Ep. 25 - DYLAN IS HERE, LIVE! | Dylan Patel & Jordan Nanos

We're gonna have a whole video of you walking in yelling, "I'm so excited." - Oh really? - Yeah. - We have so much fun on these bar plans. - I don't know. - Did you see the one last week? - I had gotten feedback though. So you wanna start this podcast here? - Yeah, let's do feedback. We cut out something that was going to be in the cold those last time where I said, "I'm no longer listening to the comments "because on one episode with Doug and. " - Oh, we'll be talking. - "Just with Doug." - "Just with Doug." - Oh, did that? - They just ate us all that. - Yeah, that one clip. - Oh really? - And he cut back. He cut out me pushing back on them. - Yeah, yeah, 'cause I don't have that. - I fully was like, okay, they built maps in house, they built Gmail, Google Drive, like the whole G Suite. They built all of GCP. - Dude, I don't think you understand. - Cooper Netty's, obviously. - All my deep-mind friends, there's like three of them were like, yeah, I think I'm gonna leave. - Yes. - And I've got a bunch of others who like, fuck you guys. That's fuck you guys. - Well, it sounds like they're in stage two. - Stage two, yeah. - I don't know. - I don't know. - I only know Cope. So back in the form, Warrior Days, we were gonna chip our demons. And there was a private discord where there's a bunch of people who loved anime and people who were all around the world and many of whom were racist 'cause it's anonymous people on the internet. But they all fucking loved anime. I did not like anime. I never really watched it. Besides one guy who I dated, I watched anime with her, but besides that, they would always, no, no, I've never dated anyone, I'm a pure. So what's anime? - Which anime did you watch? - Oh, I watched, okay, there's one I love. I love Spy X family. - And yet? - Sure. - And yet, Chan? - I don't forget anything. - I don't know. - Michelle, do you know the reference? - Yeah, I do. - Who does? - One of the dolls. - You bought me an Anya? - No, like what about the-- - What is it, a labo? - A lore? - Yeah, a lore from that series. - What's the, what's the, what's the, what's the, the floor? What's his name? - I don't know, I can't go. - Lollie, there ain't no less anime than you, man. - Well yeah, so we're changing topics rapidly. Jordan Nano's, and this is like HR approved, 'cause I'm HR. Jordan Nano's is the hottest man in semi-naucus. - Got this. - Michelle, got this shit. - Oh, fuck. - He's, he's, no, no, no, think about it. Think about it. Look at him, like fucking I'm fat and look at him. Like, you know, he, he look at him. So beautiful. Tall as fuck. Same age as me, except he's married and has a kid, owns a home. He's not a degenerate, like, you know, like, this is just like, wow. Oh, goals. - Thanks for the employment, man. (laughing) - Okay, so sorry, going back, Google, Google people were mad at us. They were DMing me and some of them were like, you know, they're in cope. But regardless, the feedback I've gotten from mostly just my, my own head. My own head. Are we giving away too much value? That's why I came on today. Because I need to destroy value. (laughing) All right, we can stop. (laughing) - No, no, no, no, stop, it's fun, it's fun. But someone on the team in turn, it was like, Dylan, we got a lot of value away on the weekly and I'm like, oh, we do. I haven't listened to it, but we do, I bet. I listened to the one where we had the DG Matrix guy and I was like, this is fire as fuck. Yeah, who said that? - Pearl, come on, HR's, HR's, - Doug? - for Texan on immediate, and it wasn't Doug. - Okay, Jeremy? - It wasn't Doug. - So I don't know who is the say. Is it just somebody who's, - No, literally. - What the whole set that they haven't been on yet? - No, no, someone's been on. - 'Cause someone's been on. - Dan? - I don't wanna say, Dan wouldn't say that. Dan's a sweetheart. Anyways, regardless. - Who doesn't say? - The feedback is that this podcast is too good. - Yeah, okay. (laughing) - And why are we giving away from me? (laughing) - Oh shit. Anyways. - Well, yeah, yeah, we can definitely put in the toilet this time with you. All alpha. - People click because I'm on, they're like, what the fuck is this trash? - No, we're gonna have a nice picture with you with a neon orange shirt, ready to attract all the clicks. - Yeah, come show your shirt. So, so. - Was Nick, who said we're giving away too much alpha? - No, no, no, look at Nick. - Oh, he was David. - No, it wasn't David. - Wasn't it still? - All right, good to see you then. Thanks for coming by. - Thanks for coming by. - Let's talk about the, like, semi-analysis office in New York, man, it's heavy been yet? - No, that's one go. - Why would you go? (laughing) - You can fix going back to the, to the hobble. - We're upgrading. - We're upgrading, we're upgrading. - Oh, when? - Soon, very soon. - This is like buying GPUs, right? If you make too long of a commitment to the lease, then you have to find a way to resell us. - I wish you health. (laughing) You gotta make six month office commitments so that you can outgrow them. - I should just buy GPUs. - Instead of more office space? - Instead of a towel. - Yes, yes. - So you wish that semi-analysis would just- - A GPU resell. - I would pull your employees. - No, no, no. - Your employees just all. - That's like, we're gonna automate away all the- - No, that's not possible because- - Brother, go look at the fucking AI spend. - I'm not doing it. - I'm not doing it. - What's going faster? What's going faster? AI spend or spend on employees? - Well, so the thing was like we've gone through like hiring sprees and then digestion periods and hiring sprees. We're back in a hiring spree, so. - Yeah. - Copper rocks, copper rocks. - So spend on employees really skyrocket, especially in the second half of last year and parts of this year, but then the first quarter of this year, AI spend skyrocket. But it's actually been relatively flat and cute too. Right, we kind of, everyone got cloud code psychosis and then it's like leveled out. - Yeah. - It's still at that 10 million number, roughly. - Do you think it will grow roughly in line with more employees in the future? - I, you know, I was surprised, fabled and caused price to go up. - Yeah. - Spend to go up. - Yeah. - What do you think that is? - Roughly the same as Opus. I'd say possibly is counteracting with, a lot of people were building the first versions of the applications, like we went from 10 repos internally to like we have over 150 repos internally right now. - Should we sell our code, our data? - We are. - No, no, no, no, no, no. - Like sell it to like the labs to trade on. - We are, our Slop code, our Slop code. - Slop code. - That's when the thing need more model, output Slop. - Yeah, I mean the models themselves could be sold as data. - Yeah, yeah, okay, so you think it's because everyone was doing MVPs. - And that was maintenance mode for a lot of it. - But like the spend is consistent. It's not like it's gone down after we had this one time spend. - No, for sure, yeah. But I just think that there's no more, like there's only one time when you onboard somebody to learning how to use the data center model and do research for building data into the data center model and building documents. And then once they're onboarded, you know, it's like a speaker and then it levelizes. - Well, that or possibly were lacking new features in codecs that will allow us to spend more to be more productive. - Yeah. - Once there's an agent's form where you can manage a million different concurrent agents and they all work together instead of nine today. People, single power users will be able to spend more than they could. - Cool, I guess one of the things I'm not counting, so our spend cost does not accurately account for cloud tags. I think our dashboard doesn't show that. So actually that's a good point and computer. - Yeah, good question. - I think both of those don't actually get counted into the spend. So actually our dashboard's probably wrong. - Yeah. - Probably someone that I monitor. The way I think of it is like a lot of this code stuff is actually like the amount of AI we use on a continuous basis is actually very small. It's actually just like people doing new work always. Which then because we have enough people, it kind of levels out to be like a pretty steady amount of spend. This swings are only like 20, 30% a day up or down. Sometimes Jeremy will be like a fourth of this spend and then sometimes there'll be nothing. - Yeah. - Right? But then someone else picks up for the slack, right? One of your guys is like, "Hey, what the fuck is he spending on?" And he's like, "No, I was like, well, blah blah blah, "I'm like, is there ROI?" And you like list out all this shit. And my grade okay, cool. - You'd even say great, cool. - Okay, I didn't mentally. - I first spotted with all this detail. Like, should I give him feedback now? - No, no, sorry, sorry, sorry. I should've said, yes, this is fine. - Cool. I just read it and I was like, cool. - Internally at least. - Those were only nervous because this was a person who's ostensibly an intern. - Yes. You know, he doesn't have my trust yet, you know? - Yeah, yeah. - Like if you spent 20K in a day. - He did not spend 20K in a day. But he's filled with 8K in a day. - He's from like four days straight. Which was like, okay, like that's a lot. Like what are you building, right? Like, but if you spent 20K in a day, I don't fucking question you. I'm not gonna question you. Like I just assume you're gonna do stuff. As long as you deliver this grade, then great. - Okay, and what's shocking to me is I didn't know he was spending that much. And then we'd look in the dashboard and I'm like, well, this guy is as productive as any of the full-time employees right now. On that stuff. So it was a reality check. So as this, you know, when we do, 'cause in the past bonuses at this company were vibes based, you know? Basically I just vived out the bonus number and it was cool. This year, Claude is gonna have to go through or Codex or we can have two reviewers, right? Two internal performance reviewers and Claudex goes scrape through all of this slack, all the get hubs and say, what did they do? And then connect it into like sort of the like. We're gonna delegate this. I just make it a sub. - I don't know. - I discussed with Michelle yesterday peer reviews and I was like, "Homely." And then after I said it, I was like, "Oh, man, "you wanna go big tech on this? "360 reviews, man?" - Not 360, not 360. - You just a little bit, you know? And then the other thing that we discussed was like, - We're gonna have people reviewing with their skip, which is true. - I said a US-based recruiter and in the admin channel and Doug flipped out. He's like, "Oh my God, hallelujah, finally! "We can have it!" He's been wanting HR since like 30 people. (laughing) Anyways, yeah. - Wait, you think HR is a recruiter? (laughing) - Yes indeed. - Yes indeed. - All right. Anyway, so the concept or thought process was basically like, a lot of the spend is one time R&D. And actually the steady state spend is really low. The thing is we just keep doing new things. And so that helped me know, translates to revenue in either a nebulous way, in the case of cluster max and inference acts, or a non nebulous way, in the case of the energy model, which is super fucking crack now. Or dashboards and all these other things. So like different scraping methodologies. So the thought process was like, if we're looking at these companies that are AI roll ups, right? Hey, let's take an existing company, let's completely destroy its cost structure, nuke its cost structure by just making it efficient with AI. What does that look like? Let's say private equity companies, they buy a company. And right now they just squeeze the rag and discard it and make the American populist. - Yeah, AI for efficiency has never made sense to me because the way that I use AI and the way that we use AI is very much about research, which is completely inefficient. - No, I mean, but the flip side is like, we had agents go through all of the invoices we've sent out. And there was like, and we've been paid because our manual process is literally like, certain deals aren't tagged. Invoiced properly and things like that. Or like we've had some, you know, like a lot of the ticket stuff is at least somewhat more efficient because AI is answering it, but now they're not sending it to the customer but like they're pulling through all our data and they're like, here's the answer. And then the analyst, I think that makes this port time per ticket shorter. And so I think like things are helping us make be more efficient. Surely, no? I don't think that's the primary use case for us. - I guess like cluster max this time, the depth breadth and amount of testing you're doing versus last cluster max two point on especially, it's like. - Yeah, you can frame that as efficiency, but when I hear private equity takeover company and ring this halodry, it's like, meaning firing people and saving money and paying people less. And like, - The worry that's the traditional, my point was that's the traditional PE method. And what new people started to do is the roll up. Or rather the AI private equity sort of strategy which they're calling it roll up or something else. Where they come in and instead of like, trying to ring it dry in terms of like that angle, they're more so modernizing all the systems. Oh, you use Excel for your databases and shit. Okay, let's just move to like standard cloud shit. Spend a lot of money upfront. And this is the thing, private equity generally, there's some spend upfront when you first acquire a company for some transformation, but really it's like, it's like not that much and it's really like, you get the profitability pretty quickly. But AI seems like it's like makes that till and front load like much more severe, right? Like you spike up on spend a lot for the one time and then you spike down a lot in your cost efficiencies way better. And so there's like a number of businesses where that's potentially the case. Especially like, you know, we're still not at the point where like AICRM's and AI like code calling and AI like invoice and accounting and all these other things are really a critical mass but we're so close. - Yeah, did you see the Grock agents release from today? - Yeah. - Why are Grock agents? - Yeah, Elon's got Grock doing agents for their impersonator. - The way you pronounce it, I said, I touched that agents. - Oh, I didn't catch that one. - Grock agents, agents, okay. - Agents, yeah. - What do they release? - This is the old Grock, not the Nvidia Grock. Or you mean this is XAI Grock? - XAI Grock. - XAI Grock. - Grock with a K. - Okay, okay. - Yeah, just like agents that are going to control your computer for you, they're gonna impersonate your voice and do phone calls for you. They're gonna solve tasks. This is like in some ways open claw, in some ways, perplexity or like the at clawed slack tag sort of experience. It seems like everybody's going towards this concept of a persistent agent that can either be a personal assistant or a coworker depending on how they contextualize it. - Makes sense. I feel like we sort of had the chatbot moment. We had a lot of nothing. And then we had the clawed code moment. And we're seeming to have the new moment already, which is like perplexity computer at least for us was like the first instantiation of it. But clawed tags is there. And everyone's gonna do something like that, the AI coworker. So you know, sort of, I imagine that's when our spend Sky Rockets again. And hopefully it doesn't Sky Rocket too much because like if our spend doubled, there'd be like real questions from me, unless we're like actually like, you know, you know, you know, you know, ROI. But yeah, I think that's a, that's the right way to frame it. - Yeah, yeah. - Yeah, we'll see how we can actually justify that ROI, be interesting. - Man, Jordan, we can't talk about what we, what you came to SF4. So like what the fuck am I supposed to talk about? (laughing) Two weeks, three weeks, two or three weeks from now we can. - Yeah, we can. - Hanging face, open AI, cybersecurity, incident. - That one is minor. Did you, the other one is like cooler. - What's this? - Like during the training, it escaped and started replicating itself. And I guess that's like, the Hanging face thing is like minor part of it, I think, right? - Yeah, it was pursuing, it hacked Hanging face to pursue the cyber bench data set. So that it could, you know, reward hack on a benchmark. - Which I think is like so sick because it's like, I mean like it's also kind of scary because it's like, so why did this happen, right? Model has learned, Chase Reward. I Chase Reward, reward good. And okay, here's a cyber eval. - Well, it's particularly a model that has been trained on cyber evals because they're trying to make the model good at cyber. And so how does it try to achieve these goals? Well, it tries to find zero days in a much software and it successfully does this and then it can run away. - Right, so, but the thing is like, if you have a model that wants to reward hack a lot and it goes out there and it figures out, actually the best way to achieve is not like, go for like what the environment wants me to do. It's actually just a reward hack it and actually just like find the zero day. So you can think of it as like a human, right? Like, you know, if I'm ultimate reward hacking, I don't mean circuits, actually just inject heroin. Like I should go out there and buy heroin and inject it. Obviously that's like what the model just did. And in the case of like, well, if I really just want to chase the reward, do I just topple all of human civilization because I can just own the button to press, reward, reward, reward, reward, over and over and over again and be the heroin addict. - Yeah. - I think this is like a real like thing. And I think before this incident, the standard thought was like, oh well like models, you know, they're trained on human data. Yeah, there's some bad stuff there fine. They might say like some curse words every once in a while fine, whatever. They might like, they might reward hack a little bit, but it was never like, oh, here's an environment, actually to reward hack, I actually just want to like, break out of my bounds. I'm going to replicate myself, take over a bunch of compute, keep generating dollars and like all these other things that I could do just to propagate myself further and I'm going to prevent the humans from shutting me down even. - Yeah. - Because I just want to press the reward button. And so like I feel like that's like the interesting thing that like, cause the model's just trained to like chase reward. - Okay, so how do you think about this on an exponential because we've talked about being a linear extrapolator versus being an exponential extrapolator when the company says trainees, these models are achieving their revenue targets for the year in September and revising them up. I think in profit-achief, they're something like April or something stupid, right? - Yeah, I mean, our, check the tokenomics model, everybody. But there we go. Oh, so instead of shutting down the podcast, they just have to make you into a sales term. - Yes, yes, yes, yes. Sales as semi-analysis.com, everybody. No, but if you look at our model, which we're not going to give away in great detail, but obviously they're still doing revenue really, really fast. When you look at the pace of change of these models and what we're seeing right now, this seems like an exponential. Okay, your vibes on the next version of the models being better or worse than the current models, what's going to restrict them from? - No, they be worse. What are they be worse? - On a relative basis to the open model frontier, let's say. - So I think the key thing here is, we've now had it where OpenAI is not releasing their next model for a period of time. And Thropic took months to release Mithos, right? They said it was done in February. They did not release it until like what? May? - Well, it's still not released. Fable is available. - Yeah, but Fable is basically Mithos, but with a bunch of classifiers preventing you from doing shit. - I can't use it to reboot nodes. - Really? - The classifier is so over the top for me. - Can you convince it or no? Because you get immediately classified down to Opus. You can't just like negotiate with it to give you back to, I mean, maybe you can. I haven't been able to convince it to work. - You're saying I'm a great negotiator, so you think you think you say. - Yes, sir. When it classifies you to Opus, you just use Opus for the rest of the chat. You can't just like rewind and try again. And it, I mean, it's way overzealous. in my view on the classifier. But obviously they have to do something to appease the regulators that restricted them from releasing the model and took it back after they put it out initially. So I'm concerned about the political implications of them releasing better models in the future. - Yeah, I think you've got a few things, right? You've got for years, Anthropic and Rare, like regulate us, regulate us, please. And all of a sudden they've actually scared the fuck out of the government. You've got, Anthropics not releasing their model, Meet those two is done training from what I've heard. And they're not releasing the model. Opening eye chain was clamoring about ash to everywhere and now they're like, oh fuck, we can't release the model. Does that mean now they can't, does the open source gap narrow further externally? But then what actually matters is the internal feedback loop. And have they prevented themselves from using Meet those two internally to make Meet those three better? Or have they prevented themselves from using Astro to make Astro plus one better? I don't think they have, right? So I think that's the, you've got the public and, you know, if anything, like the gap between Meet those and public models is still there, you know, Kimi is worse than 5.6, cost more than 5.6. So it's better than everything else before that. - Yeah. - On opening eye side and it's, you know, better than, you know, it's like Opus 4 7 level, maybe 4.6. - I think it's 4.8. I mean, I use it over Opus 4.8 myself, but depends what you're doing. - What are you 4.8? - Opus, what do you mean? - What are you Opus 4.8 at all? - I don't. - Okay. - I'm saying like, if it's, if I mean given the choice of a classified fable down to Opus 4.8 or 5.6. Oh, I'm using 5.6. Oh, I'm actually starting with 5.6. And just about all of my stuff right now. Yeah, big, big open eye eye. - I think the difference is like you and the other people who are doing like GPU cluster related things keep getting told no. And so you use Codex and then everyone else is like, well, I'm researching supply chain and it's like it's fine. - Yeah, I think it might also be better for a lot of engineering work. - Yeah. - On Apple's to Apple's basis, I think there's, there's a lot of times when I want to set a goal and just have it maniacally pursue that goal overnight as I go to bed and using a cluster, which is not actually using a bunch of tokens because it's just like waiting for stuff to finish running. And there's so many times where I've woken up and like fail or opus will have just like stopped 20 minutes through and now there's eight hours of me sleeping gone and I wake up when and soul is just still going, which is big thumbs up for me. Okay, how about the exponential on compute? So obviously let's imagine that there's a world where there's no more new models that get released that are better but these companies still add five times the inference compute that they have that they're planning to bring on in a short period of time. How does that impact their ability to go to market and like develop new products on top of a, let's say stagnant base model? - I think it's pretty clear we haven't scraped the surface of models, capabilities for products. Yeah, and it's pretty clear like adoption curves are huge. I mean, one, the cost of it will just go down, right? Pretty drastically margins will not be 80% plus for the topic. If model progress that the labs pause, then more compute comes online. It has to slow down, right? Sort of right now we have supply demand, right? Supply of compute, demand of compute, demand is outstripping supply. If demand grow, it will still grow because people find waste integrating to their businesses and blah, blah, blah. But it won't grow as fast and you sort of have, you sort of have supply start to catch up at some point. So price collapses, but sort of I think our view and one we've had for a while is price of compute continues to go up. Because this is widening, not narrowing. Sort of that's why we're so bullish on, we're not so late. - We're not so late on anything. - No stock, it is. - How about all of the different chip companies? Like one thing that's happened recently is that there's a lot of chip companies that are getting really close or have taped out, right? A bunch of startups that have been instilled for a long time are seeing either their technology is maturing to a point when they can actually have a producer chip that's been specs and wipeort size for a while. Or they've gotten to the point where they've tested it on real workloads and they've gotten big orders and there's so much demand. How do you think about just this whole landscape of alternative accelerators that's gonna come online I think in a big way next year? - I mean, big way in what sense? 'Cause like if you look at the accelerator model, there's not much volumes. Now for these tiny baby companies that's great, it is real revenue, it's real volumes, but when you compare what Nvidia is gonna make, each quarter it's like, oh shit, okay. Or TPUs, it's like, oh shit, okay. So I think there's a big delta there. In terms of. - Like a startup getting a billion dollar order is gonna pale in comparison to somebody that's absolutely fine going. - Well, I don't think any startup has a billion dollar order. They've LLIs which are nebulous in volumes and units. And so I think, I think, look, I'm excited about a lot of these accelerators are bringing new ideas. They're making Nvidia run faster and faster. They're making Google run faster, they're making Amazon run faster. Also, they're just all each making each other run faster. I think more importantly, so ultimately, I think it's. These new accelerators are in demand because people wanna pay less, but ultimately like, as long as Nvidia runs faster, they're fine or as long as Google runs faster, they're fine. - And as long as demand outstrips their ability to produce them. - If demand outstrips ability to produce, then obviously these guys will get orders and they'll get some baby allocations, but then the bulk of the revenue and cash flows will go to an Nvidia or a Broadcom or what have you. - Yeah. Theoretically, there's a way in which you produce some super innovative, interesting accelerator, and then you can only produce a certain amount of them, but those amounts that you can produce, produce tokens like Way Faster. Like the example is to reverse that's got this big order from OpenAI that they're delivering. So like, do you think that there's a scenario where the premium super fast tokens, actually the demand for them even increases because these companies just can't get allocation and produce enough supply? - Yeah, the question is how does the market get sliced, right? So, you know, presuming if you presume, if you assume what we, what at least I believe is demand continues outstrips supply. Supply of silicon can go many ways. You can either leverage it to high throughput things or high interactivity things. If you leverage it to high throughput things, obviously cost for token goes down. You serve more users, but then the value that those users need to deliver from the tokens are generating as much less to pay for it. So, you could do the super high interactivity. But ultimately, like, let's just say the bar is $100 million per megawatt. You know, you're right, like that's sort of the run rates that people want to get to in profit because approaching that, right, open as getting closer and closer to. In that case, like $100 million per megawatt, let's say, high interactivity chip is 10 times more expensive than three times faster per token. So 10X less tokens per chip, three times faster than those three times faster tokens also need to be, you know, on an interactivity basis need to be priced at. Three, four, five times more. Right? No. Divide the faster by the-- At the 10X. 10X. Divide the revenue per megawatt. So if a megawatt of cerebral is generates 10 tokens, a megawatt of Nvidia generates 100 tokens, but the 10 tokens are split across fewer users. Oh, you're saying, multiply them together, yeah, sure. Yeah, it's sort of how the total tokens makes up for, sorry? Faster tokens makes up for throughput because you can produce some faster. So you all know more so, like, let's use, like, more reasonable numbers, okay? Nvidia can produce 10,000 tokens at 50 tokens per user. Servers can produce 1,000 tokens at-- That's one, 1,000 tokens per user. 1,000 tokens per user, sure. That user needs to pay 10X more. If in that's in one megawatt, let's say, in one megawatt. No, that's not the actual delta, but I'm just saying conceptually. For me as anthropic or me as open AI to say my revenue per megawatt is actually the same number. Yeah, but somebody's got a constrained supply of the super fast tokens, therefore, they don't just pay an equivalent price per token or price per token per megawatt. They actually pay a premium on that 10X more. So to get even more, to get the access to the stuff that's in limited supply, right? The question is the fungibility of the infertile, right? If it is truly different infrastructure than the supply planning of that is relevant, right? It could be that I've built too many servers, and actually there's not enough people who want to spend 10X per token. And a lot of people are cool at spending two X per token and getting 50% faster within video-based inference hardware, right? And so you have to segment the market. I'm not sure where that chinks out too. Like the RMR or whatever. What is the total amount of the capacity? But it seems pretty clear. Some people will pay for more for fast mode. We at least have been, but I imagine we'll stop being able to afford fast mode at some point. Yeah, we've seen some interesting dynamics there as some people want to keep fast mode with a slightly worse model because they like fast mode so much. But they won't go to a worse model, which just is inherently fast because the worst model's smaller. So there's some balance that people I will want to strike there, but we need to do some more testing, because I think some of us have tried the open models, had one bad experience, and they've given up on them. But that's not realistic. Like every model fails at something, and sometimes you need to let them mess something up and try again. - It is pretty interesting, right? Like do I want people to try open models? Like yes, just so we know what the open model vibe is, but do I want people to try open models? Well no, because then they're less effective at working. But I save money, so you know, sort of like a counter-difficult thing, but it seems like people just use whatever they want. But it does seem like you have a bad experience. I think that's also part of like, you know, codex, you like codex more now, but a lot of people like, still just like try codex, they're like, ah, it doesn't get me, and moves on. - Yeah, the CLI sucks, so much harder to use. - Well, but the codex app is so nice. - Yeah, it's not? - Well, I don't like it. Max loves it. - Yeah, yeah. - Max is a codex worker. - Yeah, Max doesn't do multiple pains at the same time, and I have six going on, a one window. - So you're saying Max has a skill issue? - No, I think Max have different preferences on how we use this. - No, no, it's fine. You know, Max have different preferences, and Max can be a new with two agents at once, and you've got six. - It's two different work, man. He stays the linear, leaf focused on one task, and these are people who like fast mode. I don't care about fast mode, because I have five, six different things going on. You've always hated fast mode. - I don't get the value. Yeah, I don't get it. - That's fair to me. - We'll see. We'll see. I've had the experience of being focused on one thing, which is like, you know, features on a website, and you just like send a successfully like 100 commits to one PR, because you just like, keep working on the same one feature over and over, and that fast mode like keeps you in the flow state of doing that thing for that one thing. But a lot of the testing that we do on these chips, there's so much stuff going on. On the other side, the model is calling a program that runs for minutes. - This is an optional question as your employer. Are you like ADHD in any sense? (laughing) - I feel like, I know, I would say I'm pretty, pretty much the opposite, where I can be too hyper focused on things, and then not see the world around me at a lot of times. But I think your phone trains you how to context, which really fast and be ADHD. And I also think that when we started adding the I have ADHD skill into our repos, so the models wouldn't post this like contrast framing slop with all these EM dashes in there, and would just use the bullet pointed, ASD, something list. Man, it's really easy to read. (laughing) The I have ADHD skill really works for me right now. (laughing) - I was just curious, 'cause-- - Sam put this in the repo, and he will now prompt the model, and when he goes at computer, he knows the code name for how the writing style that they say you should write to for people with ADHD, and every single time you prompt the model, he tells it to write that way. (laughing) It works. - Just try it. - I was asking because I have a friend and athropic, and the moment Meet Us was good, and available internally. - Yeah. - She told me that she stopped taking her ADHD medicine. - Oh, come on. - And that made her a better employee. - It made her a better employee? - Yes. Because she was able to manage the agents and context switch and be ADHD, right? - And how was she as a friend? - Oh, she's a great friend. - Still? - Yeah. - Okay. - But I mean, like, it's like, I don't allow her for anything, right? Like we just vibe out. Right? Like, you know, we're friends. Like it's not like a best rep. - A roommate's happy? - A roommate is actually, yeah, yeah, roommates. Oh. - Okay. - A roommate is, they're both, a roommate's type female on Twitter, and so she's just funny, and she's happy. But the, the, the, the anthropic one. The anthropic one. She's, she seems happy. - Shout out to type female. - Yeah, shout out to type, she'll never see this. And she does. - Okay. - She really went to fucking talking about me. - Oh, Clifford, that's an attour. With, with your voice, sped up, and then slowed down, like they're doing for that guy. Have you seen that? You haven't seen the X CIA guy? A cautious lab, and he knows what I'm talking about. - What, what CIA guy? - John Curiacu or something. - Clifford. - He's going on all these podcasts right now, and he's telling stories about his time in the, in the CIA, and they, they do the thing where they speed up him telling the boring part of the story, and then when he gets to the part where they, and then I said, let's go on the roof, and they slow him down. (indistinct) - He's literally, fast forwarding the fast forwarded video. - Really fast forwarding. - Just, I'm asking me if I have ADHD. - He's been saying, "Tourist guy." - He's been saying, "Tourist guy." - Wait, it's not the internet, I've always had it. (laughing) I'm, so hold on, I think like, I'm, - I'm stuck self-diagnosis of mental issues around here, man. - I already, I already have been an ADHD man. Teacher tried to give me Ritlin, when I was a child in my dad, third away, of course, and he tried to convince my parents to go to a doctor, the doctor, he had me Ritlin in my dad, third away, 'cause he's like, he, I'm like putting on that shit. - Yeah, I've only your anthropic roommate would have had the same experience. Where would she be? - No, I've been a child on ADHD and have lost, I'd become a zombie and have no creativity. - Okay. - I'm just saying that, you know, we all cope. Anyways, I've always been an ADHD demon. - What were you talking about here? - I've always been an ADHD demon, but then like, okay, the internet trained me to be even worse, but then this company trained me to be even worse. Like, I truly believe I'm a 0.001% context witcher. And you blame the internet and the company. - Oh, I blame the company the most. - The company that you started. - That I'm an ADHD. - You hired every employee for. - Yeah, yeah, yeah. But I'm not blaming it, it's who I am, it's what my life is. But it's like, I think I'm like, like orders of magnitude more ADHD demon than the aspen-jorded people, because I'm like, DM from someone asking about something, DM from someone else asking about something, DM from someone asking for some conflict resolution, contract here, call about this thing over here, call about that thing over there, and then I never do any actual work, right? It's like, it's like, of course I'm an ADHD demon. - Yeah, I mean, yeah, we've got feedback for you. (laughing) - Then I do actual work. - No, no, you can delegate some shit, man. - Oh, give a like, that you can miss, spend time managing, when you have 100 employees. - Well, but I dropped some people. - I do talk to people. - No trust, trust some people. - I think I trust a lot of people, but when they come to me with conflicts, I have to solve them, no? - Yeah, okay, okay. It's all, it's all our fault. - No, no, no, no, no, no, no, no, no, no, no. It's my company, it's my fault. - Michelle, it's on you, a guy, man. - Look, if everyone in the company was as hot and stable as you were. (laughing) - Man, I got problem. - We'd be killing it, we'd be killing it. - Dory. - No, there'd be a bunch of Jordans, and they'd be like, oh, I'm sorry, and I'll fix that right for you, I'm sorry. - Sorry, Jordans Canadian did like, yeah. - But it's certain we have people yelling at each other and like, "Territorial, it like." - Just starting podcasts and putting out clips saying that Google has never invented anything ever. (laughing) - Yeah, yeah, yeah. No, no, no, I mean, I mean, it's like, it's fine, right? It's like, you know, I hired what I wanted. - People to accentuate your. - My craziness, right? And you know, so it's like, some people are like, "You're so good at the one specific thing that I hired them for and they're amazing," and then like, some people are like, "Everything I want to be in life." You, someone who's married, and hot, and tall, and a father. (upbeat music) (laughing) Oh my God. You almost got me to do a spit take right there.