Exclusive Interview: Coatue CIO on AI's

Participants: Jaimin Rangwalla (CIO, Coatue Public), Molly O'Shea (Host)

1. Key Themes

The "Follow the Gigawatt" Framework Replaces "Follow the GPU"

Coatue has evolved its analytical lens from tracking GPU demand to tracking gigawatts as the true atomic unit of AI growth. This reframing has massive implications for which companies and infrastructure players are worth watching.

"We're thinking about follow the gigawatts. And really, the gigawatt is almost the atomic unit of where the growth in AI is coming from. And it's one of the biggest shortages as well that's out there." — Jaimin Rangwalla 00:05:12

AI Revenue Growth Is Unprecedented and Likely Underappreciated

The speed at which AI-native companies are scaling revenue dwarfs anything seen in prior tech waves. The biggest SaaS companies in the world took 15-20 years to get where Anthropic and OpenAI already are.

"They're adding $10 billion plus or minus a month, almost $2.5 billion a week. Most of the companies in the SaaS universe don't even have $2.5 billion of ARR annually. Like they're adding that in a week." — Jaimin Rangwalla 00:00:36

"Both of them combined, let's say they're close to $60 billion. I mean, that is bigger than ServiceNow, that is bigger than Salesforce... and these guys have done it in a matter of like a handful of years, right? Versus the 15, 20 years that it took all these other companies." — Jaimin Rangwalla 00:14:29

Sellers of Shortage Are Winning; Buyers Are Being Punished

The market is bifurcating: companies supplying the constrained resources of AI buildout (memory, optical, power, CPUs) are seeing dramatic earnings expansion, while the hyperscalers spending the capex are seeing multiple compression.

"The stocks that are working and the part of the market that everyone is really excited about, optical power infrastructure, even labor stocks... memory, CPUs, GPUs, even TSMC... they're taking price too." — Jaimin Rangwalla 00:39:30

"If memory pricing goes up 100% and Microsoft has to spend two units instead of one, they're not actually getting two units of benefit... you're just seeing like-for-like price increase go up so much." — Jaimin Rangwalla 00:38:37

2. Contrarian Perspectives

Agents Launching Agents Is the Most Underappreciated Unlock in AI

Most people focus on model capability improvements, but the structural shift of agents spawning their own sub-agents — removing humans from the loop entirely — is a qualitatively different kind of breakthrough that has not been properly priced into how people think about AI's economic impact.

"Agents launching agents I think is one of the most underappreciated unlocks that has happened because you can give an agent a project now and you can just go away. And you can come back and it has done, if not all of it, really, really far along." — Jaimin Rangwalla 00:20:17

The CPU-to-GPU Ratio Is About to Flip — A 16x Market Size Opportunity

The received wisdom has been that AI is a GPU story. But the emergence of agentic workloads means the ratio is shifting from 1 CPU : 8 GPUs toward potentially 4-8 CPUs : 1 GPU — a 16x expansion in the CPU market opportunity that almost nobody is talking about.

"Now the ratio is actually moving from one CPU to four GPUs... we think it actually has a chance to flip the opposite direction, which is one GPU to four CPUs. And so some people very aggressive say it could go to one GPU to eight CPUs." — Jaimin Rangwalla 00:28:48

"You have a four to one going to one to four, which is mathematically is a 16x improvement in market size." — Jaimin Rangwalla 00:31:22

Semiconductor Companies Now Generate More Cash Flow Than Hyperscalers

This runs completely counter to the dominant narrative of the past decade where internet platform companies were seen as the most cash-generative businesses in tech.

"I never would have thought I would tell you that Samsung and Hynix together generate more cash flow than every hyperscaler. Like it's crazy." — Jaimin Rangwalla 00:32:16

Every Individual's Semiconductor Footprint Is About to Expand 1,000x

The conventional view is that AI is a data center story. But the agent economy means that each human will effectively be running thousands of virtual devices simultaneously, turning the addressable market for semiconductors and power into something that scales with the global population multiplied by agent count.

"Your footprint, your semiconductor or technology footprint as an individual is about to expand by like 10X... three or four devices is about to go to thousands of virtual devices... and each agent requires semiconductor content." — Jaimin Rangwalla 00:25:55

"If we're all going to be users and we're all going to be using hundreds and thousands of agents at a time... just imagine if we're all going to be having periods where we're running hundreds of agents... it applies to how much power each person consumes — is all going to increase significantly." — Jaimin Rangwalla 00:26:54

Negative Sentiment Has Zero Predictive Value for Market Direction

Against the consensus instinct to treat bad headlines as a market signal, Coatue's internal research found no correlation between sentiment and actual market performance.

"We did an analysis on sentiment and how indicative is sentiment of what's actually happening in the market. And it's basically a coin flip. There is no correlation between negative sentiment means the market should be down." — Jaimin Rangwalla 00:45:28

3. Companies Identified

Anthropic AI frontier lab and one of the two dominant foundation model companies. Mentioned as a core holding and primary illustration of the unprecedented revenue velocity in AI — going from $9B ARR to $30B ARR in roughly one year.

"Anthropic... it was $9 billion at the end of December. And that number was 30 billion, right? So they added 20 billion and they're adding, you know, 10 billion plus or minus a month." — Jaimin Rangwalla 00:14:00

NVIDIA Leading GPU and accelerator manufacturer. Called out as the standard-setter for the accelerator market roadmap and a key early investment for Coatue. GTC conference described as having transformed from a small 100-person event to a Taylor Swift-level cultural moment.

"NVIDIA almost sets the standard for what the accelerator market is going to look like in terms of capabilities two years down the road." — Jaimin Rangwalla 00:49:22

TSMC Global semiconductor foundry and one of Coatue's largest holdings. Identified as a "seller of shortage" benefiting from pricing power and capacity constraints.

"TSMC, which is one of our largest holdings, they're taking price too. Because they just like, we have a finite amount of capacity." — Jaimin Rangwalla 00:39:30

Intel CPU manufacturer. Identified as a contrarian catch-up opportunity given the 16x potential expansion in CPU market size relative to GPUs, a new CEO, and Elon Musk's endorsement accelerating innovation pace.

"Elon Musk giving a stamp of approval to Intel and increasing the pace of innovation... the combination of that for a stock that, yes, it's had a big move this year, but if you think about every semiconductor company over the last four years, they're all up five, seven, 10x. It's still a laggard over any period beyond 12 months." — Jaimin Rangwalla 00:30:53

Google/Alphabet Identified alongside Amazon as a hybrid "buyer and seller of shortage" given their TPU infrastructure, which gives them a differentiated position.

"Google has TPUs that they sell... maybe there's room for them as well because they would be buyers of the shortage, but they could also be sellers of the shortage. And Google has been a great stock and they're vertically integrated, which makes them great." — Jaimin Rangwalla 00:40:24

Amazon/AWS Called out specifically for having one of the most efficient CPUs (via an internal acquisition), giving them an edge as the CPU/GPU ratio shifts.

"Amazon's done a great job. They actually have one of the most efficient CPUs, but they've always kind of kept it internally. They bought a private company that helped them do this and it has helped, I think, drive AWS success, especially as we move to CPUs." — Jaimin Rangwalla 00:30:24

Exowatt Renewable energy company for data centers. Mentioned as an example of new job creation and domestic manufacturing being driven by AI infrastructure buildout.

"They're an energy company, a renewable energy company for data centers... manufacturing all in the United States... there's job accretion and new layers of the economy that are being created." — Molly O'Shea 00:34:51

4. People Identified

Philippe Laffont (Founder/CEO, Coatue) Founder of Coatue. Described as both a hall-of-fame investor and an exceptional CEO who has scaled the firm from $700M / 12 people to $80B+ / 200 people. Still operates with a startup mentality — sits in the middle of the bullpen.

"Philippe always says that when we do these presentations, we should do them more because it helps us ground our thinking again." — Jaimin Rangwalla 00:47:29

"He's an amazing investor. He's in the hall of fame of risk management investing. But what people maybe don't fully appreciate is that he's an amazing CEO too." — Jaimin Rangwalla 00:55:03

Frank (Chief AI Officer / "AI Mad Scientist," Coatue) Recently hired to serve as Coatue's internal AI implementation leader. His role is to pressure-test all new AI tools and figure out how to deploy Coatue's 20 years of proprietary data with AI to generate investment edge.

"Frank... he is trying everything out there that is new, creative on the cutting edge, figuring out how do we implement it? How do we as an organization utilize our 20 years of data that we've been collecting to give us another leg up against the competition." — Jaimin Rangwalla 00:12:09

Boris Cherney (Founder, Claude Code / Former Coatue) Former Coatue employee who went on to found Claude Code, the agentic coding environment. Cited as a live example of the "thousands of agents running overnight" use case becoming reality for sophisticated users today.

"Boris Cherney, who's former Coatue alum, but also the founder of Cloud Code... he runs some number of agents during the day, but when he goes home, he has thousands of agents running to kind of do work all through the night before he comes back in the morning." — Jaimin Rangwalla 00:26:25

Nat Freeman Mentioned as an early OpenClaw power user whose live demo at Stripe Sessions illustrated the agentic personal assistant use case: the agent proactively noticed he was dehydrated, prompted him to drink water, and rerouted his car to Whole Foods for magnesium.

"His OpenClaw saw that he was dehydrated. And so it was like, go grab a glass of water... he was also driving in his car and they were like, you need magnesium. And so the car then rerouted to Whole Foods." — Molly O'Shea 00:24:12

5. Operating Insights

Use Multiple Concurrent Agent Sessions as a Productivity Multiplier

Rather than running AI tasks sequentially, the highest-leverage users are running multiple independent agent threads simultaneously — effectively parallelizing intellectual work in the same way you'd parallelize code.

"When I use Claude, I don't just ask it one prompt or one problem. I go and say, okay, I want this done. And I open a new one. I say, I want this done. And I open a new one. And so they're all working concurrently." — Jaimin Rangwalla 00:21:16

Re-Grounding in Fundamentals During Volatility Is a Repeatable Process

Coatue uses investor update presentations not just to communicate externally, but as a discipline to force internal re-anchoring around the five to seven things that truly matter — preventing drift from noise.

"Philippe always says that when we do these presentations, we should do them more because it helps us ground our thinking again, because we kind of re-come back to, okay, what are the five or seven things that really matter in this moment?" — Jaimin Rangwalla 00:47:29

Structure Physical Space to Force Cross-Functional Collaboration

Coatue deliberately designed their new office as an open bullpen where AI researchers, data scientists, and investment analysts sit side by side — treating proximity as a driver of signal quality.

"We're much more of an open format, bullpen style where everyone is sitting next to each other. We have AI folks, data scientists, analysts, all kind of sitting next to each other. So we're really trying to push the deep insights and collaboration." — Jaimin Rangwalla 00:01:05

6. Overlooked Insights

OpenClaw (Computer Use) Is the Trojan Horse for the AI Super App

This was mentioned almost in passing, but the observation that every major Chinese internet company — ByteDance, Tencent, Alibaba — has already built their own OpenClaw equivalent is a significant competitive signal. It implies that "computer use" harnesses are becoming the new browser or the new mobile OS layer — whoever controls the harness controls the agent economy. Jaimin explicitly flagged that the current fragmented setup (manually connecting Gmail, apps, etc.) is an interim state, and whoever bundles this into a seamless super app owns the consumer relationship for the AI era.

"Even in China, like every one of the big internet companies has their own claw concept. Like ByteDance has one, Tencent has one, Baba has one. It's amazing how quick this has taken off and I don't think people fully appreciate it... There's going to be a big opportunity for someone to just say, oh, here's your phone. Here's the app. This is your super app. This now allows you to connect to everything that you have." — Jaimin Rangwalla 00:23:19

Supply Contracts Extending to 2029-2030 Signal a Structurally Different Cycle

Mentioned only briefly but enormously significant: the fact that memory suppliers are now signing guaranteed supply agreements through 2029 and 2030 is historically unprecedented. This is not a cyclical upturn — it is the largest buyers in the world (trillion-dollar companies) locking in multi-year commitments because they believe the shortage is permanent. This effectively floors demand visibility for the entire semiconductor supply chain in a way that has never existed before.

"In the case of memory, for example, you're now hearing supply agreements, like guaranteed commitments through 29, 2030. I've never seen that level of tightness and that number just keeps extending out." — Jaimin Rangwalla 00:06:08