octagon is not affiliated with the show; this page is an independent summary.
