Subscribe
Share
Share
Embed
TBPN is a live tech talk show hosted by John Coogan and Jordi Hays, streaming weekdays from 11–2 PT on X and YouTube, with full episodes posted to Spotify immediately after airing.
Described by The New York Times as “Silicon Valley’s newest obsession,” TBPN has interviewed Mark Zuckerberg, Sam Altman, Mark Cuban, and Satya Nadella. Diet TBPN delivers the best moments from each episode in under 30 minutes.
Welcome to the stream, Doug. How are you doing?
Speaker 2:I'm doing wonderful. How are you guys doing?
Speaker 1:Great to see you.
Speaker 3:Can we pull up that the video that you've that
Speaker 1:Oh, yes.
Speaker 3:John painstakingly made this morning that completely that completely flopped.
Speaker 1:I'm I'm very excited to have you on the show. It feels like that scene in Sonic the Hedgehog three, which I saw where Sonic and Shadow team up and join forces to talk about CapEx and agentic coding. What what's new in your world? Is is Claude Code still the top of mind, or are you still churning through the CapEx numbers from earnings?
Speaker 3:Are you somewhat of an agent for Claude now? Like, you work for Claude?
Speaker 2:I I do, actually. I think I mostly just move my information back and forth. You know? Mhmm. I have I have pretty much like, I think of it as, like, my manager.
Speaker 2:You know? Like, it tells me what to do, and then I go bring the information, and I bring it to my coworkers. I bring it back. Mhmm. All day, I'm just on Cloud Code.
Speaker 1:How many how many prompts are you running right now? Do you have any threads going?
Speaker 2:Okay. So I I have I have seven. I have seven threads.
Speaker 1:Seven that are running right now or or No. No. Not running. Waiting for your input.
Speaker 2:I'm waiting waiting waiting for my input.
Speaker 1:We'll let you get back to it.
Speaker 3:Why don't you just have an eighth that just Let's talk about
Speaker 1:what yeah. Orchestration. Have you played with Gastown? Are you thinking about abstracting yourself to a higher level?
Speaker 2:Oh, okay. So Gastown is is pretty intense. I don't think Ghastown's gonna work out. I think it's gonna be Agent Swarms.
Speaker 1:Okay. Oh, okay. Explain the difference between between Gastown and Agent Swarms.
Speaker 2:Okay. So Gastown is probably the most forward looking thing I've read in a It talks about how you created this self healing tool process to essentially pass all these beads across and have all these workers and ways to self repair the agent workflow process. And I read it, and I was like, dude, this is brilliant and also fucking crazy. It's it's like it it feels like the ratings of a bad man. And then I proceeded well, I was also in my bad man era.
Speaker 2:Yeah. Like, before before the New Year's, when you had two x times usage, I pretty much was, like, literally railing quad code constantly. I think I had I think I had four fourteen hour days.
Speaker 1:Wow.
Speaker 2:Yeah. It was it was beautiful.
Speaker 1:It was beautiful. Okay. So talk to us about what you're actually building because, you know, we're we're talking about SaaSpocalypse. It feels like there's a debate over build rebuild all your tools from scratch to save whatever your SaaS fees are. Yeah.
Speaker 1:And even even yesterday,
Speaker 3:it's notable. OpenAI comes out with Frontier
Speaker 1:Yeah.
Speaker 3:Which is you look at like that you gotta look at this like graphic which feels like it was made for, you know, Fortune 500 CEO to kind of or management team to kind of understand it. Mhmm. And it's like, here's more SaaS to replace your other SaaS. Right? It's like, you know, you got the system of record down here.
Speaker 3:You have a bunch of agents in between, and then you've got different applications that you're using. And meanwhile, Anthropic's just like, we're making a really smart digital guy that can do whatever you want.
Speaker 2:Yeah. So I think the two there's, like, two really interesting ways. I think OpenAI is, like, the Fortune 500 selling it from the top, if it makes sense. Mhmm. And then Anthropic is, like, here's my here's your CloudCode agent, sell 20,000 of them.
Speaker 2:Did you see the in the Accenture partnership? I think that's really interesting. No. Like, if you're if you go back there's they're doing 30,000 people at Accenture.
Speaker 1:What do you mean?
Speaker 2:So, like, thirty thirty thousand people at Accenture are going to learn how to Claude code.
Speaker 1:Oh. And then and then be
Speaker 3:deployed into different companies.
Speaker 2:Yeah. Who knows? Who knows?
Speaker 1:What else would they
Speaker 2:just replace Accenture. I mean, that's what I that's what they're gonna be doing, I think.
Speaker 1:Okay. Wait. So Yeah. So What do you mean replace if if Accenture folks are using Cloud Code, wouldn't they be using it on consulting projects internally to companies?
Speaker 2:I I I think they're going to be using it internally Mhmm. As my no. So they're gonna be using internally, and then they're gonna be doing all these consulting things. Mhmm. Because if you think about it, one of the issues is, like, you know, when you had SaaS, one of the biggest issues of, like, changing from one CRM to another was effectively being like, hey, everyone.
Speaker 2:You have to quit your jobs for, like, ten months to figure this out. Yeah. The the implementation, like, you'd have tons and tons of, like, consultants do that. Yep. And I think that that's what the Accenture partnership is.
Speaker 2:So, essentially, like, people are going to be implementing Cloud Code, and there's 30,000 people at Accenture who's gonna do it. And then on the other side, you have Frontier, which is like the Fortune 500 way of being like, here's your plan. Come to us, and we'll build this whole thing, blah blah blah. Yeah. So yeah.
Speaker 1:But is is is reimplementing your CRM really the lowest hanging fruit for America's greatest companies? It it can't possibly there must be some new ideas, new problems to solve, new tools to build. Like, why are we just going to shuffle the chips around the board instead of, like, doing something productive?
Speaker 2:Okay. So I think the system of record refresh is gonna be really awesome because, you know, like, the big lit I mean, it honestly does feel kind of boomer if you think about it. It's, like, the biggest data now. Everyone can have the big data now. Mhmm.
Speaker 2:But I think the the automation that you've always dreamed of is actually gonna happen, and the system of record is just gonna essentially have hooks out to all these other things they're gonna build on top of it, which is mostly, like, you know, the frontier thing. Yeah. And, essentially, like, all the information work is just gonna be, like, all on the agent, and everything else is gonna be, like, place where it lives and is stored for. So instead of me having someone let let me use my personal stack as seminalysis. We use HubSpot, for example.
Speaker 2:Sure. So, hey. The sales this quarter, we need this, like, quota or who did what or what products are selling better or, like, you know, what's How many more podcasts
Speaker 3:should Dylan do this month if we wanna hit our goal?
Speaker 2:That's totally different. We don't actually have, like, our our, like, you know, the big yeah. Sure. How many more podcasts? Shit like that.
Speaker 1:Yeah.
Speaker 2:Yeah. I could just vibe code it. I was just like, hey. Can you run this analysis for me? And in a perfect futuristic world, it'll go into the CRM Yep.
Speaker 2:Pull all the information of the all of our inbounds Yeah. Make it'd be like, hey. The day after Dylan goes on a podcast, there's, like, 25 people who who come in. The conversion rate is x. You could price it at this.
Speaker 2:Dylan, quit your you know, stop working and effectively just, like, hit the podcast. There we go. You know? Like yeah. Yeah.
Speaker 2:Exactly. So so you could do this with anything, though. Like, it's just information, man. Like, it's it's it's gonna be pretty sick. But I think all the SaaS companies are going to essentially just become hooks for all the crap they've built on top of
Speaker 1:it.
Speaker 3:Yeah. Yeah. So Did you see Jensen yesterday was was kind of defending some companies like SAP and ServiceNow and saying, hey, if I was a if I was a really smart humanoid out doing work in the world and I needed a screwdriver, would I just invent a new screwdriver, or would I just take one off the shelf? And so that was, like, his defense. Tyler here was took took the other side of it and just said, there's gonna be a lot of situations where, especially in a software only environment, it's easier to just build a very specific workflow that you need that you would have gotten from a SaaS provider versus, you know, you don't need to actually rebuild the entire platform.
Speaker 2:Yeah. I think, we're gonna be building a lot of screwdrivers. Like, things like, the thing that's important is, like, okay. You're not gonna rent a truck. Right?
Speaker 2:Like, like, you're not gonna build your own truck, but your own screwdriver a 100%. Like like, you're doing this big ginormous job. You need a hammer. You'd be like, okay. Pull it out of my belt.
Speaker 2:Okay? Mhmm. But you're not gonna be like, I'm about to move 700 tons of, like, here to here. I need to rent a truck. You're you're not gonna build the truck.
Speaker 2:And so that's what I think the system records are gonna look like. They're gonna look like places where, like, actual data that cannot be, like like, cannot be vibed effectively. Like, what's your inventory cannot have any fucking hallucinations. Right? Like, your ERP.
Speaker 2:But all of that will just be hooks for everything else. Mhmm. Because, like, all the information is just, pulling, retrieving, making the correlation, running the the charts, like Okay. But even fixing that all the time.
Speaker 3:But then but then how do you square the fact that a system of record is way less sticky if you have agents that can work around the clock to switch you over to a different system of record? Like, that still ends up putting massive pricing pressure.
Speaker 2:So so to be clear, I don't think it's good for everyone. Like, I think my my my favorite analogy of this is, like, there actually is a very old school type of software that's, like, existed for a long time. All the shit on mainframes. It's all it's all out there. Yeah.
Speaker 2:And, you know, like, funny enough, mainframes still grew, like, 6% a year or whatever. Someone has the real number from, like, 2,002 Yeah. To to 2020. It's crazy. So, like, they're they're gonna grow, but it's just gonna be, like, a very different vision of the world that I don't think people are ready for.
Speaker 2:Yeah. And and the adjustment period is the big problem because all the stocks are priced like they're not gonna be mainframes. Yeah. And and and also just for context, mainframes, there's like, hey. There's one of each company now.
Speaker 1:Yeah.
Speaker 2:There's not there's not like 10.
Speaker 1:Yep.
Speaker 2:There's one each.
Speaker 3:Yeah. So I think with all these all these whether you're a system of record or you're, you know, some vertical software, you're gonna need to show insane revenue growth in a truly AI native Mhmm. Product. Otherwise, investors, I think, are gonna continue to not be able to create a super compelling narrative why why you should own it during this period of uncertainty.
Speaker 2:Yeah. I mean, pretty much what happens and we're gonna go like investor brain. When anything goes x growth, the multiple goes massively down.
Speaker 1:Yeah. Yeah. That makes sense.
Speaker 2:Eight times earnings.
Speaker 1:Can you talk a little bit more about what you're actually coding, what you're building, like what the software is? Because it it from the demos that I've seen that you've posted, it feels much more like you have an agent that can do knowledge retrieval, data transformation, build dashboards, charts, and and, like, knowledge work as opposed to truly replacing software tools at this point. But do you have you built anything that's, like, long lived and and runs, like, daily? Or is, like, something you keep revisiting because it's now a piece of software that does the job
Speaker 2:So the Cloud Code Commits the Cloud Code Commits is now software that lives and and runs every single day.
Speaker 1:Okay.
Speaker 2:That's like a scraper. Right? And then, that that, like, lives in a database, and that that will run forever. Okay. There's, like, a lot of other tracking price data tool stuff.
Speaker 2:Like, a lot of the scraping that we're, like, that is not, like, publicly available. Like Yeah. We do, like, a lot of that. Like, we had a data team just do that. Yeah.
Speaker 2:And now, effectively, we can, like, really accelerate that so everyone can do that. Sure. Sure. Sure. There are other, like, little things that I think are, like, heuristics.
Speaker 2:Yeah. Like, little skills of, like, I have blind spots that I consistently make over and over. Yeah. And I'm like, hey. I know this blind spot's an issue, blah blah blah.
Speaker 2:You should, like, consider this in this case. I don't think it's like the Galaxy Brain software, and we're very far from there. Because if you actually play with these tools a lot, context fraud is real. Yeah. So
Speaker 1:I they is it is but it's is it how fast is it getting better? Because it feels like we're seeing the meter graph
Speaker 2:Scary fast.
Speaker 1:Scary fast.
Speaker 2:Scary. Scary. Scary. So I started vibe coding with Quad four. Yeah.
Speaker 2:And it just wasn't or the sorry. Opus four on Quad Code. And it just could not one shot websites in the way that four point five and four point six can.
Speaker 1:Yep.
Speaker 2:And if it just marginally improves from here, it feels like, why would I pay for, like, any kind of UI UX if it just could be generated at a good enough quality?
Speaker 1:Yeah. How did you process the new models this week? Four six, five three. What's the review?
Speaker 3:If you don't if you can't immediately notice the difference between four five and four six, start polishing your resume. You are cooked.
Speaker 2:Yeah. Yeah. You just got automated by an agent.
Speaker 1:Oh, my god.
Speaker 2:I think four six was a little disappointing, if we're honest with
Speaker 1:you. Mhmm.
Speaker 2:I I think it might have been Sonnet five.
Speaker 1:Oh, that's what people are saying. That's the conspiracy theory. Right? But what what does that mean?
Speaker 2:Original Sonnet five leaks were that it's like as good as Opus 4.5, but with 1,000,000 context window Yep. And specifically trained for Asian swarms.
Speaker 1:Sure. So well, yeah. But does it just mean, like, same quality but faster, cheaper, at least for Endoptic?
Speaker 2:And then they may and they make more money.
Speaker 1:Oh, yeah. Yeah. Yeah. But, yeah, better margins. How are the margins looking for the the labs right now?
Speaker 1:There's been there was a bunch of, like, FUD around it, but it seems like from all the leaks, it's been, like, fifty, sixty, 70% pretty good.
Speaker 2:Yeah. If if you x all the free users, it's always really good. Right?
Speaker 1:Oh, yeah.
Speaker 2:Yeah. Yeah. Honestly, Anthropic has no free users or, like, on a relative basis. So their margins are ironically, like like, kind of on a like to like basis kind of not as good as you think. Yeah.
Speaker 1:Can you break down a little bit more of the thesis of the Claude code is an inflection point article? What what the key takeaway? Who you're speaking to? What update you wanted to share? And then I wanna go into some of the some of the pushback and your response to that.
Speaker 2:Yeah. Sure. So first, I think the thing that makes me really excited is the first time since Shana thought, I feel like we have a new scaling that feels very, very different and hardcore, and I can actually see my entire life day to day change. Mhmm. I think I can expect some version of a Cloud Code harness to be effectively all my information work from now till the future.
Speaker 1:Mhmm.
Speaker 2:I am a daily user. I was not a daily dailies before, and I expect to continue to be one.
Speaker 1:And that's kind of what happened with the reasoning models. It went from, like, you could ask it stuff, but it might hallucinate to, like, the answers are good. Like, you can pretty much rely on and there's gonna be citations, and, like, it's gonna be 99.999%, like, percent, like, usable for things. So you just have a question, you get that. It may be not great at certain things, but in general, like, it delivered on the initial, like, chat experience that I think a lot of people are looking for.
Speaker 1:And then they became. Yeah.
Speaker 3:How much, do you think Anthropic cares more about winning in consumer than they've led on to date?
Speaker 2:No. I don't think so. Mhmm. Everyone everyone who works there is exactly, like like, what you think it is.
Speaker 3:Yeah.
Speaker 2:They're exactly who they say they are.
Speaker 1:Yeah. They're software singularity build.
Speaker 2:Yeah. And then I think co work is what they're really excited about.
Speaker 1:Sure. Makes sense. Yeah. Yeah. What what what this person
Speaker 3:they're even they're not even thinking about
Speaker 1:Mhmm.
Speaker 3:The scenario where a bunch of people are using Claude in the work in a work setting and say, hey, this is pretty great and, yeah, ChatGPT has ads. I'm happy to pay $20 a month. I'll use it personally. Because I just think there's like an iPhone like, I think the game to get to like 3,000,000,000 users is like over when you just look at the traction of like Gemini and ChatGPT and the fact that norm normal people aren't caring that much about the Nuance. Maybe that don't have that much to automate in their life.
Speaker 3:But there's like an iPhone size market. Mhmm. Like the iPhone wasn't the first smartphone to launch. And it's possible, like, when I see this, like, when I see this, like, the the Super Bowl ad, the sort of like trust nuke, I was calling it. Right?
Speaker 3:Just like, hey, like, it's really funny they're like, you know, rage baiting OpenAI, but at the same time, they're just destroying trust around ads and LMs potentially, like, permanently. Right? Because people, even when they start seeing ads that are more like display ads, they'll start thinking, well, was the result influenced too? You know, it just like, it hurts the trust. And so I think there I think My my theory is that any product that like really catches on in the workplace could very well trickle over into into life, and Anthropic could someday have a pretty big, you know, they could have like a Netflix sized subscriber base for people that just want an ad free AI experience.
Speaker 2:Yeah. That sounds completely right to me. Mhmm. You're saying for your point
Speaker 3:is like, it's just secondary to them. They're like, it's a nice to have, but like, we don't. That's not that's not our intention. I
Speaker 2:think okay. So singularity pill, but I also I think you have to pay for the singularity, and I think it's gonna be enterprise that does it. Makes sense. Yeah.
Speaker 1:Yeah. I mean, the the other the other take on like, you could wind up being like the Apple and like the premium, you know, privacy focused, or you could wind up being, like, the DuckDuckGo, which was, like
Speaker 2:Yeah.
Speaker 1:You know, the yeah. It was a counter to Google, but it never got to any meaningful scale. Right?
Speaker 2:Yeah. But I still think I still think OpenAI is the Apple. Like Yeah. Like, Apple was synonymous with smartphones when it really took off. Like, what is the other smartphone?
Speaker 2:Nokia. You could argue this is like a Blackberry. Yeah. I can't name it. Yeah.
Speaker 2:Well, like, Blackberry. Right? Yeah. Was known for work. Yeah.
Speaker 2:And then, obviously, like, it swapped over. So I still think OpenAI is, like, the cognitive referent. And, honestly, 5.3 cooks.
Speaker 1:Yeah. Faster or just better or both?
Speaker 2:Faster and better.
Speaker 1:Faster and better. Okay. Talk about 5.2 token efficiency. Rune was pushing back on the article saying you're making you're making assertion that 5.2 token efficiency ruins long horizon planning, and yet 5.2 tops the meter chart for long horizon planning, half baked. What's the explanation there?
Speaker 2:Didn't someone completely mock that argument? But he's kind of a he's like, sorry. I gotta find the guy. But it's like, I don't know what task is being done here. Is True.
Speaker 2:Are they the same hardness? Yes. Did you just spam it to infinity and, like, you finish, like, a sufficiently long task to completion versus, like like, okay, let's just say we have two kids taking the SAT Mhmm. And one does a better job and finishes first, and one does like almost as good of a job and took seven times as long, and you're like, wow, that one's a smart kid.
Speaker 1:Yeah.
Speaker 2:No, dude. That doesn't make any sense.
Speaker 1:Yeah. Yeah. Yeah. Yeah. That makes sense.
Speaker 1:Yeah. Yeah. An elegant solution delivered faster is uniformly better.
Speaker 2:Yeah. A 100%.
Speaker 1:That makes sense.
Speaker 2:Yeah. And so if if you spam more tokens, like, and you win and you're like, oh, look. I mobbed them. And it's like, dude, what if you just use less tokens?
Speaker 1:I think the benchmark is supposed to be for for they they have they have a reference class of of projects that are supposed to take x amount of time. They would take a human developer six hours, and then they have all the models compete. And if you can compete the six hour task, then you get put at the six hour mark. It's not, did you run for six hours? It it it it's makes sense.
Speaker 1:So so so it could be, like, implement a CRM product or, you know, write a very complicated, you know, database or something. It would take, you know, a a talented software developer six hours, two hours, one hour, and they have different tasks and then you're trying to climb that hurdle.
Speaker 2:Oh, yeah. And then it's it it okay. It climbs higher and higher.
Speaker 1:I I think I think that's loosely it it it's because because obviously, you could just say, okay. Just reasoning count to 1,000,000,000 and just go as slow as and it works for days and that's not impressive. Yeah. I think that's not how it works.
Speaker 2:The yeah. I mean, it's okay. So, yeah, you're right. The the different the scaling thing, but, like okay. So one the other thing I was doing, like, now that we have vibe coding available to everyone, you can just have it do the same task and do, like, a b I've been doing a lot of internal benchmarking.
Speaker 2:Everyone could benchmark, guys. Dude, Codex 5.2 took so long and just never built for me. Oh, interesting. All the Codex hype during it just never worked for me, man. It never one shot projects like Opus 4.5 did.
Speaker 2:Mhmm. And I'm just like, this feels like complete foot. But that being said, Codex 5.3 cooks. Like Yeah. I think I take everything back off about 5.2 for 5.3.
Speaker 1:Nice. Total reversal, classic AI narrative, just day by day complete switching of the narrative. Talk about the NPM downloads. Because you said that you're now scraping them every day trying to understand how many commits on GitHub are related to Cloud Code. And and the pushback from Roon was that this counts NPM downloads as authoritative when Cloud Code numbers are hugely inflated because GitHub actions does automatic Cloud Code download every time continuous integration CI runs versus Codex compute cloud?
Speaker 1:So maybe it's not apples to apples. What's more nuance on the the the fast takeoff of cloud code? Because, honestly, when you said 20% of commits by the end of the year, I was like, that feels extremely low. I would expect, like, 70%, and I would expect Codex to be at 30% and no more human commits because it's working.
Speaker 2:So I I wanted to make sure we have, like, a high standard like, a high Sure. 95 plus. Sure. I don't think like, sure. If it continues to grow on a week on week basis, like, yeah, it's, like Yeah.
Speaker 2:You know, a 100% by June or something like that.
Speaker 1:Well, there's also the there's also the fact that, like, you could be writing code and still just, like, almost be using Cloud Code as, like, your linter or, like, your your interface to GitHub. And if there's an abstraction layer there that people adopt, you're gonna see the commits go through the roof even if they're still here in the loop meaningfully.
Speaker 2:Look. Look. Look. Look. Look.
Speaker 2:Yeah. I'm not gonna pretend, like, the Cloud Code commits thing is, like, the cleanest way Totally. There's a lot of way there's a lot of ways to fuck the data up.
Speaker 1:Yeah.
Speaker 2:For example, people who use who you could just say don't do this. Yeah. And and won't do it. Number two, like, private on a ratio is like five times bigger. Sure.
Speaker 2:That matters way more. Yep. And then like, I also think that like the way you consume it, like this doesn't count for cursor. People have been clearly using AI for like a long while and it doesn't show up. Yep.
Speaker 2:This is just the example I can like say, it's like, chart goes up really quickly.
Speaker 1:Yeah. Yeah. Yeah. Yeah. Yeah.
Speaker 1:It's so cool.
Speaker 2:It's not perfect. Yeah. It's a data set that I could create No. Absolutely. In relatively short amount of time.
Speaker 2:And I'm like, well, no. Seems pretty cool.
Speaker 1:Yeah. Yeah. No. It it is. Are we
Speaker 3:talking about Amazon?
Speaker 1:Yeah. Yeah. Let's move over to hyperscalers.
Speaker 3:Reaction was the number two low. Dude. Yeah.
Speaker 1:Yeah. Seriously. Yeah.
Speaker 2:What what Future billing was crazy. Yeah. That was crazy to
Speaker 1:me. Okay. Why?
Speaker 2:Yeah. That was really shocking. You know, we do a lot of data center tracking, and we do a lot of accelerator tracking, and we were too low.
Speaker 1:Okay. Are they are they are they trying to play some sort of hype game where they're throwing out the biggest number and they're not actually even going to be able to buy enough equipment to spend it even if they are signaling to the market we're going to be hearing like, well, we wanted to buy this many NVIDIA chips, but we couldn't get them, or we have a delay at this data center because of regulation. And so they're just trying to project strength because they're sort of behind on the AI narrative a little bit. They don't have the big position that Microsoft does in OpenAI. They don't have, you know, a deep mind level team.
Speaker 1:And so they're saying, we're gonna go biggest on the dollar front, but then maybe they don't deliver on it? Or do you think at the end of the year, we'll be like, yeah, they spent 200,000,000,000?
Speaker 2:I think at the end of the year, they're gonna be like, yeah, they spent 200,000,000,000.
Speaker 1:Let's go.
Speaker 2:They they are the single biggest provider of power in the entire world, I think. Wow. Like, the incremental. And the AWS, like, supply chain can ramp a lot quicker than anyone else. And every every example that we track in the data center, like, the data center team, they are on time and can scale to, like, levels that are crazy.
Speaker 1:Yeah.
Speaker 2:Like, Ranir's ramp is just, like, out of this world fast compared to everyone else. Every other gigawatt project is essentially delayed, and they're gonna be, like, ish on time.
Speaker 1:Wow. So isn't that extremely just, like, good for Amazon? Like, they're properly positioned. They're properly transitioning. Like, they yeah.
Speaker 1:Like, who knows what happens to the rest of But eventually, this thing is just, like, part of the
Speaker 3:We were talking earlier.
Speaker 1:Yeah.
Speaker 3:Like, Jassy didn't exactly paint this, like, incredibly exciting vision
Speaker 2:Mhmm.
Speaker 3:And share like, hey, we you guys are actually underestimating
Speaker 1:Yeah.
Speaker 3:Demand. Still, even if you're bullish on AI, you're underestimating demand. And we're in a position to actually try to get a more accurate read here, and that's why we're investing.
Speaker 2:Yeah. And it's funny because they could have said one thing that would have made the AWS call better, it'd be like, yeah. We see high twenties, and, like, the stock would have
Speaker 1:ripped. Mhmm.
Speaker 2:But they're like, we continue to project to see this level of growth.
Speaker 3:Mhmm. Percentage what percentage of the 200,000,000,000 do you think will actually flow to NVIDIA? Because NVIDIA is rallying today. That's why we're wearing white suits, but it didn't rally immediately in after hours.
Speaker 2:I think a meaningful amount. Mhmm. I I definitely cannot disclose what somebody else's thinks. Yep. But I think they're gonna run out of training and and the answer is, like, what's the biggest amount of supply chain that's, like, locked up?
Speaker 2:It's it's it's NVIDIA.
Speaker 1:Yeah. That makes a ton of sense.
Speaker 3:Can you get, Macron a free Semi Analysis plan? Because he came out this week with his big new initiative, €30,000,000 for AI research. France is gonna be the home of of research.
Speaker 1:How do you think all the hyperscows will respond?
Speaker 2:You you you know what's crazy is people have been trying to do a lot of work in France for a long time because they have this giant Nuclear. Nuclear power plant. Yeah. Yeah. It's kinda stranded.
Speaker 2:Weird. And no one uses it. And, like, everyone wants to be like, dude, I can get a giggle out here. And then they, like, try to start building, and they're just like, yeah. This is never gonna happen.
Speaker 2:I'm just gonna go back to United States. Even though United States is, like, all fucked up, it's like, I can I can start there? And they're like, no. No. No.
Speaker 2:No. We'll start in, like, five years. Yeah. And I can think of two specific projects that essentially did the same thing where it's like, oh my god. All this France data center power.
Speaker 2:And then, like, they started and, like, never mind.
Speaker 1:Yeah. That's funny. No.
Speaker 3:Why do you think Grok is climbing the charts right now? Any any insight? It's like number three after get free cash and after ChatGPT in the overall app store.
Speaker 1:The iOS app store.
Speaker 2:I dude, actually, one, this tells you how locked in I've been with Claude Code. I had no idea.
Speaker 3:Yeah. You're so locked in with Claude researching the AI race that you
Speaker 1:It's just yeah. It's interesting. I I I you know, the the App Store is based on, like, acceleration, but, you know, the the the Grock hype cycle of, like, you know, let's push all the Twitter users or the x users there. Like, that sort of already happened. Like, I don't know how this is happening because there isn't much hype about
Speaker 3:the yeah. It's happening off x.
Speaker 1:Yeah. And a lot of people were like, yeah, like you can talk to Ani and Valentine, but like is that really popular? Might be.
Speaker 3:Don't know. Ani singular.
Speaker 2:The real singular is lonely people. Yeah. Oh, my god. Maybe. Yeah.
Speaker 2:I I did see a video of, like, the Stormlight archive thing, and that I feel like like, hit a broader audience in terms of video generation. And I think video generation, like, that always kind of wins. Yeah. We actually did an analysis Yeah.
Speaker 1:Which is why you
Speaker 3:you need to be bullish on the Disney OpenAI deal.
Speaker 1:Think I so.
Speaker 3:Because we've we've we've seen the Nano Banana bump Yep. With Gemini. Yep. You and and this feels like it could be on an entirely different level.
Speaker 1:Yeah.
Speaker 2:Well, you know, my favorite thing is Gemini wasn't what actually, like, made it rip. It was Banana Banana.
Speaker 1:Yeah.
Speaker 2:Like, the ratio, like, really improved in terms of OpenAI to Gemini Totally. And, like, way before. And then, like, Gemini, like, slightly helped, but I would say it's, like, 90% is Banana Banana.
Speaker 3:No. No. No. No. Put Doug in a bikini.
Speaker 1:You can yeah. You can just share an image, and it's immediately apparent what is going on. It's a unique capability that you can't get anywhere else. They've cornered the market, specifically on, like, the image editing, not just the diffusion, but the like, being able to Yeah. Take a photo, change the background, and have it actually look like your face or have the text look great.
Speaker 1:Like, it was it was a unique unique product, really, beyond a model.
Speaker 2:Yeah. Oh. Oh. This is by the way, this is my final steaming hot take in COD Code. Please.
Speaker 2:The reason why you should actually pay attention so much is because this is the first time like, image models essentially always gain share, video models always gain share, like Studio Ghibli moment, and then obviously, Chatuchi BT. This is the first, like, new moment. Yeah. It's an in a new modality, being the agent, and it's, like, actually kicking off.
Speaker 1:Yeah. I mean, how how important do you think the co work, like, desktop app mobile functionality is to that? Because, like, the like, you can have One one second. Truly magic
Speaker 3:Chad has some insight. A lot of people using Grok video to compete for a $1,000,000 Oh, that's right. Content. Free money.
Speaker 1:You have Free money. Okay. Yeah. So it's it's free cash and then free money. The the the top two apps and the top three give you cash.
Speaker 1:That is a good fact check. Thank you, Chad. But, yeah. My my question about, like like, can you have a Studio Ghibli like moment if you have to open up a terminal just because there are so many normies that would just will never open the terminal no matter how magical the, you know, AI god is behind the terminal. It's just too much to go type one line of command.
Speaker 2:That's why co work and Codex are gonna, like, probably what actually happens. Yep. I think it's really fun to, like, play around in the, like, whatever 1% adopter.
Speaker 1:Yeah.
Speaker 2:And I'm really enjoying it, but I just don't think Yeah. Like like, yeah, it's gonna be Cowork or Codex. And Codex is actually pretty good. Codex is, I think, a slightly more polished experience than co worker.
Speaker 1:Yeah. Last question for me. Take me on the journey of what's going on with Microsoft, what you predicted, the how that's changed, what how their strategy has changed. Like, give me the the proper way to understand Microsoft these days.
Speaker 2:Yeah. Microsoft's not in the race, bro.
Speaker 1:Why? They Where where are they?
Speaker 2:They they I know, but they're getting owned.
Speaker 1:They have all of the IP. I I I don't know. I agree with you. I don't understand why it's not like, oh, 5.3 launches? Microsoft's announcing it the same day, and it's actually integrated, and people are using it, like, on day one.
Speaker 1:It takes time.
Speaker 2:They it's a skill issue. Yeah. It's clearly something's going on. And and, honestly, the thing that makes me most bearish that is the fact that Satya is like, I'm not the CEO. I'm the I'm the product manager of Copilot because I'm so boned if I don't get this figured out.
Speaker 1:Yeah.
Speaker 2:Like, you can argue it is the it is now existential. Mhmm. He's decided, like, hey. My CEO job is getting this one thing right. Otherwise, we're screwed.
Speaker 2:And that is kinda worrying.
Speaker 1:It does feel like they could potentially
Speaker 3:it feels like Google went over the past. At the beginning of last year, kind of the the the the quick pause
Speaker 2:Mhmm.
Speaker 3:Is now looking silly in the context of
Speaker 1:It was on
Speaker 3:the coming in now and saying, yeah. We're everybody's on board now. Yeah. We'll see.
Speaker 2:Yeah. Yeah. We'll see. I think they have the most to lose.
Speaker 3:Mhmm. What about utilization? Brad Gerstner was hosting CNBC today, was very cool. Mhmm. Talking about how, you know, inthe.com build out the the dark fiber with something only, like, 7% of fiber that was being laid was actually being used.
Speaker 3:It was, like, obvious even at the time Mhmm. That and yet now, we're seeing GPU utilization rates, you know, maxed out.
Speaker 2:Yeah. Yeah. I think that's a pretty good counterpoint to anyone who's like, oh, whatever. Like, at this point, h 100 pricing has massively firmed up. B 200 pricing definitely has super firmed up.
Speaker 2:And, like, hey. There's there's clearly demand. I mean, you know, whatever they're doing on the other side of it, that's like that's the customer's issues. But, like, I I mean, I still think like like, honestly, man, the codex or sorry. My brain's all messed up.
Speaker 2:Cloud Cloud Code has been the most magical moment in technology for me in, like, my entire time, I think.
Speaker 1:Yeah. I
Speaker 2:it just feels awesome, man.
Speaker 3:Since the Game Boy.
Speaker 2:Dude, this is better than Game Boy's for me. I'm an information addict, though.
Speaker 1:Yeah. Makes sense.
Speaker 2:I I am. So
Speaker 1:Well, we appreciate you taking the time to come chat with us and writing about it and everything that you do. If you're listening, hop on Semi Analysis.
Speaker 3:Up for the 10,000,000 a year plan.
Speaker 2:Do the 10
Speaker 1:get
Speaker 3:million dollar a year Doug's phone number, you can text him.
Speaker 2:You actually can, actually.
Speaker 3:Yeah. I know. I know.
Speaker 1:You should just do it though.
Speaker 3:I mean, it's underpriced. You're giving it away at at it's taking away your time from your all your different agents. So you gotta price it right.
Speaker 2:I mean That's right. You know? My manager will hate that.
Speaker 1:Yeah. Have a good Oh.
Speaker 3:Rest of you,
Speaker 1:Doug. Day. Have a good weekend. We'll talk to you soon.