Dylan Patel Explains the AI War While Cooking | In-Context Cooking

[00:00] I'm not crying because of the because of the AI researchers leaving. I'm crying cuz onion. I promise. I swear to God if Uncle Roger finds this video, I'm going to cry. It's about to be sweeter than Panda Express. [laughter] >> Wait, what do you vote? What do you >> Hey guys, welcome to In Context Cooking, where we take one dish, taste it, and try to recreate it with minimal help. Today we have a very special guest, the founder and CEO of Semi analysis, Dylan Patel. Welcome, Dylan. >> Hello. Thanks for having me. Yeah, thank you for being here. I guess the question
[00:31] to start off is how would you rate yourself on a scale of 1 to 10 of one being awful and 10? Okay, so you're an amazing cook. Looks looks [laughter] >> as a cook, how would you rate yourself on a scale of 1 to 10? >> Uh, probably like a five or six. >> Okay, five or six is not bad. I feel like >> maybe three. Maybe three. Let me let me revise. [laughter] >> So, low expectations and then overd delivering. I'm going to overd deliver. >> The presentation is great, just not the plate. >> Okay. Honestly, like we'll work with three or five, whatever. But yeah, I guess in front of us we'll we have a bunch of ingredients. Do you have an
[01:01] idea of what we're going to make? I feel like it's kind of obvious just looking off of it. >> Eggsum >> eggs, [music] rice, chicken. This is like >> very >> This is like very like >> Yeah. Anything anything. But then the right side, you know, like the peas really throw me off. >> Yeah. >> The ginger though. So And soy sauce. Okay. This is like fried rice. >> Okay. Yeah. So today we have chicken fried rice which we'll be recreating. >> This is restaurant chicken fried rice. So, we'll try to recreate it as close as possible based off tasting it. So, here's a spoon for you. >> Cheers.
[01:34] >> M. That was That was very good. >> I'm shocked it's still warm. >> I assume this has been sitting here for an hour. >> Yeah, we nuked it a lot and then kept it as is. So, yeah. Just to get started, do you want to introduce yourself? I know that um you went to Minnesota briefly after and was it you were a beekeeper for two years. You did a lot of these kind of like side quests. you have like a lot of [laughter] these states, >> but now you're, you know, the leading voice on um chips and everything that, you know, whether it be hedge funds or even people in AI. >> So, I did live in Minnesota briefly after college. Um I'm from rural
[02:04] Georgia. Um yeah, we uh I did beekeep for like a year and a half basically. I feel like I've just kind of sort of went through a lot of life uh just next step, next step, next step. Doesn't seem like there's a clear immediate path. >> Yeah. uh looking back I can spin a narrative like oh obviously I [laughter] wouldn't be doing this because my interest when I was eight was this and my interest when I was 12 was this but uh you know like moderating forums related to chips but you know I thought it was just like a serendipitous thing you know >> um >> but then eventually sort of everything culminated
[02:34] >> and like blogging and doing consulting and doing research being interested in AI and data science being interested in chips and then it all sort of like culminated in like oh my god everything blew up all together and and so I guess right time right moment um maybe some foresight to have your passion be the thing that everyone cares about now. >> Yeah. No, I think it turned out very great, you know, doing amazing. And rumor has it you started your Substack because you read Doug's and thought you could do a lot better and he was like, "Hey, you should start a Substack." Is that true or is there more behind that
[03:04] story? >> Yeah. So, what happened was I um I had an anonymous blog >> on the internet for many years. Yeah. Uh I was moderating Reddit and all these things anonymously for around like hardware, Nvidia, Intel, AMD, this kind of stuff. >> I was posting all this stuff >> and I had an anonymous Twitter in the sil in silicon Twitter, right? Which people on like teapot and like tech Twitter don't understand. Yeah. >> Um >> and so I was doing this anonymously. Doug Doug started posting and I was like, "Wow, this is interesting, but like I think I could do a lot better."
[03:34] [laughter] And then he's like, "Dude, why are you like why are you like posting on WordPress, right? Like do it do it non- anonymously. Do it on Substack and start charging for it." So then but then I was like too I was like I was like I'm not going to charge for this. I'm so great. And then and then one day I was like you know what screw this I'm going to start charging for it because Doug told me too many many times. >> Um and then a few years go by and then Doug Doug joins the company. Um so it was really a great moment that he he told me to uh do it because otherwise it might still just be like a a niche anonymous blog where I'm still just doing random consulting rather than actually uh you know a company with 60
[04:06] people. >> Gotcha. Okay. So, at that time you were still doing consulting, right? And you just had a separate blog. That was >> It was consulting related to the blog and industry. >> Gotcha. Okay. Yeah. That's amazing. I guess now we could kind of look at the ingredients and maybe try a little. So, here we have ginger, garlic, carrots. >> We're just going to eat the whole garlic. >> Yeah, we could all take a piece. And then this I think is And we also have tasting spoons of you want to try. Um I think this is sugar. Yeah. [laughter] >> Like baby. >> So sweet. Okay, let's see.
[04:36] >> I really hope this is a salt. >> Okay, it's sugar. Yeah. So, this is sugar. We'll know cuz everything will be laid out for you exactly like this. Um, so we'll have to eyeball and we'll be going fast. This I don't think we need taste. I'm pretty sure this is baking powder. >> Oh, >> yes. >> What kind of What kind of fried rice does have MSG? >> Yeah, that's true. That's true. >> Uncle Roger. Yes. But yeah, we have soy sauce here, baking powder, sugar, onion, of course. I think this is >> If he discovers this, he's going to roast the [ __ ] out of >> Yeah, he will. But hopefully, you know,
[05:06] whole give us some grace. [laughter] >> But okay. And then eggs, salt. Yeah. >> Wait, is that the tactic? Intentionally don't put MS in your rice. >> Yeah. >> So, rage bait, Uncle Roger. >> Yeah. This is also short grain rice, which is also a red flag, but we just have it. And this is day old. So, this is the important ingredient to have a day old, not fresh. >> Why does that matter? >> Um, it's a little drier and it'll make the fried rice just like the grains be a little more separated compared to being a mushy mess. But yeah, I think we're good. Are you ready to show off your cooking skills?
[05:36] >> Sure. >> Great. Let's get started. >> Dylan, are you ready to make some fried rice? >> Absolutely. >> Okay, so the first thing, let's grab the chicken and then we just need to marinate it. So, or velvet it rather. So, take the cornstarch. It's one of the white powders that we saw before and the baking powder. And then add like a spoonful, maybe half a spoon of cornstarch um to the chicken. And then just a hint of the baking powder.
[06:07] Um, and so yeah, once we add that, >> a pinch or what we talking? >> I just did like Yeah, a pinch or like a very small amount with my spoon. And then we can add a little bit of salt and soy sauce. I think this is salt. Make sure it's salt and not sugar. >> Yeah. >> Okay. There's a little bit of salt and then soy sauce. And so what we want to do is we want to mix it kind of vigorously so it just looks like >> you said salt and soy sauce all over. >> Yeah. Um just like a little bit of each
[06:37] to give it some flavor cuz we're going to marinate it. But yeah, we just want to mix it vigorously so that it looks like a slop of chicken. But I guess one question on my mind that you've also discussed about is endgame scenarios with Arachus or in this case Taiwan to TSMC. I guess like have you thought about what that would look like? >> Bro wants to make fried rice and talk about China politics. Let's go. >> Indeed. >> Very topic. >> That's vaguely racist. [laughter]
[07:08] >> Just just vaguely. Just vaguely. >> Vaguely racist. Yeah. Um Okay. So, so endgame scenarios are are kind of insane, right? Like um >> there there's there's a variety of things that one could uh that could happen. >> Yeah. um >> such as >> so so in in in some cases right like it's like you know status quo is the best right um you know no war happens I'm done with the uh chicken by the way no war happens >> and just start cutting the onion to like small dices but continue >> we have no war no invasion there's no
[07:38] blockade sort of status quo China continues to China continues to you know industrialize itself with the industry that Taiwan has but there's no there's no major event that occurs >> you That's sort of option one. Option two, which is what a lot of people seem to like think is the best, which is Taiwan actually like more stringently claims it's independent, at least a lot of like Westerners. and and then that seems like the like actually a poor option um just because that that uh potentially causes China to uh move move
[08:09] much more um aggressively on Taiwan and sort of like there's like another set of options which is like okay well what if what if China moves close or Taiwan moves closer to China >> um in a political way right and so like that's an example of that is like hey like they elect the KMT um and the KMT ends up uh winning so there's two parties in Taiwan the DPP KMT yeah um if the DPP wins, they're sort they they've won for the last decade and they're sort of more pro- US, more anti-China, more more Taiwan
[08:39] independence. >> Um >> and then there's sort of like the last scenario, which is just a full-on invasion. >> Yeah. >> Um or or at least a a a political coup or takeover of some sort. And so there's like a variety of options possible. Um what I think is most likely um is that there is some sort of political coup or action that um destabilizes Taiwan in some way >> um but doesn't act doesn't necessarily lead to an invascale um invasion. And so sort of this is the
[09:10] best of both worlds for at least China, right? like they don't have to actually enter an invasion, but they get to continually creep more and more on Taiwan without actually having to deal with the repercussions of a war and subsequent blockades and so on and so forth. >> The like kind of galaxy brain thing for an American to want is for actually the Taiwanese um government uh the pro- US party to lose, right? >> Oh, interesting. >> Um so you want KMT to win. Um and if KMT wins then that means China will be placated more and you don't have Taiwan
[09:41] fully move into China's orbit. >> Mhm. >> Um but you do have the government sort of placated in China. Um at the same time uh even if even if the KMT wins, it's not like TSMC starts disobeying American export restrictions because the way American export restrictions are upheld is that Taiwan utilizes American banking systems, American equipment industries, and so they'll still have to uphold any US export restrictions. So sort of like you know China's placated by the fact that they have a friendly government in power and yet China doesn't actually have any of the chips
[10:11] and the US continues to get to access the chips. >> I guess for export controls as well. So I know that right now there's like stances for example Daario right is very anti uh China having access to things. Um and I think supposedly one of the top researchers Shunyu also left because of that. There could be maybe >> he left he left uh anthropic. >> He left one of the top labs. >> Got it. Got it. I guess do you have opinions on like if the US >> how much should I dice? >> You just cut a dice like half and then after that just start cutting the
[10:42] carrots to like a similar size. Do you have any thoughts on you know the danger of maybe a lot of Chinese talent fleeing due to the very staunchly anti-Chinese stance that a lot of these AI labs or top figures may take or do you think this is a like a non-issue? >> You know um I'm I'm not crying because of the because of the AI researchers leaving. I'm crying cuz I promise. No, but it would be it'd be a travesty, right? If like a lot of like Chinese researchers left American labs, you know what? I think it's probably like like >> half or at least a third of researchers
[11:14] at labs are Chinese. >> Um so obviously there's like a level of like you know why are you ant don't antagonize too much. >> Yeah. Um at the same time there is a level of like you know this is this is the greatest uh technology to ever uh fall into humanity's hands. >> Um you know obviously we think we're the good guys and by we I mean Americans. Um and so Americans think that they should control it and maybe maybe Anthropic thinks they're the good guys not Americans as a whole. Um, but whatever the whatever the moral justification is,
[11:46] like, you know, whether it's, hey, we're the good guys and AI is going to be super powerful or and we're the only ones who can steward it or hey, AI would be a great weapon, so let's make sure we're the only one with that weapon. Um, there's certainly some level of control that needs to be had, right? Now, the question is like where do you where does that control like start and stop, right? Um because one could say, "Okay, well, let's just control the chips." Um or let's just control the AI, but then it's like, "Okay, well then you let them buy the chips and they're able to they're able to do everything they want to do anyways." Yeah.
[12:16] >> Right. And this is this has existed across many technologies, right? Um you know, China China has great engineering. You only remove one piece of the puzzle, they're able to re-engineer that last uh bit of the puzzle. You know, that's that's one context, right? And so so that that is like some people's argument, right? So for example, if you look at Nvidia or you look at like David Saxs, I think their argument is like, you know, hey, um, you know, we should not let them have the models, but we should let them have the chips and everything else because then uh they're still relying on American ecosystem, American talent, American, um, you know,
[12:48] technology, um, American platform. The other argument is like hey like look this stuff um you know if if they don't have access to our models but they have access to our chips that alone gives them um shoot >> that alone gives them all the control and capabilities they need uh to to basically um be on par or just slightly behind right and so we see that in the current regime today right yeah China has effectively uh great access to chips
[13:19] um not as much not completely unfiltered access, right? They either have to rent it or they have to smuggle it or there's some chips that are allowed. Um, and and China's not that far behind, right? You look at like Kim K25 uh agent swarms and it's like, well, this is like >> not worse than codeex by a marginal amount. Maybe 5.3, right? But 5.2, it's like not worse than codeex by an like a large amount. Mhm. >> Um and so there is there is like an argument to be made that like hey current regime does not have the US
[13:49] leading in AI by enough >> right >> um and and that's sort of the argument that like Adario would make I think is that look like current regime of export controls China is is still way caught up uh China is still um you know not that far behind um and it would be disastrous if they are >> um in in in in that in their eyes right and so the question is how does one how does one um write that circle right so one could either a antagonize China more ban more things um does that risk a
[14:20] Taiwan invasion or does that um does that lead to something let's say say catastrophic does that does that alienate researchers in in America um that's one argument another one is like well look this is like we've got like two years right you know AI27 bros right um I'm not fully there right on terms of AI27 but I'm pretty bullish on AI generally and so then the other argument is like look we only have a few years until the um capabilities of AI accelerate GDP growth. And it's it's a By the way, am I cooking something? Am I cutting something else? >> Um yeah. [laughter]
[14:51] >> Okay. >> Oh, you did mash kicks, huh? >> So, you you can start cutting. We need to mince the garlic and then the ginger. Um but garlic and the ginger has to be a lot finer. >> Okay. >> Um >> um >> but yeah, continue. I felt I felt a little awkward just [laughter] talking and not doing anything. And I feel like >> so we're just cutting all the vegetables to >> Sounds good. Sounds good. The one stance that like a lot of like um folks have is that you know Yeah. like let's let's be more anti-China um
[15:22] >> cuz there's only a few years left and so like you know until like super powerful AI systems and those super powerful AI systems will make next generation AI systems, right? Mhm. >> And that's like that's like a lot of the um argument that like let's say the Darios of the world would make and I'm sympathetic to that argument to some degree as well, right? Um if we start looking at like >> hey um >> this year Google spending uh Amazon spending $200 billion, Google spending $180 billion on on AI infrastructure primarily, right? Yeah.
[15:52] >> Um this is, you know, 4x what they were doing just not too long ago. Mhm. >> Um and if they get returns on that of any degree, then we're talking about trillions of dollars of economic value being added um in in just the next handful of years. >> Yeah. >> And so, you know, the the risk here is that like, okay, um whatever AI is capable of doing, it's obviously adding hugely to to the economy, and >> you want to do anything and everything you can to slow China down. We've never had an explos explosive growth like this. We're on the cusp of it, right?
[16:23] Right. Right now we only have like you know um you know the AI industry maybe does 50 bill of revenue >> um you know AC and but like we're seeing it explode right anthropics you know adding two three billion of revenue a month now versus they were just adding a few hundred million of revenue a month earlier um so clearly we're in the takeoff period >> um and so so the argument there is that like let's just limit them completely >> um so they don't end up with um all the all these super powerful AI systems. I I I I'm I'm I'm a little bit like, you
[16:53] know, it's it's hard to rationalize, you know, every specific like argument just because like, you know, I could I could I could argue or I could steal man uh any of the arguments, right? Yeah. Actually, okay. No, maybe maybe maybe China is um should be sold our chips because at the end of the day, they're still relying on our chips now for their AI systems. Um and and then they have less incentive to invade Taiwan. And if they invade Taiwan, then, you know, all of a sudden um the whole party stops and we can't do anything. And when you think about what's the capability of uh China
[17:23] versus the US to um to to have a a vertical supply chain in um chips or AI or in really anything uh the China China has by far the most advanced uh supply chain in semiconductors if you just look at China itself, right? In which case like if the the lifeblood of AI is compute to some extent u to a large extent then China would win if we didn't have Taiwan, right? >> Yeah. And maybe that time scale would be way longer because Taiwan is so far ahead in the production capacity and
[17:54] China doesn't have the equipment ecosystem like the rest of the world does. Um but at the end of the day that's that's exactly like sort of the argument that that one would make is is hey like you know if you push China too far now they might invade um they might invade Taiwan and that ends up with with this catastrophic scenario. >> Yeah. Double clicking on something you mentioned already with these big AI labs and even hyperscalers somewhat overextending on um future spend with these data centers. Would there ever be
[18:25] something that would prompt you to get a little concerned because it seems right now we're in a very um acceleration type of moment for AI where still adoption for what even like cloud code and a lot of these great tools um isn't as mainstream. But are there things that if you see happening in the next what 6 months or year that would kind of concern you in terms of where the future lies of these companies maybe overextending a little too much and whether concerns of being in a bubble may actually have some merit to it.
[18:55] >> I mean I think I think sort of not answering your question but talking about something else I want to talk about which I love to do. >> Um >> is is sort of like you mentioned like these doomsdayers and these believers. I think the biggest risk is actually just like the general public hates AI, >> right? Um, you know, I think I think if you go literally anywhere, the general public absolutely hates AI so much. They they have literally no um they don't Yeah. I mean, like, you know, you go to your random artist and they like hate AI. You go to your random like
[19:26] person in rural America, they're like, "Screw AI. It's like, you know, taking all the water." Um, you know, completely like nonsense arguments, but like it doesn't matter. Um, so when you look across the ecosystem, you've got that problem. >> Um, and so like, and then you've got the doomsdayers, right? There's also the general public which like doesn't quite understand AI or maybe they do. Um, you know, and and and they're just like so worried about AI like taking taking our derbs and like all sorts of other things, right? Um, and so >> I think I think that's that's another
[19:57] aspect of this that's like quite interesting is um, you know, general public hates it. Um, the doomsdayers, you know, look, I'm I'm I'm I'm a I'm a bit of a too uh live in the- moment person to like think, you know, hey, what exactly is alignment mean? And and does AI kill us all? Um, you know, obviously there's huge risk to that. Uh, that's not what I'm an expert in. So, I don't I don't really care to opine too much. Um, >> but um, as far as like, you know, is is it a bubble? Are people are people are
[20:28] we are we doing too much? Like, what's going on? Um, is there a bubble? Are we doing too much? Um, you know, the question is, um, you know, if if AI model progress slows down, then of course we're in a bubble. That's that's obvious. Um, but you know, the the the way that progress is accelerating, uh, so fast, right? Um, you know, you you you see it uh month on month, right? I mean, you know, the new models are coming out every month, every week almost it feels like nowadays. Um,
[20:58] or new capabilities, right? Just think about like like this year so far, right? >> Um you know, obviously Cloud 4.5 came out and cloud code came out uh last year. >> Mhm. >> But adoption really upticked in the beginning of this year. Um and so we saw a huge uptick uh just in this month uh just in January. It went from uh 4% of um or 2% of uh commits on GitHub to 4% of GitHub commits were done by cloud code. Right? That does not mean you know a lot of people use cloud code without having cloud commit for them. So you're
[21:28] like you're talking about like or and then they use codeex they use cognition they use you know or sorry Devon they use all these other platforms they use GitHub copilot they use codeex you know we're probably at 10% or so of total code is being committed um or written by AI if not more >> um but at least for cloud code itself it went from 2 to 4% in one month >> right um >> this this acceleration is is like >> I think it's the thing that we've all been waiting for um cuz like fine people are using chat GPT that's great find people are using like image genen,
[21:58] that's great. These were not like things that add trillions of dollars to the economy. These are social networks. These are like, you know, hell, help me on my homework. These are like chill things. >> Yeah. >> Um but now we're we're in the stage where it's like, no, no, no, like these are trillions of dollars of uh economic value that could be added. Um and if you think about worldwide software developer wages, $2 trillion in wages, you end up with a a pretty incredible amount of um spend that could happen. Um, I feel like it's my worst interview ever because I'm I'm I'm like so focused on chopping
[22:30] garlic and not cutting over off my fingers. Um, I wonder if people are going to like trash me for my lack of claw like grip [laughter] cuz like I've tried to do it but then I've failed and I've done like more dangerous cutting grips. >> Mhm. >> I I do wonder how how how vicious is is the audience. >> We'll see. Um but yeah, you're talking about cloud code adoption and GitHub commits being significantly more especially this uh past like couple months and like the uh value that cloud code
[23:00] will have towards economy. >> Do you want to elaborate more on that or was that a finish? >> Yeah. Yeah. I mean I think I think like there's there's quite a bit of um I'm definitely claw gripping now. You know this is I got to make up for it now that I'm cognizant. Uh so I think I think um the adoption of AI has been so accelerated over the last month. You know just just think about everything that's happened in the last month. We've had Claude code happen. We had Claudebot, we had uh Maltbook. Uh
[23:30] now we have uh you know we had Kimmy K2.5 Swarms. We had codeex um 5.3 which is you know a significant step up as well um and [snorts] seems better in some specific areas and and these are these are there's so many areas right and it's like um at least internally like at my company we've we've completely um flipped over right like about a third of the company's engineers about third of the company's hedge fund people and about third of the company is like passionate individuals um you know so the ex-hedge fund people they are they're all in on cloud code now too
[24:00] right they scrape data [clears throat] they do pro they do financial modeling they proform of financial a modeling with cloud code as the assistant. Um, and so there's a variety of like sort of like um, you know, I think I think like we we've hit like sort of escape velocity and all these things. And so, you know, over the last month or over the last two weeks, we've had u, you know, the hyperscalers report earnings. >> Um, and everyone's stocks have gone down, right? Um, Google announced 180 billion of capex, the stock went down,
[24:30] and then Amazon announced 200 billion of capex and their stock went down like I want to say like 10%. So the market hates it, but they don't realize like, you know, these capex decisions are because they see the light at the end of the tub tunnel, if you will, right? Um the amount of the amount of adoption is just insane. So now that we've had these companies like report earnings and they've disclosed what their plans are for the year, and they're much higher than almost anyone predicted, >> um you you you've got you've got >> you've got the market just hating it. Um
[25:02] >> and and and and and so that that that brings like this interesting conundrum which is that like okay the market is mad they're spending all this money on compute capex but these companies know much better than than than you right like and by you I mean the investor right um in reality they're spending this much because they see insane amounts of demand right um you know anthropic doesn't just add $2 billion of revenue in one month uh you know without having you know huge demand And they're doing it at positive margins, right?
[25:32] >> Yeah. >> Um and they're doing it, you know, and and and when they go to everyone, they're like, "Look, guys, we need more comput. We need more comput. We need more comput." Um and so so it's it's, you know, after 3 years straight of every of the AI labs saying, "We need more compute." The hyperscalers are now saying it's not just we need more compute because we want to train bigger models and we want to do more research. It's actually we need more compute because we need we need to serve our users, right? um we need to add hundreds of millions of dollars or billions actually billions of dollars of compute right if anthropic added two and a half billion of revenue
[26:02] um and their gross margin is 40% they added like $1.5 billion of compute in one month >> right just to serve that you extrapolate that line out a little bit and you're like holy crap you know they actually need hundreds of billions of dollars of compute um okay fine well we need to build this all right because it's a front run of you know you build it and then they can rent it >> um >> so the bet here is that growth will continue to accelerate and the amount of money that Anthropic and Open Eye makes will just continue to go up. >> Yeah, exactly. And and and I don't you
[26:33] know like look the party can stop at some point like you know that's for sure. >> Um at any point you know I think my favorite thing was uh I I tweet I tweeted about like people were like oh who expected this capex and then I was like well we did right and then someone replies the whale watcher told you that you're going to see whales. Wow surprising. [laughter] I was like, "Wow, that's a pretty good uh uh reply to me," you know, but like anyways, like it's like, >> you know, you know, obviously obviously I'm the whale watcher here. Um >> you know, we see the capex coming.
[27:03] >> Yeah. >> So, the market doesn't like it, but it's clearly obvious that it's needed. Um >> and and and now the discussion is sort of like you I think a couple years ago I said the hyperscalers would have no free cash flow, right? I.e. they would not have any uh they would not be generating profits and buying back their stock in a short amount of time. You can also transfer your onions and some veg to the bowl, white bowl on your left, I believe, or right. >> Yes. >> Um, so on my cutting board, I just have garlic, ginger, and the green part of the scallions. So, leave the green part for a garnish for later. So, just keep
[27:34] it on. Um, I guess at what point do you think the market would be um satisfied or okay with this? >> I think I think the market is going to get really mad at the hyperscalers. um they haven't really yet. Um but we saw that we saw the signs of it at the beginning of about mid last year, right? >> Okay. >> Um for example, Oracle peaked when they announced within a week after announcing uh that they were going to do, you know, 300 plus billion dollars of deals with
[28:04] uh Open AI. >> Um and the market really uh peaked uh around then. >> Mhm. Um and and and then like since then they've gone down and and like other darlings that were like doing AI infrastructure like Cororeweave uh have also peaked, right? So now we've got this like interesting um >> conundrum where now the hyperscalers are starting to say how much capex they're going to do. >> Yeah. Um
[28:36] and and and and so when we when we think about like hey what's going to what's going to end up happening is you know these hyperscalers are going to keep spending right what is their biggest advantage right it's it's it's that they can build the most infrastructure in the world they've built the organization to build uh infrastructure faster than anyone else um >> wow this like got k cooked super fast >> show the camera >> it's just eggs bro [laughter] Um I just don't generally use an uh induction, right?
[29:07] >> Yeah. They heat up very fast. So just as a heads up. >> Yeah. Um so so so the hyperscalers are, you know, have been the most profitable companies to ever exist in the human in humanity, right? Whether it's Meta through ads, uh whether it's um Google through search, uh whether it's Amazon through AWS and Amazon.com, um so on and so forth, right? uh Microsoft through, you know, Windows plus Office 365 plus Azure, right? Um they they've all been the most profitable companies. Am I going to continue with the onions and such?
[29:38] >> Yeah. Or so are the eggs cooked? >> Okay. You could put the eggs on the plate. >> Yeah. Already done. >> Okay. And then now add some oil. And then Yeah. Add all the veggies. Yeah. >> But make sure not to add too many onions cuz there'll probably more onions and carrots and other things. Just try to have like an even balance um of vegetables. >> Yeah. Yeah. And then Yeah. So, just get some color on the veg. >> Um, do you do you uh tend to um do you
[30:08] tend to do the carrots and onions at the exact same time? >> Yeah. Like doing them all together is probably just easiest. So, carrots, onions, and the white part of the scallions. >> Okay. Um so so the hyperscalers have been the most profitable companies ever. Yeah. >> Um and now they're they're about to face like sort of this like interesting conundrum, right? There's a huge innovator's dilemma here. Whether it's you know Meta Meta doesn't own the
[30:38] platform. Um and uh wherever people's eyeballs are is where people are going to like uh spend their cash. Um or in in in the case of Google, right? AI can disrupt search or in the case of Microsoft, right? Productivity uh suite is where they make all the money, right? Um Office 365, Windows, etc. But things like Cloud Code, Cloudbot and and future iterations of it, just take a little imagination, we'll displace those immediately, right? Um you know, you've got like, you know, same with Amazon, right? AWS is is is a general purpose uh
[31:12] you know, AI infrastructure or sort of infrastructure play. Uh but there's a lot of risk with everything else um with everyone getting disrupted quite heavily. And so they've got this this this dilemma where they could get um disrupted heavily. Um at the same time they've also got uh they've also got this challenge with regards to um potentially you know being beaten, right? So they they have to invest hugely in AI. They have to try and win AI. Yeah. >> Um and if they don't then they're really really screwed. Um but right now the
[31:42] demand for AI is insatiable and they can get pretty good returns just by building infrastructure and renting it out to the labs. Uh but they'll obviously get way better returns if they uh if they have AI models in house. So they need to spend like crazy to do this. Um and at the same time um if everyone else is like sort of it's like um it's Pascal's wager, right? Um if I don't spend like crazy and others do, I lose, right? If I don't believe in God, right? digital god coming then and others do and it happens then I'm I'm a loser right yeah um and
[32:13] so they've all got this dilemma and the only solution is I have to spend more and more you know until to to keep up in the race and so you know this year's announcements of $180 billion of capex from Google and 200 from Amazon right which is you know 4x what they were just doing a few years ago um is is quite in intense uh but in addition to that we're looking at um we're looking at this like skyrocketing in the next few years, right? There's no reason why Google will have any profit uh in 27 at all, right?
[32:45] In terms of cash flow, they will just spend every dollar they make on on AI infrastructure. Um and I think that's at least my belief. Um and and AI models and so on and so forth because that that's basically my belief and the market hasn't fully woken up to this realization. Um we've been saying it for a couple years. Um in fact we even did a piece uh last year which was like how much debt can the hyperscalers borrow right because at some point you know they they have to lever up on building capacity right so an example of this is Meta right Meta is not as large as
[33:15] Google and Amazon but they want to be in the race yeah >> um and so they've already started taking some debt on to build their data centers now obviously they have a tremendously profitable business that could pay it off uh they just have to stop spending the money um on capex that's not necessary uh but but Zuckerberg has woken up and fully realizes Yes. How much do I want to cook the onions and carrots? >> Just get some color on it and then just put it in the plate with the eggs. >> Okay. Um and so so you've seen some hyperscalers such as uh Meta, they've
[33:45] they're already taking debt on for their largest AI cluster in Louisiana. >> Um you know, they're taking like $40 billion of debt on for that. Uh but they're in the market to take on much much more. Um Google and Amazon haven't taken on debt yet for AI infrastructure, but they will. Right. >> Yeah. And and so I think I think people really realize and panic and and and probably this year when they see all of these companies doing exactly this, right? Are they going to you know what ends up happening uh when the most profitable companies that have ever existed which have compounded at double
[34:15] digits for over like a decade and a half now all of a sudden say we're not going to we don't care about profit anymore. We're just building pixie dust, right? We're building digital god. Um and if you believe in it, great. If you don't then, >> you know, tough luck. Um and and all this capex predates infant revenue, right? Cuz you need to have spent the capex, brought on the clusters and all that before you can have the um before you can ever have the um the revenue come online, right? And then the revenue starts off at lower margin, right? >> Yeah. Also, it's time to cook the
[34:46] chicken. So, just add some oil and then we'll cook chicken. >> Okay. Sounds great. >> Done. Yeah. >> Yeah. So, they're saying investment before the revenue comes online. >> Yeah. Yeah. Yeah. So, so there's like a timeline lag when revenue comes online. Uh there's a timeline lag in terms of like when you rent the infrastructure versus like hey, you have to train the model before you can ever start to um um before you can ever start to actually get the like uh AI service revenue, right?
[35:16] >> Um and so you've seen this like with like all the vendors, right? like you know there there's huge spend for openthropic and Google on training models um and and the others are doing it too like uh Amazon and such uh before they ever end up with enough revenue uh generating revenue from the models from the services that they sell on top and so the market is just going to really hate this. Yeah. >> And I feel like that's going to lead to despite the fact everyone in San Francisco is going to see revenue skyrocketing. They're going to see all the amazing capabilities. Uh but they we
[35:47] live in a bubble, right? If you told we we put out some research I was like hey 4% of commits on GitHub are cloud code and everyone at SF is like that's too low right like it's like 100% of mine >> maybe it's 50% for all the boomers right and it's like no no no like you know we've got a lot of adoption to go um and so people are going to like sort of like see all this amazing model progress and revenue growth and adoption in SF but then like in New York and in London and like Hong Kong and other financial capitals of the world Singapore etc people are going to see the exact
[36:18] opposite, right? They're going to see they're going to see the most profitable companies ever are destroying their business model to build um capacity in something that maybe ne necessarily doesn't have returns. >> Yeah. A big bet. >> And so that I think is going to and then and likewise, they're going to see the general public [ __ ] hates AI. And you're going to you likely see like a real backlash to AI from both the financial class and the normal people of the world. Um, >> does it matter though if the public hates AI if it provides a lot of value towards enterprise and companies? Like
[36:48] isn't the main value and profit coming from enterprises more so than the general public? >> Exactly. And I think I think that that like is like the big fear, right? You know, we've already had like decades of like people being like, hey, income inequality is bad. Um, and the value of labor has been has been falling, right? Uh, the value of labor used to be way way higher, right? um you know as a percentage of the economy. But as we've recognized as capital has become more and more important as machinery has grown um we've sort of had this major change
[37:19] which is that um you know capital is taking more and more of the uh uh earnings uh from of of uh capital's taking more and more of their earnings >> and and and so people are really mad about that and now we're going to start seeing huge uh job loss too, right? like, hey, like turns out there's shitloads of software developers just out of school uh who can't get jobs. Okay, fine. But what about like the 2 million people who drive cars for a living? Okay, well works well. Tesla
[37:51] robo taxi starting to be deployed. Zukes is starting to be deployed. Um you know, we're starting to see really the beginnings of all that. >> Um we're going to see, you know, the stock market maybe does well or the econ the GDP is going to look good, but then normal people aren't going to be acrewing much value from it. And so more and more, you know, and eventually like the financial markets will not do too well either because software is imploding because hyperscalers are going to invest all their capital and you're going to end up with this like major major uh weird fear and worry for everyone in the industry.
[38:21] >> Um or sorry, everyone in the world um and and there's like an AI backlash, right? Um and and I think that's going to be like the hottest button issue of like the next election, right? Um if not the midterms, right? Um, and it seems obvious to me that like any party that wants to win should just become the anti-AI party. >> Um, because life as we know it is changing. Taking a little detour, where do you think the alpha is or like the bet is because you said that Nvidia is kind of
[38:51] covering their basis with the Ruben CPX, uh the Gro chips, standard GPUs, um and there's a lot more startups out there that are very specialized and even like a lot of YC companies, right, are like popping up and kind of tackling this industry. Do you think it'll kind of turn out to be a play where like Nvidia at the end of the day still reigns and um like crushes every other company that tries to take away the market share or do you think there will actually be a lot of value crew to these more specialized smaller companies?
[39:21] >> Um in in the in the in the chip space specifically. >> Yeah. Um yes I think that's a really strong debate that uh people are having um is how much how much value occurs to Nvidia how much value acres to the model companies um how much do they start to um really um you know do do smaller chip companies take charge and uh win um and it is really an innovator's dilemma in the sense that like you know hey why did why
[39:51] did Intel and AMD not win in AIG fuse? because they were they were making money off of CPUs and Nvidia was focused on parallel computing. >> Um and now you've got got sort of the same question which is um you know will will Nvidia be able to innovate on all the things that needs to be innovated um or or will there's a lot of scrape stuff on the bottom. >> Yeah. Yeah. Well, it's like there's stuff stuck in the pan, you know, so I'm
[40:22] trying to scrape it off. We could also get a new pan. I have two new pans. >> Actually, that would be amazing. >> The final thing, we just got to now add everything together. So, first add some oil to the pan once it's dry and then add the garlic ginger, but have it kind of lower. >> Have it what? >> Have like the temperature be lower, like not too high. >> Oh, really? Okay. >> Yeah. Nvidia is um they've kind of got this like innovator's dilemma. The nice thing is they embody Silicon Valley spirit more than maybe any other
[40:52] company. Uh which is Andy Grove, right? Andy Grove uh from Intel. >> Only the paranoid survive. >> Only the paranoid survive, right? And and I think Jensen Hong is like one of the most paranoid people in the industry, right? Um >> he's he's constantly like freaking out changing internal things like you know in a good way though, right? Like truly founder mode. Um >> all right. Uh the aromatics are very aromatic. >> Okay. Then just add um all the veg. Don't add everything um cuz like proportions, but add like the onions,
[41:23] carrots, eggs, peas, and then chicken. And then once you have that, then add the rice. And like should mix everything together. And then at the very end, you're going to add soy sauce. And then some sugar and salt to adjust. >> Yeah. But >> Jensen Jensen's very paranoid. Um and
[41:54] that makes him like an amazing founder. >> Mhm. >> Um and CEO. Um, and so you you have all these people freaking out, but it's like the moment he sniffed wind of the OpenAI Cerebrous deal, he immediately went out and was like, "Okay, I don't actually I didn't actually wasn't building this technology because I didn't believe in it, but now I do because OpenAI is trying to is trying to use Cerebrus, so I'm just going to go acquire Grock, right?" Like, you know, it's like that's like why he did it, right? So it's like, you know, there's there's a bit of like um you know, the moment he he sniffs anything, he changes course and tune,
[42:25] updates his priors. Um and I think that's like really impressive. And so as you step forward to like, hey, what about um >> you step forward to like, okay, well, what does that mean for his hardware road map? Well, before he was like making one one, you know, just a few kinds of architectures and chips. Uh but primarily it was all like very similar, right? It was a large GPGPU. Um, and it was like having it was like the best memory, the best networking, everything sort of the best as possible. Um, and sort of like one sizefits-all, uh, with
[42:56] the main line of like A100, H100, B200, right? Um, but as we look to Reuben and beyond, right, Jensen is really like starting to fully embrace heterogeneity, right? Um, you know, much like this fried rice, right? There's no one individual ingredient that shines above all, right? You've kind of got to have a little bit of everything. Um and so so this is like this is like sort of like what Jensen's believing here. So he's got you know he's got this CPX chip. Yeah. >> Right. Which is made for um context
[43:27] processing prefill. It's pretty good at video and image gen as well. Um but it's not really good at latency sensitive applications. You know they've of course got their main line of GPUs and now they've got this these Gro chips right. So, so, uh, you know, they've Nvidia's sort of got every single, uh, aspect, uh, or type of chip possible now within his company. Um, and he's he's continuing to try and like, uh, innovate and and move as fast as possible in all these things. >> Yeah. Um and so when we think about like hey what ends up happening um
[44:00] with Nvidia in in this case it's um it's you know Nvidia knows they will lose because they have a business uh model deficit right Google Amazon they get to vertically integr integrate and vertical integration always saves tons of money um so he has to be better than everyone um by not just like a little bit by by a ton to justify his margins otherwise the vertical integration of of his competitors will win out. >> Mhm. >> Um and and and so this is sort of like I think the big challenge. Um and I think
[44:31] I think the story is not finished, right? Nvidia will remain on top this year and next year um based on what we see, but others will gain some ground. Uh and the question is what happens in the long term. Um and and honestly, you know, the the the the cards are up in the air, right? Um no one has the the right to win. No one has a destiny to win. uh things are moving so fast whoever whoever does the you know innovates the hardest will win not not necessarily like oh you know and and and I think moes are as shallow as they've ever been right uh because how fast
[45:03] things are moving the the the size of the numbers that are being thrown around now right it's hundreds of billions of dollars for each indiv major hyperscaler the size of the numbers are so large that you can just go and justify hiring anyone any talent the moes become much smaller um and This this is sort of like pretty pretty big deal with regards to, you know, does Nvidia win or not, right? >> Yeah. I guess what do you think's the biggest bottleneck for speed to keep us from going as fast as possible? Is it
[45:33] memory? What do you think is the main bottleneck? >> You know, I think I have zero, by the way. >> Yeah. Um I don't think I'll have any walk with me. >> I swear to God, if Uncle Roger finds this video, I'm going to cry. I'm like he's like he's going to be like no MSG. Hi. No. Hi. Induction furnace. What are you doing? [laughter] Um, no. But anyway, sorry. Um, you know, what's the biggest bottleneck to speed? You know, I think you cook fried rice much faster if you have a walk. So, okay.
[46:04] [laughter] All right. I'm done. I'm done. I'm not good enough to justify that. There's like a hundred other mistakes I made, you know. Um, but like I think the biggest bottleneck to like, hey, why why only $200 billion this year for Amazon? Why not 500, right? Yeah. Um I think I think there's like a number of limiting factors and it's sort of like year by year it's been different right yeah >> in 2023 it was definitely all um related to chips right semiconductors co-as uh which is chip one wafer on substrate um driving up production of this was very very difficult and then as we step
[46:35] forward to 2020 um four as you step forward to 2020 oh I didn't throw any sugar or soy sauce no wonder >> oh did you serve it already >> no I didn't I almost did >> okay Yeah. Final step, just sugar, soy sauce to taste. So, taste it as you go. >> I don't know how much sugar to do, but >> dude, this shit's about to be sweeter than Panda Express. >> Oh [ __ ] I completely forgot about the soy sauce. Um,
[47:05] >> anyways, um, you know, in 2023 it was co-s, it was semiconductor supply chains. As we step forward to 24, 25, it started to become data centers. Um energy is a bigger deal in 25 26. Um but as we step forward, right, you know, supply chains are fast and they react quickly, right? So this this current like whole thing of like oh data centers are the shortage. Yes, data centers are a shortage. Yes, power is a shortage. Um at the end of the day actually there's a lot of other shortages around too, right?
[47:35] um you know and and and and when you think about power, okay, well like if you were not creative, right, and you just relied on grid power, well, there's only three companies that make dual combine cycle reactors. But as you step forward to like, oh, okay, well, what if I make uh what if I want um what if I what if I take something else, right? What if I take um aerodynam? Okay, there's a few more vendors. What if I take industrial gas turbines? There's a few more vendors. What if I take um
[48:06] What if I take uh medium speed reciprocating engines, right? Um which are these like sort of like any any company that makes diesel engines, there's dozens of them. Uh they can make medium speed reciprocating engines and I can use I could connect those up to make um power for the data center, right? So when I look at when I look at like hey what did who who who sort of broke these bounds, right? Elon was the first one to sort of say well no I don't care about the actual rules. Let me just like let me just put power generation on site with lowquality mobile turbines, right? Not turbines even, right? Industrial gas
[48:36] uh engines, um you know, reciprocating engines, etc., etc. Um so, so Elon broke all these rules and now the the whole industry's reacted fast enough because there's so many suppliers, right? And the lead time to like ramp up production of these things >> is ultimately not nearly as long as it is in the semiconductor supply chain. So, going back to your question of like, hey, what's the big bottleneck? Um well there's there the big bottleneck is now back again to semiconductors right semiconductors are extremely cyclical. Uh the the the buildings that chips are made in are the most complicated
[49:06] buildings people make. Um you know they are they are um you know they have multi-year timelines. uh they require not not just all the complexity of like electricians and plumbers that uh that data centers do, but they actually require a lot more complex um because there's there's all sorts of uh chemicals and and precursors and so on and so forth that are going through uh the data center, right? Or or through the fab. And so, you know, people have just not built enough fabs. And then, you know, that's that's ignoring all the complicated tools, right? These tools
[49:36] cost hundreds of millions of dollars in some cases. Yeah. Um they're the most complicated thing people make. Um and so you end up with wow. So the the most the the the challenge here is not just um it's it's it's ramping up production of semiconductors >> and and so now we've entered an age in especially in 26 but as we go into 2728 um you know and when we look in 2026 Google would buy a lot more TPUs but they can't ramp production fast enough right and so they have to buy tons of GPUs and we go to 27 it applies again right Google simply cannot buy enough
[50:06] TPUs and they have to buy tons of GPUs um and and when you look across um the entire supply chain no one is getting enough capacity of semiconductor Um, yes, they can put them in data centers. Yes, they could get the energy uh through maybe ghetto methods like putting reciprocating engines, right? Uh, diesel reciprocating engines or gas engines, but like you know, may maybe not the most cleaner uh efficient thing, but they can do it. And so you end up with, oh, okay, semiconductors are the shortage, but what's the bottleneck to building more fabs or to to building more chips is more fabs.
[50:36] >> And people just have not built these fabs yet, right? And that's I think the big bottleneck now. And that's going to persist through the end of the decade or until AI, you know, sort of slows down. >> How are you feeling, Dylan? >> Uh, I I turned around and I saw yours briefly, so now I'm trying to like wipe the edges of my bowl so it looks beautiful, you know. >> Yeah. No need to worry too much. >> No, no, no, no. People are going to judge me hardcore. >> How was that, Dylan? Was that fun? >> Uh, it it was quite fun. I think there was the right amount of stress and uh involved, you know. I think I maybe didn't share my thoughts as well as
[51:06] normal, but >> maybe that's more natural and fun. >> Yeah. Like a mix of things. Great. And this is yours. Okay, let's try yours first and then we'll or let's try the restaurant actually first and then >> control. Control. >> Yeah, control. Okay, cheers. >> Maybe a little cold. >> I always feel like you got to cheers the food. >> M. That's good. >> Mhm. >> Okay, we'll try yours now. >> Yeah. Okay. >> A lot of meat. I love it. I'll try it as well. Cheers. >> M. Yeah.
[51:36] >> I like my more. >> Yeah. [laughter] No, it's very it's it's like a lot stronger. Like it's very deep. >> Mine's probably going to be bland compared to that. >> So I I I think um you know there's this like debate in the world, right? French people, they don't season their food so much. >> Mhm. >> They're all about the ingredients shining. Then you got like you know equator uh equatorial people, right? slop like Indian food and like Caribbean >> Southeast Asian food just like throw on
[52:06] the spice, throw on the sugar, throw on the like everything. >> And it's like, okay, well then is this and and sort of like the the elitist French would say that's cuz your ingredients suck. You have to throw all [laughter] the slop on there, right? >> Overdo it. >> Yeah. But I don't know. I'm I'm I'm a slap son, you know? >> Yeah. I mean, it is very tasty. So, let's try this. The pallets are Yeah. Yeah, I think compared to yours, it's like a lot blander. It's very light. Not going to lie,
[52:37] >> you like your yours the better, right? I could tell. >> Honestly, yours is like >> there's a there's a mess up of technique, right? The bottom of my uh pan was getting burnt. >> Mhm. >> And you know, we can't get a walk a because this is induction cooktop, but because the bottom was getting burnt, the smoke flavor, I think, was importing into the rice [laughter] and I got a walk. It's intentional. >> Yeah. Yeah, you can kind of taste it here, but it's like hard. Very good. Okay. Are there any call outs, things that you want people to know? Are you hiring um
[53:07] clients? >> Yeah, I am I am hiring. Um we're 60 people now. Um we work with all the top companies in the world, major AI labs, major hyperscalers, um semiconductor companies, data center companies, etc. We cover the entire swath from um AI infrastructure, AI models, uh tokconomics, right? usage of AI models um where who who's using them, what are they using them for, what's the cost of running it, all these sorts of things. So that's the area we're really expanding into this year um and last year. And so I think that's the audience that also matches sort of this this
[53:37] audience that you have here. So if anyone wants to track those things, right, usage of AI, the person who was working on this before, unfortunately for them, but also a call out, they got hired by Enthropic, right? You know, so sort of like, you know, the person who was working on this got got hired by Anthropic as a as a sort of like, you know, please don't poach anymore by people. Um, but I think it's it's a good pathway, right? It's like it's like showing your work is public. >> You know, if you're going to kill it, you're going to kill it. You have all the resources behind you of knowing and understanding infrastructure. I think that's the big call out is like we're hiring for that role. We pay well. We have healthcare. We hire globally. We're
[54:08] in eight or 10 different countries like US, Japan, Taiwan, Singapore, France, Germany, Israel, yeah, Canada, um, UK, right? So, we're everywhere. Um, yeah, I think that's that's that's the allure. We get to work with all the coolest people. >> Great. That's amazing. Well, thank you so much, Dylan. Hope it was a great time. >> Well, thank you for having me. Yes. >> Yes. >> I really care about the quality of the chicken. >> Okay. This is like chicken egg fried rice, right? >> Yeah. >> What's your How do you assess? >> I think mine is a little intense.
[54:39] >> Yours is intense. There's probably too much soy sauce. >> That's fair. I forgot to throw it in and then I just threw it all in. >> Yoloed. How much did you have? See, he had he left like >> He lit left a little bit. Yeah. >> Yeah. Yeah. I would say, you know, like I I would take I would take Allan. I would take the soy sauce from me. I don't know. But >> I like Dylan's. >> Okay. >> There's one each. Brandon, you're a tiebreaker. >> I'm on the side of the slot. >> Yes. >> Wait, what do you vote? What do you >> You vote for me. No, I >> I [laughter]
[55:10] >> You worked in a restaurant,