May 15, 2026 · 57m · sourcery
Chips, Gigawatts, AI, & Agents: Coatue's Public Markets Update | CIO Jaimin Rangwalla · Sourcery with Molly O'Shea
0:00 / 0:00
▶ Watch on YouTube →
gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions
In this episode of Sourcery, host Molly O'Shea interviews Coatue CIO Jaimin Rangwalla to examine the hyper-exponential scaling of artificial intelligence, critical supply chain bottlenecks from GPUs to gigawatts, and Coatue's public market investment strategies.
How this conversation actually went
Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. Molly holds 19.4% of the talking time here. How this is scored →
Molly as informed peer 4.4 Guest teaching 5.2 Guest disagreement 0.2 Molly pushing back 0.2
speaking balance: gold is Molly, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 41:30 Reframing single bottleneck premise
Jaimin contrarily reframes Molly's inquiry about the single biggest bottleneck by asserting that the real bottleneck is the simultaneous existence of multiple interconnected bottlenecks across power, memory, optical, and labor.
Hardest push from Molly ▶ 8:51 Direct challenge on public investing simplicityMolly directly pushes back with 'Is that true?' when Jaimin claims that early public market investors only need to find a couple of good ideas a year to be wildly successful.
Biggest teaching moment ▶ 37:42 Sellers vs. buyers of shortage mechanicsJaimin delivers a thorough conceptual masterclass explaining why fixed capacity and pricing power give semiconductor providers massive operating leverage while compressing hyperscaler multiples.
Molly holds their own ▶ 24:07 Nat Friedman OpenClaw real-world case studyMolly demonstrates first-hand knowledge of the developer ecosystem by introducing Nat Friedman's Stripe Sessions demo where OpenClaw autonomously monitored hydration and rerouted vehicle navigation.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | Molly as informed peer | Guest teaching | Guest disagreement | Molly pushing back | Why |
|---|---|---|---|---|---|---|
| Welcome to Sourcery and Coatue's New Office Layout | 4 | 3 | 1 | 1 | Molly opens with casual banter about Coatue's new office layout and insider details before pivoting to the rapid pace of the AI tech cycle. Jaimin enthusiastically describes the open bullpen culture and the competitive office soccer league. | |
| Mapping the AI Supply Chain: From GPUs to Gigawatts | 4 | 6 | 1 | 0 | Molly asks how Coatue maps out AI categories and what has surprised Jaimin most. Jaimin details their conceptual shift from following the GPU to following gigawatts and notes unprecedented multi-year supply lockups. | |
| Jaimin Rangwalla's Career Background and Investment Philosophy | 4 | 4 | 1 | 2 | Molly explores Jaimin's background and briefly challenges his premise about public investing only requiring a couple of big ideas a year with a sharp 'Is that true?'. Jaimin explains how Philippe Lafont hired him right before the 2008 crash. | |
| Massive Scale: Comparing Private AI Leaders to Historical IPOs | 5 | 5 | 0 | 0 | Molly sets up the comparison between historic Mag Seven IPOs and today's private AI giants. Jaimin contextualizes how SpaceX, OpenAI, and Anthropic have reached top-25 global valuation scales while remaining private. | |
| Exponential Revenue Scaling: OpenAI and Anthropic | 5 | 6 | 0 | 0 | Molly cites the slide showing OpenAI and Anthropic reaching $25B ARR at unprecedented speeds and asks for a breakdown of the revenue sources. Jaimin explains the economics of developer productivity gains and IT budget reallocation. | |
| Defining the Token Economy and Exponential Growth | 4 | 6 | 0 | 0 | Molly prompts Jaimin to define the token economy for listeners unfamiliar with the concept. Jaimin defines a token as the atomic unit of artificial intelligence thought and explains how quantifying intelligence changes business modeling. | |
| Mid-Episode Sponsor Interlude: Turing AI Infrastructure | 3 | 6 | 0 | 0 | Following the sponsor read, Molly asks about autonomous agents as the major technological unlock. Jaimin explains the fundamental shift from human-in-the-loop chat interfaces to recursive agents spawning sub-agents. | |
| OpenClaw Harness and Mobile Remote-Control Workflows | 7 | 4 | 0 | 0 | Molly demonstrates strong domain currency by bringing up Nat Friedman's live OpenClaw demonstration at Stripe Sessions. Jaimin unpacks the harness architecture and the emergence of phone-as-remote-control workflows. | |
| Expanding Individual Semiconductor and Memory Footprint | 4 | 6 | 0 | 0 | Molly asks about the exploding memory TAM per person. Jaimin explains how individual users will scale from having three physical devices to running thousands of virtual agent devices concurrently. | |
| AI Architecture Shift: Persistent Memory and CPU Demand | 5 | 6 | 0 | 0 | Molly asks about behavioral shifts driving architectural changes in AI hardware. Jaimin uses a '50 First Dates' amnesia analogy to explain persistent memory and the rising necessity of CPUs for serial task orchestration. | |
| Key CPU Market Players and the Intel Revival Thesis | 5 | 5 | 0 | 0 | Molly asks about CPU market players and offers a colorful kitchen analogy about semiconductor stocks being like 'the nerdy kid who became cool.' Jaimin reflects on spending the first decade of his career shorting cyclicals before semis became cash cows. | |
| Data Center Buildouts, Power Bottlenecks, and Labor Creation | 6 | 4 | 0 | 0 | Molly introduces her direct observations from visiting clean energy data center startup Exowatt in Miami. Jaimin agrees and frames the broader macroeconomic dynamic of skilled labor bottlenecks and new business creation. | |
| Sellers of Shortage vs. Buyers of Shortage Framework | 4 | 7 | 0 | 0 | Molly prompts Jaimin to explain Coatue's framework distinguishing between 'sellers of shortage' and 'buyers of shortage.' Jaimin gives a masterclass on pricing power, fixed cost operating leverage, and hyperscaler multiple compression. | |
| Unexpected Breakout Sectors and Interconnected Bottlenecks | 4 | 6 | 1 | 0 | Molly asks about surprise breakouts and primary bottlenecks. Jaimin highlights optical infrastructure and reframes the bottleneck question by stating that the true bottleneck is having multiple compounding bottlenecks across power, memory, and labor. | |
| Mid-Episode Sponsor Interlude: VCX by Fundrise | 5 | 6 | 0 | 0 | Following mid-roll advertisements, Molly questions the divergence between negative online headlines and resilient public market performance. Jaimin explains Coatue's historical data showing zero correlation between sentiment and market direction. | |
| Managing Market Volatility and Staying Focused on Fundamentals | 4 | 5 | 0 | 0 | Molly asks how Coatue manages rapid flash volatility and narrative shifts. Jaimin details how grounding in fundamental earnings growth and deck preparation helps filter daily price noise. | |
| Identifying Market Leaders and Technological Transitions | 4 | 5 | 0 | 0 | Molly asks how Coatue identifies emerging category leaders and evaluates impending trillion-dollar private IPOs. Jaimin highlights technological inflections like copper-to-optical transition and historical tech leadership turnover. | |
| Key Risks and Market Cycles in Tech Investing | 4 | 6 | 0 | 0 | Molly asks about the biggest risks facing public AI portfolios. Jaimin identifies algorithmic efficiency breakthroughs (referencing DeepSeek) triggering Jevons paradox, alongside broader 10-year macro correction cycles. | |
| Outlook for the Year and Leadership Mindset at Coatue | 3 | 3 | 0 | 0 | Molly wraps the interview asking about Coatue's culture and founder Philippe Lafont's longevity. Jaimin attributes success to an unrelenting mindset of tracking change and playing padel tennis. |
Statements from this episode (43)
Rangwalla: OpenAI Reached Almost 1B Users Fastest in History
“How OpenAI reached, you know, almost a billion users in the fastest time in history.”
Rangwalla: Gigawatts are the atomic unit of AI growth and a major bottleneck
“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.”
Rangwalla: Memory supply agreements feature guaranteed commitments through 2029–2030
“And now, you know, in the case of memory, for example, you're now hearing Supply agreements, like, guaranteed commitments through 29, twenty-thirty.”
Rangwalla: Big tech AI capex is not a cyclical peak-and-trough bet
“The buyers of everything are all trillion-dollar-plus companies, right, for the majority of it. They are not stupid. They are not investing hundreds and hundreds of billions of dollars growing at these high rates into something that they think is just going to…”
Rangwalla: Meta was the largest Mag Seven IPO at $100B
“So the biggest of the Mag-Seven was Meta in 2012, IPO'd around a hundred billion dollars. All of the predecessors had IPO'd in the nineties, you know, in the 10 to fifty billion dollar range, which for that time was very large.”
Rangwalla: OpenAI's latest funding round valued it over $800B
“OpenAI's most recent round was 800 plus billion dollars.”
Rangwalla: SpaceX reached a $1.25T valuation in xAI transaction
“SpaceX is most recent, you know, when they did the transaction with XAI was 1.25 trillion dollars.”
Rangwalla: Anthropic's last round was in the high $300B range
“Anthropix last round was, you know, high 300 of billions.”
Rangwalla: Private AI leaders are breaking into the global top 25
“Private companies that are breaking into pretty much the top 25 in the world before they even go public. It's an unprecedented time, but I think that's what makes AI so exciting, is that you have these private companies that are capturing so much value and so …”
Coatue: Most Mag Seven Stocks Underperformed Nasdaq, Three Down YTD
“Outside of Alphabet, so many of the Mac sevens have underperformed the Nasdaq by a lot. And in fact, three of them are down year to date.”
Coatue: The Magnificent Seven's Pecking Order Is Rapidly Reordering
“To me, it's a further indication that even within the Mac seven, the reordering of the pecking order is happening quickly.”
Rangwalla: Anthropic Grew ARR From $9B in December to $30B
“It was nine billion dollars at the end of December, and that number was thirty billion, right? So they added twenty billion, they're adding, you know, ten billion plus or minus a month, almost two and a half billion dollars a week.”
Rangwalla: Combined OpenAI and Anthropic ARR Exceeds Salesforce and ServiceNow
“And then when you think about both of them combined, let's say they're close to sixty billion dollars. I mean, that is bigger than ServiceNow, that is bigger than Salesforce, that is bigger than any big name SaaS platform that you would think about. And these …”
Rangwalla: AI Spending Does Not Require 1-to-1 Labor Reductions
“I don't think it's a straight one for one, like, oh, if you spend a hundred dollars on Anthropic, I have to reduce my labor by a hundred dollars because I think that there's productivity gains that are achieved.”
Rangwalla: AI Gives Top Developers 2x to 3x Productivity Gains
“Even if you listen to the top coders and the top developers, they say increase in productivity is like two X, three X, and so most companies, at least in the technology world, you're not you're not limited by the you're not limited in terms of what you're able…”
Rangwalla: Token Economy Quantifies AI Decisions Unlike Human Intelligence
“And so the token economy, when people talk about it is really the proliferation of agents and decisions made by artificial intelligence. And those decisions can actually be quantified in a number of tokens. And so it's like very different than human intelligen…”
Rangwalla: AI agents launching other agents is an underappreciated breakthrough
“Agents, 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.”
Rangwalla: ByteDance, Tencent, and Alibaba Have Built Agent Harnesses
“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.”
Rangwalla: Agent Harnesses Are Transitory Before Mobile AI Super Apps
“To me, it's just an intermittent state. I don't think that you're going to have it where You got to go home. You got to hook up all these connections to your claw because today you still have to give it access to Gmail and this and that. There's going to be a …”
Rangwalla: Individuals will run hundreds to thousands of AI agents within two years
“If you or me in, in a year or two, we're gonna have a claw that's running hundreds of agents at all time. I'm gonna have a similar impact at work. I'm gonna have hundreds, maybe thousands of agents running at all time.”
Rangwalla: Individual semiconductor footprints will expand 10x due to AI agents
“Your footprint, you're like a semiconductor or technology footprint as an individual is about to expand by like 10 X because you used to have a couple of devices. Maybe you're someone has an iPad too. So you have X a little more, but like three or four devices…”
Rangwalla: Next-gen AI accelerators will eliminate agent amnesia with persistent memory
“What's changing is now you're getting persistent memory with all of the next evolution of accelerators. And so I think that you're seeing memory content is increasing. The agent is not going to have amnesia anymore. The agent is going to be able to remember.”
Rangwalla: Rise of AI agents is driving massive CPU and memory demand
“And so I think that that is just creating a massive need for more CPUs and more memory. And we're seeing that we're actually just at the beginning of that. I think agents is really what is driving that.”
Rangwalla: AI Hardware Ratios May Flip to 4 CPUs per GPU
“Now, the ratio is actually moving from, to one CPU to four GPUs, and so it's improved kind of by two x already, and we think it actually has a chance to flip the opposite direction, which is one GPU to four CPUs, and so, and some people, very aggressive, say t…”
Rangwalla: Intel Has Significant Catch-Up Room vs Other Chipmakers
“If you think about every semiconductor company over the last four years in this heyday of AI, they're all up five, seven, 10 X. Like, it's still a laggard over any period beyond 12 months, and so, to me, there's still a lot of catch up for that company.”
Rangwalla: Samsung and SK Hynix Generate More Cash Flow Than Hyperscalers
“Samsung and Hynix together generate more cash flow than every hyperscaler.”
Rangwalla: US new business formation is near record highs
“Even today, people don't talk about it, but we're actually at one of the highest new business formations that we've ever been at.”
Rangwalla: AI buildout will put skilled trade jobs in pole position
“There's a lot of companies that we look at that are massive shortages the amount of labor that they have, and it's not one type of labor, it's a lot of different types of skilled labor, but which usually means that the salaries for those jobs are gonna go up t…”
Rangwalla: Microsoft, Amazon, and Meta multiples compressed as capex grew
“And so if you look at the multiples of Microsoft, of Amazon, of Meta, they've actually all compressed in the last couple of years because their capex is going up.”
Coatue: TSMC is one of our largest holdings
“Even TSMC, which is one of our largest holdings they're taking price too, because they just, like, we have a finite amount of capacity, and so we have to try to figure out how are we gonna make sure that we're generating enough profits, and knowing that if we'…”
Rangwalla: Coatue favors Amazon and Google as hybrid shortage players
“I actually think Amazon and Google are an unusual camp, and those are the two companies that we really like because they are also a bit of a hybrid, in that Google has TPUs that they sell, and Amazon has talked about selling Tranium, or at least openly discuss…”
Coatue's Rangwalla: Optical networking is becoming critical future infrastructure
“You know, optical for years was viewed as just an area that had constant oversupply, undersupply, oversupply, undersupply. And now I think it's becoming more of a critical part of the infrastructure of the future. And I think that that is the one that probably…”
Rangwalla: AI buildout is hindered by unprecedented simultaneous bottlenecks across sectors
“I mean, it's, I've never seen so many bottlenecks at the same time, and I think that is what makes it so hard, because if it's just one vertical slice that's the bottleneck, you can say, okay, well, if they put enough dollars in, they'll be able to solve the b…”
Coatue: Markets reward sellers of AI shortages, not buyers
“You can see that the market is really rewarding the sellers of the shortage, but it is not rewarding the buyers of the shortage because the profits are so high that the market is really focused on grabbing the short term profits.”
Rangwalla: Market Sentiment Has Zero Correlation With Stock Performance
“We did a 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 like correlation between negative sentiment means the market should be down or it's contrary”
Rangwalla: S&P 500 Earnings Growth Accelerates to 18% Through 2026
“The S&P, 500 earnings growth through 26 is like 15% accelerating to 18%.”
Rangwalla: Stock Multiples Remain Consistent Because Earnings Growth Outpaces Prices
“The multiples for the market actually haven't gone up that much. The multiples for the market are actually staying kind of consistent because the growth is so high. And so the market is up, you know, whatever, seven, eight percent this year, but earnings growt…”
Rangwalla: AI is accelerating at a faster pace today than 1-2 years ago
“Ultimately what we've concluded at the moment is that the fundamentals are really strong, the market is really resilient, and AI is accelerating its pace. Even though it's hard for me to say that, and maybe someone says you're crazy, like, AI is all everyone's…”
Rangwalla: AI Accelerators Will Adopt More Optical Networking in 2–3 Years
“There's a scenario two, three years from now where there's going to be less copper and more optical, or at least like maybe not less copper, but more optical, right? Because it can transmit data faster, but there's limitations. It needs to be longer distance a…”
Rangwalla: Tech leadership turns over 25% every five years
“You know, we've done the studies on it, like, almost plus or minus every five years, a quarter of the top 25 in tech usually come out, and a new four, five, six come into the top 25.”
Rangwalla: Trillion-dollar AI IPOs will not trigger Mag Seven selloff
“I don't think that there's this concern in my mind of, oh, that means you got to sell mag seven or this and that, because, you know, the S and P is 60 or 70 trillion dollars of market cap. We're talking about one trillion dollar, three one trillion dollar comp…”
Rangwalla: AI Efficiency Breakthroughs Will Drive Adoption via Jevons Paradox
“If let's say that the deep seek moment that happened you know, last year, another moment like that happens where someone figures out a model that can do all the calculations with less power, less Semiconductors, less memory, all of that. If that happens, it is…”