Invest Like the Best with Patrick O'Shaughnessy
· Patrick O'Shaughnessy
[00:00:00] S0: My guest today is Gavin Baker, the founding partner and CIO of Atreides Management, and this is our sixth conversation. The central theme is watts and wafers, the two physical constraints that in Gavin's view will dictate the next phase of AI. On power, he thinks the near-term shortage starts to ease in 2027 and '28 as new sources of energy come online, and that orbital compute helps solve this problem in the long term. On wafers, he explains what is different this time from the dot-com bubble and why TSMC's capacity decisions may be the single most important variable to watch. We also discuss Elon's Terrafab, the disaggregation of GPUs, the role of new chip companies, and whether the economic value of AI will keep accruing to the frontier models. Enjoy this awesome sixth conversation with Gavin Baker. All right, so this is our sixth time doing this, if you can believe it, which puts you back into first place — or at least tied for first place with Gurley. Back in esteemed territory. Always my favorite conversation about markets and everything going on. Even since the last time we did this, which was so exciting and spectacular — I think we're in an even more interesting time now. Maybe just start by riffing on how it felt for you living through March and April of this year. It felt to me just like a completely unique economic, technology, and market environment — and you're the biggest student of history and of these times. So what did it feel like?
· Gavin Baker, Patrick O'Shaughnessy
[00:00:00] S1: Broadly speaking, there are two kinds of drawdowns. There are drawdowns where you're wrong, the company missed estimates, your hypothesis was invalidated — and you have to take your medicine and crystallize that loss. And then there are drawdowns or periods of underperformance where you're underperforming because of companies you know really, really well, and where you profoundly disagree with the price action and you can lean in. Instead of crystallizing negative performance, you can build pent-up alpha, pent-up future performance. For me, that is what March felt like. The NASDAQ was selling off, and at the same time, what was happening in AI was the most extraordinary moment in the history of capitalism, the history of American business. What I mean by that is Anthropic added $11 billion of ARR. What is astonishing to me about this, is the SaaS and cloud revolution created, we'll call it, between $5 and $10 trillion of value. Arguably the 3 highest-profile SaaS companies in the last 10 to 12 years are Palantir, Snowflake, and Databricks. These 3 companies employ thousands of people, tens of thousands collectively. They've all spent 10 years building their businesses, and Anthropic added their combined businesses in 1 month. Nothing like that has ever happened in the history of capitalism — forget my career, just the flat-out history of capitalism, the history of business. It's wild. And then Krishna comes on this show and shares some stats. 500% NDR . [00:00:00] S0: You do the math on that for 3 years. Insanity.
· Gavin Baker
[00:00:00] S1: So there's just no precedent for this. We tech investors hear a lot of discussions about S-curves and investing in exponentials. I've just never seen an exponential like this. It felt even more extreme than DeepSeek, which was a very similar setup. They happened at about the same time. If we go back to '25, there's a huge sell-off on DeepSeek, which was very strange because the paper gets published 7 days before DeepSeek Monday. It got published, I believe, on a Monday that was a holiday in America. I read it and thought, "Hmm, this feels like it might not read that positively for the AI trade." I took action, and then we had DeepSeek Monday, where AI really imploded a week later. That was really strange because by DeepSeek Monday, it was super clear that this was going to be the most positive thing that had ever happened to compute demand. Prices in the AWS availability zones in Asia had already doubled. You were seeing GPU availability go down. This was the first time we saw how much more compute-hungry reasoning models are during inference than non-reasoning models. So that was a similar setup. You had to do some work to see that. It's not that hard to say, "Oh wow, stocks are selling off, the price of DRAM is going vertical, the price of GPUs in Asia is going vertical, GPU availability is going down." And then 2 or 3 days later, GPU rental prices in America started going up. All you had to do in March was simply observe what was happening to Anthropic. There are all these people who seem to regret not buying during '22, not buying during COVID, not buying during DeepSeek. You had the same valuation setup at the beginning of April, and an even clearer AI inflection. So there have been all these chances to buy into AI. And then of course, what complicated it was the Strait of Hormuz. I became a believer, and am a believer, that maybe one thing the market was mispricing — I'm no macro expert. I do a lot of pro-national security investing, so I have access to people who are experts and are excited to share their thoughts and opinions with me. The Strait of Hormuz being closed is actually relatively awesome for America.
· Patrick O'Shaughnessy, Gavin Baker
[00:00:00] S0: Why? [00:00:00] S1: Particularly for the goals of the current administration. Electricity is a very important industrial or manufacturing input. The key input into American electricity prices, which feeds into AI, is NG1 — Natural Gas 1 on Bloomberg — and that was down 20%. While natural gas in Asia, Europe, and everywhere else doubled or tripled. Our relative manufacturing competitiveness improved overnight. And for better or worse, that is what the Trump administration seems to care about. They are very focused on America's relative position. A lot of people had memories of the 1970s. What made the '70s so traumatic was that it wasn't just that prices went up — it's that there were actual gas shortages. But the US economy is dramatically less energy intensive than it was. The United States is now the world's largest producer of oil and gas, and we've now become the world's largest exporter of oil and gas. On top of that, there's this relative manufacturing advantage. That made it easier to stay focused on AI fundamentals, stay focused on what were historically attractive valuations. On a relative basis, tech essentially got as cheap as it's been versus the rest of the market as at any point over the last 10 years. Just think about that in the context of market efficiency. We have the most extraordinary moment in the history of capitalism, that's wildly bullish for AI, and you get a chance to buy AI at really attractive valuations.
· Patrick O'Shaughnessy, Gavin Baker
[00:00:00] S0: What do you make of the multiples that specifically Anthropic and OpenAI — which in my mind are the reference assets, the most pure-play takes on this trend — really being not that crazy? If you just look at the sales multiple and compare it to maybe where Databricks and Snowflake and these companies traded at their peak, how do you process it? How do you make sense of it? [00:00:00] S1: OpenAI and Anthropic are pretty different animals from a capital efficiency perspective. Anthropic clearly has a dramatically lower cost per token than OpenAI. They just do. You can see that in the amount of money they have burned to get to a roughly similar revenue scale. I think they've burned maybe 80% less than OpenAI. So as businesses, they clearly have very different structural ROICs. I think Sarah Friar is one of the most exceptional CFOs. They're doing a lot of things to try to improve this. [00:00:00] S0: And they've secured a lot of compute.
· Patrick O'Shaughnessy, Gavin Baker
[00:00:00] S0: And they've secured a lot of compute. [00:00:00] S1: They've secured a lot of compute. That's another big difference. It turns out being aggressive really paid off. Anthropic at $900 billion for $50 billion in ARR, growing at ridiculous rates — a true statement is that if Anthropic could just wave a magic wand and get all the compute they wanted, they'd probably be doing well north of $100 billion today. Maybe $150 billion. They have clearly deprecated the intelligence of Claude. Claude is generating, even on Opus, 70% less tokens for the exact same question. As we talked about last time, token quantity equals quality of answer and quality of thinking at some level. And there is an intelligence density per token that also matters. I've felt that as a user. So they would be doing materially more — $100, $150, maybe $200 billion. So you might be buying it at more like 5 times unconstrained — I'm going to make up a new number — URR. Unconstrained run-rate revenue. [00:00:00] S0: Why do you think they don't raise $100 billion at a $3 trillion valuation or something like this? If you were the Anthropic CFO — Krishna's awesome, we just had him on — or if you're Sarah, it seems to me, if the inbound I received following the Krishna episode is any indication, everyone I've ever met is trying to invest in both these companies.