This crap makes mistakes all the time. All the time. It is still just like a... like I think of it once again as like a junior analyst, right? The analyst goes and does all this like really pain in the ass information. You bring it all together to make a good decision at the top. Historically, what happens is that junior analyst who I once was went and gathered all that information and after doing this enough times, there's a meta level thinking that's happening where it's like, okay, here is what I really understand and how this type of analysis I'm an expert in. Actually, I'm very good at. I consistently have a hit rate. Now I'm the expert, right? I don't think that meta level learning is there yet. Um, we'll see if L1s do it, right? Everyone who's spending one quadrillion dollars in the world thinks it will. It better, it better happen, right? If you're spending, you know, a trillion dollars and there's not meta level learning. But for me in our firm, that massively amplifies everyone who is an expert because like you have to still do something. You can't just like slop it up. It's very obvious to me what it's slopping.
Doug, welcome to the podcast.
Yeah, thank you for having me. Yeah. Um, I after all this time, I just... Is it okay if I just call you Swiss? I feel that's where my brain is. So, I've known you for so long. You can call me Mule if you want. I'm, um, you know. Yeah. Yeah. I mean, it's been, it's been a long time.
It's been, it's been a long time coming. I think I first met you at like New Orleans or like one of the, one of the...
Yeah. Yeah. Yeah. I mentioned one of the... I think it was Vancouver, right? Yeah. I think it was like some party.
Yeah. Yeah. Yeah. And you were like, 'Hey, like who's this tall dude?' I'm like, 'Oh, okay.'
Yeah. Yeah. Well, I mean, it's just like I, I knew about you and we, we've like been internet, you know, pen pals for a long time, so it was like cool meeting in person. Yeah. Yeah. I think that was the first time I ever met you in person. So, yeah. Amazing. I, I didn't go to the New Orleans one. I really wish I did. I love New Orleans, obviously.
So, we have two New Orleanses in a row. And um, yeah, honestly, we should go back there.
Um, are you guys going to Melbourne or the Australia one this year?
I, um, have... I don't even think that far out, but that sounds pretty interesting to me. I think, um...
I can't remember which one. There's a, there's something in, in Korea this year, right?
ICML. I think I'm going to try to go to ICML in Korea. And I know NeurIPS is...
I don't know, man. There's so many conferences. I honestly, I hate to say it, I'm not much of a travel guy.
Well, yeah. I mean, I'm glad to catch you. Um, I mean, I, I am traveling to you.
Yeah. Thank you. I really... Yeah, it was fun.
Yeah. Yeah. Uh, I did not know that I'd be caught in a snowstorm.
Uh, it's, it's funny. I feel like people recently have been coming and they keep getting stuck in these snowstorms. So, yeah. Uh, first blizzard in four years or something like that. Thank you for coming.
Yeah. Yeah. It's a pleasure. Um, and so you go back... You, you used to be anonymous. You used to be Value Mule, which is how I know you.
You know what's funny is that Value Mule was like the very first one. That's the... Yeah. Do you know how I noticed you? I was just like, 'Oh, this guy seems smart.'
Yeah. I don't know, dude. I mean, I, I remember noticing you, too. So, it's like, you know, this was in the early, like the primordial days of Twitter. Um, honestly, I miss those the most. It was like 2017, 18, something like that. Yeah.
But yeah. Yeah, I remember from Value Mule. So, if that's like the deepest cut, if you are even aware of what that is, that is like the deepest cut that you'll possibly have. Um, and then yeah, I have another account and I actually have a third account which is my, my main account these days.
Yeah. So, wait. Oh, which one is that? I don't have one.
Okay. I don't want to dox your other account.
Oh, okay. Just semi, semi-dox.
Yeah. Yeah. So, so, so it is there. That's... it's not... that's okay. That's like my oldest finance account. I think of it as my legacy account. Um, I, you know, I want to have some privacy. I feel like, uh...
Yeah. Yeah. So, so now you've gone all in on the brand and everything.
Yeah. Yeah. I got brand and everything. Yeah. Cool. Cool. Cool. Um, same profile pic, you know. So, yeah. So, um, let's, let's do a little bit of the Doug story because a lot of people hear about Dylan. Uh, and I wanted to just make this the Doug story. Made the Fabricated Knowledge story. You used to be a value investor. That's kind of how you, you were Value Mule and you had a mentor or something that nerd-sniped you into semis. Is that the, the story?
Um, no. Actually, I solo nerd-sniped myself. So, I, I wouldn't say value, because, uh, well, well, for, for everyone who's listening to this podcast, it might as well be value, right? Maybe quality-focused back in the day, but we had this whole thing where we wanted to buy quality compounder companies. And, um, the one I found that nerd-sniped me, all the, like single-shot me, is I found ASML and I, like, fell in love with it. And then I, like, after ASML, I just, like, read about all this stuff, how complicated it is to make these, who are the people who are able to make them, and then I, you know, semiconductors, the whole downstream, all from there. But it started with ASML in 2018. I really fell, fell in love with it and then I read like textbooks and I just, like, kept going deeper. And my favorite part about...
I was going to pull out the Asianometry.
Yeah. That's perfect. That's a perfect one. Be amazing. John's a monster, honestly. This one, right? Uh, yeah. Yeah. Um, I mean, the thing that's crazy is he has, um, I don't know, he has a whole playlist about it. Every single aspect of what goes into it. And what, what's truly great about it...
It's all science fiction. Like that's my favorite thing is like science fiction exists other than, you know, the talking perfectly intelligent robot, whatever, information LLM. Yeah, ASML is all science fiction. So the semiconductor stuff's always been science fiction. Always loved it. Always thought it was cool. Thought it was the most important thing that we ever made. And, uh, yeah, kind of followed from that.
Yeah. I don't know if you know, but obviously, you know, I used to be an analyst myself. I covered TMT, which is a freaking huge sector to cover. It's absurdly huge.
Like I was covering Sprint.
And got like, uh, you know, Viacom. Yep. And then there's ASML and... Yeah. Yeah.
Yeah. The, the T, I feel like, uh, the T and the M and the T are actually three completely separate industries. Um, but once upon a time, I think in 2000, they were kind of really close together, right? Uh, but, but ever since then, it's really split off. Yeah.
Yeah. Well, I mean, I just, my reflection is like I used to be in... I used to be, I guess, our tech sector guy and like I did the flights to Taiwan, and I took those meetings with like Credit Suisse and all those guys that would, you know, tour you around and all those. I never really felt like I got it, uh, because I was always being filtered through like investor relations and all that. And, um, I think you have to do what you did where you sort of go muck mode into like textbooks and stuff and like actually learn about the, the tech, but then you, you hard... it's like really hard as an investor to like make the connection to, okay, wow, what does that mean for this, this quarter, like, or at least this, this year, right? Because like...
One, there's like just so much foundational knowledge and then, and then, then you're like, well, okay, everything here is taken for granted. It's already priced in. So like, yeah, you, you assume that all the Taiwanese people who are buying and selling the rumors of capacity are pretty well informed. You assume all the people who are TMT investors in the United States are pretty well informed. I think the thing that was like the foundational difference for me is like, you know, real thesis around, um, one, I think being young and brash and believing in yourself to be like, no, this is something that really matters and everyone else doesn't see it, really helps. But for me, the thing that was like, I guess, radicalizing was, um, I really believed Moore's law was dead. Um, and I was like, 'Oh my god, not only is it this cool new technology, it's super hard to make and very interesting and technologically very fun to understand and, and like I get it intuitively, but, um, also everything, all the old playbook is about to be thrown out because it's been like, this is a super mature industry.' You read these primers about it, like that's how you learned about things back before ChatGPT knew everything, or, um, you had to go and read these primers of all this information. They're like, oh, it's a very mature industry, and mature... they used to be really immature in the 80s and 90s and 2000s, but now you're consolidated, growth doesn't go up a lot, and everyone kind of had this old playbook from the early 2000s. A lot of people hated hardware. Um, there's just this perception that semiconductors weren't valuable, weren't as valuable. Actually, software was the most valuable thing. Now software is getting [__] on, but that's like outside of the scope of this. But, um, people just thought it was this old mature business that had nothing new under the sun. Meanwhile, every single day, just making a new chip was like science fiction. Um, people took that for granted. And when the science fiction ends because you can't make the chips as small as you, you could, all of a sudden all those free gains you got go away and you have to think about it. And what happened for semiconductors specifically is it created a lot of pricing power or value for everyone who knew how to make a good chip. So Nvidia is probably the best case. You could talk about parallel compute and all that stuff, but it's not just like they know every aspect of it from the chip to the networking to the design to the scale-up, the whole thing. It, it is like, you know, versus in the past it was just CPU gets better, right? And so I think that I had a really deep belief that...
That, uh, Moore's law was, uh, Moore's law was ending and everything would change. And so coming in with that, like, thesis at the top level just like made me want to attack every little assumption and something that really changed as well. And dude, this is honestly my, my favorite post I've ever written. It's like 20, uh, it's like ChatGPT-3 and the writing on the wall in like 2020. You know, my early... I get an early pitch to, um, for Fabricated Knowledge and I'm like, hey, you know, I'm going to make a really, you know, Moore's law's over, scaling laws seem like a big deal. If you simplify it all the way through, it's like, okay, supply, uh, you know, supply, divide, uh, you know, demand, right? Yeah, demand is growing a lot because of scaling laws. Uh, supply is actually slowing down because Moore's law is completely screwed. Um, that's probably really good for semiconductors and parallel compute is going to be a big deal, blah, blah, blah, blah, blah. Um, my conclusion then was, um, you should just, like, Nvidia is pretty much the only one who's going to benefit. Um, and, and, you know, so that's my, my, my, like, good long-range prediction. I feel like I just... Yeah. I just don't think, I don't think I would have expected the magnitude. I think that that's been the kind of the craziest part about this whole story is like I had all these beliefs and thesis and like I really, really, really believed. The reason why I met Dylan is he was the only person who was as semiconductored in the entire world as me, is how I felt. Um, and so I remember like yelling at him, arguing about all these kinds of things and like our, our DMs and stuff like that.
Online. We met in person and... No, no, no. I've, I've actually only been to Taiwan with him one time, I think. Um, so, so I mean, like, look, we, we just met in person. We yapped. We went to conferences. Um, but I think that that's like kind of we were both really early to the thesis, kind of have a different background and perspective. Dylan is technology first and, you know, obviously the technology matters. I have a little bit more of a financial background, but always around him. And it was just like, you know, he's the only one guy who like cared to the same level. Um, so yeah, this, the, the thing that's crazy is like we called it, we were right, blah, blah, blah, blah, blah. But like the thing that I think that still shocks me all the time is the magnitude of how right we are, you know, like be like, oh, Nvidia was good, right? Nvidia is pretty good. Um, and then it's like, no, Nvidia is now the most valuable company in the world. And I think if you had me read that and like truly, hey, I wrote that, I believed it. Um, I still wouldn't have put that together or like I wouldn't have believed it if you...
This is one of many theses at the time.
Exactly. Yeah. Yeah. Like there's so many things...
Like what else are you writing at the time, right, that didn't work out, you know? Um, yeah. We can look, we can look back, but...
I, I, I'm pretty happy with my, my long-term track record. I really am. Um, but yeah, I'm just really surprised the magnitude of how everything happened. Like it's crazy to me that like CoWoS is like a, a, not a household term, but like relatively well-known. It was like an exotic technology. So all this stuff has been this like learning journey, really believing where technology is going, uh, why chips are so important, and then obviously understanding the big scheme of all the things, putting it together. And so that's the... Yeah, that was like the early days and it's, I think it's all been downstream of that, like, you know, one goated insight pretty much.
Yeah. I mean, and, uh, probably like a career maker right there, you know, and, and I just like, I love those kinds of like sort of quarterbacking those career decisions for other people who are also weighing a bunch of things, right? Like I have ADD and like I, I just chase like whatever is interesting, but at some point you just have to like really choose and...
Yeah. I, I think one of my skills have always been like trend following and trend watching. I think when, you know, if we're talking like on my account, like Value Mule or, you know, full all the time, like I was always pretty good at trends, like being relatively early. I remember loving and being obsessed with TikTok in like 20, uh, 2019 and everyone's like, why are you so obsessed with the dancing music show, like stuff like that. I feel like I've always been decent with the trends, but I think the thing was, um, when you see a really big wave that you have a lot of conviction in, it's worth going all in. And that's, that's kind of what it came down to. It's like, wow, I see this really big wave. It's worth going all in. And so, um, I reoriented my life around it. Um, yeah.
Cool. Um, we're going to talk about other trends that you've spotted, primarily like the sort of memory cycle, which, uh, but also, uh, optics, which amazing story. Uh, but we wanted to sort of focus this for the Claude Code launch, Claude Code anniversary, and you've been a big Claude Code show.
Yeah, I, I am. Where's the chart with the four, uh, 4% of code?
Uh, it's, it's actually go to the top left. This one. Yeah. Yeah. Android. Oh, you know what's really crazy is we've updated that chart. I think it's like five now. I mean, and, and like as you know, it's really easy to generate code now. So like that, that number will continue to climb, but it's like just staggering the rate at which this is happening. So, so let's recap for people who, let's say, I think this is one of the most important pieces I, I've read in a long time. Um, and, uh, you know, you, you let it, and it's, and it's, it's weird because I think of you as like an analyst, right? Like, um, you, like one of SemiAnalysis' alphas is that you're kind of like the fun millennial semiconductor firm when everyone else is super boring and old. Yep. Um, but like what are you doing, you know, getting so into Claude Code? Like, you know, shouldn't you be reading reports and stuff, you know, like tell the story of your, uh, Claude Code psychosis?
So yeah, I think here's the thing is, um, if you want to be good at any game, we're, we're tool users at the end of the day, right? If you are good, if you want to be like, and, and obviously this is like outside of my job as SemiAnalysis, like I have all these other things I need to do to, to grow and make SemiAnalysis the best research firm ever. But like, let's say you're a fund manager or an analyst, right? Your job is to find information edges and like new ways to put information together that no one else has done. Um, and so like I've always thought it's really important to know the most important weapons-grade, uh, tool that you can do all the time, which is essentially Claude, Anthropic, all this kind of stuff. And I've been pretty like, I'm a, I'm an early adopter in tools as much as I can be. And, um, like for example, I've been running the, our, our case study that we have into Claude Code since it first came out. Like, you know, I think over a year. Like, you know, I want to say March, April, I started to...
Uh, so the case study for people when we, we're hiring like a financial analyst, like a, our core research seat or something. Hey, um, you know, can you, can you take this company and do some analysis, blah, blah, blah, give us this format back. Um, and I've been running it through like the agentic things, like, hey, what, what, when agents really come around, they should be able to one-shot multi-step hard things to do, things that would take a human 24 hours to do, right? And I always wondered because I, you know, there's some good submissions and there's some bad submissions. We pride ourselves in the case study and being good. And, and honestly, I always joked like, well, you know, they're going to start to beat the worst submissions. And so like that was, that was always my base level. I have a base level of is it better than a ChatGPT agent mode or Anthropic's Claude Code or Gemini CLI, whatever. And so I started running these benchmarks a little bit. And so I was very familiar with how good it could be, but then I was like, ah, it isn't quite there. Um, I vibe-coded some stuff on Opus 4 for sure, but like, you know, it was like kind of interesting projects on the side. It was really hard. It took a lot of feedback. They would, it would mess up. It just didn't. And then, um, you know, everyone was freaking out about Claude Code 4.5 and I like took it for a spin, especially around the holidays. I had some free time and then I was like, 'Okay, well, like how good is this?' And it just like, one, it started like one-shotting everything, right? Like all these MVPs that like, you know, you have to be like, 'Well, the UI is whatever.' It's like, 'No, it just one-shots it.' And then you ask it to do something better and explains what you're doing. Like that's actually really good. And so I was like, 'Wow, generalized easily one-shot MVP of these like projects and able to like really build things on top of it because you can trust what it's doing to a certain extent.' And it felt like some level of capability was beaten. It was very different than what I've done in the past. Oh, I also tried Codex 2 before this, like, um, like Windsurf 5.2. Never really got it to work in the way seamlessly agentically like...
Of course. So this was, this was recent.
Oh, no, no, no. So, so, uh, this, uh, my most recent when I was like, oh man, the, the awakening, um, probably December 27th.
You know, it's a day... something like that. Something like that. I'm thinking because it's between the days and I, I got home from Christmas and I was like, uh, um, my fiance wasn't feeling so well, so I had some time to mess around just by myself. And, and then also there's 2x usage limits. Oh my god, I miss those days. But I mean, now I'm addicted to fast. Um, but, but look, I, I was playing around with these coding agents just like everyone else should or, or should in the space. And like Claude Code versus Codex, I was like doing, you know, simple testing to see if they can make a thing and it never really like one-shot like a total idiot's thing. And then 4.5 just started one-shotting stuff and that to me was like a huge difference. And so, um, I was like, wow, it could just like one-shot stuff. I have all these interesting ideas.
Uh, is it Excel, uh, sheets primarily?
No, no, no, not Excel, uh, sheets primarily. I would say it's usually a mix of like a dashboard or Excel or something like that. But a good example where I like, I, I think Excel, it's moderately okay at like, let's say, one-shotting a basic financial model or like just taking and, and putting information from one place to another. It's not a human level, but honestly, if you know much about investing and being in the business, it's like, is your model, you know, being 5% more accurate really going to ever make a good investment decision or not? No. Never. Not once. Like no one's saying, oh yeah, my estimate is always one cent more tighter than everyone else and that's why I'm good at stocks. No, it, it doesn't matter.
It's like sell-side is ridiculous because like everyone's like, I'm bullish because my EPS estimate is like 10% higher than, than the street. Yeah.
And I'm like, 'Oh, who cares?'
Well, I mean, as you know, sell-side, if we're going to do this, like shots across the bow on, on sell-side, I mean, look, one of the reasons why SemiAnalysis has such like a successful business is because I think sell-side as a concept is very broken. If you're talking about waves and things that are changing, uh, sell-side in a lot of ways is this hereditary child of like, let's say, 30 or 40 years of banking where you had, um, you know, a company go public. So you needed someone to talk about it to issue securities and selling the stock. You're literally selling the stock. You have, but you have to be independent-ish. So your ratings, buy, sell, hold. Uh, one of the biggest sales you could do is like when your, when your company IPOs, um, we'll talk about you so people know who you are. That's the, the core original part of the sell-side, right? And the problem is like all the research kind of has just like really kind of fallen apart. It's just not different. A lot of banking regulations has changed. And so like the primary information process, it's like a 40-year-old business model on its last legs. And so I mean, that's one of the reasons why Semi is so good is because we are not focused on being a one-cent EPS thing, which I would argue isn't exactly skill. It's just mechanical maintenance. Um, we are really good at understanding when technology changes and how that impacts everything, right? Um, because it doesn't really matter if one EPS is slightly higher or lower. It does matter if like, I'm, I'm just giving an example of AMD's Helios rack is super on time and is like out at the gate ready to make tokens on this day because that's going to be billions of dollars of difference in revenue for AMD, right? Or some networking technology or something like that, some bottleneck, being really right on the timing and the magnitude of those inflection points will make a huge difference in the stocks. And so that's our business. We're a research firm. We're independent and, um, we've had a really good hit rate and we, you know, we care deeply about the technology.
Exactly. Yeah. You know, I, I didn't mean to characterize you as like... No, you are young and fun, but also you're extremely damn good. Yeah, it's like, it's almost like a triple threat. And I always, almost always wonder if it's like, okay, it's like one, you have like deep understanding of the tech. Two, maybe you're like sort of financially sort of, uh, literate, but also two, three, there's like this like X factor that is like, well, focus on things that matter. [__] everything else. And, and I don't know what that is, but that, that obviously is the alpha.
Yeah. 100%, 100%. That's, yeah, that, that, that's always been the analyst PM conversation. It's like, hey, you know, there really is only one or like three things that actually matter, right? Find me those three things. Find, find me those three things, right? And then there's all this information. What's actually what, you know, that's the hard part. But yeah, we, I think the thing is like we're really focused on finding the things that actually matter, right? Like the things that like, hey, this 30 is better than this 30. This case doesn't matter. This one actually matters because now you have a giant opportunity. Um, and so that's, that's what the game is all about, I think, in terms of the research. Yeah. And like a, you know, finance perspective. But on top of that too is just like when you do so much research, all these different little industry parts are so hard to understand, man. Like you go to some networking conference and you're talking to a guy who works at a company with, they're talking about their new EML versus whatchamacallit laser, um, you know, I can't even remember what, what EML is replacing, blah, blah, blah, and you're like talking about all this stuff and they have PhDs and you don't. Okay, everyone has a PhD at the deepest level and they're all doing... so you have to understand all these deep understandings of these parts of these, um, supply chains, but you also have to have a big understanding too because you know this little part at the bottom of this supply chain is actually going to impact this giant, you know, business at the top because it's all interconnected, but it's so complicated. Just paying the tuition to show up is very expensive. So I, I think one of the ways I'll bridge this for listeners is that, um, this is the complexity of the problem domain that's, there's extreme depth, there's extreme width, and, uh, you have to kind of throw human attention at all of it to find what matters. And you're, you're saying you noticed some kind of breakthrough in December where it was suddenly clicking for you. I, I just really wanted to figure out like the tasks, the task that I was nailing and the task that it still is not great at.
Yeah. So let me specifically talk about my use case because, hey, I am still a stock guy. I can't trade or do anything in semiconductor or, or AI world. But, you know, I do still really enjoy stocks. It's one of the reasons like I'm passionate about it and it's probably my, my defining skill. What makes me good or bad at stocks, quote unquote, you know, the people who are really like stocks, they're like lifers. They just love this [__]. Um, it's, it's like an addiction. Okay. Um, so I, I'm like, 'Hey, you know, um, here's like all my positions and like here's some like thoughts on it. Can you just like kind of like start copy-pasting some notes over and putting all together?' It's like, yeah, it does. That's why you give Claude Code. Yeah. Okay. So I started doing this and then I'm like, okay, like add it, make the portfolio run some basic risk stuff. It's like, yeah, sure, fine, whatever. And then also like everything you do is perfect. I'm like, okay, well, like actually can we like make an investment framework for my investment style and start to grade all this stuff and then like attack it and do stuff like that. You could just do like iterative work. And then I was like, whoa, whoa, whoa. This is like a crazy useful tool that systemized how I think really quickly. Like, okay, what else can I do with it? And the answer is like [__] anything, right? And my joke on the, the podcast is it's all a skill issue now. Um, and so I, I've been, I've been doing this systematically for every aspect that I can think of. Like, hey, now it's so much easier. Like I was, um, actually perfect example is this, is this chart right here. Hey, Claude Code is a really big deal. Everything's one-shotting. I'm reading everyone going into psychosis like me at the same time on the internet. How do I actually know what's real and what's...
I wonder, right? I wonder, right?
So I'm like, 'Okay, uh, I heard about the fact that Claude Code has the commits, right, onto the public, uh, onto your, your commit. It says, 'Hey, signed off with Claude.' Well, why, Claude Code, scrape me all the commits, right?' And, uh, you know what? Lo and behold, it pretty much did. Like, and it's like, 'Okay, well, like I'm looking for this signature right here.' Copy-paste was like, 'How would you systematically go about doing it?' Did like a big query pull for all the stuff. Pulls all the, uh, like every single day. The API is relatively open. And then I'm like, oh my god, let's see how much this is growing. And it's like, okay, chart go up. And you're like, how big is it as a percentage of GitHub? You're like, chart go up. It's a huge deal. And I'm just like watching your, uh, you know, I have like a cron job updating it every single day, blah, blah, blah. And I'm like, this is a huge deal. Like this is the, the biggest deal. I, I love watching trends. I love watching exponential trends and I've never seen one even remotely at this rate. You would, you know, 4% in like two weeks or something.
It's a, it's a previous attempt prior to you, but somehow they didn't, they didn't, uh, they didn't talk about, they just talk about merge rates, but they didn't, they don't, they don't plot it as nicely as you do.
Yeah. Well, and also you want to... Okay.
You, you asked, you have to, you have to, you asked the question of what is this as a percentage of GitHub and this, this guy didn't.
Yeah. And, and also, um, I mean the other thing too is, yeah, I have a lot of those as well. Um, but, but I thought the Claude Code, because I'm just trying to really, really, really, really focus on that specifically. Yeah. Well, and also you want to give an example. Um, bro, I didn't make that chart.
Yeah. Or, or I think 4.6. I'm like, 'Hey, um, I want you to do in this style. This is the SemiAnalysis color scheme. This, uh, I like summarized books about visualization and like put in here little style tips.' Yeah. Here's some... Yeah. Um, I don't, I don't even know, man. It has like, has like, I had it go read like 70 books or something. I'm like, give me like, you know, the like...
It's probably a waste, like...
Part of it is a waste. Look, tokens are free. The, the cost of doing this is nothing. That's the part that's so amazing.
The cost of doing this is nothing. The information gathering and synthesis is like, hey, if it cost effectively the same doing 70 as three, who cares, right? And so I like whatever. And the answer I'm like, oh, this is too many tokens. You better like really summarize this into like 90 tokens or something like that, a really basic whatever. And then you have all the skill. But like, okay, now you can put all that into a skill of how to make charts, um, in the SemiAnalysis format using any kind of data. And then you can systematically just push this out again. I'm like, hey, data analyst, uh, please consider all the relationships you can and you generate information. Like I think it's, um, that one was not chat, that was not generated, that was not generated, which I hate honestly. I don't like that much, that one as much as it doesn't have the guidelines. Yeah.
And, um, and so you can just... that was, that was generated. And so you can just, what you can do is you just ask it to do is like, hey, here's all the data that we have. Can you like visually brainstorm with me a way to better represent this information? It's like, yeah, actually I'm going to generate you a timeline. Um, you can just do things. Um, and I, I mean it's, that, that is your catchphrase, right?
Yeah, that is my catchphrase right now. You can just do things. And, um, so people were looking at this from the perspective of people who are coding and they're like, hey, just programming is automated, right? But like all information work is, you know, I would argue coding is a big subset of all information work. I think there's a Brian Hobart tweet or something forever ago. He's like, you know, coding and financial, you know, finance people actually are very like different types of abstraction, but you know, you are doing abstraction. Excel is a ginormous abstraction. You're building these relationships and you're describing what you think a financial thing is worth, right? Um, I think coding is a little harder if I'm being honest with you. And you're telling me the hard one got automated. Why can't the easy one get automated? So I started to ask myself how much can we do? And the answer is it feels like a skill issue. It makes issue, it makes errors on the, on the margin, but you can kind of force it into like, for me, I love using rubrics, right? Hey, um, I care about XYZ, uh, out of 10, uh, score this. And then you can really do multiple things. It helps with the stochastic dance.
Do you, do you put it all in one prompt, like the, the, the task and the rubric for the task, or do you put the rubric after all the task is done?
I, I actually have two versions of this. I'm like, hey, you can pull all this stuff together, just run the, uh, the rubric or whatever. Or you can, uh, do the task and the rubric. It just depends on how you want to do it.
Yeah. Because obviously if you put it task and a rubric, then it can iterate itself. But if you put it after, then it's probably more like you pay attention to the, the rubric.
Yeah, exactly. And well, in the other part, part of it too, um, yeah, it will iterate, but like the context rot doesn't matter. I kind of like it to be separate because the thing is it's like, okay, it needs to be this like fresh look at it. You have to think of it kind of like it would perceive anything anywhere, right? It just each context window is just opening it up. And I think sometimes, um, if you have done, if you do it together, it co-mingles the information to the point where it becomes biased or susceptible. Opus 4.6, as you know, is like super sycophantic. Like it loves to like say yes, okay, yeah, I'll do this for you. Um, I think having it separate keeps it like, keeps some of that drift kind of away. And that's like one of the things that I've really personally, I like the results better. But it's, it's just complicated. Like part of this is really weird because I am, I'm weirdly now opinionated on taste in terms of how you should design things because you can, like for example, the context rot thing, until someone explained it, I was like, 'Oh my god, thank god someone said it. This is a huge deal.' Um, there's this like meme where it's like, uh, these guys, um, well, do you see the meme? It's like, um, Of Mice and Men, and at the end of, um, you know, at the end of the book, I can't remember which character shoots the other. I never read it.
Yeah. So, so one character shoots the other guy and it's like some guy made a meme about it being like, oh, this is after your, um, after your Claude Code is garbled, you know, 5 million tokens, you're like, okay, it's time to put you down. Um, because the context rot is huge. Um, so yeah, this, yeah, this is an example where... So what, what are your compaction practices? Do you sort of aggressively compact manually or...
Um, so I personally with the one new 1 mil, it, I feel like I try to do it all in one compact window. Um, I'm not doing ginormous projects.
The one is very new, right? Yeah, 1 mil very new. Okay. Very, uh, but it's a big deal too because, because your skills and whatever your CLAUDE.md as a percentage of 1 mil is so much smaller. So you just get so much more oomph, right? Because, um, the, the 200ks are just wiping over and over and over. Um, that's a big deal. I think it's a huge deal. And, and also with how the agents are working, the sub-agents will have their own context window and then the pasting kind of like really saves that, that big, you know, the 1 million. You just want a really high-quality, uh, project within that. That's the best in my opinion. Um, compacts just kind of start the compression of the noise. So, yeah. Yeah.
Uh, mentioning sub-agents and multi. So first of all, I wanted to give a shout-out to this thing from Anthropic Research where they were like, here's our production traffic. Um, and they, uh, they did a, they did a report that was kind of like their, their equivalent the meter chart. Um, and there's a lot of people saying that, oh, you should, you know, software engineering has PMF, but here's, here's the, the next list of everything else. But what if they're all also just software engineering, right? Like, like software engineering is like 50% right now, but what, like there's nothing stopping it from continuing to go to 80.
I think maybe what's going to happen, um, this like a, maybe a giant take.
It has like data analysis in here, which that's what you were doing. I...
Yeah, that's in my opinion, that is downstream of like, that is, um, so, so I think how we should think about it is software engineering might all be downstream of chips, which is downstream, like, like chips is upstream and then it's AI and then it's software engineering. It is all the extension of that same compute hierarchy. And I think the, like, you know, teaching where machine and, uh, code kind of inter, or and the world intermingle right now is code. And so that's just going to be the bleeding language that's used to, to figure out everything else. Um, that's, that's my belief. Like it doesn't make sense to build, like for example, this is a perfect example of this is like Excel, Claude for Excel is much worse than Claude Code using Python to use the Excel skills to then deposit into it. It's all much worse.
...even all the work they're doing.
Yes, 100%. Because if you think about it, it's a legacy. Why make a car engine fit into a horse carriage? It should just be in a car. It's like a backwards compatibility thing where it does work because LLMs are relatively generalizable like this, but why bother? Because that same abstraction of information on Excel is just that — because it's human-formatted for us to understand. And I think that's the important distinction. All of this information stuff, all this software stuff, is just to be consumed by humans. Doesn't matter. If they're just as good at putting the data together, we should be much more concerned about machine-focused software consumption. And so the LLM and the agents can put together and synthesize all the information and deposit it god knows however you want it to be. I don't need to make a chart in PowerPoint or Excel. It will just deposit matplotlib — matplotlib lib — matplotlib lib in a chart to me in an image. Fine. That...
Oh, you're trying to use matplotlib.
Why? Why? You know, it's better. It's better understanding that code.
So why ever make a chart again? If it — if it's better...
It's just like it could be inconsistent with the other charts that you do.
Yeah, I don't care that much about... I don't think we would care that much, but I think, one, our new charts are better than our old charts. And number two, I think if it increases the speed of information, that matters a lot. And so I think the new charts will outweigh the old charts because it'll just grow. So yeah, I think it is a little inconsistent. We have the same watermarking. Honestly, I think it's better than our old formatting anyways.
Well, the first thing this reminds me of is Bloomberg. I was like, you guys are just becoming Bloomberg, which is a nice view. A couple of things I wanted to sort of double-click on, because this is just a Claude Code brain dump for one of the biggest Claude Code shows in the world, which is subagents and agent swarms. I don't know if you've tried...
Pick either one you want.
I have a controversial opinion that Claude does not do RL on agent swarms or agent teams or something.
Yeah, it's just an experiment.
It's just an experiment. Thank you. Thank you. Because no — exactly. Because it's just via prompt and it's actually very bad. I think subagents are okay because they usually have a CLAUDE.md to go do whatever. But the agent team is actually really...
Okay, well we can't knock it, it's experimental. So — yeah, what did you try it on?
Well, it was like some big data analysis of many, many different companies with different KPIs into a dashboard, all in one. I was like, hey, can you just make this all — whatever — split up the teams, you know. Speaking of that though, you say that, but Kimi, Kimi 2.5's agent swarm is actually good. I have also tried that. It is actually really good. So I did some, like, examples of things that were never available to me, like internal benchmarking of these models, and be like, hey, here's a set of problems I'd like you to do 20 times. Can you do them, and then I can measure the performance between them and then do qualitative — like, what's the difference between X and Y? That was completely out of the hands of me, a normal guy, like three months ago. Okay. Now it is completely available to me. That's awesome. I care about this stuff and now I have the tools that are able to automate and do a lot of this stuff, because hey, all of software engineering is like partially automated. And so my experience is the 2.5 swarm actually improves the model's performance meaningfully. The agent team makes it meaningfully worse because there's clearly not RL done, so it isn't context-aware of what's the best thing to be done. And so I think it's interesting. I like subagents because it's usually a little bit cleaner on a task to go do it and then come back, but the agent team is just very...
They had some post about how they did stuff where it was... Yeah. There's a bunch of RL for this and I tried it myself. I thought it was like pretty — it's cute how they do all these little games and stuff.
Yeah. Also, it's crazy how the setup you have to — it's a lot of compute to just run the swarm. I think it's like a 16-node cluster of H100s. And you're just like, dang. So you and I are not going to be running that. And I'm sure there's concurrency available. But yeah, I think it's really cool and that's like — I think that's the sign of what's next, because you know these agents are going to get better to a certain extent. It's another benchmark to hill-climb, right? But then it's going to be how many of these together in a bigger chain can you get to work. You could argue it's kind of like a scale-out of the reasoning problem too. Hey, how do you get this one agent to essentially get a verified whatever, put it into a bigger process and do more information work? That's the next thing, and it's important to have context windows that don't garble up into random stuff and are able to do just good enough with token efficiency, I think is a huge part of that.
So yeah, that's kind of what our experiments have shown at least in terms of the agent swarm versus not. I think it's very clear the agent team out of Claude is an experiment. But Kimi — better, they'll definitely do better. But the Kimi 2.5 tells you that this is already, boom, perfectly great new place to do more work on, completely available to us right now. I think it's huge, because if these agents get any better, like, I don't know, I'm never going to sleep again.
Honestly, it's very interesting, this sort of Moonshot AI, and this is a tangent — we're not really going to focus on this very much — but you know how the AI tigers out of China were DeepSeek and Qwen, and then you were like, well, who are these Kimi guys, and these sort of newer names, I guess MiniMax as well would matter, and Z.ai has been around longer but only recently much more active. So I noticed that Kimi is much more in the productization phase, as opposed to the Qwens of the world, the DeepSeeks of the world, who don't really care that much.
I mean, Qwen, because of how it's attached to Alibaba, right? They have a way to productize it, but it's like the Gemini version — they have so much stuff to do elsewhere, right? But yeah, Kimi's pretty interesting.
They're pushing so hard. They got everything.
They got Kimi Manis. Kimi Claw. Kimi Claw.
Yeah, I know. Kimi Claw. Yeah, dude, I was gonna say — have you messed around with OpenClaw? Because I did. Oh, I remember, what was it first called?
Clawbot. Yeah, dude, I was going to say it was really, really euphoric. I was having it read all my emails and my calendar and do all this stuff and I was like, wait, this is really, really prompt-injectable, and this is pretty secure and important stuff. So I was like, you know, Claude Code psychosis is good enough for me at this point in time.
I mean, so what I do is I just have multiple emails, right? And there's a safer email to give to bots, and I can let it use that, and if it impresses me, then I can upgrade it. But Clawbot didn't impress me at the...
I'm honest with you, I wasn't impressed either.
That was the reason why people were freaking out about this model. I was like, bro, have you actually used this? Because right now on Claude Code in a relatively focused terminal, it will be like, oh, blah blah blah. Like dude, in the env there is an API I told you to use for this subclass of problems, and it's in your CLAUDE.md, like please focus up. It still is making mistakes. This is not truly AGI, and there is a harness, you still have to wrangle this thing, but it's not a perfect skill follower, and the context in each attention window is going to change, and sometimes it'll be lazy, sometimes it won't be, but it's definitely good enough to do a lot of information with.
Yeah, I was going to — so I use our Discord as basically a way to just bring information in and out. I just saw this too, where basically a lot of people are just setting up things that they could have done in Zapier with Clawbot, because they're like, well, now I'm like AGI, but actually they could just have done it more securely with Zapier.
Okay. I think it's kind of interesting. I guess I do think it's kind of interesting. But the difference though is Zapier. I mean, I remember I've tried to use Zapier before.
Yeah, and it's also not very...
It's also not very good. The difference though is — and that's okay. It's okay to be early to something and just wrong because you weren't the one that made it happen, right? Clawbot, Claude Code, Clawbot, whatever, all this stuff — the reason why it's so powerful is it gets to completion, right? And like, okay, Zapier, maybe you can get to completion all the time, but man, it probably took you like eight hours of clicking through things and copy-pasting crap to make sure it all works and it's all secure. And it's like, well, Clawbot did it, or Claude Code did it in like four and a half minutes, and that's good enough for me. That's a faster achievement. And so it's totally okay that they were right, but they were just not the right mechanism, right? You see this happen in information, like in the history of computing.
I think there's also an innovator's dilemma thing where Zapier, as a pre-existing business, had this view of the world of automations as very strict, on-rails workflow type things that their giant user base already uses. They couldn't really pivot that much.
So that's why I think, you know, one of the co-founders left because they were like, well, I can't exist within this — the box will control you. It's your golden handcuff. Yeah, it's just like your cage. You know, you're going to act like how you are in the cage. And so yeah, that sucks. Honestly, I feel like that sucks for...
The framing I have is like your priors become your prison.
That's pretty good. That's pretty good. That's pretty good. Your priors —
I haven't blogged that yet, but I should.
You should, you should. Your priors become your prison. I like that a lot.
Coming back to Claude Code. I also want to make this like the sort of Claude Code —
No, no, no. I want to indulge because that's how natural conversation goes, and I think people enjoy that, right? And probably that's the only time we'll talk about Kimi.
So do you use hooks? Do you give me the Doug Claude Code setup?
I have just essentially a few base skills, and then I have a lot of APIs, and then we've also made sure to work — and this is all work in progress as well — to have APIs for some of the SemiAnalysis information out as well. And so that way we have an internal server that is accessed by people with an API, so that all the SemiAnalysis researchers are able to hit some basic level of context, because I think context is really what matters. I'm too dumb to be really smart in order to have — well, I guess I do have some hooks, if it makes sense, in terms of like...
I think hooks are very underrated, right? Because you can do like a Ralph loop just with a hook.
Yeah, I feel like I underutilize hooks. I think that is true. But I do run some version of them on skill calls effectively — like, hey, on this, then you have to start pulling all this stuff. But I think in the beginning I tried to do all this hook stuff and compounded stuff like that, and I found that, you know, the Gastown Ralph loop era, it is a sign of what will come, but I just don't think there's enough fidelity to make crazy multi-turn something happen. So like, okay, actually less is more. Try to have a strong set of smaller skills with a good amount of context information to be pulled in, and then at the beginning of every session ask and focus on what you want to do, so that it prompts the — not like a Claude within a Claude, whatever — so here's the goal to finish within this single context window, and then get it done.
And this is like my generalized research thing. Hey, I want to look at the price of NAND since 1984 or something like that. This is what I want to — actually no, let me just give you the best example that is probably not going to work. I would like to fine-tune a time-series foundation model to predict NAND and DRAM prices. Okay, I'm going to first start by gathering as much information as possible from all this stuff, blah blah blah, and then we're going to fine-tune it, evaluate which ones we're going to do. I chose Kronos 2 because of covariants, blah blah blah, try to set this whole project up. We'll make a Vercel dashboard internally for SemiAnalysis. Maybe we'll external it if we want, if it's a good enough product.
Okay, so then it does all this stuff and then I just start plowing away. Hey, can you go research — here's a search API, Serper or Exa or whatever you want to use — to go look for all these different information sources and then bring it together, right? So this agent goes and gathers all this information. This agent goes and works on considering the fact that the price isn't perfect to do all this fine-tuning on, and then we throw it in. I also had it like, well, what do I use? It showed me which GPU, whatever, we're renting on an hourly basis. And so yeah, we just pull all this stuff together and we fine-tune it, and I'm like, okay, cool, how did this work? And then we just have this constant iterative loop until I try to finish something. I got to the point where I was like, okay, this time-series LLM is probably not going to work, unfortunately.
You said it was because of regimes or something else.
I think so, regimes, yeah, there's no way — it's so messed up. For a lot of people who are new to finance, this is why I have an issue with all these kids doing stock trading games with LLMs. They have no idea. They've never studied finance. Past does predict the future a lot until something fundamental changes, and the macro shifts, like risk-on versus risk-off. They've never heard those terms. I had to explain it to people at Cognition, and yeah, the rules invert, they completely invert. What used to work is exactly the opposite of what you need to do when you have a regime change.
Exactly. And it's very, very, very hard because — and the other thing too is, you'd be like, okay, each of these are almost like a one-off onto their own...
Right, which reduces your sample size.
Yeah, which reduces your sample size. And so then at the end of the day, you end up being like, well, it kind of just — I guess it's, here's some heuristics, good luck, have fun, right? Here's your checklist to see it might be over, but you really don't know anything until then. So, but like, okay, an example of where this project was helpful is, okay, I'm not going to have the magic LLM tell me what the price of memory is going to be. Hey, it was a good weekend project and I did burn quite a few tokens, but I do happen to have, after all this information synthesis and analysis, all of the memory prices of everything I could possibly find, plus the things behind API that we paid for, plus enhanced data sources, and I have all the covariants, so hey, what was the consumer sentiment, every macro thing of all time.
And you know what's really interesting is I'm going to just be like, okay, well now can you go make a summary of each and every memory regime and what it looked like and what created the beginning, middle, end, and put that in a dashboard so it's relatable and easy, sharable, consumable within my firm and company. Yes. I'll probably be done with that today. And okay, so you're like, well, that's just gathering, doing information stuff, like you don't understand. No one's ever done that in the history of time. I know for a fact as the guy who is the cycle semiconductor guy, I've written and done more work on the cycles than I think anyone else has at this point, especially for the older ones like the '80s and '90s and 2000s and 2010s.
And when I did it the first time, the human grind is I went and I read these old annual reports and I put it together and I tried to string a narrative through it and I brought through all — I'm like, okay, what was GDP this year? What was this year? What was all this stuff? You have to make all this giant sheet to whatever and then make the narratives. No, none of that, dude. I mean, this is too much information to gather. It's a lifetime of work. It's like a PhD project. I did it in a day. Two days.
Yeah. I mean, I think the pushback would be that then you don't have enough expert information to criticize the reasoning that went into the report that you're slopping out, you know.
There is some slop. I definitely agree with the sloppiness. So I think of it once. So, right now —
By the way, that's also essential for you guys. If you get caught putting on some slop to your clients, right? You have to — at one point be extremely AI-pilled and, you know, number one in the world and applying AI to your productivity. Great. But also you got to, you have to —
So I think the thing that's really interesting is this whole thing is like a game of hygiene now, because I think this is really hard, and I think about it all the time. I feel very comfortable with doing all this work because the thing is, at the end of the day, since I've done the work, I have a lot of embeddings in my brain, a lot of information. The vibes that have got me in here is actually tons and tons of information, set-up scenarios, and pattern recognition, right? But yeah, you're right. This crap makes mistakes all the time. All the time. It is still just — I think of it once again as a junior analyst, right? The analyst goes and does all this really pain-in-the-ass information, and you bring it all together to make a good decision at the top.
But the problem is, historically what happens is that junior analyst, who I once was, went and gathered all that information, and after doing this enough times there's a meta-level thinking that's happening where it's like, okay, here is what I really understand and how this type of analysis I'm an expert in, actually I'm very good at, I consistently have a hit rate, now I'm the expert, right? I don't think that meta-level learning is there yet. We'll see if LLMs do it, right? Everyone who's spending one quadrillion dollars in the world thinks it will. It better happen, right? If you're spending a trillion dollars and there's not meta-level learning. But for me, in our firm, that massively amplifies everyone who is an expert, right? And we are a firm filled with experts. And so it's this hard part where I wonder if new people — we will be less lenient in terms of how much AI tools...
Are you like junior or new to the firm?
Junior. And because you have to still do some of that. You can't just slop it up. It's very obvious to me what is slopped, right? When it's slopped and there's no cognition, then whatever the artisanal last 5% is, that really matters. But for me, I know inherently what the 5% is. I can write it away with some really easy heuristics and time and be like, okay, well this is the last 5% you fix. This is what I believe. Just make up these assumptions instead. Press enter. Okay, cool, we're good to go. And so that's kind of the hard part. That's a real hard part. There is still a human in the loop right now.
One day, someday it'll be superhuman, but where we're at today, it's not there. You'll just compound all this noise and it becomes just garbled, just context rot. But in terms of the capability, the human CPU in this agentic swarm is very, very powerful now, a huge, huge, huge multiplier of what you're able to do. And for me, that was enough to be like, I feel AGI-pilled honestly, because if I define AGI as many common jobs — I'm not doing ASI, that's like religion — can it automate or change or completely shift a lot of the information work?
Yes, 100%. Like data analysis is a perfect example. Hey, every quarter I want you to just find me some examples of some information that might be interesting. I just can't imagine if I was an entry-level worker doing data analysis that an average 22-year-old would murder the hell out of a relatively well-thought-out agentic system. And so you're like, yeah, that job actually does seem at risk. And so, yeah, the 4.5 capability — we hit some level where it seems to work and do bigger information work, that's when I'm like, okay, yeah, this does change everything.
And so, yeah, there's all kinds of mistakes. It's a new level of hygiene that we have to do. You're going to have to understand what the output of agentic work is back to you, right? I catch it making errors all the time. It doesn't always pull skills. You can definitely tell context windows get dumber over time. It's not AGI today, but it can do these crazy long tasks, and as long as you finish it at the end and deposit it as information work, that's very valuable.
Yeah, amazing. So you do a lot of client visits obviously. By the way, Transistor Radio, amazing for understanding what your world is like.
Are you also Claude-coding your analysts and, you know, on the other side?
I've definitely Claude-coded the analysts. Everyone in the New York office must try it. I really tried pilling.
Not — I mean, not your SemiAnalysis customers and all that. My perception is they don't adopt any of this stuff.
Okay. So yes and no. Some people are interested, but you have to remember it's relatively more conservative. But if you ask any analyst if they're using AI, every single one of them will tell you yes, I use it every single day. Of course, how could I not? This is a vital skill. And so the basic inference I'm making is: I am a bleeding-edge adopter, I'm a relatively smart dude who knows what he's doing and if a tool is useful or not. I've evaluated the tool and I'm like, wow, this is an amazing tool that I would literally pry out of my dead cold hands. Even if it makes mistakes, I will be using this for all kinds of work forever. Then I look around to everyone else and I'm like, most of these guys are enough like me that if they have an opportunity and an edge, they will obviously apply it. They look at this tool and they start to use it. If they're thinking like me, they're going to obviously adopt it. I'm like, well, I don't understand why everyone doesn't adopt it. I would argue — we'll see in the 24-month view — it will be a base level. I think Claude Code, cowork, whatever, is going to be a base level of all information work very soon.
And you know, you see one — my friend was telling me how his portfolio manager found cowork and he's getting it to read his emails and he's like, oh my god, I love this, right? Everyone's moment is going to be a little different, but I think my moment — it feels like GPT 3.5 or 4 for me, where there's that first time where you're like, okay, I know it made some stuff up, but this is better than if I went for hours searching, putting information together. And then it also has the analogy power, you know, where you can say, hey, this is the setup, can you describe it in these really strong pattern-matching skills that are really powerful? I just think it hits some level of capability. I can't tell you what it is. It is my personal taste where I'm like, oh wow, this is completely over the chasm of what needs to happen for it to be a very, very powerful tool. And so yeah, that's my Claude Code moment.
There's some kind of automation chart that, you know, X has this automation and I think we need a version of this that is the Claude Code like —
It needs to be much — but what's crazy is this Claude Code thing murders the axis.
Exactly. It just shifts everything, right off the axis.
Yeah. But also, what I was trying to figure out is, well, okay, it is maybe dumber, less human attention, but because you can spin it up so quickly and it can spin parallel so quickly and it gets done, you get more turns at the wheel.
Whereas as a human you get one turn. You get one turn, but with Claude maybe you get three turns and the sort of review process is the thinking.
And you just need to get very good at review or hygiene.
Yeah, I think of it as hygiene. The thing that's really going to be painful though is a lot of my expert opinion has been built by pre-phones — your attention span, you know, the children are cooked, okay, the attention spans are really bad, all this stuff. I read this really sad thing like, oh, we're getting dumber or something, first generation. I don't know, I'm not going to — maybe that's like...
You see the Coinbase earnings, all the coins.
So like you have this thing where it's like, okay, and it's cute and all, but it's such an addictive technology that I feel very grateful that I'm like, well, I understand what I'm doing, have this history of doing stuff and able to apply a tool. But people who are riding this curve, it's going to be very dangerous. It's like giving everyone right now — I'm strong —
I think you should just do that. Well, we do deal with some of the SemiAnalysis memes, you know, and the thing is, you say some of this brain rot is so bad, which it is terrible, but some of it is also hitting some attention mechanism in my deep primordial monkey brain.
Yeah, it's stimming me and you're like, I can't look away from the Subway Surfers.
You couldn't look away. I had to pause it.
Yeah, yeah. Hey, have you ever been at a bar where they play like these weird — there'll be like TikTok videos for lack of whatever, and you just watch — TikTok bars in New York?
No, not TikTok bars. Not TikTok. Okay.
It's like there's essentially a B-roll channel that they'll sometimes play in public spaces and you will just find yourself engaged with it. Like there are certain things that just work. So yeah, sorry, that's completely off. But I wonder — this Claude Code pill is very powerful for me. I believe it will shift all of that over massively, the chart. But it's just really weird because if you didn't pay any human cognition to get there, I don't think you're going to be a great reviewer. One of the reasons what makes that human feed that loop well is because once upon a time you did that and you could make that — like, yeah, idiot, you're not thinking about this problem in this way. You're missing this. You're not considering this 90%, the 10% tail, something like that.
And so it's like, yeah, I know you said this, but I know that I told you the valuation is the only thing that matters, but it's also a fraud. You can't do both, right? If you think about the analysis stuff, you have to know when your own personal embedded model is like, yeah, actually this one overwrites this one. That's through learned experience. And I wonder if we're just reviewing, we won't be building and embedding those assumptions to understand judgment.
Right, right. Because you're just checking for mistakes rather than trying to do original thought by just doing the work.
Yeah, I think that is a danger.
Yeah, and that's what hygiene sounds like to me. Like, hey, it's really addicting to be like, whatever, press the button over and over and over, but sometimes you do actually have to think. So I think that's going to be really interesting. I mean, have you tried — so the way to model the sort of meta-learning element is like once a night you do a batch job of like, look over everything I've done, extract some learnings. And OpenClaw, I think one of the interesting things I really liked about it was this heartbeat. And I'm like, people aren't excited enough about this because this is the first instance where the agents are just always on, always living, always reflecting.
Yes, and like what is it — CLAUDE.md story, I think, is much more for character and whatever, but yeah, heartbeat is heartbeat. The heartbeat is the crown.
Yeah, I mean, yeah, that's a good way to put it. And so that's the powerful thing about all this stuff, is that, okay, yes, we know that the context gets garbled. We know that OpenClaw doesn't always do everything you asked it to do initially. But you can see the design patterns — like the heartbeat CLAUDE.md is a perfect example — you can see the design patterns where it's like, well, is all of our tasks every single day actually us having this genius thing, or do we sit down in a single session, finish a single project, get up and get some coffee, then come back —
— if it's that, and you can just make the heartbeat CLAUDE.md consider the session-to-session and like, hey, meta-learning, all this stuff, and it's only specialized and focused on one form of doing something. So it actually does have a context of all the — like, I'm thinking like a customer service agent or something like that. It does have the context, in fact it can look at every single time it's ever happened. That's actually information and context no human could ever hold. And you're like, wait, that feels effectively good enough to do a huge information test and have enough context and be able to fetch it, and maybe there would be some verification to make sure it doesn't just totally mess it up. But that to me feels like a design pattern that you can build something on.
And so that's the vibe, is that we've hit some capability that you can build these much bigger blocks now. And those bigger blocks are not just this single line of code. It might actually be a business. It's kind of crazy. I wouldn't have put myself as AGI-pilled. I think 4.5 is like actually —
I think my own timelines have moved up a lot.
Are you guys watching GDPval?
To the best I can, but I'm mostly just trying to —
No, no, no. So to me, when GDPval came out — I mean, I'll just — GDPval is basically a broader SWE-bench, let's call it, where it's applied on every discipline, every profession that is white-collar that you can model, and it's above something like 2 to 5% of GDP, something like that, that's why it's called GDPval. And they had human experts do the task as well as GPTs, and here's the results, where 50% is parity with an industry expert.
Exactly. Where you can see the nice increase from 4.0 to 4.1, and since then obviously 52 and Opus 4.5 have already exceeded that, we're at 70-ish now.
Which means models are consistently better than industry experts at these things.
So to me, this is the AGI definition, isn't it?
Yeah. Yeah. This is — and so I think the problem though — yeah, I would say that that is the definition. So the thing that's crazy is because there's this ASI element that people are really, really focused on.
Yeah, we're moving the goalpost.
Yeah, we're moving the goalpost. But I'm like, bro, the goalpost — I mean, we'll see if this is actually the machine god and Shoggoth will come and talk to us and vibrate on our same —
Okay. I don't — I'm being honest with you. I'm very open. I will change my mind often. This is not something I feel intuitive in my gut today. Maybe it's the next next next thing, but when it comes to the GDPval version of this, yes.
Yeah, this does white-collar work, the white-collar work which is most of the —
Knowledge — actually it's almost all — not almost all, but it's a huge portion of all work in the world. It's like now we just made — my favorite stat is like once upon a time 90% of people were farming, right now today less than 1% of people are farmers. It's kind of like this crazy shift where technology is going to massively change the relationship with all of that and it's going to be like this 99-to-1 thing. I don't know if it'll be quite that drastic or whatever. Maybe everyone's just doing leisure. So far my experience is everyone just works harder. Been my experience. But it just feels like a massive moment's happened, like the steam engine's invented and the trains are here and everything's going to change in knowledge work.
And it's kind of crazy. There's a sort of economic cycle from my macro days that I can't remember the name, I can't look it up, but it's basically like there's this stages of economic development where your economy starts out majority agriculture, then it discovers manufacturing, then it discovers white-collar work, then it builds a very mature financial sector. And these are like a layer cake all declining over time, then the new things increasing. So my theory is there's this fifth layer that has to open up that starts to happen, because I do fundamentally believe we just invent new work.
I do believe that. Yeah, 100%. Humans are very adaptable. That's my favorite thing I've learned. You're able to adapt to the coldest place in the entire world, the warmest place. Humans are in every latitude. That's in a physical sense, but I think we're going to find a way to make civilization go up. But we'll invent more work for sure. But I think the thing that's crazy is just things change so quickly, and that five-to-ten-year period, that ten-year gap can be drastic and crazy, and that's just society.
Wild. But yeah, it's happening in our lifetimes.
It's happening in our lifetime, like it's happening right now. It's just really crazy. And this is like a complete side task on —
I'm really curious when we start to see it in a much bigger way in the real economy. That's my pet —
Yeah. Why is it not showing up in GDP yet? Right?
So there's going to be, you know, some people are going to be like, oh, you know, the facts facts of the internet, same thing, information transfer, whatever. I think I'm actually scared for a third worst thing, which is like now now this is a complete crackpot theory. Please don't hold me to this internet. But what if AI is massively deflationary? And and also I think one of the more interesting conversations I've had in a bit is like what was GDP was invented once upon a time as a way to figure out how much we could divert, you know, normal economy away just to war during World War something like that. Okay.
My spiciest take is I feel like GDP itself is going to be very very challenged by AI because information work. Yeah. So how we how we capture it effectively is all of an economic good and then the service hours divided by hours. Okay? So there isn't like a widget to widget difference. But in theory, if we could break all of information work down into units, we're going to have a lot more information work for sure. Like more work will be done. I don't know what the value of that's going to be. Is it going to be so much increase in supply it's deflationary? That seems to be like a real concern. It's possible. Yeah. And then we'll figure out how to use it. But like there may be a great depression of AI where like we figure it out.
Yeah. Well, I I wrote this whole thing about railroad stuff cuz it's my favorite my my favorite capital. It's on fab. Yeah. Um I can't uh it's like railroad fab. Uh it's about all the railroad stuff over time. Okay. Pretty much because um we're like everyone was first looking for the internet. We've well massively passed the internet in terms of the absolute size of the buildout. It's not even close. Like we
What numbers are you thinking about?
Um I think a trillion was a trillion all in was essentially the real dollars version and I think we are well past like we like whatever this year and it's cumulative right we were well past that. I think railroads the reason why it's so interesting is because um honestly it's way crazier but but part of the problem and craziness of it too is like railroad was literally like one of the first added layers of the layer cake if you think about it before it was agriculture and railroad was like okay well how do we move this agriculture around faster um and then banking got I I kid you not like one of my big takeaways is banking effectively got invented by railroads.
Because there's no need to finance it.
Finance it yeah so much money was needed that like effectively 85% of all paper whatever was essentially just railroad debt.
Yeah. Um, one of my favorite anecdotes was before there was a Federal Bank, uh, a Federal Reserve, Andrew Carnegie was the Federal Reserve.
Yes. Yeah. There were individuals. Yes. Um, yeah. And so all this stuff, uh, so it's like the whole thing, I kind of did some work on the gilded age, all this stuff, but like my takeaway is like that was a really interesting cycle because it was so big and took so long to deploy. It actually was 45 years of like there's three cycles actually. There's three boom busts. Damn. Um I don't know if it'll be quite that long. All the cycles kind of collapse. Um that you know information moves around faster.
And so you you have all this stuff where I think it's going to happen faster but like I would be really shocked if it was all in one go.
That's my vibe. Um where it's like it's all in one instantaneous up down.
I think it's going to look like mult some multiple cycles. So yeah kind of just wrote about railroads. The there was like a baby railroad cycle then there was a huge railroad cycle. The modern rails invented out of it. That's like my favorite analogy for this because like I think it was like GDP percentage of capex each year were like high single digits for sustained for like 10 years. But what's crazy is like that amount of spend is like we're we're like well on track for that.
Did you do percent of GDP? Cuz I think that's I think the the way you make it convertible.
Uh Stargate itself 2% of US GDP. Um and uh I mean it's going to go up like
Yeah. Yeah. That's a Yeah. And it's not all going to be in one year, right? But um it's okay. So yeah, so total capex four it was 4.8% of GNP and 25% of total gross fixed capital investment.
So 25% of investment every year and four 5% of GNP.
I think we're there. You know, Stargate plus Anthropic plus whatever. We're We're right there. XAI.
Yeah. So, we're at the railroad buildout, which is like at one point like But the thing is crazy.
Yeah. No, no, not probably. Like we should, but Okay. Like this is bigger.
Okay. I I'm I'm like I I would like to say Yeah, sure. I Yeah, we we will do it. I I'm worrying. I mean, like, dude, where are we going to get all the money? That's like such a like the pedestrian concern. Yeah, it's not a pedestrian. I mean, that's what happens every capital. I'm worried like we must hands in the Middle East. You'll flip the thing. We must we must what this happens every single time. That's the reason why the like bubbles happen, right? Is like we essentially get so big where it's like this must be built. It doesn't matter the price and then all of a sudden we look at it, we're like, 'Oo, that was a steep ass price.' Um, but I think I mean the thing I think about this is like how I think about the big picture is there's a demand curve and a supply curve and we have no idea when they cross. They will cross one day and every single year the demand then we're finding that demand curve and then the supply curve we're just like we're doing our best to deploy it. And I think for me, like I don't know when that number is. I'm not I don't want to say a number go up forever because I feel like that's like intellectually dishonest, but Claude Code for me is the first time where I'm like and we're bringing you all back together where you're like demand go up so much. I am now guzzling as an individual. Like for example, I'm we're we're off I'm off Max. It's not enough. It's not even anywhere near enough. Like I mean some people buy like five Maxes and then they
Yeah. So, so I so I'm on fast I'm on fast with 1 million on API which is that is like an addiction level if any sense. Um but yeah I I I really think um it's the first time we're like okay well actually how much is this worth to me on a yearly basis? I think it's like 20 to $30,000 easily if not more. Like I don't understand like what's the like I can't price it. I have no idea the elasticity.
Yeah. You pay for a perfectly compliant junior analyst. Yeah.
Right. And so what's able to that cost like get 90k
That's able to work in parallel? Yeah. Like you can have a 100 of them. It's kind of crazy. Yeah. So
It's it's a skill issue if you cannot manage a junior analyst that is 20k a year.
Which like I mean okay like you know skill issue is like it's your fault but no like we have to learn how to do this. It's it's
Yeah. It's like it's it's two it's three months old.
Exactly. It is two month that that's the correct way to put it.
It's like it's it was definitely a skill issue that you didn't know how to get like your settings on your iPhone to work. We know at one point one point in time, but like in the very first month of us having it, no one's going to be like, 'Yeah, you idiot, you rube.' You don't know how to use your completely new technology that got birthed last month. Um I think it's just about a it's a bit of time. And it's like kind of interesting cuz you're like watching I mean it's cool is that if you're like on this absolute bleeding edge you get to see the design patterns like blossom in real time and like um we have this like really old uh older guy who's like been through the history of technology since like forever back then. He's like one of the most interesting uh intelligent people SemiAnalysis and he talks about how
Uh, Tanj. Okay. So, like you said this like we we we had this conversation one time. He's talking about like early internet how like it wasn't actually sure if the browser was going to win. It was like a remote web file service. Some people thought like, well, it just I'm just going to reach and play with someone else's web files remotely, right? Who knows, right? And that kind of, you know, it kind of is a remote web file. Who who the hell knows? They were design pattern searching back then. And I think we're at that again where all the design patterns are open. And it's like really interesting because there's many different ways this could go. And we're gonna have to kind of collectively agree what's the best set of hygiene, set of design patterns, what's the level of abstraction, and then like all the rest of how much SaaS will disrupt all everything else. Who the hell knows? But you get to watch it like front row seat right now.
Yeah. Yeah. I mean, my biggest one uh and I I I do want to bring it to Semi in a little bit, but uh is the IDE. Two months ago they uh we had Steve from Gastown talk about how 66 would be your IDE died and I like two weeks ago three weeks ago I recently was like he's absolutely right.
I'm really wondering too because like IDE so so I think that same
My like my the reason why I'm so excited about this is I get to like look I never my daily driver was never an IDE right my daily driver was like Bloomberg or Excel or something like that But I have a personal belief it's not happening yet because we're not quite there in the maturity curve like software is just going to be first. But the year like of you know Excel is dead for finance and like it's it's it's
Excel is the IDE for analysts.
