Gavin Baker - Watts and Wafers - [Invest Like the Best, EP.473]

0:02 Hello and welcome everyone. I'm Patrick O'Shaughnessy and this is Invest Like the Best. This show is an open ended exploration of markets, ideas, stories, and strategies that will help you better invest both your time and your money. If you enjoy these conversations and want to go deeper, check out Colossus, our quarterly publication with in-depth profiles of the people shaping business and investing. You can find Colossus along with all of our podcasts at Colossus.com. Patrick O'Shaughnessy is the CEO of Positive Sum. All opinions expressed by Patrick and podcast guests are solely their own opinions and do not reflect the opinion of Positive Sum. This podcast is for informational purposes only and should not be relied upon as a basis for investment decisions. Clients of Positive Sum may maintain positions in the securities discussed in this podcast. To learn more,

0:48 Visit PSUM dot VC. Um My guest today is Gavin Baker, the founding partner and CIO of A Trade's 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 two thousand twenty seven and twenty eight.

1:09 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 terapab, the disaggregation of GPUs, the role of new chip companies, and whether economic value of AI will keep accruing to the frontier models. Please enjoy this awesome sixth conversation with Gavin Baker. Alright, so this is our sixth time doing this, if you can believe it. Which puts you back into first place, or at least tied first place with Girly.

1:43 Back into Steam territory. Always my favorite conversation about markets and everything going on. Even since last time when we did this, which was So exciting and spectacular. I think we're in an even more interesting time now.

1:56 Maybe just start by riffing on how it felt for you living through March and April of this year, which 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 does it feel like? I would say broadly speaking, there are two kinds of drawdowns. There are drawdowns where

2:16 You're um Company missed estimates. Your hypothesis was invalidated. And you have to take your medicine and you crystallize that loss. And then their drawdowns are periods of underperformance.

2:29 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. And instead of crystallizing

2:41 negative performance You could kind of build pint up alpha, pint up future performance. And for me that is what March felt like. The Nasdaq was selling off. At the same time what was happening in AI

2:52 Was I think the most extraordinary moment in the history of capitalism. The history of American business. What I just mean by that is that Anthropic they added eleven billion dollars of AR. And what is astonishing to me about This

3:06 Is that This asked the cloud revolution, it created we'll call it between five and ten trillion dollars of value. I would say arguably the three highest profile. SaaS companies in the last ten Twelve years.

3:19 Our Palantir? Snowflake. And data breaks. And these three companies Employ thousands of people.

3:27 Tens of thousands collectively. They've all spent ten years building their businesses. And Anthropic added their combined businesses in one Month nothing like that has ever happened in the history of capitalism. Forget my career.

3:43 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. Five hundred percent in DR. Yeah, you do the math on that for three years. So there's just No precedent for this, and we

4:00 Tech investors, we 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 deep seek. Which was a very similar setup. They happened at about the same time. If we go back to twenty five.

4:17 There's a huge sell off at Deep Seek. Which was very strange because the paper gets published. Seven days. before Deep Seek Bud Day got published. I believe on a Monday that was a holiday.

4:30 In America. I read it. I thought hmm. This feels like it might not read that positively for the AI trade. I took action. And then

4:41 We had Deep Seat Mud Day. Well AI really imploded. A week later. That was really strange because by deep seek mud day it was super clear.

4:52 that this was going to be the most positive thing that ever happened to compute demand. Prices of the AWS availability zones in Asia. had already Doubled. You're seeing GPU availability go down.

5:05 And this was just the first time we saw. How much more compute hungry reasoning models are. During inference than non reason models. And so that was a similar setup. You had to do some work to see that. I mean, it's not that hard.

5:21 To say, oh wow, stocks are selling off, the price of DRAM's going vertical, the price of GPUs in Asia are going vertical. GPU availability's going down. And then like two or three days later GPU prices in America started going up, GPU rental prices. All you had to do in Bart.

5:36 What is Simply observe what was happening to anthropic. And there's all these people who seem to regret. Not by during twenty two, not by during covet. Not by your deep seek.

5:49 You had the same valuation set up. At the beginning of April. An even clearer AI inflection. То да бе алліз часи. To buy into AI, and of course, what complicated it was the straight of four boss.

6:05 I became a believer and am a believer. that I think maybe one thing that the market was mispricing. I'm no macro expert. I do do a lot of pro National security investing, so I do have access.

6:19 To people who are experts that are Excited to share their thoughts and opinions with me. That the Strait of Horb was being closed is actually relatively Awesome. For America.

6:29 Why? particularly for the goals of the current administration. So electricity is a very important industrial or manufacturing input. The key input into American electricity prices, which feeds into AI.

6:43 is G one. Natural gas water, blue That was down twenty percent. A natural gas in Asia. Europe.

6:51 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.

7:09 And I think a lot of people have memories of the nineteen seventies. What made the seventies so traumatic was it wasn't just that prices would drop. It's that there were actual gas shortages. Then you go through, okay, well The US economy is dramatically less energy intensive than it was.

7:26 The United States is now the world's largest producer of oil and gas. And we've become now the world's largest exporter. Oh. Oil gas. And on top of that, there's this relative manufacturing advantage.

7:40 That made it Easier to stay Focus on AI fundamentals. Stay focused on. What were historically attractive valuations, I think on a relative basis. Tech essentially got as cheap as it's been versus the rest of the market.

7:56 Has at any point Over the last ten years. And just think about that in the context of market efficiency. We have the most extraordinary moment in the history of capitalism. It's wildly bullish for AI.

8:07 And you get a chance to buy AI. At really attractive valuation. What do you make of the multiples that specifically Anthropic and Open AI, which in my mind are like the reference assets. That are the most pure play takes on this trend. really being not that crazy. Like if you just look at the sales multiple.

8:26 And compare it to maybe what Databricks and Snowflake and these companies traded at at their peak. How do you process it? How do you make sense of it? I do think open AI and Anthropic are pretty different databases from a capital efficiency perspective. And atropic. Clearly.

8:40 has a dramatically lower cost per token than open AI. They just do. And you could just see that the amount of money That they have burned. to get to a roughly similar revenue scale. I think they burned maybe

8:52 Eighty percent less than open AI? So his businesses they clearly have very different structural R O I Cs. I think Sarah Fryer is one of the most exceptional CFOs.

9:02 I think they're doing a lot of things to try to improve this. And they've secured a lot of compute. They've secured a lot of compute. That's another big difference. It turns out being aggressive really paid. Anthropic at nine hundred billion for fifty billion and

9:16 A R Growing at ridiculous rates. And I think Maybe a true statement is that Infantropic could just wave a magic wand.

9:24 And get all the compute they wanted. They'd probably be doing Well north of a hundred billion dollars today. Maybe a hundred and fifty. They have clearly deprecated the intelligence of Claude.

9:36 There's an analysis Claude is even on Opus is generating seventy percent less tokens. 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 I think they would be doing materially more. A hundred, a hundred and fifty.

9:58 Maybe two hundred billion. See might be buying it. Uh More like five times Unconstrained. I'm gonna make up a new number.

10:09 U R. Unconstrained rubber. Why do you think they don't raise a hundred billion dollars at a three trillion dollar valuation or something like this? If you were the anthropic CFO. Christian's awesome, we just had him on. Or if you're Sarah.

10:24 It seems to me like 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. I think it's wise.

10:35 The future is uncertain. You were clearly in a very capital intensive game. Even if you are Anthropic. I'm sure is at very positive gross margins on inference today.

10:46 I can drop it probably starts generating cash this year if they are not already generating cash. I think is probably the case. But still you probably want to be able to raise more capital, access more compute. The world is uncertain. Ukraine is starting to really, really win.

11:02 How is Russia going to respond? I think there's still a lot of uncertainty in Iran. All this uncertainty, I think, probably amplifies geopolitical uncertainty over time. So it's an uncertain world. If I think about Iwan

11:15 Elon has always made investors money. He treats it like a sacred covenant. And as a result, because he's made people money. For now twenty years. He has a superpower.

11:27 That is he could essentially raise as much capital. Has he wants whenever he wants. I do think being focused on making investors money. Із вайз And creates benefits that don't

11:41 just last for like a year or two. They can last for the next twenty to thirty years. And the way Elon did this was systematically underpricing SpaceX or whatever else. Like what is the actual method? Just never being greedy on valuation. Never pushing valuation. Just that simple.

11:59 My friend Antonio pointed out SpaceX compounded it. low thirty percent per year for A decade. And that was just because Elon was, I think, focused on Preserving the superpower.

12:09 And having trying to strike a fair balance between investors and employees. I think it's wise. But could anthropic raise money? At probably a at least a one hundred percent premium. To this rumored latest mark. Of course.

12:24 Let's get to the Watson wafers part of the discussion. Always my favorite thing to talk about with you. the importance of this infrastructure build out every time I feel like it's getting overheated. And then the next time I talk to you, It seems like we should have done way more than we did. You studied S curves and the steepness of those S curves a lot. And you know a lot about history.

12:44 Talk us through how you're thinking about Watson wafers today as the key to inputs into this whole thing. I think capitalism is going to solve the watts. Shortage.

12:55 Absent. big regulatory or political blowback. Which I think is a real possibility. The head of data center infra investing at one of the big PE firms. Blackstone Apollo.

13:05 KKR said it used to be Energy and chips. were our biggest gating factors. Now it's zoning. And approval.

13:13 Much more important. I think a lot of companies are waiting till after the midterms. To take. Action In terms of maybe workforce reductions.

13:23 Nobody wants to be Opinata during the midterms. You've seen a lot of companies that make Turbides announce of plans to significantly increase capacity. There's like two of these machines that can cast these big blades.

13:36 We haven't made one in eighty years in the West. We don't know how to make them anymore. All of that is true. By no means am I Minimizing the industrial engineering, magic, and artistry that goes into those, but capitalism is very good at solving problems like these over time. There's other sources of energy besides these turbines. With a longer time frame. So I think

13:57 The watch what's Shortage. Will probably begin to alleviate twenty seven, twenty eight. And then I think orbital compute will really solve that.

14:09 I do want to reframe orbital compute. Because I think when people hear data centers in space Which we discussed our last episode. They picture Pentagon size building in space. They're like, Well, we can't do that.

14:22 That's not what it is. A Blackwell rack. Weighs three thousand pounds, it's eight feet high. It's Four feet deep, three feet wide.

14:33 It rocks in space. It's BaseX has showed you an illustration. It's a rock. That's the satellite. But it's probably about the size of a Blackwell rack.

14:44 It has these solar wigs that are probably Five hundred feet long on each side. You keep it in a sud secretus orbit. So those solar panels are always at the sun. Because it's in an exactly sudden secretus orbit.

14:58 The radiator which extins behind it for hundreds of feet. This is a common criticism. Yeah. How you're gonna go over there. I've spent A lot of time at star base. over the years that I've talked to a lot of space ex engineers.

15:11 And I do think it is the most talented group of engineers on planet Earth. And they're very confident they have solved this. And they're not. Always. Car for it.

15:22 There's some Engineering that needs to happen. to turn the starship into a Mars Colonial Transporter. Will they do that? Absolutely. What are they more Focused on

15:31 I'd say probably the repair and maintenance. Those are the two big responses. The radiator and how do you repair the whatever issue goes wrong on the rack. And the answer is until you have Probably Floating Optimuses.

15:45 You don't. Starship is going to change the space economy in ways we cannot imagine. And particularly if regulation becomes a constraint to data centers, none of it's gonna matter. You're gonna sell as much orbital compute as you can make. And then obviously you link.

15:59 These racks. Using lasers travelling through vacuum. which are already on every Starlink, and it's just mind blowing to me. that SpaceX operates the world's largest satellite fleet. Which is

16:11 Ninety eight or ninety nine percent of all satellites in orbit. Every starling They're cooling it today. I think Starling V three is going to operate at twenty kilowatts. A Blackwell rack is only a

16:24 A hundred kilowatts? And people talk a lot about density. Well If you're connecting the racks with lasers to vacuum, you can make the rack. Think

16:34 Physically, you're focused on weight. Not size. In a data center on Earth where you try to connect Racks ideally using copper. Minimize length, cabling is a big cost.

16:45 You do want that rack. To be small. Copper when you can, optics when you must. But in space, you know, there's all sorts of things. That SpaceX can do.

16:54 That I think maybe some of these naysayers are not contemplating. They operate more satellites than you want. They have a twenty kilowatt satellite today. So maybe you just scale that up to sixty kilowatts to start. They seem very confident they're gonna go right to a hundred to one hundred and twenty. The same.

17:09 Company now. Also operates the largest data center on Earth. They have the world's best hardware engineers. And all sorts of people Омост Овмарнат смарт анаф, а практика ана.

17:23 Турки Спайсекс. Are these armchair skeptics? You know, I don't want to quote Larry Ellison, but somebody was being skeptical. And Larry was just like Listen, he's out there landing rockets.

17:35 I don't see anybody else landing rockets. And the reality is, ten years later. No other company is consistently capable of landing and fully reusing an orbital rocket. None of this makes sense without reusability. That means you have to land. I would like to redefine orbital compute has racks in space.

17:54 Not giant floaty hinted on size data centers in space. That's silly. What makes a data center is you're connecting these racks with lasers. So it'll be racks in space that are connected with lasers into a virtual data center. And if you think about that state of the world.

18:10 Let's say that all happens and we're really good at getting these things up economically and running. Matrix multiplication all over space. What does that mean for terrestrial? Data centers. Someone once said

18:22 America was going to Suck as hard as it can on every energy source it can get. And I just think the same is true of compute. It's why I'm probably less worried about like an a J bear case than I was.

18:36 We're going to consume as much Compute. As we can. Inference I think is very sensible for Orbital compute.

18:46 Training will be done on earth. For a long time. So I don't think that this is super bearish for terrestrial data centers. I think those are gonna be valuable for My lifetime.

18:57 But I do think if you're in this ecosystem of power production and cooling And you were massively ramping. Capacity. A lot of these capacity ramps are gonna be hitting. Just as I think.

19:12 All of the silly skeptics. start to understand that orbital compute is very real. I think it's worth thinking long and hard about that if you're one of those companies. And then all sorts of cool stuff is happening in the interim where

19:25 Getting really good at repurposing jet engines. There's that boomer space that is doing this. Capitalism is hard at work. On Watts. On wafers, though. It's just this group.

19:37 Oh twenty. Older. humans in Taiwan. who are the most important humans in Taiwan, whatever they are.

19:45 The overwhelming fraction of the country's GDP water usage електриси. They talk about the silicon shield. They all view themselves. has inheritors of Morse Chang's secret legacy.

19:59 Yeah, I vividly remember like visiting science park. More than twenty years ago. and talking to them. Do you think you could catch Intel? And they said this is such a beautiful dream, but it's a dream. For our grandchildren.

20:14 Indeed. Did it. Partly because of Intel's self inflicted wounds. They think very differently. One reason Jimson flies over there.

20:23 So much as he wants them to expand capacity. I do think it's wild that Jinson Has never had a contract with Taiwan Sumi. They do business on what seems fair in handshakes. Just fascinating. No contract. It's gonna be fair over time.

20:36 We're partners, we're gonna be fair to each other. The truth is, based on every prior market precedent for a foundational new technology like AI. You've always had a bubble. Carlotta Perez wrote this great book about this. Markets are efficient. They correctly understand that this is a foundational new technology.

20:54 There's what Mobison calls a breakdown in diversity. Everyone becomes bullish on this new technology. And I am beginning to worry a little bit about a diversity breakdown. And then You get a bubble.

21:06 That bubble. funds the build out of this new technology. But Supply gets ahead of demand. And you get a crash, and it's a particularly severe crash.

21:16 If it's a debt fueled build out like the year two thousand. And one thing's really good about the current build out is it's still overwhelmingly funded out of operating cash flows. Which is a really important fundamental difference versus the year two thousand. has his valuation.

21:30 has this the fact that every GPU is running at a hundred percent utilization. when ninety nine percent of fiber was unutilized. So there's all these fundamental differences. History doesn't repeat, but it rhymes. And as investor, we have to be very cognizant of it. And recognize that based on the last Two or three hundred years, you know, forget the internet bubble. We had a railroad bubble, a canal bubble.

21:50 Every kind of bubble. South Sea bubble. We should expect Up. That's terrifying. Nobody wants a bubble. And the reason it's terrible is if you're evaluation sensitive.

21:59 You like massively underperform. You get fired. By probably all your clients. George Vanderhuiden who is no longer with us. Great Fidelity Portfolio Manager.

22:09 He fought the bubble in ninety nine. And he retired. in early two thousand'cause I think he just couldn't take it. He knew it was wrong. His clients were deeply skeptical, George, you're out of step.

22:21 He had white hair, he's truly great man. I only overlapped with him briefly, but he was a very important mentor and friend. to my good friend and mentor, Jennifer Urig. So I have a lot of Vanderheiden DNA through her.

22:35 He was the same person who said being early is the same thing as being wrong. George retires because he can't take the under performance. And he can't take clients saying, What's wrong with you? You don't get it. And he has like Forty percent of his funded tobacco.

22:50 Forty percent did home builders. And literally he probably outperformed the Nasdaq. But I like twenty or thirty X over the next three years. And I have been

23:01 That this fundamental shortage of wafers. Which really today is controlled by Taiwan Simi. We'll prevent one. If Taiwan Cimi did what Jinsen wanted. I think NVIDIA could sell

23:14 Two trillion dollars of GPUs. In twenty six or twenty seven, maybe two and a half trillion. Maybe three trillion. О термір сомаш. He probably would be in an overbuild.

23:27 So Taiwan Semi If we don't get a bubble We need to throw a party for them. Because they will have single handedly prevented a bubble. You are starting to see

23:37 Companies go to Intel. And Samson. Let's just assume TSM stays super supply constrained versus the laden demand. What happened? The history of markets is I don't know who, but one of Intel and Samsung. They're not going to stay disciplined. They will break.

23:52 And then at some level that will force everyone else to break. I think a lot of this may come down to the degree to which Thailand Simi Can maintain A lead. Over Intel and Samsung.

24:05 You gotta remember, it's whatever it is. It's nine, twelve, fifteen months. Deleting node edge, you mean. Exactly. Yeah. pace at which they expand capacity.

24:15 If I were to watch one thing to understand where there's a bubble, it's Taiwan Sumi's capacity decisions. And I think there's a Goldilocks zone. Where they Expand enough. They make it hard for Intel or Samsung.

24:29 To really truly emerge has like a At scale second source was something well north of thirty percent market share. And yet they also Keep this fundamental constraint on wafers.

24:44 Yeah. Helps us avoid a bubble. And then obviously I think the terrafab. Is going to play into this too. Say more about that. It's a space X, I believe Tesla's involved as well.

24:55 Join Fitcher to build the world's largest fab. Here in America. I think they're going to be successful. One They have a partnership with Intel, which is very important.

25:05 Because they're getting access. to fifty years of institutional knowledge that's just nine months, a few quarters, twelve months. Three to five quarters behind the front. That's an advantage. It's also an advantage that I believe the terraf ab is going to get attention.

25:21 From the A Teams, all the Cimi Cap equipment companies. One big reason Taiwan's Simi Caught Up. Is ASML and KLA Tincor and LM Research and Applied Materials. They wanted them to catch up. They don't like having a monopsony.

25:34 The eight teams were in Taiwan working. And tell made some mistakes. And Presto. So the eight teams will be here.

25:42 Because of Elon's reputation. In hardware engineering. And then to a degree that I think is Maybe hard for people to imagine in America.

25:52 Where politics has replaced religion. I think'cause Elon had his foray into politics that makes It hard for some people in America. To see him Clearly, which is sad because I do think

26:02 Manufacturing back to America. He's revived tech. I think SpaceX is in some ways the most important defense contractor in America. He's doing a Starlink is amazing for the world. He's creating all these

26:21 blue collar manufacturing jobs, which is like a goal, I think, of a lot of liberals. And good for America. He's done more than any living human to decarbonize the world. And if you are upset about data sitters on earth for environmental reasons, well, here you go. It's sad.

26:38 But he is a living deity. In China. Taiwan. South Korea. In Japan.

26:46 Having watched him for a long time. What he's gonna do is they're gonna recruit the best people. Because the best engineers Want to work for Elon. Especially in hardware engineering.

26:58 He's gonna recruit incredible engineers. Next to Turfab, they'll be a Taiwan town. Oh, these are your favorite restaurants? I'm gonna move them. And their whole staff.

27:07 From Taiwan? Two taxes. We're gonna make everything the way they like it. And then we'll have Japan town. Same thing. We're gonna have Korea town.

27:15 We're gonna have all these things exactly. Children. То рекрут the best інженір. And that's just not the way that The people who run it tell it Song.

27:27 Think. So he's gonna have the best talent, he's gonna have the eight teams. at the wafer fab equipment companies. He has Intel, which is important. It's so good for all of any

27:38 administration's political goals. And I think it's different enough that it will not alienate Taiwan Simi. And these have long lead times, right? Terrafab is gonna be pumping out. Whatever GPs, whatever chips.

27:51 Quite a long time from now. We'll see. Elon tends to do things differently. Everybody else has taken three years to build a data center. He built one in a hundred and twenty two days. Saga had to give him an office. In their fab in Texas, because he was so unhappy.

28:05 about like the pace at which they're expanding a building. Are you surprised by You mentioned deep seek earlier. The simple reaction to that was, okay, these models are just gonna get Ninety five percent is effective for some tiny fraction of the cost to still Chinese open source models.

28:20 be able to use these for most of what we want to do. Fast forwarded a little bit of time. Two years from now. There's no reason I have to spend a million dollars a year in my small little firm on tokens or something. But then the actual reality seems quite different than this. And I'm curious why there's that dissonance in your mind. I do you think it's fascinating the returns to the frontier.

28:39 All the economic returns. To AI. At the model layer. Not all of them, but an overwhelming amount of them have been at the frontier. Which is surprising to me.

28:50 And I think it's been surprising to a lot of people. This is one of the most important questions. to be answered and you need to have a hypothesis on it as an investor. Are frontier tokens going to continue? Capturing

29:04 the overwhelming majority of economic value created at the model layer. And it is surprising. I remember when Gemini three point one pro came out. It was mind blowing to me. The so good.

29:16 Today, it's intolerable. There's probably a little bit of a dynamic where companies prototype with frontiers than when they put something into production. You're hearing a lot of people do use vertex or Open source.

29:28 But still it is a fact today that the overwhelming majority of these economic returns come from frontier tokens. And that's surprising. And whether or not it continues, I think is a very interesting question. And I'm much more open minded to that, having had the experience I've had with Gemini three point one.

29:47 And then Opus. And then I do use Croc four point three. A lot. It is on the Pareto frontier. The companies that are on the Pareto frontier.

29:55 All right, and this is by the way a big change in a consequence of what we talked about last time, Google losing They're per cost. Token leadership as a result of making very conservative design decisions with TPUPA to try and Take it away partially from Broadcom and NVIDIA.

30:10 Continuing to make aggressive choices. But Google dominated the Preto frontier, the Preto frontier being Intelligence versus cost. And I think this is the most important thing to look at to analyze AI labs. Google dominated that.

30:22 Nine months ago. Every point on the Pareto frontier. OpenAI, XAI. And anthropic, we're inside of them. Now The Pareto frontier is dominated by entropic open AI.

30:35 And then Groc four point three is on the Pareto Frontier. It's clear The best lowest cost five hundred billion parameter model. And then Gemini three point one is hanging on to the Pareto frontier and if I were to bet.

30:47 I bet that they're subsidizing that out of pride. I would just say a violation of Richard Sutton's bitter lesson is for sure the biggest risk to this trade. To all of it. The closer someone is to AI, the more skeptical they are this will. Okay.

31:01 One thing I think contributed to weakness in March. was you know a much more stupid version of Deep Seek, which was this thing called Turboquad. And Turboquad is some Google memory optimization that was written up in a paper a year ago. And then during the middle of an agreement, while Google was negotiating

31:18 with Micron, Samsung, and Heinx to sign some LTA. It would lock in really high prices for a long time. They release this. What people do is always more important than they say, and they just kinda publicize it on X. And it goes viral.

31:31 Like, Oh my God. D Ram is cooked. There's this D Rail optimization. I was unable to find a single AI engineer on planet Earth. who believe that TurboQuant would have any impact on DRAM demand.

31:44 But nonetheless, a violation of Richard Tutten's bitter lesson, you know, more compute will always outperform human algorithmic ingenuity, more compute and data, chinchilla optimal. Beyond Chill Optimal, I guess what people Increasingly do today. That's a real risk, man. The people who are building these models are skeptical of that risk.

32:03 The reason I am a little less skeptical. Is I think we're very close to ASI. And who knows if the bitter lesson holds. For four hundred IQ models. Maybe we get a

32:14 Temporary. period where these, you know, if you get to ASI. The first thing it wants is Probably to be smarter and have more resources. How does it do that? It makes itself more efficient.

32:25 I didn't know that. Isn't Actual risk. Bitter lesson. Literally, I believe includes humans in it.

32:33 So we're about to find out whether the better lesson we'll find out and if applies to three hundred IQ AIs than four hundred. Then five hundred and six hundred? And at some point We may have like a temporary violation of the bitter lesson. Based upon

32:48 AI and ASI. So I'm curious how you think about Some other parts of the Innovation around the model, continual learning and memory being two that people seem to be most focused on as things that might create yet another new paradigm that we would enter. What do you think about the role of those two things. Yeah. Well I think we've done a lot with memory through these harnesses.

33:08 And it turns out that harness engineering Is Not as important. As the model, but it really matters, and these harnesses in these models.

33:18 are increasingly being co developed. One of the big things a harness does, which you just think of as like a run time that the model operates it. And it knows where the tools are.

33:29 It creates context, memory, state. has very specific Promp or instructions. Makes a huge difference. Even simple versions. It makes an incredible difference. I think the last time I was on here or one of the other times I just said, like, hey As an investor, it's very important that you pay for the two hundred and fifty dollar a month.

33:49 Version to get your own intuitive sense. That's no longer possible. To understand what frontier AI is capable of today. Even for a non coding use case, you need to have clot code. Or Codex five.

34:03 And you need to be on an enterprise plan. And the reason for this is and this is another dynamic that's enabled by Google losing their Cost leadership. is these AI models just shifted to usage based pricing.

34:17 And if you're on that two hundred fifty or three hundred or two hundred and eighty dollar month plan or whatever it is. you are getting severely rate limited. You are getting a lobotomized version. of the AI. Because like we talked about, Claude now produces seventy percent less tokens. You want the

34:32 Tokens that claw. And its harness really think it needs to produce to get you a good answer? You need to be on a usage based plan. And by the way. This is So bullish for AI.

34:43 If we go back to Oh five to a seven. Cellular had been a great growth industry, really, for the last ten years. And the reason was You had a combination. A fixed

34:53 Pricing, you had 900 minutes for whatever it was. And then usage based pricing over that. When did cellular stop being a great growth industry? When everybody just went to all you can eat. And by the way, long distance is the same thing. AI is just shifting from all you can eat.

35:07 To pay by the drink. And it turns out people really like to talk to their friends long distance. They really like to talk to their friends on the phone. And people really like To use AI.

35:17 And particularly now that one person can have a hundred agents working. So I think the shift to usage based pricing. Is probably why you will see open AI in Anthropic. exceed well over two hundred billion dollars in ARR this year.

35:31 Not only is more compute gonna become online, but they're gonna be able to push Frontier token pricing with these usage enterprise models. It's sad, it's sad for the world, because it just means if you can't afford that, you're not at the frontier. And I think it's gonna throw off a lot of investors intuit.

35:49 Of the capabilities of AI. But yeah, continual learning, man. I mean, if we solve that. How do you conceptualize that? AI is constantly updating its weights. I mean, it may end up being something different. There's so many mysteries about the human mind.

36:02 Or such sample efficient learners. Relative to AI. Many orders of magnitude. Now we have a crude variant of continual learning today.

36:12 When something is verifiable and that's just Reinforcement learning during mid training. Continual learning is a model that dynamically adjusts it its way or adjusts in some way. In real time, like as a human. The first time I

36:27 Put my hand in a fire. I've learned I never put it in there before. That model. Today needs to put its hand in the fire a million times. And then have the designers

36:37 Effectively put a fire In the next training run or an R L Jim For it to learn. I think it has to be dynamically updating the weights. But I think people are working on really smart techniques beyond this.

36:51 But if we get that. Then we have a really fast takeoff. And people seem confident that continual learning It's kind of just around the corner, and I do think.

37:03 This is like the third big question. Bitter lesson violation as a result of ASI. Or less likely. Human ingenuity. Will Frontier tokens still command the premium they do?

37:16 And will we get continual learning and if so, when? What is the role of new chip companies in all of this? We talked a lot about NVIDIA and their relationship with the SMC and Intel and all these sorts of things. There's a thousand flowers blooming, I think literally probably a thousand flowers blooming. Trying to create a new chip to address some part of this bottleneck. I'm curious how you process.

37:37 This space, this opportunity, what role it will play. So I think this is good and healthy for the world. It's good for Jensen, too. Because a different administration might take a different view.

37:48 competition I think is good for everyone and seeing different architectures explored is good. And the reason is In tank design they talk about the iron triangle. The iron triangles take design. Is that all designers of a tank they have to make trade offs between attack

38:02 Defense and mobility. For obvious reasons. More defense you have, which is just armor, the heavier the tank is. The less mobile it is. Чи хав то лив із трианго.

38:13 And make trade offs. the Markava in Israel. Optimized for defense. Russian tanks and like the leopard. are generally more optimized for mobility.

38:22 Chip design is the same. There are these fundamental constraints imposed by the laws of physics has embedded in the Taiwan CI design rules. Do you need to live with it? You have

38:35 T P U trainium and AMD. Which are all Essentially trying to be a better GPU. And today I think probably training is doing the best. Now nobody's a better GPU.

38:47 But training is tugging. On Superman's Cape. hadn't started yet. The training three needs to ramp into production because it has a switch scale up network. But you really need to economically inference MOE models. A lot of companies have a Taurus architecture.

39:02 That's where Google was. Google's developing a switch scale network. And then AMD is like always kinda flying over behind. Yeah. AMD, we'll see the MI four fifty, we don't know yet. We'll see. We probably know more about trillium three than the MI four fifty.

39:16 Well, that's a hard game to play. So you have to Do Something. Different.

39:22 And you have to do something different that is also hard. To do. So I think the best path for these startups. My rule of thumb is one percent market share is gonna be worth a hundred billion. $100 billion is a pretty good venture outcome.

39:35 I think what Jinson would say is okay, if somebody does something different and it gets to one or two or three percent share. We'll make that chip. And that's Coming for everyone.

39:46 But if you're trying to make a better GPU, good luck. If you were doing something Different. It also needs to be hard to do. And you can make different trade offs, the disaggregation of pre-fill and inference.

39:58 really have opened the aperture. For making these different trade offs because you can make very aggressive trade offs for decode. Aggressive trade offs. For prefill. Freefield being taking in the context, eco being

40:09 You know, right the output. Yeah, I have a great colleague named Andrew Fox who said Prefill. Picture British naval ship from the eighteenth century. Prefill is loading the cannon, decode is firing.

40:20 And what pre-fill literally is, is just the model understanding the question, the prompt. And then keeping track of its own answer. And that is fundamentally a memory capacity. Bound problem. Decode is the process of generating new tokens, and that is memory bandwidth constraint. So if you're a chip designer, this gives you a richer canvas.

40:38 To paint on. But even so It needs to be hard because if you make different trade offs in that iron triangle. to optimized for memory capacity and they're not hard trade offs to make. NVIDIA is gonna make those same trade offs.

40:52 They get better prices. from talent to me than you're ever gonna get. Good luck. And they have the advantage of working with every model company. And optimizing their designs.

41:02 By the way, another very funny thing is There's this process. If you're a V C. And you're investing in semiconductor company. That is telling you they are going to have an advantage because of a Taiwan Simi process.

41:14 That they have special access to. I promise you. Де Джинсон сада просе. When it was a twinkle in Taiwan C's eyes. They know more about it.

41:26 then this little company with Two hundred people. Can imagine. Taiwan C everybody in the supply chain is showing Jensen. Everything. The same way they're showing

41:36 Amazon everything. AMD everything. T PU everything and that's another reason. Don't go try to make a better GPU. So you could do something different. You could paint in the pre-fill canvas, you can paint in the decode canvas. But you also have to do something hard. Because if it gets to scale

41:52 You're gonna have those four companies. has very fast followers. My firm was a Venture investor and Cerebrus. What Cerebrus has done is something hard and fundamentally different.

42:03 Wafer scale computing. It comes with a set of trade offs. But that Architectural decision they made was hard. And lets them do something that no one else can do. And we'll find out.

42:16 How big that is. They're working on really cool things. One of the problems Cerebrus has. Is Once you start needing to glue a lot of chips together and scale up networks or scale out networks.

42:28 You need a lot of I.O. And I o is bound by what's called the shoreline. The sides of the chip. Cerebrus has an overwhelming ratio of on-chip computed memory. Relative to shoreline IO.

42:41 Well, they're really smart people. They did something really hard. They're trying to see if they can put an optical wafer right on top of that. And then that solves that problem. I'm sure they're looking at hybrid bonding of D RAM. On X.

42:54 these alleged limitations that are not true. A Cerebus machine can theoretically run any size model. They're sizes of models where they're much better than other sizes. So Cerebros, what I think is interesting is they did something different that's hard to do. Really hard to do. Waiferscale computing. I do think there's a role for these.

43:11 I just encourage them all. Make a different trade off? Try to do something hard. Everybody's gonna get funded after this three per CPO. It's not gonna be a problem.

43:21 But it took Cerebrus. Three generations of chips. To get it right. Andrew Feldman. The CEO? You can just s

43:30 How hard it was. What he did, and that whole team did. To get where they are today. And they need to have the grit to do that, the resilience. This first chip is a failure. It happens. Can you come back and make a second chip? Well the one I see on this topic.

43:46 Is This is gonna be amazing for the useful lives of GPUs. and may single handedly save private credit. Say more about that. What do you mean by the private credit? Well just private credit, they're in pain from these SaaS loans. And however much they're marked down, they probably need to be marked down more.

44:01 'Cause if the public companies are struggling to adapt. How's like a debt laden company going to adapt? Then invest in what is a very different margin structure business. There's a lot of private credit and GPUs too. And they were underwriting that to, I think, three or four years.

44:16 The disaggregation of inference means That I think these GPUs are going to have ten or fifteen year lives. The AI skeptics are like, Oh, these companies are all cooking their books. The useful life of GPU is only a year or two. The useful life of CPU is only four years because the rapid technological change. No.

44:33 What rapid technological changes done with the disaggregation of pre-fill and inference. Is mean that you can put a Cerebro system or Grok LPUs that NVIDIA acquired. in effectively in front of a hopper or even an ampere, use that hopper and ampere for prefo. And extend the useful life of that GPU.

44:51 Until it melts. They do melt, so they have a time, but you know, maybe you d you don't have to run'em. Has fast. This is gonna be really good for the whole private credit industry. It's gonna help finance the AI build out. 'Cause if you can start to finance GPUs, it more like

45:06 five percent or six percent instead of I think Corway's lowest financing was like low sevens. That actually mathematically changes the cost to finance this build out. We have this technological innovation that's gonna lower the cost of financing. still the useful life of compute on earth. And then I do think the one last thing that's interesting about that.

45:23 Is My friend Jamin from Kotou just did a podcast and Ko two had a deck. And they talked about hey Sellers of shortage are doing so much better than the buyers of shortage, buyers shortage being The hyperscalers.

45:35 But if you own a giant installed base of what is currently in shortage. That's also a very good place to be. And we're hearing CPUs are way more important than they were in an agency world. They do all these things around orchestration, tool calls. The biggest CPU fleets in the world sit at the hyperscalers. may catch up a little bit to the sellers of shortage. I wanna talk about this idea of different and hard.

45:59 applied outside of the infrastructure. piece of this. Now you're starting to interact with new founders, existing CEOs and founders that have to adjust to this new world. What are you seeing the most AI native founders that aren't building chips or infrastructure or models? But just people using this technology to build other stuff. How do they feel

46:19 The most different to you if you've observed differences. I do think this is just for chip design. To me, it's always been a fundamental question for venture. So there are different ideas. That are obvious to everyone on planet Earth as soon as they hear it. And if that's where you are in venture.

46:34 If it's not hard to do. If it becomes obvious to the world. Before you have built scale. Scale is the ultimate advantage. You're in trouble. And the great thing Amazon had

46:45 Was I think it was obvious to a lot of people, but it wasn't obvious to the retail CEOs. Amazon, they were very smart. Any e commerce company that VCs invested in, they would destroy. They'd be like, Oh, that's so cute. We're gonna take our margins of that to negative ten thousand percent.

47:01 And that's like the guys at Wayfair, they did something hard. Amazon tried to kill them and they failed. Those are like tough operationally Really competent CEOs. For me in venture, I always look. Is this gonna be obvious to the world?

47:14 before this company could build scale. Or is this both not obvious different? And really hard to do. I think a lot of founders are really struggling. With this.

47:25 In AI. I think people are Becoming worried Today in Jinson's five layer cake of AI, the profits. Recruiting to energy.

47:35 Green to data centers. Gran just chips. They're crewing to models. Not really accruing to the applications. I think cursor and cognition.

47:45 Got to a scale. They focused on coding. eighteen months ago that people were focusing on coding. Open AI was doing everything under the sun. People focus on coding where cursor cognition. Anthropic. And it was really right to focus on code.

47:57 I'm Jad Masad, the founder of Raplet. tweeted something that I thought was so smart. Just it was something like Bitter less than adjacent. Is the fact that coding might be the shortest path to ASI.

48:08 And useful AI. 'Cause if you're really good at coding, you can write yourself code to do anything. So I think it was really smart of those companies to focus intensely on cody. They all probably got to A scale where they have a place. I think cognition is doing something really, really different.

48:23 But I think a lot of founders are really struggling, man. I think they're trying to get confidence that in niche areas they won't get steam they can get to them and get like a Data moat. before the model companies get to that niche. Or that it's a small enough niche.

48:40 That the model companies won't do it themselves, but it can still produce their venture outcome. Is this related to what you would call like the token path? I know you've used that phrase for. Yeah, I think it comes from a guy that altimeter Jamin Ball, but he just said if you're a software company or an AI company of any kind, you have to be in the token pack. So data breaks that's in the token path. Comparable companies are in the token bath.

49:01 You're not in the token path. In you're not in some really niche. Thing. Life may be hard, and even for these vertical niches. I think if you talk to the people

49:14 At the model companies. They're even skeptical of some of these. Because all of the data that's being generated in these niches I'm from humans. But then you're betting that you're able to use that proprietary data in this narrow vertical.

49:28 Two Train a model that's lower cost. than the Frontier Labs can ever get to. And maybe that's a good bet. But I just think you have to be Very, very careful. Now on the other hand

49:39 If the returns to these frontier tokens relative to other tokens come down. There's going to be an explosion in value creation at the application layer. And I think another really important Point is

49:53 I have a belief. That whenever he wants Jensen can probably get pretty close to the frontier. With his own mouth. With his own model.

50:02 I don't think he wants to do that. But that is what Open AI. And Anthropic are kind of trying to do to him.

50:12 Unsuccessfully, he's a very logical thinker. This is the logical counter move. І усі да опен сорс фронтір. Which today consists of Chinese models. With stolen. American tokens.

50:25 Somebody told me that like deep seek Maybe the original one was only a hundred and fifty thousand reasoning traces. There's many ways to launder this if you're a Chinese company. You can hit All these different APIs you can make it hard now.

50:38 The American labs are working really hard on anti distillation technology, but I just think Chinese open source. They're doing really impressive things in a very resource constrained way. But there's a lot of distillation. And this is why. I think in addition to there not being enough compute to serve Mythos.

50:54 They did not want it to be distilled. They wanted to use Mythos. just still it themselves use it to RL their next model, whatever it is. And then I think what they and eventually I think OpenAI, anyone on the frontier will do is just say, There's going to be some very interesting game theory because it's a new kind of prisoners' dilemma. You know, we talked about the old prisoners' dilemma. being just around like hey you're at a prisoners dilemma where you have to spend

51:17 The new prisoners' dilemma is going to be if you are at the frontier. Do you release that model via API or not? If everyone at the frontier agrees Not to do that. Then Chinese open source.

51:30 If one person defects They're gonna have the best model. They're gonna have a lot of revenue and cash flow, and then of course, resources equal intelligence. So they'll start to pull ahead and then that will lead to Everybody else releasing it. So it's a new game theory. It's kind of the same game theory that you have with Taiwan Cimi.

51:46 CM Sunk and Intel. The reality is if a company like NVIDIA or AMD were to ever really really use one of these other foundries, that foundry would get better. Really quickly.

51:58 So I do think Jensen is going to keep Open source. A certain time frame behind the frontier. I think that's

52:07 Gonna be a very interesting thing to watch. And then by the way. Open source gets monetized. There's this misnomer that open source is free. Open source tokens, they cost energy to produce. You need to make up on GPUs. And the open source model companies almost always get a revenue share. How are you preparing a trades?

52:23 For the world of Mythos three, mythos four. We're just trying to over invest in cybersecurity. But I really believe. Everybody needs to have a safe word.

52:33 Everybody needs to go. Leave your digital devices behind. Literally go to the ocean. And have a family safe for it. Or company safe for it. And it can't be one that can be like socially engineered.

52:44 And this is just to avoid cybercrime where what looks like Your son or your daughter, or your grandparents, or your parents, or whatever. FaceTime's you. It's Utterly.

52:54 Accurate. Simulation of them. They know. Everything and can extrapolate based on what they're likely to say. And says, you know, wire me a million bucks.

53:03 So doing everything we can with cybersecurity. That's defensive. What about Analytical or processing. What will you still be able to do that it won't be able to do, I guess. So it's a good question. I just watched the last CMRI and I asked

53:16 people at my firm to watch it. And the last CMRI, if you haven't seen it, I highly recommend watching it. It's actually a movie that's aged really well. It's Tom Cruise movie from twenty years ago, you know, the conceit is Tom Cruise, it's this better washed up. Civil War veteran who's actually a very good

53:31 Soldier. When he's bitter and washed up'cause he feels like He participated in negative actions against the native Americans. It's during the Mi G restoration.

53:40 And he's hired by the modern elements of the Japanese government to train like an army of peasants. How to fight the samurai. There's a first battle, of course the CMRI win, even though they don't have guns. He fights valiantly, so the samurai decide not to kill him, take him to their village. He becomes a samurai. It feels like the civil war to him. So he fights on the side of the samurai.

54:00 At the end of it, he's massacred by a peasant with a machine gun. The machine gun is here. If we do not all Become masters of the machine gun. We're mastered, so I am trying to become a master of the machine gun and then

54:13 I'm optimistic. There's a long period of time. Where just like if you were a fifty year old samurai veteran of many wars. I fought many wars, Master Dwarf. You will have advantages using the machine gun. I'm optimistic as a lifelong student of investing. I'm gonna be able to master the machine gun, this new technology.

54:34 Integrated into my own process, integrated into our firm's process. In ways that let me contribute value as a human being. For a long time. Like everyone, I have agents running all the time now. What's your most useful agent?

54:47 My single most useful agent is a really good summary of the points that would be interesting to me. From podcasts. There's just six hours a day of stuff. that I feel like it's in my job description to watch. Every time somebody from OpenAI

55:03 X AI Google. Cursor Fireworks, base chin. To say nothing of Jensen, Elon, Dario.

55:13 I feel compelled to watch and I just Don't have that much time. And there's some real needles and haystacks. That is what I would say for me is the most useful. I do think there's a set of things. That I always like to see.

55:28 Like I'm very sensitive to management compensation. What are they incentive to do? Gf stupid RSUs. Or do they have PSUs? And if they have PSUs, what are those PSUs and sent them to do? And we now have

55:41 systems that do a very good first pass at that. That saves people a lot of time. It frees them up for more creative work. Then like going through the proxy. Pulling the PSU thing.

55:53 Looking at how it's changed versus all the proxies. Because there's signal in that. That's very labor intensive and that's so good for an AI. And there's obviously all sorts of same things within investing. pressuring the organization in those ways.

56:06 I think it's been helpful. This is the most exciting, thrilling time to be an investor. I'm getting a little bit worried. The diversity breakdown thing. Yeah. Say just like a little bit more about the kinds of people that are I don't know anyone like me who's not really bullish. Uh D RAM.

56:22 There's all these interesting things happening with AI right now. One is cross-sectionally the valuations do not make sense. They just flat out do not make scenes. They cannot. All be true. In other words.

56:33 You have CimiCap equipment companies trading at forty times next quarter's annualized earnings, and D RAM companies trading at Mid single digit. At the peak of the last cycle that was Five verse twelve. At one point it was like three versus forty five.

56:46 Those can't both be true, and yes. Semiconductor CapEx business models have improved more than the memory business models. We don't know how much HBM. is going to improve memory business models yet. Yes, they have some element of recurring revenue with parts and maintenance.

57:02 But it's not worth a thousand percent multiple cap. I think it's hard to square the valuation of something like NVIDIA. Which is still in early April was essentially as cheap as it gets relative to the market, like in the last ten or twelve years or whatever it is. And very cheap absolute.

57:16 Very s hard to square that valuation with something like G e Vernova's valuation. 'Cause it builds in An unfathomable amount of share loss for NVIDIA. So valuations cross sectionally are really different. Because we are in shortages.

57:32 The lowest quality companies are doing the best. So if you're an oil and gas investor. Mighty divester natural resources. investor and you're well versed in thinking of costs, this is very intuitive to you. And a real bull market for a commodity.

57:47 The commodity suppliers with the highest Costs go up the most. Because it's the most beneficial to them. They go from on the verge of bankruptcy to gushing cash. And this is I think one reason commodity investing is really, really hard.

58:00 Because quality outperforms during the cycles, but you get All of the outperformance during the downturns. When the high cost guys that mooned during the shortages and the commodity bull markets go bankrupt or whatever. You're seeing that happen in every industry.

58:14 The lowest quality players Companies that are hated. And detested. By the hyperscalers and the buyers. 'Cause they have high costs, they're unreliable, the parts fail at a high rate. They're sold out and raising prices.

58:29 And then that activity gets The interest of these retail accounts on X and these stocks get bid to the moon. Whereas some of the higher quality expressions Have like actually really underperformed. А за невестор і сад безю.

58:44 Within a Shadow of a doubt. That that thing that's moon tin acts. In three months or six months. is gonna go right back down.

58:54 Subject to what they do with all the cash. And so it worries me a little bit that people who are very skeptical a year ago are no longer skeptical. But then I just contrast that with evaluations of these High quality. Companies which are just not extended.

59:10 And it makes me feel better. I always thought it was funny in twenty four and twenty five that anyone asked about an AI bubble or talked about it. You have this nuclear bubble and this quantum bubble right here, right in front of you. What are we talking about? This is so real. Some of that nuclear quantum silliness is maybe spread into more speculative, lower quality, smaller cap names.

59:32 Where if you have a big presence on X or Reddit, it's easy to move um. And that frightens me a little bit. But I just wish there were more AI bears. I wish there were more memory bears. Astera is a stock I've been close to a long time. There's a lot of bears on that.

59:47 I love that. Great. I first invested in the C. Good luck thinking that's a copper loser. And then there's also You can feel the baskets in the market and the leverage baskets.

59:59 And what baskets you're in is really important, you know, copper, optical, DRAM, man. In a very interesting thing that's happened this year. Is in twenty four and twenty five the AI trade traded together. You could be long. GPU compute scale up networking and optical scale across.

1:00:16 And short power or whatever it was. that trade worked from like a risk management sense, cause, you know, I'm very factor aware. That all blew out in January of this year. Scale up networking would go crazy while scale out was going down or D Rims. Massively underperforming NAND and HDDs, which had not ha happened.

1:00:36 So these cross sectional correlations within AI. really fell apart and you had to get very fine grade. You couldn't hedge your memory. Anymore with Some semi cap equipment or

1:00:50 Na Everything cross sectionally. really changed a very interesting way in January. And I think maybe one reason for that was AI got to a quality Where it was all of a sudden really easy people to get really smart.

1:01:07 on these different sub sectors start trading them. And then they get put into baskets. And those baskets in fluids. Yeah, exactly. I think some of the biggest opportunities outside of these higher quality names that I think can come pound for a long time. And they're safe on like these low quality names, which are terrifying.

1:01:25 Is in names that are miscategorized. Castro was in a lot of copper loser baskets. Astera, their biggest product is going to be a switch. You use both copper and optics. To connect switches?

1:01:38 Two accelerators. Definitionally. If you're a switch company Or an accelerator company, you cannot be. A copper loser.

1:01:46 Because you're gonna be on the other side of that connection. I wonder if you could rift just for like a sentence or two on each of the major companies. Google, Microsoft, Amazon the major players that are public. All the conversation is centered around these exciting new companies. Maybe run through them and ref.

1:02:00 Google was incredible last year because they had that TPU advantage, which is now gone. The reason I think they're still in a great position. It's just they have the most compute of everyone. We talked about the value of installed bases being higher as a result of shortages. They have the biggest installed base of compute. Google I. O is

1:02:17 This week. If they don't release something. That even slightly leapfrogs. Open AI. And or Claude.

1:02:27 That's interesting. It's not a disaster for Google. It's just interesting and it just means this NVIDIA effect we discussed is even more powerful than maybe I'd imagined. But I'm very curious to see what the Pareto frontier looks like. Literally in five days after Google's Announced its new stuff.

1:02:45 This is a big card for them, but Google between the amount of data they have and the YouTube data is actually really genuinely valuable. It is valuable in a world of robotics. the amount of compute they have, the search business they have. Google's never not gonna be in a good position. And then you see that with G C P going crazy.

1:03:02 You gotta give Zuckerberg a mitzcredit. What he's done in terms of making Meta an AI first company internally. And I do think he is the only one of those true internet giants to have done that. I give him a lot of credit for that. I give him a lot of credit. for paying up.

1:03:19 when he did for that building out of talent. And Muse, I think, was a really big upside surprise. Was the first model from MSL. And it's not on the Pareto frontier.

1:03:32 with you know x a i google's one entrant And then open AI and Claude, but it's pretty close. That was very impressive to me. So I think Meta is in a

1:03:42 Better position. Still not as strong of an absolute position as Google, but like their better position and rates of change matter more than level, as you know, in markets. particularly over short three year time frames. Long time frame's level of competitive advantages tends to dominate. But even within that.

1:03:58 Changes really matter. Amazon, I think, is in a really strong position'cause of training. I do think you're gonna see real PL efficiencies from robotics over the next eighteen months in their retail business. I actually think Nova, their internal models are not where Muses, but they're better than they get credit for.

1:04:17 then Microsoft I like Satya, I admire him. I think he's an exceptional CEO. And I give him a lot of credit for the decisions he's made, but He did go from we're gonna make Google dance.

1:04:29 To being the product manager of Copilot. In like three years. I would love to know during the coup attempt against open AI. De Satu regret his decisions. Does Satya wish that he had supported Ilya instead of Sam?

1:04:44 And that Ilya And Mira. We're really running open AI today. It is heart of hearts. I would love to know.

1:04:53 'cause I think the Microsoft Open AI partnership might look very Different. In that world. I think that's a very interesting question that we'll never know the answer to. But I give him a lot of credit.

1:05:05 But he is doing now. He's taking risk. This goes to the decisions you have to make in that cone of uncertainty are not only How much you spend. Well what you're gonna spend it on, I think Microsoft.

1:05:16 Flinched. For like a moment. and early twenty five. They have this algorithm, we spend this much capex dollars, we get this return. That algorithm was kinda off.

1:05:27 And if you flinch, you lose position. You lose all these allocations, and it's difficult to get it back. So they flinched. And now the decision Satya is making, which the market has punished him for, but I think is the right decision. I mean, who knows how fast Azure could be growing if they're willing to just

1:05:43 sell GPUs to open AI. We're gonna use our compute. Internally to make our own products better. One reason Copilot is so bad or has been so bad is just one enough compute available. They're fixing that. He's the product manager at Copilot. I do think he's a great CEO.

1:05:59 They're trying to use their compute to train their own models. I am a little skeptical that they have the right team to succeed there. But Just like that, they can afford it Ford.

1:06:08 To hire maybe a different team. But I is making Good decisions. That are risky decisions. To position Microsoft for this world.

1:06:18 Where frontier models are no longer API accessible. And I think it's a really courageous decision that I give him a lot of credit for. And he is forgoing. I mean, Microsoft probably be an eight hundred dollar stock today. if they were using their GPUs to serve solely OpenAI and Anthropics capacity instead of using them for their own products.

1:06:36 So I give him a lot of credit for making a great decision. I think what's really interesting Is the degree to which these companies are outward facing. In their Decisions.

1:06:47 The two companies who are the most deeply engaged with startups. or Amazon and NVIDIA by a mile. Then there's a really intense engagement with Google. They're next most intense. Broadcom is engaged in a Different.

1:07:02 They're just everybody's favorite ASIC supplier. If you're a startup that's considered like a level up if you get to work with Broadcom for your second gen chip. And it's considered mana from heaven if Broadcom works with you for their first gin chip. And then you see essentially Zero.

1:07:17 Engagement with startups. From AMD, Microsoft, and Meta. When I say zero, it's a little. And I just wonder about that decision. Because some of the best

1:07:29 Teams Are no longer a big public company is There are at these smaller startups. And I think it's going to end up being a pretty big advantage for NVIDIA, AMD, Google right behind them. To have this engagement.

1:07:44 That you just don't see from these other Hyper scalers. As we wrap up, I'm curious for you to riff on any other out there knock on effects that you've started to think about for this giant trend. We've talked about the specific companies in a lot of detail that this most impacts. We talked a little bit about the application layer and what would have to happen for there to be more value occurring to that layer of the stack.

1:08:06 I'm curious, any other just fun knock on things that you've been thinking about as this world changes so quickly. And it is wild. I mean at the application layer, forget value accruing, just value has been destroyed. AI has net destroyed, even if you count cursor cognition, the most successful. AI Natives. Trillions of dollars of value has been destroyed by uh the application layer. And just in this context.

1:08:26 The companies that are doing the best. Today. That are seeing their values increase the most, that are Creating economic value. Are the companies with the highest effective ratio?

1:08:38 of utilized GPUs per human. Maybe this just means that every human's gonna get a lot of GPUs. But I think that's an interesting fact that we kind of need to be cognizant of. I will just say, and maybe this is a little dark. I am more and more worried about personal safety.

1:08:54 And I worry about this a lot more for People who Have a much bigger public presence and are much more associated with AI. I hope nothing tragic happens. There is this upsurge in political violence here in America.

1:09:06 And as AI increasingly becomes political, I worry that's gonna get directed at more and more. Yeah, political leaders. Whatever I may think or may not think of open AI. I think it is terrible that someone threw Melotov cocktails. At Sam Altman's house.

1:09:21 I am worried that we are headed into a higher variance. Higher beta higher risk world because of AI. And that's for me as an individual, and then for people who are Big players on the chessboard.

1:09:36 Think about what it means geopolitically. We're watching the Ukrainians are really starting to win. And the reason they're winning, I think is not really because they have better drones. I think they do have better drones. That's part of it. Ain't the reason Ukraine is really winning? Is they have the best battlefield AI. Outside of probably America and Israel.

1:09:54 And Has China As our adversaries begin to process that. How do they respond? If the United States, because of its edge in AI

1:10:05 It's great if you're America. But it is destabilizing for the rest of the world. Something I think a lot about is creating a charity to just educate the world on how awesome the West has been. Slavery was endemic to essentially Almost every civilization and slavery was really ended by the British Empire.

1:10:21 Tell that story. But America after nineteen forty five. We had the nuclear bomb, no one else had it. We could have controlled the world forever. Instead, we rebuilt Germany and Japan.

1:10:33 Who are America's most reliable allies. Israel, South Korea, that's a testament to like the American spirit in our country. We didn't take over them. the world. There were these fears that were documented at the time that the American generals MacArthur was a little bit of an American emperor in Japan.

1:10:49 We're just gonna take over the world. And they could have. And they didn't. They call it. We demilitarized.

1:10:55 And then you had this period of great global stability between, you know, a scary they were too American. Yeah, yeah, the Pax Americana. So maybe it's not destabilizing, maybe it leads to another Pax Americana. And I'm so optimistic that AI It's gonna be amazing for the world.

1:11:14 There's Someone like me whose daughter was diagnosed with a very rare disease, there's no cure. He was able to assemble a lot of resources. He was able to get a lot of compute from the labs. We were made aware of what was happening. Spent up an immense amount of agents.

1:11:29 Came up. Using AI with a drug on the market. That can actually impact his daughter's disease. And then has spun up a company. To cure it.

1:11:39 Hm. Life is already immeasurably different because of AI. So I'm like an AI optimist maximalist, but I also just acknowledge it's like an event horizon. It for sure, I think, is gonna be a discontinuity. We need to navigate as society. I think the Luddites are gonna be wrong.

1:11:56 But we need to be like really thoughtful in how we address their concerns. We need to make sure. That it's good for everyone. Like it is a little dystopian that now the best AI is only available to people with a lot of money. We need to solve that. We need to approach this with humility, recognize there's a lot of uncertainty, and be thoughtful. When I do this with you, I tell people afterwards, I'm like, may you find something that you love as much as Gavin loves markets and companies and capitalism and history on display today, as always. Gavin, thanks so much for your time. Thank you. Thanks, Patrick.

1:12:27 If you enjoyed this episode, visit Colossus.com. You'll find every episode of this podcast complete with hand edited transcripts. You can also subscribe to Colossus, our quarterly print, digital, and private audio publication featuring in-depth profiles of the founders, investors, and companies that we admire most. Lear more at Colossus.com/slash subscribe.

What you see above is a preview of the first minutes. One unlock costs 10 credits and covers this episode forever: full segment and word-level timestamps on this page, plus .txt, .srt, .vtt and word-level JSON downloads, as many times as you like.