TBPN - Microsoft Cloud Outpaces AI, Disney’s AI Dilemma | Samir Chaudry, Dan Wright, Flo Crivello, David Tuttle, Will Ahmed, Aneesh Dhawan

Episode Date: August 4, 2025

(03:34) - Microsoft Cloud Outpaces AI (33:41) - AI Coming For Consultants (01:03:35) - Disney’s AI Dilemma (01:17:52) - Timeline (01:42:09) - Dan Wright, Co-Founder and CEO of Armada, discusses the company's mission to build hyperscale data centers for edge computing, targeting the 70% of the world lacking AI infrastructure. He highlights their work with critical industries like energy, mining, and defense, emphasizing the deployment of mobile, full-stack AI data centers that can rapidly utilize stranded energy sources, such as natural gas, to meet the growing energy demands of AI. Wright also mentions partnerships with major entities like Microsoft and the U.S. Navy, underscoring Armada's role in providing distributed compute solutions in remote locations. (02:05:58) - Samir Chaudry is an American content creator and entrepreneur, best known for co-hosting "The Colin and Samir Show," a YouTube channel and podcast that explores the creator economy through interviews and industry analysis. In the conversation, he discusses the challenges and opportunities presented by AI in content creation, emphasizing the importance of human connection and community in the evolving digital landscape. He also shares personal experiences, including the loss of his home in a fire and the birth of his child, highlighting resilience and adaptability in both personal and professional spheres. (02:33:14) - Flo Crivello, founder and CEO of Lindy, discusses the recent launch of Lindy 3.0, emphasizing its enhanced capabilities in creating powerful AI agents with ease. He highlights the introduction of templates for various functions like sales, customer support, and engineering, allowing users to set up agents in under two minutes. Additionally, Crivello introduces the 'autopilot' feature, enabling agents to operate their own computers and handle tasks requiring login credentials, such as managing support inboxes or cross-posting content between platforms. (02:43:49) - David Tuttle, co-founder and CEO of Rune Technologies, discusses the company's recent $24 million Series A funding led by Human Capital, with participation from existing investors like Andreessen Horowitz and Point72 Ventures. He highlights the deployment of their AI-driven logistics software, TyrOS, in military exercises, emphasizing its role in modernizing military logistics by replacing outdated manual processes with intelligent, autonomous systems. Tuttle also underscores the importance of integrating advanced technology into military logistics to enhance operational efficiency and readiness. (02:49:59) - Will Ahmed, founder and CEO of WHOOP, a Boston-based wearable technology company, discusses the recent launch of WHOOP 5.0 and WHOOP MG, introducing new hardware, membership tiers, and features like the Healthspan tool, developed in collaboration with the Buck Institute, to provide users with insights into their biological age. He highlights the company's efforts to navigate FDA regulations, particularly concerning the blood pressure monitoring feature, emphasizing WHOOP's commitment to delivering wellness-focused innovations while addressing regulatory challenges. Additionally, Ahmed announces the upcoming Advanced Labs feature, enabling users to integrate and analyze blood test data within the WHOOP platform, further enhancing personalized health monitoring. (03:02:28) - Aneesh Dhawan, co-founder and CEO of Knit, discusses the company's recent $16.1 million Series A funding and its mission to revolutionize enterprise consumer research through AI-driven solutions. Knit's platform enables businesses to obtain actionable insights in days rather than weeks, significantly reducing costs and time. By integrating AI with human expertise, Knit ensures high-quality, context-rich research outputs, positioning itself as a leader in the evolving market research industry. (03:09:26) - Timeline TBPN.com is made possible by: Ramp - https://ramp.comFigma - https://figma.comVanta - https://vanta.comLinear - https://linear.appEight Sleep - https://eightsleep.com/tbpnWander - https://wander.com/tbpnPublic - https://public.comAdQuick - https://adquick.comBezel - https://getbezel.com Numeral - https://www.numeralhq.comPolymarket - https://polymarket.comAttio - https://attio.com/tbpnFin - https://fin.ai/tbpnGraphite - https://graphite.devRestream - https://restream.ioFollow TBPN: https://TBPN.comhttps://x.com/tbpnhttps://open.spotify.com/show/2L6WMqY3GUPCGBD0dX6p00?si=674252d53acf4231https://podcasts.apple.com/us/podcast/technology-brothers/id1772360235https://www.youtube.com/@TBPNLive

You're watching TVPN. Today is Monday, August 4th, 2025. We are live from the TBPN Ultradome. That's right. The Temple of Technology, the Fortress of Finance, the capital of capital. We were recently featured in the New York Times podcast, their headlines podcast, and they quoted that segment of the stream,

and we snuck in an air horn. So we're bringing the air horn to Legacy Media. It was very satisfying. What a week last week. Absolutely insane. We did the Odd Lots podcast. That's out now. You can go listen to that.

Bloomberg Odlots with Tracy Allaway and Joe Wisenthal. Had a great chat with them about the AI talent wars and who Mark Zuckerberg's white whale is in the Moby Dick analogy. We kind of settled on Ilya Sutskever, but we're going to read the book. We're going to have Joe Wisenthal on the show later and get to the bottom of the full metaphor of if we are living in Moby Dick world, who's who. Who's who? Who's Ahab? Who's the whale? There's a whole bunch of funny side characters that I was trying to map out to different AI feeble and I only got like halfway there, but I want to do that. Anyway, big news from Microsoft This was interesting. So this is from the Wall Street Journal. Microsoft is an AI darling, but its core

businesses are booming too. Did a bunch of reading into this also. Let me tell you about ramp time as money say both easy use corporate cards, bill payments, accounting and a whole lot more all in one place I love this one you know what John switch your business to ramp dot com love that anyway quick note please Microsoft yes 4T 4 trillion they did it hit that gong Jordy hit that gong jordie hit that gong jordy hayes we found a video of somebody playing a gong like much much more professionally than we do where they like have you different tools and they can bring out different songs they can make it sing and so you know late night has the live band I think a live gong player might be in our future I think that I think if we ever if it's ever

discover that TBPN goes bankrupt because of our incredibly high cost structure it will be because we have the world's foremost gong athlete yes player yes gong artists truly the greatest gong player of all time will be on staff but you know why the Microsoft news or hitting that number is meaningful. Why? It's because everybody has always said Sotia is more of a more of a two to three guy. People were saying that. Yeah, he's really good in that range. He's in that range. I don't know if he's in the four trillion club in the four club. Got it. Got it. There's not a lot of guys in that club. There's one guy. Yeah. Jensen Wong. He's a founder. Founder mode. So this is, yeah. So is Sotcha in

founder mode or is founder mode defeated? Because if a manager mode can take company to $4 trillion, you know, All bets are off. Put your company in manager mode. As long as you can recruit Sachinadella to be your CEO, I think it's okay to step back. Manager run companies can excel. Exactly. Yeah. So if you're like, I'm a little burned out as a founder.

I want to take a break. I'm going to step back. I'm going to get a manager in. Normally very bearish. Just give Sachinadella a call and you'll be okay. You'll be okay. In that case. Anyway, this stream is made possible by re-stream one live stream, 30 plus destinations,

multi-stream and reach your audience wherever they are. Very exciting company. to partner with them. Anyway, so from the Wall Street Journal, Microsoft's blockbuster earnings last week. They blew out earnings. It was very exciting. They cemented its status as one of the biggest winners in the artificial intelligence boom. We knew this. Sotcha had carved out just a massive amount of territory. GitHub co-pilot. First, real major breakthrough product that was monetizing on top of GPD 3.5. Great product, obviously led by Nat Friedman when he was there. Last we heard it was at around half a billion in ARR, but that was two months ago, so it's probably

billions now. Yeah, probably bigger. And then the crazy Open AI deal, they got in very early. They have this massive revenue share. They have an ownership. They get a copy of any software that Open AI writes basically or acquires. That's kind of crazy. And so.

Sotia did seemingly one of the best deals of all time for Microsoft. Potentially. Yeah. I think so. It's hard to, I mean, it's an awkward situation now. And it's a really hard decision. Sam is, you know, renegotiating the deal. Yep.

And Ben Thompson was noodling on this like, should they take the money now or should they play it more like a venture style bet? What is the role of the CEO of a $4 trillion public company where the shareholders have different expectations? They're not seeing, if you hold Microsoft stock, you're not feeling like you're an LP in a venture fund. Yeah. So if the CEO comes to you and says, hey, look, we're taking the cash flow now. flow now we're going to divot in some of this out we're selling down the position we're we're thinking strategically about this as opposed to just we want the IS multiple we want to ring the gong on the deal that could be reasonable so

obviously we're we're tracking where that goes but um they also have clipy generational precursor to potentially all the AI agents the original AI agent clipy potentially make a comeback the original super intelligence the original super intelligence and that was the first time they had to move the goalposts and This is why we're worried about getting paper clipped. Because Clivey will become too strong. I mean, with the power of Open AI and Microsoft Azure, anything's possible there.

So anyway, outside the AI race, Microsoft is minting money from corporate customers spending on regular technology. Long a sweet spot for the company. Many companies are shifting from buying their own IT equipment to renting it from Microsoft through its cloud's computing service. They are also renting more standard issue computing stuff, hard drives for data storage, for example, to support their AI efforts.

And so this is the key. stat from the earnings call that the Wall Street Journal is highlighting and then we'll kind of dig into this number and what it means because there's a lot of different explanations for what could what could be going on. But Microsoft and the CFO and CEO didn't necessarily give all the context that we'd like to set this definitively, but I think there's some really good theories floating out there. So the quote from the Wall Street Journal is a large chunk of the recent strong growth in Microsoft's cloud business called Azure stems from that. More than half of Azure's 33% revenue jump in the company's March quarter came from non-AI services.

While the company didn't give a comparable breakdown, a comparable breakdown of the cloud unit's 39% growth in its June quarter, it said that the core infrastructure business, Microsoft's lingo for its non-AI cloud business, was the driver. A massive win for enterprise SaaS, just good old-fashioned SaaS. No, no. So just the thing, it's not SaaS. Yeah, so it's people migrating from on-prem to... It's what's called infrastructure as a service.

Yeah. So there's software as a service. That's when you go and get teams or you go and get a subscription to Excel in the cloud or Outlook in the cloud. That's software as a service. I was saying more traditionally someone else using... Yes, yes. Using it as infrastructure as a service.

Then there's also platform as a service. That's like Heroku on ADUS where you go and you deploy an app. You could think of maybe even like a replet as like a platform as a service almost where they're hosting you, but they're not just providing you the raw infrastructure. Azure's Azure's core infrastructure business is essentially infrastructure as a service. IAAS is the term. And that means, oh, you want some CPUs and some hard drives and some Ethernet cables and moving stuff around data transfer. You had an interesting thesis offline earlier. You comped to people in the internet era buying a computer.

hear about the internet they're like hey I think this might be a thing I should get a computer yes yes yes and just buy a computer yes and now you can see something you know where companies say hey this AI thing might be big we should get on the cloud yeah yeah you want to set yourself up for it and I think if you have a whole bunch of data in some sort of on-prem you know you have a data center for all of your data in a bunch of hard drives and you have CPUs that can do the do different workloads and data workloads and maybe you're using some SaaS on top of that, but you realize that you're never going to be in a position to buy

100,000 H-100s, and you're going to wind up being a leaser of that for some small fine-toeing run. Or just using other people's application layer products. Yeah, and so the integration that comes from being in the Azure ecosystem, that could be a driver. There's a few others. When I think about the core, the AI Azure services, I think of that. Almost as, you know, it is SaaS.

Like if you go to Azure and you say, I'd like to, you know, put my credit card down and I want to be able to use the GPT4 API. And I also want to be able to use Lama 3. And I also want to be able to be able to use deep seek. And I want to be able to call all these APIs within my product. I almost think of that as tokens as a service. Like it is SaaS, but it's something else.

And I think these token factories, I think this idea of how much revenue are you generating from your token generation business. is really what we're talking about when we talk about Azure's core AI products versus the infrastructure. But there's a bunch of interesting wrinkles that could be going on within the classic core infrastructure business. So this is of course is virtual machines. You just want a Linux box with a CPU to host a website.

That's something that you do on Azure. Storage, networking. Okay, I want to store all my data. You could go to AWS. stored in S3, you could go to Google, stored in BigQuery. You could also go to Azure and fire up any sort of storage database or just raw hard drives, and then networking, moving stuff around. So this is the infrastructure as a service

versus tokens as a service, their higher level AI APIs. But so GPUs are, so the question is like, why is their infrastructure as a service growing faster than their tokens as a service product? You would think that, Microsoft is going to their enterprise clients and saying, like, you need to build AI, you need to bring AI into your products,

and we have all the best APIs, so just buy tokens from us. And you think that that would be the boom. Well, couldn't the other factor here be that they are massively supply constrained on the GPU side? And they have this multi, you know, multi-tens of billions or hundreds of billions of dollars of backlog, but they can't fulfill.

Meanwhile, they had a, you know, more kind of like predictability on the, the traditional data center cloud side that they were able to scale up to. Yep. So that feels like a potentially like a pretty big driver here. Yes, definitely. But it's still shocking. And so as companies come out and they say, okay, we are scaling our, you know, hardware

and software, our technology footprint broadly going into this AI era. We're excited about this stuff. Well, we're also going to need more databases. We're going to need to put more data in those databases. We're going to need more CPU workloads. We're going to need more of everything. And Microsoft's like, yeah, of course, we can definitely get you a whole bunch more hard drives and a whole bunch more CPUs. We're not constrained on that at all.

And the KAPX is keeping up so they're able to service that. There's also an interesting thing where tons of AI stuff can technically be happening inside the core infrastructure bucket. You just don't necessarily know what's in there. So some examples are like if you're, let's say you're a pure AI company, or you have a new AI workload, you could go to Azure and say, hey, I'm gonna do my own AI thing, but I need a ton of storage for data

because I'm gonna be training on it. I'm gonna need a bunch of networking to move that data around when I do a training run. So that could be driving core infra up. And then also, if a bank hypothetically spins up a huge cluster of H100 GPU virtual machines to fine tune an open source model like MetaSala 3, this would show up as core infrastructure,

not Azure AI services. And so that's like textbook AI, boom, but it's just happening in the wrong bucket. And then I also saw a post that potentially chat GPT counts as Azure core infrastructure because they're not serving chat GPT through the Azure AI API. It's not this like snake eating its tail or a boros. It's like open AI just came and said, give us a whole bunch of them.

The headline itself ends up becoming pretty misleafy misleading. Exactly. And again, this is this was probably, you know, a lot of people were reading into AWS's growth, you know, the reports that, you know, Andy or the comments Andy had given on AWS last week. And the big thing that AWS is missing is having a chat GPT building on top of. Yeah. Yeah. Yeah. There's no dominant consumer product, at least at that scale. ChatGPT I think as of this morning, somebody who's estimating getting to a billion weekly actives this year, which, again, those types of products don't, you know, the power law is like extreme, right?

Yeah. So if you, I mean, I believe Anthropic is pretty tightly hitched to Amazon, and I think the next big cluster from Anthropic will be powered by Amazon for the most part. And so they're getting there, certainly on the, if they're building all the, all the infrastructure Yeah, but again, that would show up on core. Token generation side, like more of the... No, no.

So if Anthropic goes to AWS and says, we want you to build a huge data center to serve cloud code for us, that's going on an infrastructure. Yeah, not actually APIs. But when you go to AWS and you're just some random company and you say, I need a database and I need some storage and I need a web server

and I also need a bunch of tokens from whatever model you can serve me, like we got Claude and then you're like, yeah, let's pull the Claude tokens into my app. It's a good model. That's token as a service. What? It's a good model, sir. It's a good model, sir. Exactly.

One thing that stood out to me, that has stood out to me across this year with Microsoft is they've done more layoffs this year than the past three years before that combined. So 2022, 2023 and 2024. That's crazy. So this just shows the level that Satya is operating at is like the company has been on a tear. this year performing exceptionally well, and he's still thinking about how do we get more and more fit. And so.

So to be clear, literally every piece of Microsoft business is growing and at a very solid clip. So Microsoft 365 commercial cloud business, which houses remotely access versions of Word, Excel, other productivity software, that grew at 16% from a year earlier. So that's, I mean, it's not the most insane growth rate, but that's still crazy because you think about like, who doesn't have Excel that needs it?

Like who are these people who are like, you know what, 2025 is the year that my company's getting on Excel. We're doing it. Well, it's just crazy when you compare it to AWS growing at 19%. Obviously very different scales. Yeah, totally. You would think you don't wanna be in the same ballpark as.

And so that was the news of the AWS, if you missed it. They beat earnings, they did very well, but they weren't growing as fast as the other the other cloud platforms, Google and Microsoft. And so the Amazon stock traded down. And I mean, the narrative around AWS is different because Microsoft has Open AI, Microsoft Research, and has GitHub co-pilot, and is like really moving things

forward in the AI world. And Satcha is seen as someone who goes in the Dorcasch podcast and talks about AI. And it's clearly like really on top of it. And obviously Google has Gemini and a million different products and and strategies around rolling that out and staying on the frontier. I mean, they have a frontier lab internally and Amazon's just not there either on the partnership side or on the core like training frontier lab side.

And so it's a little bit, it's a little bit of both. Anyway, let me tell you about Figma. Think bigger, build faster. Figma helps design and development teams build great products together. We had a great week in New York celebrating Figma and the IPO. Stock's been up, stock's been down, wild ride, but we are still very happy to be partnered with Dylan Field and the Figma team. And I'm super excited to see what happens next because truly.

Thursday was such an incredible day, just getting the, if you didn't get a chance to listen, we talked with every the seed lead investor all the way through, of course, Andrew Reed, who led the sea. Yep. And then kept it off with Dylan and also got to speak with Lynn Martin, president of New York Stock Exchange, as well as Chris, the CTO of Figma. So really incredible day and it's just very proud of the Figma team. Yeah, it was awesome.

In one sense, investors might prefer to see AI businesses driving growth. That, after all, is what has driven the company's valuation through the roof. But tech companies' stocks arguably hinge on too much on AI to the extent that they can keep increasing other revenue streams. They are on more solid financial ground. Of course, none of that means anything if all of the growth is coming from Open AI, but at the same time, does anyone really think ChatGBT BT is Yahoo anymore? Like, you know, they're generating what a billion dollars a month in revenue at this point. Everyone uses the app.

It's installed everywhere. Like that, that token demand is not going anywhere. That infrastructure demand is not going anywhere. On the other side, Amazon is down just over 9% since Thursday. Since Thursday. earnings. And then this morning, unrelated, they announced that they're shutting down Wondery, the podcast studio. They acquired in late 2020. Wait, why did Amazon? Why did, the market's way up. Wow. Nasdaq's up 1.8% today after a brief sell off on Friday. Good news. Fair market is over.

Yeah, we were so over, but we're already so back. It's fantastic. Love them that happens. Markets go up, markets do go down. But the march of technological progress, the arrow points. in but one direction. That's right. And its march is relentless. There is another silver lining for Microsoft. Non-AI sales can be substantially more lucrative than AI ones. Non-AI gross margins within Azure were around 73%.

Wow. That compares to 30 to 40% gross margin for AI. He estimated because of the huge cost of setting up AI infrastructure that makes sense. You get more margin on just a bunch of CPUs and databases that you've harnessed and built everything around. And also, you know, you still have that interesting dynamic where it seems like all of the cloud, the hyperscalers, like, don't really compete on price because they're all pretty comparable, but they seem to all have really good margin. That's what it seems like. Yeah. I'm not exactly sure if there's something else that's more fundamental going on.

And the Coca-Cola dynamic is like, I don't know if we, it's hard to tell what conversations were off air on air, but John last week, forget when was describing how you would think. that Coca-Cola or Pepsi would decide to get aggressive on price to try to gain market share and get people to switch. But ultimately, that would just lead to a price war with both companies, you know, massively eroding their, you know, margins. And then... Did all the RC Cola coded then? Yeah. Every...

So if you have a fierce competitor, consider entering an unspoken gentleman. That's the big. Maybe a gentleman's agreement. It's not, it's not even a gentleman's agreement. It's not, it is unspoken, but it is a, it is a natural game theoretic Nash Equalibor. Like it is the natural state of things that both sides understand that to go to war would be mutually assured destruction. And so they don't even need to talk about it. And so instead they both agree to keep prices where they are and instead compete on marketing.

Compete on marketing. Compete on marketing, really. Yeah. Yeah. I mean, they don't form, they don't reform. often. They mostly compete on on on marketing. And that allows them to have this like continually compounding business. And that's why it's in the the Warren Buffett portfolio. Coca-Cola. He's been in there for a long time. And Pepsi's been doing one too. And he was a DAU, of course. He was a DAU of Coke.

Still. Diet Coke or Coca-Cola? I think Coca-Cola. No, Diet Coke. Oh, he's a Diet Coke guy. I feel like he was a Coca-Cola guy. Warren? Yeah, look that up. I want to know. Anyway, I'll keep from this Wall Street Journal report. Luckily for Microsoft demand for lucrative non-AI services appears to be reasonably strong. Measures of broad IT spending. There we go. You know why?

He was a regular Coca-Cola guy. You know why? It's because it has corn syrup in it. It's literally corn grown from Mother Nature from the Earth. And then syrup, it's what you put on pancakes. Like, it's the most wholesome combination of foods you could imagine. Maze. This is something that's been grown in America for generations.

And corn is so popular in Nebraska, right? It's grown everywhere. It's culturally significant. And the syrup that you put on pancakes, it's the most American, most wholesome ingredients. Not this like refined sugar, this crazy stuff from somewhere else. No, it's American. American. Corn syrup.

There's nothing. It's Lindy. Corn syrup. Yeah. Ced oil. Ced oil. Cato oil. Catero.

Catero. Cateroal. That we need to return. to corn syrup. None of this, none of whatever's in this Coke Zero. Yeah, your grandpa was, was drinking corn syrup. Yeah. Oh, and it's too good for you because you read a couple posts on X.com. Think you understand something better than corn, delicious corn, corn, corn on the comm. Something you have on a barbecue? Oh, now it's too good for you. Can I possibly have corn?

What's next? No apple pie? What's next? No, no, no, no, no, no, No rotisserie chicken. No turkey on Thanksgiving. Somebody's going to listen to the show for the first time today and just go raging for you promoting corn syrup consumption. It's as American as apple pie. And Warren Buffett knows best. He's doing great. And of one study, sort of a Brian Johnson.

Yeah. He's sort of the Brian Johnson. He's the original don't die. Yeah. And he's been doing fantastically on that front. Yep. He's great. Anyway.

Measures of broad IT spending were fairly muted at the start of the year as companies pondered the impact of Donald Trump's tariffs and concerns bubbled about the health of the global economy. Attitudes appear to have improved somewhat in the second quarter, though. A UBS survey of cloud computing customers in July showed a clear improvement in tone about spending. Most were moving forward with efforts to migrate computing work to the cloud. They're like, this internet thing is real. It's real. We got to put the data in the cloud. We held back as long as we could.

We could. But it's 2025. We have no more excuses. The tariffs that's come and gone. Now's the time. We were resisting the 21st century. Yeah. But we're a quarter the way through.

Put the data online. It's not going away. So let's use the computer online, in the cloud. Put the docks in the cloud. Put it in the cloud. Just put the docs in the cloud. Put the fries in the bag. Anyway.

And put your. compliance process on Vanta. Automate compliance manage risk and prove trust. Vantta's trust management platform takes the manual work out of your security and compliance process and replaces it with continuous automation whether you're pursuing your first framework or managing a complex program. In the longer term, there is a little question that

cloud computing is going to grow in ways that play to Microsoft strengths. It's rivals, mainly Amazon.com, and Google are growing quickly too, but don't have all of Microsoft's broad corporate software offerings that enhance its cloud footprint, even outside of AI. Amazon on Thursday said it's cloud, It grew at 17.5% in the June quarter, disappointing investors and forcing CEO Andy Jesse to answer questions to answer questions about Azure's outperformance. Why are you getting beaten?

You created this category. You created this product. Why is Sotchenadella getting you? You are the cloud. You are the cowboy of the cloud. You're getting put out to pasture. Recent quarterly earnings in Azure's favor were really just moments in time. he said chart John the uh oh yeah wow that that's worse than I thought I wasn't sure we

had that pulled I mean it would have been hard to predict five years ago that we'd be sitting here with with Microsoft at four trillion and and uh what's the Amazon it think 2.2 trillion 2.2 2.2 2.27 okay so almost double still magnificent but so magnificent, but you got to keep fighting. Got to keep fighting. The company's stock fell 8% Friday, and it looks like it still slide down. The question for Microsoft's investors then is less about its prospects than its

valuation. The company's stock is up nearly 40% since the beginning of April, pushing its forward price earnings multiple above 33. That's a bit richer than Amazon and a large margin above Google, which is trading at a multiple of roughly 18 times. That should be easier for investors to digest because, Because while Microsoft's AI growth is real, it is far from the only thing going right at the software giant because they got Excel. They got core infrastructure.

They got AI, APIs. They got tokens of the service. Infrastructure as a service and software as a service. They got the Royal flush. It's going well over at Microsoft. They won't need to call McKinsey, but maybe, maybe Amazon.com will. And that brings us to our next story. Well, before we got to jump in,

quickly. Graphite, code review for the age of AI. Graphite helps teams on GitHub, shift higher quality software faster. I'd love to do an ad read, George. It's very clean. We got to hang with Merrill. Yes, we got to hang with him. Last Thursday, which was fun.

In New York, which is great. He came with a tux. He looked fantastic. He looked incredibly sharp. Yes, it was great. Great sign of respect in our culture. Well, American Eagle Outfitters, Inc. Is up 23% today.

No way. after Trump came out with a lot of excitement for the advertisement this morning. I'll let people go read it, but he did describe it as the hottest ad out there. And it certainly struck a chord last week. But anyways, let's jump into this next story. Really quickly, before we go that,

the battle for the Mag 7 is heating up. Microsoft is now in the 4 trillion club. Let's go to the Polymarket. for the largest company at the end of 2025. Nvidia has a solid handle on it. As of December 31st, they are expected to be the largest company in the world at 63%. Microsoft is at 28%, but climbing up from 22% earlier.

Apple has been falling. They were the largest company, and now they are falling down to just 6% chance. And of course, Tesla's hanging out there at 3%. Anything could happen with Tesla. one crazy move, one naturally aspirated V12 cyber truck, and boom. That's all it would take. It's a $5 trillion stock.

What about a V12 optimist? Like a gas powered. Okay, I have a pitch for you. Richard Mill of cars. A Richard Mill for the streets. Richard Mill for the streets. A racing machine. Basically the pitch is that now it's routine, I would say,

Most people, if they put their minds to it, they'll be making $100 million a year. Seems like that's kind of the benchmark for like you did okay in life. $100 million a year. Even a million dollar Ferrari is not really gonna stand out. That's three days of income, right? So we need a car brand to come out

where the floor price is like $20 million. Yeah, Bugatti doesn't quite get you. And the thing with the Bugatti is that it's like this hypercariness. 2,000 horsepower. It's so extreme. The whole point of the Richard Mill is that you can wear it to the gym, but it's $250,000. So I'm looking for something that's like a daily $20 million dollar, yeah, $20 million sedan, $20 million dollar. Pickup truck. Just crossover, just mid-size SUV, $30 million. This is what we're going to do in here. This is what we're going to do. Because a lot of these, a lot of these AI researchers, they don't want to pull up in a Bugatti. Yeah. But they have the money to spend.

Something a bit more understated that still says to the right audience, I spent tens of millions of dollars on my daily. Exactly. Exactly. That's definitely what they're going for over at MSL. If you pull up to meta-superintelligence lab, like you don't want to flex, but you want people to know. We need satellite imagery of the MSL parking lot. Yeah. That would be funny if it just looks like a car meet. Yeah. Everybody has like a SVJ,

Konigseg, La Ferrari, La Ferrari. F40. Somehow I think it won't be that at all. It won't be that at all. But do you think that there's actually an opportunity to start a new car company that it takes the Richard Mill approach and is 10 times what the luxury cars are?

So like maybe it's not 20 million, but like a G-Wagon is a functional car in the $200,000 range. something at two million. The thing with RM is they made something that was light and incredibly durable and that was effectively...

I don't think you can make a light and durable car for $2 million, Doran? What are you talking about? World's lightest. I mean, yeah, that would that be interesting. It just drives like in a go-kart. It's just totally insane. But it feels like it came out with...

I don't know that you necessarily need to optimize performance. It was aesthetically in an entirely new category. It was it had the performance, it knew levels of performance. Yeah. Within the category broadly from a, from a durability standpoint. So I think, I think with the RM, it does stand out.

It looks completely different, but it's still in this weird watch world where to just a completely average person, it's not as bold as a cyber truck. It doesn't, it's a, it's still an if you know, you know type accessory, right? And so that's where you need to target. It needs to look kind of like, is that a BMW or an Audi or Model Y? It's like, it blends in a little bit. I think one reason, not to crush your dreams, one reason potentially many that this is not as exciting as it sounds.

I mean, I want to drive it, to be clear. But I think that the automotive regulations around, you need to have a seatbelt like this, the seats need to work like this. It needs to be able to handle this type of, Okay. Maybe there's some sort of like, you know how Steve Jobs famously bought a new 9-11 every single year? It was every 90 days, I thought. Every three, yeah, I think you're right.

Every three months, Steve Jobs would buy a new Porsche 9-11 so that he didn't ever need to have a license plate on it. Just for anonymity, I suppose. Yeah. So he would never be identifiable. Which is potentially an incredible excuse for being just like, I like to drive brand new cars. What is what is breaking in three months on a 9-11 the new car smell goes away by that point maybe he just really love new car smell he's like I could never do a new car smell air freshener that's that simply won't do so after my new way and then later in life I believe he switched to

Mercedes and I think he got the SL 63 or something equivalent to that so like the 9-11 equivalent of in Mercedes world but yeah maybe that's the hack so so you find something that would never be approved, but you do some crazy thing where you bring it in on show and display and you can only drive it a few thousand miles a year, but then you're swapping them out. And so you're kind of doing like a subscription. And so your real cost to own is in like the $5 million range.

But the beauty is that you're getting something that's just complete regulatory and run around the rules so that you wind up with this very unique, very different, experience, but only made possible at the extra high end. I think at that point, helicopter. Maybe. Maybe helicopter's the move. Let's get into the next story.

The journal says AI is coming for the consultants. Inside McKenzie, they're saying this. McKinsey? Somebody's got a case of the Mondays. McKinsey. They say this is existential. This is wild to put on the record. Yeah, I was talking with a buddy.

We were talking yesterday, or not yesterday, last week. And he said, word for word, adopting too much AI will hurt our bottom line. That's right. And that totally makes sense, but it's not something that lawyers are actively saying out loud because you don't want to be the law firm that's, anti-tech. If you go out as a big, if you're, you know, one of the top 100 law firms

and you're like, yeah, we're excited about the potential of AI, but we won't be adopting it internally because it, you know, naturally it's going to hurt our profits, which you can just translate to, we're just not going to be able to bill our clients for the same number of hours. And so that statement made me bullish on, why am I blanking on the name of this company?

We had this company on, not Harvey, the other one that is doing just like contract, like NDAs and a few other types of contracts, but they're an actual law firm. Oh, interesting, yeah. And the Atrium, Clear Spire model. Similar to that. Fascinating model. I don't know if you remember this, but Justin Kahn when he started Atrium had a very complex corporate structure

where the company was half a law firm, an LLC, which is the traditional corporate structure for a law firm and employed lawyers. And then there was a tech startup, C-Corp that was bound together with a master service agreement or something like that. So these two companies were combined to be a tech-enabled law firm. And it was probably a good idea just a little bit too early. But interestingly, there was another attempt to do the exact same thing 10 years earlier called Clear Spire.

And that didn't go well either. And it was started by some All-Star lawyers and some technology. and they came together and they were like, let's build this AI enabled law firm. Maybe now is the right time to build the AI native law firm, but I don't know that they need to be doing technology development. I think that the next world might be you have a power law winner in legal AI tooling, maybe it's Harvey,

maybe it's someone else or a combination of firms, and then you have a very, very small team of high EQ deals guys lawyers who are phenomenal at the human element thinking creatively coming up with crazy structures understanding are you familiar with the battle of the forms this idea in law firm in like Lee in the legal world but I can guess what it what it looks like basically if you talk to most lawyers they will tell you that you can express almost the exact same legally binding theories or like rules in a contract or bullet points in a contract, everything you want in a

contract can be expressed both in one page or 100 pages. And so which one you choose to send to your counterparty is like an aesthetic choice. It's like for certain deals, you want to send the 100 page contract because you want to send the message that like, we're taking this really seriously we're thinking of everything but for other people you're more like we want to show that we're moving quickly we want to use the safe we want to use and and structurally it can be the exact same deal and so a good lawyer will know when to when to send a lot when to send a little when to slow a deal down when to speed things up and so you get a bunch of folks who are great and that type of legal thinking the creative thinking the human element the the relationship building the golden retriever mode

And then you empower them with the best in class legal research and analysis tools to actually instantiate and do all the grunt work. And then separately you have a company like Harvey that's building the tooling that then sells to the team of humans. And that law firm will have a very different cost structure, very different dynamic. Might even look like an older school law firm where it's just a few partners. One thing I know, lawyers are not going down without a fight. They're like, we're making 90% less per case. We're going to initiate 10 times as many lawsuits. That might be the equilibrium here, for sure, for sure.

Well, if you're planning to manage your lawsuit or build a legal AI company, get on linear. Linear is a purpose-built tool for planning and building products. Meet the system for modern software development, streamline issues, projects, and product roadmaps. Good transition, not great. We're not recommending you use linear to manage your lawsuit, but if you are building use it for everything. Yeah, John, John's very in favor of using linear for everything, including workouts. Haven't you seen those people that do art in Excel?

Then they zoom the grid way out and they color each, each cell is a pixel. Linear really is a canvas. I think it's a product artist. I think it should be. Anyway, let's go back to McKinsey. Companies paid dearly for McKinsey's human expertise and for nearly a sense. they have had good reason. The elite firm's armies of consultants have helped generations of CEOs navigate the thornyest of challenges, synthesizing complex information

and mapping out what to do next. Now, McKinsey is trying to steer through its own existential transformation. Artificial intelligence can increasingly do the work done by the firm's highly paid consultants, often within minutes. That reality is pushing the firm to rewire its business. AI is now a topic of conversation at every meeting of McKinsey. These boards, their board of directors. The firm's global managing partner has said this. The Bob Sternfels, I wonder the nomin of determinism there. Do you think he's a very stern individual or do you think he's really friendly and fluffy?

Should get him on the show. Just chop it up with him and just throw them tons and tons of jokes. I don't need to have them on the show to know. No, stern. This is like, you guys should be good doing so much better. You should be more serious. Just immediately stern. Stern mocks us.

The technology is changing the, ways McKinsey works with clients, how it hires, and even what projects to take on, and McKinsey is rapidly deploying thousands of AI agents. Those bots now assist consultants in building PowerPoint decks, taking notes and summing up interviews and research documents for clients. What do you think about building PowerPoint decks using purely AI? We were talking about this this morning. I think that AI can do a tremendous job, making you a generic deck template or just market analysis and turning that into page by page kind of analysis that you could generate slides against. But I think just grounding it in something

more specific, I think if you asked if you were just trying to build a pre-seed or seed or series A or series B, really any type of fundraising materials that are, or even like, a deck for a venture fund, raise, it's going to produce something that will not have you looked at very seriously. It will put together something that maybe like the most basic McKinsey associate would put together in terms of like, these are all the different types of slides you can have and, you know, here's data and and here's research that support these things, but the best decks take you on a journey. And my framework for making decks is you basically want,

in a fundraising context, you want every single slide to have like a very short sentence that takes, like, walk, that's walking somebody down this path that you're taking them on, where they just agree, agree. agree agree agree so they get to the end and if at any point in that deck they've read a headline and they're like I don't agree with that or that's dumb you're screwed yep but usually if you've you know thought through your or I'm confused or I'm bored confused yeah if you're bored

or confused yeah that's C tier but B tier is disagree decks uh decks just get looked at um it's not about like the supporting content it's about like the high level like narrative arc through the deck yeah and getting somebody to click through enough pages to actually care about what you're doing. And I think that maybe there's some really incredible prompting that you could do to try to come up with something like that, but I haven't been able to. Yeah, I just think the models aren't there yet in terms of like putting disparate information together coming up with novel insights, contrarian thinking, independent thinking, all of that stuff

is where the models fall down. But that's fine because they're fantastic and a bunch of other stuff and anything that they can RL on is fantastic. And so I would think about it like, you know, we already did the Figma ad read, but we're going to wind up doing another one. But, I mean, they have a tool for building decks. And obviously you've built decks and Figma for years. And when I think about as that tool becomes more AI native, I don't necessarily want it to try and one shot a prompt from build me a deck that convinces you that TBPN is a growing media company or something like that. I want a prompt to be like go through and make sure all the text is centered.

And then it just clicks through every deck and does the grunt work of centering every piece of text. Or there's so many times. Or you put all the content in and you say generate other variations of this slide. Yeah, yeah. Or make sure all the black font is actually the correct color of black because I've updated that, like the style guide. And then pull in this data and then keep the data up to date. Like all the grunt work that happens, that should be the domain of AI. And for a long time, that's been the domain of the low-level management consultant.

The person who is at a wedding and has to leave because their boss told them, hey, you've got to update slide seven. Yeah, hey, yeah, there's a missed comma in this deck. And I'm about to go present and it needs to be flawless. And so McKinsey is continuing saying they've reduced the head count from 45,000 people in 23 to 40,000 through layoffs and attrition. in part to correct from an aggressive pandemic hiring spree. It's also roughly rolled out roughly 12,000 AI agents, which is a very, like, that's not apples to apples at all. Like, you cannot just, you can not just.

It's also funny to be thinking about like the whole point of agents or that like anybody in a team can, like, spin them up and get value and shut them down. Yeah. And they're not necessarily perpetually running. So just to come out and be like, yeah, we have 12,000 agents on our team. Yeah. It's like how many... We rolled out 12,000.

It's like, okay, what percentage you're doing? How many Google Chrome tabs do you have open? Tell me that too, because that's useful. Like, yesterday I used agent mode to do some like really deep research and then I wound up and then it like was kind of working. It had this cool interaction where it asked me a question and I just wasn't online for that and it just typed for me, continue and then it just kept going. And so it wound up putting together really, really thorough result, but it took a long time.

But in the meantime... You sure it was agent? You sure it wasn't a ghost in the machine? It was a ghost in the machine. But it was weird. that it was like typing for me, but it was actually a great U.S. because I would have forgotten about it. But then I just went to 4-0, like, in the interim

and got like basically the same answer. Because like I didn't ask that complex of a question. And so I kind of like overestimate. I was like, I need the nuclear bomb for this one. This is the most brilliant question you could possibly ask. And it was like something that like probably Googlable. I think the thing with management consulting. Yes.

Is that it will the question, this question has come up forever. is the value of hiring McKinsey actually the strategy that you get or the advice that you get? Or is it outsourcing critical decision-making or backing up critical decision-making so that the management can say... I don't want to do the layoffs, but McKinsey told me I had to. Yeah. It's a lot of cover and that... Or I didn't want to go into the cloud of McKinsey.

Yeah, yeah, yeah. Yeah, yeah. It's a lot of defensibility. It's a lot of like internal office politics, it feels like. But I do think that there is, I think that that narrative is a little bit overblown. And I do think some of the top McKinsey folks are actually great at business strategy. It's just that when you think of McKinsey, you think of kind of the mid-tier junior associate who's still just kind of learning. And of course, you're not going to get that much out of that person.

So my question for this. So the key quote in here is that Katie Smage, Smage. a senior partner that Sternfels tapped to lead McKinsey's AI efforts early this year said, do I think that this is existential for our profession? Yes, I do. I think it's an existential good for us. But of course, that read as they think it's an existential threat, potentially, if they don't get it right.

And so consulting is emerging as an early and high-profile test case for how dramatically an industry might must shift to stay relevant in the AI era, McKinsey, like its rivals, grew by hiring professionals from top universities, throwing them at projects for clients, and then billing companies based in part on the scope and duration of the products. So my analysis of this was that, okay, so clearly McKinsey's going to change. The question is, like, in the medium term, like pre-super intelligence, like, how does it change? And what's interesting is that, like, when you look at the hordes of, like, 80-hour weeks from, you know, junior analysts on a McKinsey team, like the typical, like, Harvard, Yale,

Stanford, Ivy League grad, doesn't exactly know what to do maybe, goes into the big three, Bain, BCG McKinsey as just kind of a junior consultant. And then after like two years, it's like go to go get your MBA and then maybe come back or do something else. Those 80 hour weeks, like what they are doing during that time, there's a lot of hurry up and wait, but there is a lot of actually hard work, pulling data, doing deep research, doing stuff that can be done by AI. But it also serves a, as basically like an extended interview. It serves as a benchmark for is this person a grinder? Are they a creative thinker? Like they have to do the baseline. They have to do, they have to be very detail oriented. But then also it's a question of while you're there, can you actually develop a relationship

with a Fortune 500 CEO? And that happens. Like I remember this crazy, crazy story of a woman who was at, this is a crazy story. So she was, I think I've told you this before. So she was at, I think McKinsey, and she went to do some consulting in a Middle Eastern country. Maybe like, what was it? It was something in, I don't want to be too offensive, but it's one of like the Borat countries. I think it was like Kazakhstan. Is that the one where he's from?

That's where Borat is where Borat is the character. The characters from Kazakhstan. So she goes to Kazakhstan. Sasha Barakowin, that's where he went. Yes, yes, yes. So she goes to Kazakhstan. And Kazakhstan has a number, of state-owned assets.

And one of the things they own, I believe, was a cigarette-making facility. So they built a factory for making sense here. We build a heater factory. National tobacco. We need to secure heater production. It's critical to national security. Yeah. No, no.

I mean, basically, like, when you are a big tobacco company in America, everything is privatized except for, like, the roads here, basically. We're, like, the most privatized country. But when you go to a foreign country and you're like, let's set up, you know, we'd love to set up business in your country. You have a bunch of hard workers who can work on this manufacturing project. The government just says like, okay, we will be the customer. Like we will be the service provider. We will be the counterparty to your request for certain services.

We will provide those and we will own this and that will be tax revenue for our government. And so that's what had happened years ago. this country had built up this cigarette manufacturing business and then once you build up the business the privatizing that can also be very lucrative because then you get all the cash up front. And so if you want to raise a bunch of money for your government and then maybe build some roads, one thing that you can do is you can sell the assets that you've developed. And so they brought in McKinsey to discuss how could they package this cigarette manufacturing building or factory or series of factories into an asset that could be sold

to either private equity firm or to big tobacco. So, basically. I mean, this is what happened with Saudi Ramco was going to go public. Where do you think Kazakhstan's nationalized cigarette infrastructure? What do you think it would trade at if it went out on?

Probably a pretty low multiple, honestly. Very real meme stock potential there. Yeah, yeah. Can you imagine that? It would be crazy. Anyway, so, they bring in McKinsey and this woman's on the team and she does the whole analysis and says that like,

okay, if you package it up this way, you could get this multiple, you could sell this, and then that would bring in this amount of cash that you could go and use and you could reinvest in roads, infrastructure. You could build an airport. Data centers.

Yeah, you could build anything, right? Foundation model, lack. This is pre all that. This is maybe like 2005 or 2010. So she does the deal. It's massively successful. Comes back. Applies to Harvard Business School.

and she needs a letter of recommendation. So she goes to like the prime minister of Kazakhstan and is like, hey, I helped you with that deal. I helped you privatize your cigarette factory. Can you write me a letter of recommendation? And I had the chance to read this letter of recommendation. And it was one of the funniest things I've ever read because it presupposes. It was a glowing endorsement. It was like, wow, she was incredible.

She helped us with this deal. Does it just assume that she's getting in? No, it doesn't assume that she's, getting in but it is worded in the craziest way possible it basically presupposes that it was unthinkable that a woman could do this and no way oh huge cultural bias because you know it's like developing nation and so this country this prime minister is like I was very surprised when she joined the team because I'd never worked with a woman before and I didn't think 2005 yeah no and I didn't

think that she would bore at the whole the I don't want to say it was Kazakhstan because it could have been any other country in that area, but you can just think of it as generic developing nation. But Sasha Berra Cohen was making jokes like that. Exactly. Exactly. And so, and so I saw, I saw a real instantiation of like the joke that he was telling, basically. And so the, the, the letter of recommendation Harvard Business School, it starts with like, with like, it was mind blowing that, like, a woman showed up to this meeting.

I didn't think that was possible. And I, and I didn't think that she would be, like, actually, But then she was really helpful and this like surprised me and so like you should definitely let her into Harvard business school and then it ends on something that sounds like a vague threat He's like great great great great great great sorrow will befall Cambridge if you don't let her in And it's like it was clearly meant to just be like like you're you'd be making a mistake You'd be a missed opportunity but it was worded in this very aggressive way that sounded like a threat. It was wild I mean I guess anybody that's trying to get in to HBS, get a prime minister of a foreign government,

to make a threat to the college in a letter of recommendation. You can do that, you can do anything. So you could read into it as like, OK, this country is developing. Like they clearly have different cultural norms. They're not the most forward thinking in terms of women in the workplace. But instead of getting dragged down in that, HBS clearly saw that she basically went to a developing nation

and like RISD them. like insanely and got a phenomenal deal done and made a ton of money for her firm and also for the country. And it was just this like win win take private. And she got in and had a very successful career. And yeah, this is a funny story of what can go right with McKinsey. But it brings me back to the question of, you know, there's a world where she would just be the like the PowerPoint person. And that's clearly not what she did with that opportunity. She went there and actually interacted with the prime minister and the top, top economic advisors to this developing country and wound up driving a lot of value.

And I'm sure there were other folks on the team on the client side. But she developed such a relationship that she was able to get this like silly but glowing recommendation letter. And so and so. Recommendation threat. Recommendation threat. Yeah. Yeah. That's the real alpha.

If you, if you're, don't write me a recommendation letter. Threaten the person that I'm applying the job for. I would write a recommendation threat. For anyone on the team. For anyone on the team. Yeah. For sure. Great.

Your next training run will fail unless you hire. If you do not hire Tyler Cosgrove, you will effectively curse your bloodline. I would happily put pen to paper to say that. Anyway. So the 80 hours a week, it's not just AI agent work. Like you could not have an AI agent do what. do what she, that example, like do what she did, fly to a developing nation and actually sit in the room with someone who was just on the cusp of learning what a take private was and the value that could be created and convince them that this was a good idea. And so I think that these 80 hour weeks, they're gauntlets and they're used to identify top performers that can still lead the firm.

And the question is, if you don't have people doing the hard work and you're not in, and like a lot of these things, we were talking to Danny Rimer at Index, we were like, is venture, is your model a mentorship business? And so even if you can like outsource to AI agents these low level tasks, you still need to like build the next generation of talent in your firm and figure out who the top performers are. Now maybe you can collapse that to an e-val. Like the reason that she got hired at McKinsey, I'm sure, was because she went to Harvard undergrad or something and had like perfectly SAT scores and like a bunch of, and like all of that might be able to be distilled such that, you know, if you just ask Goldman or McKinsey to just say, take the top 10% of your class, don't hire the bottom 90% and then immediately put them on the executive track and they'll be your managing directors. Like maybe that's, maybe it's all predictable, but you do have a lot of churn where some of your top performers are going to leave and start companies. Some of them are going to go to big tech. Some of them are going to go to big tech. Some of them are going to.

going to just burn out and want to be like teachers or something random. They might want to change trajectories entirely. And so even your top performers aren't necessarily going to matriculate into the partnership leadership. Like leading the firm is a separate thing. Yeah, it will be interesting to see like if there's new archetypes of people that thrive in management consulting, right? Because like the right person that thrives in the first whatever it is, five years of just insane grunt work and just like toiling over documents and stuff like that, the person that thrives in that oftentimes clearly can like rise above it and excel and client side, firm-wide kind of, you know, running the firm itself, et cetera. But there's also

probably other people that would never make it through the first five years, but would just absolutely crush, you know, for the sort of the majority of the majority of the career and I think that does happen every once in a while there are people that lateral from entrepreneurship or tech into a big three consulting firm at the top so they're never the low grade analysts grinding they might have grind somewhere else though like pretty much everyone grinds at some point and I think that's kind of the the part of the takeaway of founder mode is that the founder has has been grinding in the exact context and so you talk about like Dylan field working it on WebGL in the trenches of Figma.

And then, you know, that enables him to speak with a different level of authority when he's managing a massive company that's now public. And so there's there's, there's a world where like the grind is relevant to what you wind up doing. But the big question is just like if, like there is a risk. Like how do you, how do you identify these killers early on in the interview process? What does it mean for retention? And then the other question is like today, if you leave McKinsey to build a new consulting firm,

you might be able to take some of your top clients with you. Like if they put you on the Coca-Cola account and you wound up being really, really impressive as an analyst, and you spent a lot of time with the CFO and the CMO or whatever you were working on, and you built that relationship and they see you as a critical consultant, even though the managing director or the head was the one who closed the deal, they might be willing to take a shot on you if you leave and start your own firm, but then you still have to spin up all your own analysts to actually do all the work that you were doing before with your team.

And so that takes time to hire and train. But in the future, if you leave, if you're really high performing and you leave, and you can just be like, let me get an account on chat GPT Pro or whatever. And then I effectively have 40 consultants on my team on day one. 12,000. 12,000, exactly. I would imagine that leaving to start your own firm and going like sort of so low and just leveraging just the relationships, just the ability to just your innate ability could

be like extremely high leverage much earlier in your career. And so because you don't have to do the the thing where like you leave and you say, I'm taking everybody with me. You know, you can just leave and be like, I'm taking the AI agents with me because they We live in Azure and I'm just spinning up new ones. Yeah, yeah, the real question will be, is there real IP to develop around the agents or can you use chat GPT out of the box to say, pretend like you're a McKinsey analyst and make me this, you know, 20 page report.

I need sections on this, this, this, this and this. Use the entire internet as a data source as well as these materials from the client. and give me this report and then, yeah. Well, here's an idea for a consulting firm. Leave McKinsey, go to companies. Tell them to implement fin.a.i. The number one AI agent for customer service, number one in performance benchmarks,

number one in competitive bakeoffs, number one ranking on G2. That's basically your hero slide. You just show them the website and you say, pay me a million dollars and I'll help you implement it. Just kidding. Of course, you can implement this yourself. You could start a free trial.

McKinsey is private, but it's Accenture is not. Sure. And they have traded down 35% in the past six months. Bookings are slowing. That's crazy. They had a big government. They were impacted by Doge moderately.

But McKinsey does have the benefit of being private and being able to make hopefully some longer-term decision. The scale of Accenture is insane. Guess how many employees they have? 40,000 close okay how many of that have 791,000 what 700,000 they're they're 20 times bigger than mackenzie wow is that a real number there's more people that work at Accenture than the population of san francisco where do they hold their Christmas party like yeah i don't even know where a million people Gunger Gate. Like, is that like Mecca? They have to go to Mecca? Have you seen those like a million people praying in one area? I think that's where they have to go. Or I've heard about these like-C. Coachella has a daily capacity of 125,000 people. They can't even do Coachella. Where can a million people? I mean, I guess there was the million man march. So they could march on Washington, D.C. That's what they could do. They could take over all of D.C. This is their new. They just do rotations and they basically do like a daily walking ultramarathon.

where they just walk through and they just completely shut down towns and they all wear the company merch they probably never been in the same place they'd probably never had a christmas party with everyone they should use they should use restream for their all hands they should some real infrastructure there um anyway that is absolutely wild how many if you're selling something i have some consulting advice for you this one's free get on numeral hq dot com sales tax and autopilot spend less than five minutes per month on sales tax compliance go to numeral hq.com the team just disrupted in applause. We love it.

We love it. What else we got, John? We got Disney. My question, is AI a sustaining innovation for Disney or is it a disruptive innovation? Where will Disney be in 10 years in the medium term? Obviously, you're going to be able to generate a lot of AI slop. You might be able to infringe on their IP. They might get paid by Google VO when you generate a Mickey Mouse AI slop edit.

they might get a couple pennies, but will it be good or bad for them? Children's, you know, I could see them making a product that allows you to make a, you know, book or a story for your kid that actually, you know, and they get, they get some type of revenue. They should get revenue. And I think that they will through the courts, especially because as like big companies, like you're not so, it's not this crazy, oh, there's like a bunch of kids doing random things. Like they never had to go after.

the street artist on Venice Beach that would draw a picture of Mickey Mouse and sell it to you for 20 bucks. That was never material to their business. They had to go after Napster. I had to go after Spotify. I went to Venice Beach. Yes. I got them to draw me as Mickey Mouse. Really?

And then I performed a citizen's arrest because I respect IP. I have an IP respecter. Yes, yes, yes. Citizens arrest is underrated. Citizens arrest. Citizens arrest. You're going to jail, buddy. This is for Bob.

This is for Bob Eiger. Yeah, so big question. This feels like a moment where you want to be in founder mode. Bob Eager is one of the greatest CEOs of all time. How will he navigate this? The Wall Street Journal says, is it still? He's not the founder. The founder died in 1966, Walt Disney.

But you're claiming that he died. He might be able to turn it on. Sotchenadella certainly was able to do it. He's navigating the AI shift flawlessly. We'll see what happens with Bob Eager and Disney. Wall Street Journal says, is it still Disney magic if it's AI? The stakes are especially high for the studio caught between how to use artificial intelligence

in the filmmaking process and how to protect its famed characters against it. So there's a little anecdote that we'll kick it off with. When Disney began working on its new live action version of its hit cartoon Malana, executives started to ponder whether they should clone its star Dwayne Johnson. The actor was reprising his role in the movie as Maui.

A barrel-chested demi-god. Have you seen Malana? No, you have not, of course. we know this for certain days on set Disney had a plan in place that wouldn't require Johnson to be there at all under the plan they devised Johnson's similarly buff cousin Tanoa Reed who is six foot three two hundred fifty pounds would fill in his body double for a small number of shots Disney would work with AI company metaphysics to create deep fakes of Johnson's face that would be layered on top of Reed's performance in the footage a digital twin essentially digital double that effectively allowed Johnson to be in two places at once obviously that's better. Calls up his cousin, you want a job?

I need you to be me. Yeah, hit the gym, buddy. Better be jab. Get on a cycle. But yeah, I mean, these movie schedules are famously tight, three months in and out, crazy schedules. And then you move on to the next one. There's that famous Henry Cavill story where he filmed Superman,

wrapped, moved on to another movie where he had to grow out a mustache. He grew out a mustache and a beard. and then they said, hey, we got to do some re-shoots. You got to come back to Superman, and he came back, but he couldn't shave his mustache and beard, so they had to change it in CGI, and it looked terrible. Probably not a problem now with deep fakes.

That's actually a good use of AI, and something that probably shouldn't be very controversial, but obviously everything in AI is controversial right now. But we will continue with Disney. They say, what happened next was evidence that Hollywood's must discuss, much-discussed, much-feared AI revolution, won't be an overnight robot.

Takeover. Johnson approved the plan, but the use of a new technology had Disney attorneys hammering out details over how it could be deployed, what security precautions would protect the data and a host of other concerns. They worried that the studio ultimately couldn't claim ownership over every, that the studio couldn't ultimately claim ownership over every element in the film if AI generated parts were in there. So if there's AI training data from a DreamWorks film in there, and they use the Dreamworks training data to make a Disney film, film even if it looks like Dwayne the Rock Johnson, DreamWorks might come knock it and say, hey, give us a royalty. I think that's the risk. But the lawyers are having fun, maybe. Full employment for lawyers over at Disney clearly. Disney and Metaph has spent 18 months negotiating on and off over the terms of the contract to work on the digital double, but none of the

footage will be in the final film when it's released next summer. They went and shot it. A deep fake, Dwayne Johnson is just one part of a broader technological earthquake hitting Hollywood. Studios are scrambling to figure out how simultaneously they can use AI in the filmmaking process. and how to protect themselves against it. Is it sustaining or disruptive or both? Can't be both, but we'll see. While executives see a future where the technology shaves tens of millions of dollars off a movie's budget,

they are grappling with a present, with a present, filled with legal uncertainty, fan backlash, and a wariness toward embracing tools that summon Silicon Valley view as their next century replacement. And if you are trying to sell an AI tool into Hollywood, you gotta get on AdiTool. customer relationship magic adio is the AI Native CRM that builds scales and grows your

company to the next level and you can get started for free so the Academy of Motion Picture Arts and Sciences is surveying members on how they use the technology studio chiefs are shutting down efforts to experiment for fear of angering show business unions on the eve of another contract negotiation and no studio stands to gain or lose more in the outcome than Disney the home of Donald Duck Bell Buzz Lightyear Stitch and countless others which has churned out some of the most valuable and protected creative works in over the past century. So my take on this, so two years ago I was hanging out with the founder of a very large generative AI image generation company. And he was telling me that by 2025, anyone with a

laptop and an internet connection could generate a full Hollywood movie about anything they want with a single prompt. And it was a hilarious conversation because we were on a Zoom call and his internet wasn't working. And it was the classic example of like the technology is amazing, but we got a lot of stuff to iron out. So anyway, extremely aggressive timeline, but obviously things are going to change for Hollywood. So my question is, will Disney benefit? It feels like a moment to be in founder mode, but Walt Disney died in 1966. So basically everyone believes that meta will benefit from AI, even if they miss the train on owning the next dominant consumer tech platform. But if Disney got really AI-pilled, what would that look like? I don't think they need to train their

own foundation model just like they don't need to train their own they don't need to build their own cinema cameras they can just use iMacs when they the time calls for iMacs they can use blender when the time calls for blender they can use hoody when the time calls for when the shot calls for some high level VFX yeah but they do need to rethink how they structure their business and negotiate with unions and underwrite content they might need to go more risk on not just from a brand risk position but taking more smaller bets we're in this weird barbell world where everything seems like it's either a hundred bucks and it's shot on an iPhone, it's a viral like TikTok or it's a hundred million dollar blockbuster with like 50 million dollars of

VFX. Actually that's kind of a low number. It's usually like 300 million dollar production with a hundred and fifty million dollar VFX and then no one sees it and it's a flop, but then they hit every once in a while and they're great when they're good, but it's this weird like venture style betting at the high end, but there's nothing in the middle and maybe if that's like the death of the art house film, but I'm just wondering if in the age of AI like maybe there's this interim step where Disney ladders down a little bit and gives like 10 filmmakers, $10 million each and says, hey, you're still required to deliver a 90 minute full film, but you're doing it for 10 mil, and it's not quite Blair Witch level production, one notch up.

You've got to be created. This is what me 24 is doing. Yeah, I'm really bullish on this idea that the, you know, historically like a TV show would film a pilot episode. and they would use that to get the budget to shoot an entire season. And you can imagine now, you can, you know, even for film, you can just make, you know, make the trailer ahead of time with AI. The other advantage that I think Disney has that's very real just going in and why they're just broadly seem to be positioned very well here is that I think that, I think that broadly

like content customization will probably take off. because Disney can make a film now, and then you could make millions of different variations of it that become interactive with the underlying fan. So imagine you're watching Moana, but your kid is in the film, is a character in the film, and you can now do that at scale, right? And how much more would you pay as a parent

to have something like that? I have a funny story about this. But last thing I'd say, so customization and just like democratizing, like basically making, being able to make variations of films. Yep. I think it's going to be big. I think just the time, it's the same way.

It's so difficult to make a new luxury brand. Yep. It's so difficult to create, it's easy to create IP. It's hard, extremely difficult and time intensive to create valuable IP. And Disney's advantage is they have this sort of like 360 in that they can make a film and they can bring it to Disneyland. They can bring it to a cruise. They can create physical products around it.

And so they develop IP in a way that new entrants are not, you know, they don't have the benefit of like having a Disneyland where they can make new experiences that that increase the value of that IP, right? So I think. So two things. One, if I were to go back, one of my favorite Disney properties is Star Wars, A New Hope, the very first film. If I went back and was like, let's customize that for me, I don't know that I would make any changes.

Like, do I really want a scene where, Hans Solo breaks the fourth wall and says like, hey John, like I'm about to go, you know, save Luke at the Death Star. Like that doesn't improve the product for me. I actually like that it's just the vision of George Lucas. So I don't know about customization being better for me. I don't know what I would change. But here's an example of how to make like a magical experience for a kid. Like imagine after a movie ends.

Yep. A kid could interact with a character and ask it questions. in real time like about about the story or that's cool conversation with them and that's what I'm talking about like bringing that IP to life totally right now that exists and I was obsessed with this when I was a kid I would I would I would watch Star Wars and then I would read the the books that in the expanded universe and I remember even having books that were just like encyclopedias of every single ship this you could read into this a little bit more but we'll leave that where it is

but I but I would learn every single Every single ship, this is what a Star Destroyer does, and it would have all this backlog and stuff. And so, yes, I agree with you. That's very cool. You finish the movie. And then you can just interview Luke about how does the lightsaber actually work? And he can talk about the Khyber crystals. That would be very cool.

On the flip side, I had a very funny experience. I was watching a horror film in high school with a couple friends. And I grew up in Pasadena. And this horror film just happened to take place in Pasadena. Like that's where they set the film because Pasadena, California is just like a place where you set films. It's just a real place. So if you've seen Kill Bill, which I know you haven't, but Kill Bill by Quentin Tarantino, there's a scene where Uma Thurman just shows up and it says at the bottom, Pasadena, California, because like that's where the character went. It exists in the real world.

But this horror film that took place in Pasadena was terrifying because I was watching it. And I was like, this is happening here and it's night and it's dark outside. And like, now I'm so much more immersed. And so I was thinking back then that you could use like the IP address of a connected. I think we were watching it on like a PS3, like a DVD player. Like you could use the DVD player to dynamically change the location of the establishing shot. So you're like, this whole horror film is going to take place inside of like one house where there's like a monster in the house. And it's going to be like the usual like they're upstairs, they're downstairs, there's blood, there's, you know, someone's running and chasing.

Like it could be any town USA, but they usually, they usually establishes like this is happening in Amityville. This is happening in, you know, some random town, Lake Placid, right? But they could easily just dynamically change that with a few establishing shots to just show you, okay, it's happening in Malibu. And now it's a lot scarier for you. So I think there's something interesting there. But again, it has to be the work of like an option. It's yeah, it's not something you can copy and paste either because it works for some shows, but then others like the place is so obviously the place that it would throw you off as a view.

Exactly. And so I do think that the that the fully AI generated fully custom content, that will exist, but it will exist on independent third party platforms. It won't be the domain of Disney. This will be something where if you in the future, if you really want to see like, you know, AI generated stories about surfing in Malibu, like there will be an, endless stream of those and you will be able to go and experience that particular content. And you can already kind of experience that because there's probably some Instagram person who makes really great content about surfing in Malibu. And if you follow them, you get that vibe and that might be what you're into.

And you can kind of like, and it's handled just by the great democratization of creativity. Yeah. And we should talk to Samir about this later. He's coming on the show. So we can talk about the future of content creation and whatnot. But anyway, let's talk about public.com. Investing for those that take it seriously. They got multi-asset investing, industry leading yields,

and they're trusted by millions. Go check it out. Anyway, let's hit the timeline. Let's go through the hottest posts on the internet. Ray Dalio. So last Friday, after we wrap the show, it was a timeline in turmoil over in DC. I believe the head of the Bureau of Labor Statistics

was fired. This was after a negative job report. It was very controversial. People were not very happy about this because this is the data that we used to judge the health of the economy. Insane, insane whiplash too, because big tech, you know, Thursday, printing,

Rip bang. Figma. Massive, massive, you know, biggest gain in a very long time. And then it turns out the real economy, or at least the statistics coming out of the Bureau of Labor Statistics, were not looking great. And the prior two months

got revised down. And so we are still adding jobs, but mostly in health care. We're at Accenture. Although I think they're cutting as well. Yeah. And so broadly, the U.S. economy feels strong, but we are having some unemployment uptics.

And so that's always a canary in the coal mine. But Ray Dalio chimes in with his take. Also, I believe, like, tourism. is down pretty much across the board. Yeah, and there's a lot of debates about people immediately were digging into what's driving the job losses. Is it, is it, does it have anything to do with Trump's immigration policy?

There were a lot of different, a lot of different ideas as to why the, why the job numbers are what they are and how bad are they? Obviously you can see in the market, the market tanked on Friday, but then is up today. And so there's, you know, There's not necessarily consensus now that we're like entering a recession or anything like that. But it's certainly something to watch and something to dig into.

And obviously we want high quality data about everything from GDP to CPI to unemployment. And so Ray Dalio is breaking down his argument for firing the head of the Bureau of Labor Statistics. He says that's because it's process for making estimates is obviously obsolete and error prone. And if there's no good plan in the works and there's no good plan in the works for fix, it. The huge revisions in Friday's unemployment numbers are symptomatic of this, especially because the revisions brought the numbers toward private estimates that were, in fact, much better. I assure you that this is something that I know a lot about because of how I use data to follow the economy and bet on where it's going. Of course, if the way most people in the media

are conveying President Trump's motivation for the firing is correct, such as the New York Times saying, quote, when President Trump didn't like the weak jobs numbers that were released on Friday, he fired the person responsible for producing them. That would be a big problem because leaders manipulating numbers that distort the truth to suit their political objectives is a classic sign of the loss of a functioning system with the rule of law and checks and balance. And the loss of these things leads to the loss of confidence that underpins our whole economic and political system. So it'd be good if President Trump made his thinking clear. In any case, we do need big renovations,

uh, renovations to the ways the government estimates what's going on in the economy to make them more, not less accurate. And so this is interesting because Ray Dalio is not a perma bull on America by any stretch of the imagination. He literally wrote a book about the changing world order that was basically like America's cooked. I mean, to some extent, he was like talking about the rise and fall of nations and is, and he's been very bullish on China. and sees China's a rising nation. So he is not someone that is just constantly ringing the gong for America's unrelenting progress.

And so this is interesting take from him. And then I mean, I have personal experience with how the government collects data. I didn't work at the Bureau of Labor Statistics, but I worked at the Census Department. From the people that brought you that brought you the Department of Motor Vehicles comes labor statistics.

Exactly. And so I always, In the year 2000, the economy was ripping, and I believe the census was done by volunteers because it wasn't, it was just kind of like, oh, let's get a poll with people. But then in 2010, the economy was in a recession after the great global financial crisis,

and it was seen as a stimulus program. So people were making like $20 to $25 an hour, which is very good back then to go around and survey people. But the survey data is a little messy, and I was always wondering. Have you told this story on the show before? I don't think we have time today, but we should at some point. But basically my takeaway was always, was always like, couldn't we just ask Google?

Like I have a feeling that the big tech companies know not just exactly how many Americans are in each jurisdiction, but they know everything about them. Their IP addresses and their geo locations. Like the big tech companies truly know so much. And I know you're going to say, like, well, not everyone uses Facebook. And that's true. But if you aggregate data across most of the cell networks and most of the web services, you could get a very accurate reading much more so than just randomly going door to door and seeing like, oh, this person opened the door and talk to me, which is like how the

census currently works. And so it is interesting that the unemployment numbers were, the government numbers were less accurate than the private estimates, which are just like, because obviously every hedge fund wants the data and they want the data earlier than the government puts it out. And some traders obviously trade on top of the BLS statistics. And those are important, but the private estimates are better. And so interesting to see where this goes, how much modernization can be done, how much. And then there's always the question of like privacy if you're like tracking everyone.

Because it's like there's a lot of people out there who might say, I don't want the, I don't want the government to know if I have a job or two jobs. data, Palantir announcing earnings, I guess, I'm assuming in 30 minutes. Another post here from Bucco, Bucko, Capital, Bloch. He says, Trump, we need to cut. BLS. Turns out the job market is weak. Trump, no, not like that. Another post, interesting data point here, Derek Thompson says,

spending on services, especially travel and tourism, has declined for three straight months for the first time since 2008. And expanding when Trump, the U.S. economy is showing strains of weakness. Outside of the pandemic, the employment rate for college graduates over 25 is rising fast and is higher now than any year since 2014. Job growth in the past three months is the lowest since 2010. And spending on services has declined for three straight months for the first time since 2008.

How are companies, individuals, and the government are supposed to deal with an economy slipping toward recession if we're if we're smashing the tools we need to know how fast we're falling in a tempest it's unwise to ransack the navigation system that is Derek thompson's take uh what else we got we got ain't sleep sleep dot com get a pod five they have a five year warranty 30 net risk free trial free returns free shipping continue jordy it was absolutely brutal having to cover the figma IPO uh without sleeping on an eight eight snap and i'm And ramps, and ramps series E2. These are big days.

But green shoots for us. Code TBPN, of course, if you want to make the switch. Benji Taylor says people are finally starting to realize how few truly great designers there are and how disproportionately valuable they are. If you missed our interview with Dylan on Thursday, he said that design is becoming an enduring edge for companies. We've seen this with companies like Linear, which came into. to a crowded category and is dominated by just building a truly thoughtful product.

Yeah. And this is something. People love to like reduce design to nothing. Oh, cursor, win-ser, if these are just forks of VS code. And it's like, yes, but like if it was truly perfectly competitive, why is there a split in revenue at all? Why isn't it just 50-50? Well, like, there's probably some design.

and differentiation in the product that you don't realize until you're actually a DAU of these. And that is what's driving the adoption. Yeah, but yeah, I have worked with hundreds of designers across my career and there's only a handful that when people come to me and say, you know, I want to recruit a founding designer or something like that. Jacoby.

Jacoby. Jacoby in the comment section. Yeah. he's there. No, no, no, no. He's in the comment of Benji's post. Yeah, yeah, yeah. Anyways, uh, on Janjane. Well, if you're looking to design something, design a billboard, put it on adquick. Dot combs, out of ad hocciful. Say goodbye to the headaches of out of home advertising. Only ad quick combines technology, out of home expertise and data to enable efficient

seamless ad buying across the club. We didn't make it to our billboards in New York. We didn't. But I felt like I'd been there because so many people had shared them. Yeah. Still low hanging fruit in billboards. On, on Janay says Mistral, 11 Labs, Black Forest Labs, Lovable, Helsing, Granola, European AI scene is on fire. So I just wanted to give some credit. Like this top comment. Not to mention lovable. And then Ange just says, reread the list.

Like, it's literally the fourth one. That's very funny. And I know all these companies. This is bait, though. This is bait for us. Because you know I refuse to step foot on the old continent. And so by posting something like this, this is just this is trying to get. get me out of golden retriever mode out of being nice. This is trying to bait a dunk from me

because nothing makes me more upset than anything that's not America. Well, one thing I would say, I think you can be happy that the cap tables of these companies are majority U.S. Yes, yes. I'd like the cap tables to get 100% American owned and then I'd like the companies to relocate to America. What about what about the teams? Everyone and all the customers. Just move everyone to America. 11 Labs is also 11 labs from my understanding is primarily based in New York. A little truth zone going on. But I think Anjane would know better considering Andreessen is an investor.

Rune says, it's true what the misinformation is saying about us. We indeed have Lilith, progenitor of human life locked up under the Mission Bay headquarters. Don't read the comments on this one. We're skipping over the comments. Very funny. But anyway, yeah, there was a GPT5 article that went out and this is Rune kind of putting in the truth zone and saying like, hey, we're just grinding and we're building and, you know,

tell your story a different way, basically. Bay's Lord says, I propose we create the Terence Tao Institute for Math and Science and fund it with tech money. Terence Tao got his funding at UCLA cut not specifically but but funding was cut broadly due to UCLA's reaction to yeah he was sort of collateral damage yeah in in a in a political fight between UCLA and the current administration and so I think someone should step up and keep funding him I think there's certainly value in having him do extremely undirected research. He could land at a lab and do cool stuff.

I was saying that he should join Vlad's... That would be very cool. Yeah. So Vlad Tenev, founder of Robin Hood, has a new foundation model company or AI lab focused on math, and he's got the money to provide Terence Tao with everything he needs and more,

and I think this would be a great use. And if it doesn't happen soon, we're going to start negging every billy. into doing it. We're going to create the shame on you list for billionaires who have not donated to Terrence now. But Noah Smith says we now have a big story about why this time is different, rapidly increasing debt funding one single sector, an opaque corner of the financial system that has recently grown, systemically important lenders and meshed in the new sector. And I just want to say I really...

I feel like calling top signals was like a two weeks ago thing. I think we're back at this point. And I don't know why he's calling top on everything right now. I just hope that this time the music never stops. Me too. Because I like dancing. And I like dancing around. And who doesn't like a little game of musical chairs?

But also, I mean, maybe this time there's enough chairs. Everyone just takes a seat. No one said if you're playing musical chairs, you have to stop the music. Also, yeah, in musical chairs, you could still land on a chair. There's only one person gets eliminated. So just don't be that person. Don't be that guy, pal. Yeah, don't put yourself in the FTX bucket.

Do we have time to go through this? We have 10 minutes. Okay. But I think there's probably more stuff. Noah Smith on no opinion. Will data centers crash the economy? This time, let's think about a financial crisis before it happens. The U.S. economic data for the last few months is looking decidedly, meh.

the latest employment numbers were so bad that Trump actually fired the head of bureau of labor statistics. We don't know that that's why, but it's a reasonable take. Accusing her. Yeah, his take would be, you know, she's been messing up the numbers a lot and we need a new approach, you know, kind of. Yeah, but accusing her of manipulating the numbers to make him look bad, but there's one huge bright spot amid the gloom, an incredible AI Denistator building, boom. Let's go.

Let's go. Look at the chart. Pull up the chart, scroll down. capital expenditures quarterly, we're so close to touching 100 billion across meta, Google, Microsoft, and Amazon. You add in the independent CAPEX that's going on and you're easily in 100 billion a quarter. You'll love to see it. A little disappointing Q1, but Q2, we came back strong, strong. Yeah, no one's really talking about the self in Q1. No, what says inference compute now represents most of the cost of running advanced AI models

and increases in inference computer responsible for many of the ongoing performance gains. So compute needs are probably only going to grow as AI keeps getting better. Whoever provides this compute is going to make a huge amount of revenue. Whether that means they'll make a lot of profit is another question. But let's table that for right now. We reported earlier that Microsoft's margin on the token and inferencing side with something like 30 to 40%. I would just zoom out and look at it as like, what's Capax? we're talking about $100 billion a quarter.

What's token-related revenue? So add open AI, they're making $3 billion a quarter, Anthropics making a billion a quarter. You know, obviously there's, you know, core AI stuff happening in all the hypers. But you add all that up. We're making like maybe $10 billion a quarter in revenue right now on AI stuff and we're investing $100 billion.

And so like we're not right side up on the trade generally, but there's no. sign that that that chat GPT will ever make less than a billion dollars a month like there's just no sign that that's not going to that's going to fall off a cliff yeah who's unsubscribing the other I mean the same dynamic right if you're building if you build a if you build a factory do you expect to to generate profit that month that quarter that year the next year yeah potentially now I mean this stuff depreciates like quickly because it doesn't stay on the frontier but like the base workload for just

generating tokens for GPT4O, like reasonable questions and just asking just basic stuff. Like that workload could stay on those H-100s for a decade and you could make that money back. Yeah, it is interesting that Apple we covered after their earnings. Tim Cook called out that they're using third party like lenders to finance a lot of their CAPEX and they're choosing not to deploy too much capital themselves. So Noah says, roughly speaking, Apple is choosing the former while the big software companies, Google Meta, Microsoft, and Amazon are choosing the latter. These spending numbers are pretty incredible. We covered this chart already.

For Microsoft and Meta, this capital expenditure is now more than a third of their total sales. Let's go. Let's go. If you're looking to deploy a capital expenditure of your own, get on Bezell. Get Bezell.com. Your Bezell concierge is available now to source you any watch on the planet. Seriously, any watch. It goes on the balance sheet. It's CAPX. It's personal CAPEX. Do it. Here's Chris Mims of the Wall Street Journal.

Mag 7 tech firms have collectively spent a record $102.5 billion on CAPX in their most recent quarters, nearly all for meta-alphabet, Microsoft, and Amazon. Apple and Via and Tesla together contributed a mere $6.7 billion. Investor and tech pundit Paul Kedroski says that as a percentage of gross domestic products, spending on AI infrastructure has already exceeded spending on telecom and internet infrastructure from the dot-com boom, and it's still growing. He also argues that one explanation for the U.S. economy's ongoing strength, despite tariffs, is that spending on IT infrastructure is so big that it's acting as a sort of private sector stimulus program. KAPEX spending for AI contributed more to growth in the U.S. economy in the past two quarters than all of consumer spending, says Neil Duda, head of economic research at Renaissance Macro Research, citing data from the Bureau of Economic Analysis.

Here's that chart from Kodroski, who has been doing an excellent job following this story as it unfolds. So we can pull this up right now. Infrastructure, CAPEX as a percentage of US GDP by era. And you could see railroads. We went pretty hard. What's weird is that he's showing telecom CAPEX in 2020 when, And that was not like the boom time for telecom. Obviously, 5G and fiber infrastructure are being built out.

But this doesn't surprise me. We got to get this number up. We got to get to 6% of GDP, for sure. 50% of G. 100% of GDP. Has it Dorcasch talked about this? Yes, yes, yes. If you play out the scale is all you need thing, you wind up investing like 200% of global

GDP eventually. And it's just like the math doesn't work. Moss is like, I can come up with the money. I'm good for it. I'm good for it. I mean, technically you could invest more than 100% of it. of global GDP with enough debt like you can or you could just save up for a decade like the humanity could save up and then just deploy it all at once like like wealth is not the same

as GDP GDP is an income figure and wealth is a is a balance sheet figure and CAPEX is a deployment from from the balance income not in yeah I mean loosely you you should you should benchmark GDP to to revenue or something like that more more than just like Like, GDP, like, the U.S.'s GDP is not how much money the U.S. has right now or can spend. It's just how much money is flowing through the economy. Yeah. Anyway. Noah says I think it's important to look at the telecom boom of the 1990s rather than the one in the 2010s

because the former led to a gigantic crash. The railroad boom led to a gigantic crash two in 1873 before the investment peak on Kedroski's chart. In both cases, companies built too much infrastructure outrunning growth. in demand for that infrastructure and suffered a devastating bust as expectations reset and loans couldn't be paid back. And doesn't Tyler Cowen also have a bunch of writing on why these sort of booms and bubbles are actually can have it tremendously. Tyler Cowan has written about it, but I think you're thinking of Bern Hobart who wrote the

book, Boom for Stripe Press, fantastic book. You should go check it out. But basically it's just that you get, yes, you do get these gyrations, these market corrections, but the booms create new opportunity at lower prices because you have all this dark fiber. Who bought up that dark fiber? Google. And so we got Google post crash, actually.

And we got Facebook post crash because the infrastructure was there and the railroads had been built. And you can see the same thing happening where there's a whole bunch of inference capacity that's out there. And then it's like, yeah, we're really not using it all

because we're kind of good with GPT4 class models and we don't need to influence. at the 4.5 level and there's monster models because it's kind of not worth it. I'm good with my McKinsey analyst. I'm good with this one. But then in the next 10 years while we're kind of licking our wounds from the sell-off, somebody goes and figures out how to actually use that crazy, crazy higher amount of inference and the capacities there, which is very cool.

But anyway, if you're looking to take a vacation and go visit a data center, you got to book a wander, so you're sleeping in luxury while you're going out to the data center. That's right. You can find your happy place. You can book a wander with inspiring views, hotel grade amenities, dreamy beds, top tier cleaning, and 24-7 concedure service. We should make our own map of wanders. It's a vacation home, but better. Yes, that are closest to the data centers. There's a big data center. If you want to do some data center tourism, sort of like a

Pinoa says in the context of the telecom boom and the railroad boom, in both cases, the big cap expenders weren't wrong. They were just early. Eventually, we ended up using all those railroads and all those telecom fibers and much more. This has led a lot of people to speculate the big investment bubbles might be actually be beneficial to the economy. since manias leave behind a surplus of cheap infrastructure that can be used to power future technological advances and new business models. But for anyone who gets caught up in the crash, the future benefits to society are of a cold comfort. So a lot of people are worrying that there's going to be a crash in the AI data center industry and thus in big tech in general. If AI industry revenue doesn't grow fast enough to keep up with the CAPEX boom over the next few years,

a data center bus would mean that big tech shareholders would lose a lot of money. like dot-com shareholders, you're not losing if you never sell. Like dot-com shareholders in 2000, it would also slow the economy directly because big tech companies would stop investing. But the scariest possibility is that it would cause a financial crisis. Financial crises tend to involve bank debt when a financial bubble and crash is mostly a fall in the value of stocks and bonds. Everyone takes losses and then just sort of walks away a bit poorer like in 2000. Jordan, Schullerick, and Taylor survey, the history of bubbles and crashes, and they find that debt also called credit and leverage is a key predictor of whether a bubble ends up hurting the real economy. So I think we need to dig into this more.

We need to learn more about private credit, have some folks on who are financing this stuff, analysts who are covering these. It's really been a golden era of private credit. And I think that there's another three hours of, you know, discussion here. But we have our next guest. We had one of the founders of Core Weave on. Yeah. We should ask him to introduce us to a couple of the lenders that kind of finance that business and ended up printing. Yeah.

Yeah. I mean, just off the top of my head, I've been talking to a bunch of folks about this. And it's not quite the same because you can hold private credit for a much longer time. And there's a lot of this stuff is being funded off the balance sheets. They're just cutting a dividend or drawing down on a cash balance if you're a mag seven company. So there are a lot of differences, but definitely worth digging into. so we will cover this more in the coming days. But we have our first guest to the show, Dan from Armada.

Welcome to the stream. How are you doing, Dan? Hey guys, I'm doing great. How are you? I'm good. Welcome to the show. Thanks so much for joining. Can you kick us off with an introduction on Armada and how you're explaining the business

these days? Yeah, so Armada is building the hyper-scaler for the edge. Look at the map of the world. There's about 30% of the world where you have these big hyperscale data centers. And I mean, 70% of the world doesn't have AI infrastructure today. And so we are filling the gaps in infrastructure of building that hyperscalor for the edge. Got it. How much of that is military versus, you know, that we hear about the oil and gas industry. They need a data center on site. They're in some remote location versus just energy price arbitrage.

Let's go where the energy is cheap, put a data center there. Crusoe was doing that early on in a little bit less of a portable package. But how are you thinking about the, the market? and kind of the most tangible opportunities that you can explain the business through? Yeah, it's both. So we're working with critical industries, a lot of energy companies, oil and gas, mining,

manufacturing, logistics. And then in the public sector, we're working with the DOD as well as allies. And then also doing a lot with emergency response, anything that requires sort of split second decision-making, processing of large amounts of data in more remote locations. That's a real focus for us.

And then what we've seen is that now that the kind of blocker for AI is moving from just the chips to energy. And I think Anthropic just came out with a white paper last week. We either are talking about when you've unlocked 50 gigawatts of energy by 2028 for AI. And so we actually just launched a new product, which is called Leviathan, which is megawatt scale AI data

centers that can rapidly co-locate with areas where there's stranded land of energy. And you can have a sort of a plug in play AI factory in weeks versus years with the traditional data center. And that's important because we've got about six gigawatts of stranded energy around the country. China's really fast. We need to unlock all that energy in order to win.

Yeah, walk me through kind of the shape and scale of the products. I've seen the shipping container. Sure. What does stranded energy actually look like? I have a bunch of guests. is, you know, you know, infrastructure was overbuilt in a specific area. It's like beached whales.

If there's beached whales, you go, you extract the whale oil right there. Yeah. And then you can burn it and generate your power for your data center. Just like that. Just like that. No, I mean, there is a ton of stranded energy around the country and then also, you know, an allied parts of the world like the Middle East. If you looked at like the top energy sources, what's top of the list, it's natural gas,

right so you have a lot of stranded natural gas in places like you know north dakota uh texas we're already working with developers that have a lot of that land and stranded energy and um you know they're looking to utilize it for a i infrastructure and if you look at um you know the the AI action plan that just came that came out i think part of what we're trying to do is say okay we have all the stranded energy we're going to need hyperscale data centers you know you sort of gigawatt scale data centers we need to build those really rapidly but we're also going to need to build distributed infrastructure distributed data centers that can be deployed very quickly to use that stranded power so uh when i'm when i hear stranded natural gas does that mean like some exploration

company has gone and and done fracking experiments or research to understand that there's natural gas there but maybe there's just no pipeline to extract it so you can get it out of the ground, but it's very expensive to move somewhere else that's more valuable? Or is it something else that's going on? Like, how can natural gas be stranded, basically? Yeah, that's definitely, you know, some of the scenarios. Other scenarios are like there are major energy companies around the country that have, you know, sort of flare gas that's just being released today in the atmosphere. And so they're, they're under pressure to do something with that gas. Yep. And so what they're now doing is they're actually using turbines to convert it into electricity so that you can power.

infrastructure, some for their own internal use, but then also they're looking at this as, hey, I can create a new revenue stream. And so it's a mix of different situations, but those are common ones that we see. Makes sense. Let's go back to the actual product, what it looks like. I've seen shipping container renderings. What are the different scales and products that you think are most important to push forward right now? Yeah. So again, we're building the hyperscaler for the edge. And what's unique about us is we're doing it full stack. So it's, the hardware, the software, and the AI. And the easiest mental model to have is like think about one of the current cloud providers.

Think about, you know, Azure, AWS, GCP. What are they? They're, you know, really just full stack infrastructure companies. They focused on building these large, hyper-scale data centers in, you know, areas that had a lot of data, major metro areas. But again, that's only 30% of the world. So basically what we're doing is focusing on the remaining 70%, but otherwise it's a similar type of approach. We're building full stack.

We provide everything to customers as a service in a way that's totally termed key. And I'd say that the biggest difference is we see a world where it's going to be distributed compute. So we're building rapidly deployable mobile infrastructure that can be co-located with not only energy but data. because what we've found is that you've got, you know, some of these energy companies that have oil rigs that are generating one to five terabytes of data every day, similar situations with mining companies, you know, similar situations on a battleship, right, where you might have just huge amounts of data being generated from drones, autonomous technology in the middle of the ocean. Now you have connectivity via Starlink or Star Shield and other types of connectivity, but you also need to compute there locally to do any sort of real-time data processing at the edge. Are you seeing more demand for kind of like tokens as a service managed models or just pure infrastructure, bare metal, let me do whatever I want on it? It's all, so again, it kind of comes back to think about like the cloud provider.

Like there's a reason why Amazon Web Services is called Amazon Web Services, right? It's the services that the people care about. And so for us, we also sell services or so we call them solutions where it's like, there's a specific thing that the customer is trying to do, whether it is rapidly standing up an AI factory in some location where they have stranded energy, or it's a scenario where they have lots of unstructured data, let's say, from drones, and they want to process all that data in real time versus having hours or sometimes days of delay so that they can do a specific thing. A good example is we're working with the state of Alaska, their Department of Transportation,

transportation and they're using our entire platform. They're managing all of their connected assets, meaning like Starlink terminals, Skydeo drones, SD-WAN devices. They used to have over a day of latency to process unstructured data from the drones for response to avalanches in remote areas. Wow.

And you know, we're helping them get to near real time to respond to those types of things. And so that's what we're seeing is there's different specific problems that these customers have at the edge and then we're unique in that we actually solve them full stack. Jordy? I'm curious how many how many sort of like systems do you want to have like what are this sort of like more near-term goals in terms of deploying systems when I think of when I think

of the word armada, you know, obviously you think of like the fleet you know a fleet that's stretching as far as the I can see obviously these are spread. battery pack where I can figure like one and it's a couple grand or I can spend a hundred ninety two million dollars or something if I buy a ton of them but it's really just like a lattice yes super yeah and even like how you're managing kind of like how to prioritizing different customers when I'm sure you get calls all day long and somebody's like hey I need a data center in Antarctica and you're like well like we're more focused on you know the northern hemisphere

The IMOs next week. I need a data center next. Yeah, exactly. No, we're exactly right. We're focusing on the largest sort of mission aligned entities in the world. And what I mean by that is we're working with some of the largest energy players in the world. We're working also with the DOD. It's public.

We're working with the U.S. Navy. It's kind of nice that the company's called Armada because it goes in hand with that. But the way that we approach it, we really focus on customer. that we can see that they will have thousands of what we call connected assets. So, you know, take the Navy, have a galleon, we call our modular data centers galleon sticking with the Armada theme, have one on every ship, you know, have one in strategic locations around the world where we have bases or maybe more of a mission scenario, being able to support that.

Similar situation, if you think about like a really large energy company, like we were We're in the news with a RAMCO and Microsoft where we've already deployed our technology. We're starting with one site, but they have rigs and refineries all over the world, onshore, offshore. And so the idea is that this is distributed infrastructure for all of their operations. And what's really interesting is if you open up our operating system, we call it we call our platform the Armada Edge platform AEP, you see a fleet map and it shows all of your galleons as well as any other connected assets that you have, all in a single place. And so what we want to be is that single partner for the most important organizations in the world,

you know, obviously aligned with the U.S. for all of their operations. Can you talk to me about the history and development of Starlink that feels like a foundational technology that unlocks a lot of what you're doing? How big is Starlink? How useful is it? Where is it going? Where do you expect the capability unlock points to be? Is there a framework where I should be thinking about Starlink in the same rollout as, you know, 3G, 4G, 5G, something like that? I imagine that we're still in, you know, the exponential rollout phase of that technology. Yeah, absolutely. So Starlink only launched in public beta in November of 2020.

So it's a relatively new technology. The first couple of years, they were focused on consumer business. And then they started selling to enterprise and then also working with the government with Star Shield. So you're really only talking about a few years and they're now in well over 140 countries all over the world. And what's been really interesting, we've been working with Starlink from the beginning of Armada. I've been working with them since the inception of the company. And we have a great partnership. But the technology is continuing to get better and better and better.

When we first started working with them, a lot of people where we're using it in like pockets, take it like a really large energy company. They might use it as a backup form of connectivity. But they wouldn't use it as a primary. What's ended up happening is that everything that Starlink does is getting better really, really fast, including, you know, new generations of the birds, the satellites up in space, which are getting better performance. There's new types of terminals.

There's even a new kind of like priority lane. They call it multi KU where you can get better performance a little more expensive. So what's happened is a lot of these organizations, they started to use it as backup, and now they're actually starting to use it as a primary source of connectivity. And what we expect is over the next five years, it's going to cover the entire globe. and you're going to have, you know, Leo connectivity that's as good as having fiber in a lot of these locations. And so we're building for that world today

and sort of completing the rest of the compute stack for that world. In terms of fiber, I mean, that seems like speed, but is there a world where when you're in the data center space, like you need the pipe to be extremely big and we're hoping that, you know, the rollout of Starlink will enable bigger and bigger pipes? Is that the correct, like mental model? Yeah, what's interesting is that a lot of times people will not want to send all the data back.

So they only want to send the metadata back. So what's very kind of like synergistic between what Armada does and what Starlink does is we can work with them. The performance, you know, will continue to get better and better from a latency standpoint. point, we will handle all of the local data processing and then we'll use Starlink to send the metadata back. Right? Yeah. And there's really interesting things also as we're continuing to evolve what we do.

So we started out with inference. That was the primary focus was like, okay, you take an AI model that's been training the cloud, you run it at the edge on our infrastructure. Now with this Leviathan launch that we recently did, we've kind of completed the picture for a AI factory that can be deployed. in weeks versus, you know, years with a traditional data center and it's still mobile. So you can relocate it as needed. And what is interesting there is that you can do very cool things with like federated

learning. People are going to want to take these models, sort of fine tune them for different scenarios that and run them at the edge, but they're going to want to leverage the improvements in the model across all of their different sites. And so the way that we're evolving, the way that Starlink's evolving, we think we're going continue to unlock a lot of new high-value use cases that way. Yeah, can you walk me through some of the, some of the enterprise, like, infrastructure that you bring to a customer that's not out of the box with Starlink,

because I know that SpaceX has a fantastic consumer product, but there's a certain level when you get into the enterprise workloads that you need another partner in the stack to actually deliver capabilities. Yeah, we're seeing that customers want to use multiple connected assets. We call all these things connected assets. And it's, you know, Leo and geo satellite connectivity. So they sometimes will use Starlink, but they'll use Starlink plus like a Hughes or a Viasata geo satellite connectivity.

They'll use SDWAN devices like Cradlepoint and PepLink. They'll use drones like we have partnership with Skydeo on the drones. So what these customers want is they want some. somebody to help bring it all together to solve their problems at the edge. And that's what we do. And so a lot of times when we'll get involved, you know, working on opportunities with Starlink, they are already working with Starlink, but then there's something else that they want to do in addition to that.

And we help that. And then the other thing, there's, you know, different things that customers want just in terms of optimizing for different scenarios based on not just performance, but for cost. And so we help with a lot of those types of use cases. Yeah. What's the scale of the business these days? So we've been growing really fast. We've only been around for a little over two and a half years.

Wow. We spent the first year in Stel. Ray is now well over $200 million from a bunch of great investors like Congratulations. One of the really exciting things that happened last year is that Microsoft ended up in making investment in the company. They invested in this round. And then I met with Sachin Adela.

He got excited about what we're doing. And so he ended up talking about us on stage at ignite their big annual customer conference. And we're going to market really globally now with Microsoft and SpaceX, among other partners. and we're working with some of the largest companies in the world. I mentioned to Ramco. There's a bunch of other large oil and gas companies. We're working with some of the largest mining companies. And then, you know, we're working with the Navy that's public and then with states.

So we're kind of just continuing to focus on ones where like we were talking about, we can see fleets of connected assets, like thousands and tens of thousands of connected assets being deployed all over the world and then, you know, scaling to become their single partner for the edge. Yeah, do you think the relationship with Microsoft looks more like in the future I go to Azure and I need to do a workload and leveraging stranded energy just gets me a better price? Or do you think it's more like a channel partner where a really big enterprise client goes to Microsoft and they can route to you as a specific solution for a very like, you know, I don't know, like off the beaten track use case. Yeah, I think what we're already seeing is that customers want to optimize more and more workloads between the cloud and the edge.

And I think that's going to continue. Like the cloud is always going to be there. But more workloads are going to also be run at the edge. And so the more of those things work hand in hand, the better it is for the customer. Yeah. And so how it's working today is most. of our customers are already working with Microsoft. They have really large agreements, they have Azure credits. They can actually buy all of our products under their agreements with Microsoft

and using those credits. That's interesting. And then Microsoft sellers also get some credits. So it kind of all works together. Yeah, yeah. But what's nice for the customer is there's different capabilities that they already are using from Microsoft that have never been available in these more remote sites. And we're making them available in these more remote sites. And that can range from the latest AI model that's been trained in Azure, that somebody wants to run locally at the edge, to things that are more basic. Like an example is in the galleons, you can have what's called Azure Stack, which is a piece

of hardware that Microsoft produces. Oh, interesting. And a lot of applications that these companies like to use are designed to run, and they already run on that, but that's never been available at the edge. So we had a partnership that we announced with Microsoft and also caliber. where there's just a series of applications for like production automation that these large energy companies are already using and they're more connected sites. We're just extending that to the more remote sites. So you don't have to completely re-platform and go down to, okay, I got to make sure it runs on Linux or runs on Kuda.

Like you can assume that you have access to all the niceties of Azure but at the edge. That's right. That's awesome. Congratulations. This is a fascinating business. And I really appreciate you taking the time hopping on. I would love to talk to you again. Your travel schedule must be absolutely insane.

Yeah, next time you're out at some stranded energy, call in from the field from your phone, presuming you have Starlink or at least cell service. And we'd love to see what it looks like to see some natural gas plant flaring for the last time before it becomes tokens. Let's do it. I'll put my Starlink Mini in the backpack. Fantastic.

Make it happen. Amazing. Thank you so much for hopping on. We will talk to you soon. Have a good way. Thanks again. Up next, we have Samir from the Colin and Samir YouTube channel, of course. Second time on the show, we ran into him at the Coinbase. Base launch event very briefly chatted in the stairwell.

We were moving past each other. Very excited to welcome him to the stream. Samir, how you doing? Looking professional. Looking extremely professional. I don't know if we have your audio yet. I think we missed your audio. That might be on our team's side.

Let's check. We are not getting audio, but it doesn't matter because you look fantastic. Yeah, the, the, the, clearly hitting the gym. Yeah. Ben, is this on our side or, or? Okay, I think you're muted. Let's try and throw that on and we'll bring him in and, uh, and continue the, the glazinated 3000.

The white balance on that camera, phenomenal. The exposure to die for. The lighting, very soft, very neutral. Let's see. Can you hear something? Samir welcome to the stream nothing yet let's work on that we will bring I mean what what a what a what a moment you know even even creators at the absolute top of the top of their field the apex content creators still still

learning yes yes let's let's do one timeline post and then we will move back into Samir this interesting data point to calculate the grind score of a company. Have you heard of this metric before? So Figma is in the top 15 companies in terms of grind score. Grind score is the percentage of employees that have a positive business outlook divided by the work life balance. So you're working really, really hard, but you are extremely optimistic about the company. And topping the charts are applied intuition who we talked to, Kayser Yunus, we talked to him at Hill and Valley. The boring company, course, it's an Elon company. You're going to be working very hard. But if you believe that you're

going to drill tunnels all over the world, you're probably pretty optimistic. I had a roommate when I first moved to L.A. that worked a buddy of mine that worked at the boring company. And he, I thought I was working hard, but he would often leave before I was up and come back after I went to bed. It's an extreme, extreme company. I've heard of stories of investors who were trying to get allocation in the company and a prerequisite was basically that you go do a full month of full-time work to, you know, advance the company. It's like, oh, you think you're going to value, you think you're going to add value? Yeah, come out here for a month. Just be in the office every single day. I love to see Ramp up there and Vanta. And ideally, you know, we can get all of our

sponsors into the top 15. Shield AI. Yeah, it's a good crew. Also, Gautier, I'm pronouncing it correctly finally six months into reading his posts on the stream uh i just vibe coded this it works flawlessly and it's a picture of a plane with an engine where the plane should i actually think this plane should exist would this fly it might actually work i don't know you got me now samir yeah we got you are live let's go we're just looking at a random image of a plane uh we got the i mean the the other setup uh we got we'll have you on again uh very soon to get the full experience, but great, great to see you. Great to see you.

How are you guys? I'm sorry I'm giving you this, this non-premium experience, but I mean, it's still better than the nature of stream, maybe. We're live. We're live. So yeah, this is fantastic. Yeah, what's new in your world? What's the latest and greatest?

Give us the little update and then I want to go into just AI, creator economy, Disney, a bunch of stuff. We'll try. Amazing. New in my world. I mean, dude, there's, there, this has been the craziest year of my life. I don't know if the viewers know. the viewers know, but both mine and Colin's homes burned down in January, which is a crazy

experience. And then 12 days later, I had a baby. Colin and his wife just had a baby about three weeks ago. Well, congrats on the baby. Sorry about the house. Very, very tumultuous time, for sure. Actually, it was a conversation with Reed Hoffman. That was really interesting. He, on his way out, we were chatting mostly about AI and the world of creators in AI. and then on the way out, he turned around and looked at Colin and I because he knew about what had happened earlier this year. And he just looked at us and he goes, never waste a good crisis. And then walked out.

And I was like, whoa, that is some amazing advice. So trying to, what's new in my world is trying to understand that statement more and what that means to me. A bit of what that means is we've always wanted to do in-person events. We're starting to do in-person events now. We have a big one in New York on September 4th. So, you know, on one side of the spectrum is like content AI digital. The other side of the spectrum is just like getting back out, being in front of people.

Yeah. Being with the community. I know you've done a lot of public speaking. What is a good in-person event look like for you? Because I can imagine, like you could just, you have such a dedicated fan base. I remember running into you at one of those YouTube vidcons or something. And you were getting swarmed. I imagine you could just do Q&A for two hours and everyone would be satisfied.

At the same time, you could define. you could define like a very tight, you know, 45 minute presentation with slides, take people through the history of the creator economy where things are going. You could also just do interviews, an interview top to your people the entire time. How do you think you wanna play it?

So we actually, the first thing you do is like you ask. So this is an application only event. Okay. And in the application, we just asked, what is this, what is this event have to do to be successful in your mind? Like, what does it mean to be successful? Also, like, the two questions that Tim Ferriss gave us.

One is, what are you here to give? Like, what do you feel like you could give at this event? And then what are you here to receive? What, like, piece of advice? So I would say, like, overwhelmingly, right? Like, there's, there's 380 confirmed applications for this event represents over 300 million followers. Wow. I would say 90% of people just said that being a creator is low.

and I want to meet other people in the industry and like other like-minded individuals. So I would say most events probably undervalue free time and space and prompted networking or prompted conversation and overvalue what happens on stage. I'll tell you like not one application said man I can't wait for a panel. Yeah, yeah. All of that is window dressing. It's always window dressing just to like an excuse to get out there and then you're just like hanging out at the cocktail part of the entire time. But that's the best stuff that actually makes it good. I feel like for me even like when I've been on panels like to have a to have a mediocre panel is probably the easiest thing in the world.

Oh yeah. Right. And so like to make it good, I think exceptionally challenging to make it worth your time. So we're working hard to do that. But mostly like people want to hang out with each other. Like people are lonely. People spend a lot of time at their computer. And like being in a curated group of like minded individuals is fun. So I think we just should have a location. or philosophy around what makes what makes for a good spot? I think New York City is like the greatest, right? Like I just think coming to New York is really fun. 62% of our attendees are traveling in, which is really interesting.

And, you know, the venue is at the refinery Domino Sugar factory, which is like this big glass building that has like a views of the city. So I think what was important to me was like, You walk in and you're already like, this is cool. I'm excited to be here. Yeah. So yeah. Yeah, that makes sense.

We were debating this earlier. There's an article in the Wall Street Journal about how Disney is wrestling with the way they use AI. A lot of companies are going through disruption. Maybe it's a sustaining innovation. Like Microsoft's doing great, right? And it's not disrupting their business. They're just printing money all over the place. Everything's growing from Excel.

all the way to their cloud business, to their AI business. But for Disney, it's a lot more complicated. And they were giving this anecdote of the Rock and using a deep fake to essentially create a digital twin using his cousin, apparently, where his cousin go to the shoot and they put a deep fake over his cousin. Obviously, they pay his cousin. They pay Dwayne the Rock Johnson. But there's still some backlash from the fans, potentially, for using AI at all.

And I've just been noodling on this question of, you know, will Disney, be disrupted by AI? Will they just continue on unabated? Will it be really good for them? So maybe we can just kind of noodle on it for a little bit and just kind of get your initial reactions. And then we can kind of take it in a bunch of different directions. Okay.

So I think my first POV on AI right now is that I think we don't talk about AI and prompting as entertainment enough. And what I mean by that is not like its ability to produce entertainment. but the fact that it is entertaining to chat with an LLM. And I think for me as a content creator, you know, all of us are in the business of attention. Like you guys are in the business too. Like we're looking to create something meaningful enough to capture someone's attention.

And I think Sam Altman has created something that is meaningful enough to capture a lot of attention. And I think a lot of creators think about this as like, oh man, if AI can produce graphics in VFX, then people are out of jobs. And that is totally true. Like that is happening. But at the same time, us as creators, we might be out of jobs if AI itself is entertaining. So I think on the, you know, no matter what,

we live in a landscape and a media environment where like creating stuff is really competitive. It's really hard. It's gotten very expensive. Like studios are going to use AI to supplement production, Creators are going to use AI to say. I mean, almost every creator probably uses it to supplement production in some way, shape, or form. Yeah, totally.

But I think if we think about it from the consumer side and what's entertaining about learning a new recipe or figuring something out on ChatGPT or writing with ChatGPT or asking questions with ChatGPT, that's where I think about like time spent on the internet and is more time going to be spent over here. Are you guys familiar with Showrunner? It's coined as the Netflix of AI. I have heard of it. I'm not familiar beyond just seeing like one headline. So Amazon just backed it. And I think it's an interesting one to look at. It's essentially like they give you a show. So there's a show called Exit Valley. It's kind of like this their version of family guy. Sure. But you can choose your own adventure within the show and put yourself into the show as well. So you can kind of prompt what happens next. And I think it's an interesting thing to look at. out of like making content playable, right, making it interactive.

Because if we look at the most culturally relevant stuff from history as well as now, video games are wildly culturally relevant. And those are stories that you can put yourself inside of, right? That you can prompt what happens next inside of that story. And you look at live streamers right now, Kaisenau, I Show Speed. As the viewer, you actually have impact over what happens in that story. Yep. So I've just rambled.

part a bit about this stuff, but I think that we should really evaluate the like consumer and viewer interest in having a say into what happens next and also having interaction with the internet. Yeah, that interaction thing is so fascinating. I haven't thought about it in that frame, but when I go to chat GPT and I get it to generate an image, it's definitely a little bit like hitting a slot machine. Totally. Because I prompt it and I'm like, I have something in my mind and it's unblurring like it's in the like it's in the dial-up era and it's loading it's like is this the right image the previous generation is you you know talk with the designer and you give them an idea and they're like okay I'm gonna I'll come back to you in a day yeah waiting and you're waiting

you're waiting they send you the files and you're like but it's totally a dopamine hit but compressing that down uh yeah that makes like the dopamine yeah and you can see in the in the charts i think uh chat chbt is up at 30 minutes a day for the daily active users and that's getting into TikTok numbers that and that's getting into real's numbers and and so maybe it's not immediately disruptive but it's certainly taking time from somewhere and if you go back to ted sarandos at netflix saying like our competition is fortnight well you know yeah your competition is chat chippy now if people are having fun on there even if they're not watching you know movies yet on there just the interaction even if it's text based

It's all in the same media. It's all in the same screen time. Yeah. Casey Nystatt came on, I think, a week and a half ago and was saying that he's just like feeling like it's like this tidal wave that's going to wash over everyone and nothing will be the same. And it was almost, I wasn't say it was like super bearish or blackpilled, but it was just like, get ready, it's coming.

It's going to disrupt a lot of different verticals. I'm curious what percentage of creators that you talk to are, more excited about the potential of AI than they are scared. The fear versus greed index. Yeah, exactly. What they say in the public markets. So that's really interesting. I think like the aspiring creator, the young creator is fearful and angry.

And like we get a lot of anger when we talk about AI. Sure. You know, I think for me, I don't feel that way. I think it's, there's an inevitability to it that's like if you are, if you do feel angry, If you are a young creative and you're like, this thing is making logos and people are using it to do their branding instead of hiring designers and you feel angry about it. It's like it's upsetting, but there's an inevitability to it. Like it's not going to stop. It's this is the worst they'll ever be.

So I think the, uh, from the, the professional creative perspective, this is the best it'll ever be. No, the worst it'll ever be. This is the worst it'll ever be. It's just getting better and better. Oh, you're saying that the AI is the best. Yeah. saying it's the best it's ever going to be for you if you're just angry that AI is doing work. No, no. Like the AI right now, what you can do with AI, this is, this will seem comical to us in five years.

Totally. And so it's only getting better. And I think every professional creator uses it in some way, shape, or form. I think one of one of the, maybe the nerdiest things that I ask my friends and that I tell my friends is like if I have a favorite prompt or if I ask them what their prompts are. But I ask. always if we write something, I ask it to challenge based on being a different persona. Sure. So, you know, I've done this as Casey before. Hey, imagine you're Casey and I stat. Give me feedback on this. Sure, sure. Imagine your Johnny Harris.

Give me feedback on this intro. So that's been a really, like, that's probably the simplest, most effective way. But all the way up to like, we use it to mock up thumbnails now. Yeah, I was going to ask you about the thumbnail. Because when, so Mr. Beast had his like AI thumbnail. generator and then he got some some pushback extreme backlash but it seemed like it was from a a bunch of a bunch of it was my point of view was that he should have said like I understand that this is

disruptive to like a small number of people that that do thumbnail art as a profession but I care more about democratizing access to great thumbnails by making anybody generate them and I'm going to keep this in the product. Your point was that he reacting out there was pro creator anti-thumbnail artist and that was like the trade-off that he should have taken. And so he in my view he you know and I'm sure he has a bunch more context but in my view he was siding with the 10,000 people in the world that are that make a living doing thumbnails instead of the millions of people that want to create. I think actually the the core issue with announcement was the face swap. So what he did was he showed two thumbnails and he said it's it's

it's very simple to put your face on this thumbnail. Now look that happens on YouTube all the time, right? Like the reality is that when one creator makes something that's like really successful another creator just goes oh cool that's almost like public domain that idea and I'll just take that and like reinterpret it for my own channel and sometimes people are so egregious to just put their own face on someone else's design. We made a video about this called the copy and paste culture of YouTube because it is a big culture. And I think what a lot of creators were feeling was that Jimmy was making that process very easy. Thumbail designers as a cottage industry of YouTube are some of like the most powerful people on the platform. They literally are the gatekeepers between creators and like millions

of views. Attention. Yeah. Totally. They are very, very talented. Even for us like, what, What happens for us is we're all visual people, so we mock it up with AI, but I don't think anyone, you know, I think Dude Perfect did one thumbnail that was pure play from an AI generated thumbnail, which was an airplane with a bunch of basketballs coming out of it. That was like an impossible photo to take or an impossible, you know, thing for them to mock up. And the thing they got from AI, they're like, this represents the video concept. People were upset about that. But I think mostly how creators operate is they go, here's the concept. How we operate as we go, here's the concept. Okay, here's how it looks visually.

And now we hire a designer to bring that to life. And we still do that. I mean, I think thumbnail designers are like wildly talented and they operate more as like strategists as well. But it's a fascinating cottage industry that people looking into YouTube probably have no idea exists. Do you think that there's a role in the future for like thumbnail concepter because I've made thumbnails and they like they live and die I mean obviously the last step of the Photoshop work is extremely important but you have this like 20 minute video where you talk

about so many different things and there's so many different visual elements and then you're trying to distill it down into like the three element rule or just like one thing and picking that one thing that like the plane I can imagine what that video is about right now just based on what you say yeah but that could have been botched so many times on the concept level and it doesn't matter if you had you know the the best photographer in the world working with the best Photoshop artist and you know the entire AI team from all the different foundation model lab companies working on it like the wrong concept's just not going to break through because people are going to be like this is confusing like I don't want to watch this you can't fix you can't

you can't fix a bad idea exactly you just can't yeah but I think what makes a good idea is really to decipher today. Because I think on the internet, we live in a world of two extremes. One is if I'm going to click on something that has to be something I've never seen before on one extreme, right? So Mr. Beast, let's take that and go, he just raced an F1 car and a I've literally never seen a cheetah race an F1 car. So fine, Jimmy, I'll click on that because I've never seen it before. On the other side of this attention spectrum, I think it's show me something that impacts my day-to-day,

life, either with my friends and family or like is interactive. So your guys show is a great example of something I think lives on that, which is like I get to turn this on and it impacts my life because I get to be a part of a water cooler conversation. I can turn to my friends and I have more things to talk about with my friends. Love Island is massively culturally relevant. I think that is something that, you know, even around my social circles, people talk about. And it, they get to connect with their wife or with their girlfriend or with their buddies. So I think we live in those two extremes. Anything in the middle is really hard. Right? Like educational stuff, very hard right now. I think viewership is going to go down in the middle. It doesn't mean it's less impactful. It's still

impactful. But I think it's interesting that the educational content is it used to be if you wanted to quickly understand a complicated topic, finding a short YouTube video on it was super effective because you're like, I don't have to Google and I don't have to find this summary and I don't have to piece together this Reddit thread and that blog post and this news article and I'm just going to watch this like three-minute video and have a good understanding.

Now you can just prompt the model and say, how does this work? And then you can ask follow-up questions and you can ask follow-up questions and you can click in to see the source. And I feel like that middle is just getting hollowed out and we're not that far from OpenAI just being like, do you want to generate a 10, you know,

10 minute video explaining this? I'll turn this report into a video, right? And that's that kind of like... I mean, you can already do it with podcasting. Mustafa Solomon from Microsoft, the CEO of Microsoft AI, on our show prompted a podcast that was really compelling using co-pilot. And I think, though, have you guys ever used the like kind of FaceTime feature with Chat Chappet? It's the voice mode, right? Yeah.

It's voice mode, but it's video too. Oh, does video now? Yeah, yeah. So I did that in the kitchen. Like three nights ago, I made, you know, I saw some, like everything on Instagram on my Instagram feed is like cottage cheese, right? It's just like put cottage cheese in something. So it was like these like cottage protein brownies. Awesome.

And I was like, all right, I'll make these. And I, in the middle of it, I like had a question. And so I opened that up and I had like the video. And I was like, all right, wait, I only have one banana. Like, what do I do next? Oh, interesting. And it was just live prompting me. And I think the obvious.

thing you recognize is that that won't be our phone soon, right? Like maybe it'll be glasses, maybe it'll be contact lenses. Maybe it would be like contact lenses in an earpiece, but there will be like an omnipresent nature. And that's very dystopian. But I think this all connects to that question of education. Like educational content will be real time personalized in my ear and in my eyes. At least I think. What did you think about the launch of waves, the new glasses, initially for live streamers. It felt like the most polarizing launch where if you were a live streamer that does IRL stuff,

you were thinking, this is amazing. I want to get this as fast as possible. And if you were a member of the public that values your privacy, you were saying, like, please stop this before it launches. Yeah, I think it's, uh okay i think on the live streamer side i mean my perspective is pov is not that interesting maybe i'm wrong but i think what's interesting is like a filmer filming speed doing something

uh you know not necessarily pov the only guy who nails it is you know that guy in new york who like plays the uh the keys and like someone raps with him yep yep that's like the dopest kind of solo dude i assume he's solo maybe there's a cameraman with it but i mean it seems like it's just like one of many niches and there will be creators that are shoot wide angle shoot drone footage of themselves doing things shoot yeah you know like like cinematic stuff it's just like one of the tools in the tool chest but sure people really like humans and they really like seeing speed or kaisenops and they like that relationship totally yeah and i think um feeling like like you don't like you don't like you don't like for example like the idea like the idea you don't like the idea

have an NBA player like wearing glasses. Oh yeah. That might be cool for a second. Yeah, where they're dribbling down. Yeah. Like shooting, pulling up. But then you probably want to just go back to like feeling like you're actually there in the stadium.

I would get bored quickly with that. I think on the privacy side, like, you know, you guys are in L.A., right? Like you see Waymo's all the time. I think there's a lot of concern about that too of like, okay, now we have more cars on the street that just have like tons of cameras on them. And I think just generally there's, you know, there's a lot of concern about that. But it is like a natural evolution to us just unloading our lives and data on the internet. Like people are concerned about that.

And they're the concern about their chat, QPT search history or question prompt history. Like, I think Scott Galloway said it. Like if you want real anarchy in the world, just release everyone's Google search history. Like that would end civilization. And I think it actually would or do you think everyone would just be like, okay, we're all equally weird. Yeah, or it would be this like cool moment of just like the veils lifted. Yeah. John's search history.

How much creatine can I take? Legitine. Legally. For losing my life. I don't know. I don't know how this stuff's going to play on. I just think that, you know, I did get, Colin and I did get the opportunity to spend some time with, with Sam Altman. And I think, you know, I, I, I, my curiosity with having a lot of the, like, like, like,

Mustafa on the show or Reid Hoffman on the show is to understand the incentives and the curiosity of the creators of these programs. And even with someone like Sam, like, it's actually hard for me to fully tap into. I understand the desire to create and to innovate and to be number one and to push things into a place you can never imagine. But I think the implications of this are pretty different than anything else. And I think the implications of advances in AI are going to change human society in a way of that, like, I think Casey's right. It's going to change things in ways we can't imagine. Yeah, one thing that's interesting when I think about creators that I've watched over the last decade, you think about in music like Anthony Fantano. Yeah.

There will be an AI creator that just instantly reviews a new Drake album. and it can come up with some analysis of the song and it'll tell you what this lyric meant and it'll say this sounds similar to a song or an unreleased track from this period and it'll give you this context but people will still want to hear what Anthony Fantano thinks about the album.

I think we look at it in terms of like efficiency and inefficiency. There's certain times where we want efficient media. So like let's say I'm going into a meeting and I'm like, wait, I didn't prep for this. I need something really efficient. Like when I was in high school, that was Spark Notes. Do you remember Spark Notes? You guys are too young for Spark.

No, no. I use Spark Notes, of course. Spark Notes was like, if I didn't do the required reading of class, which I rarely did, I would just- The original clearly. Yeah, exactly. I would go like, all right, I got the Spark Notes. So I think there's like efficient media. But then if you look at, there's also people who just love reading a book.

So there's going to be both ends of the spectrum, right? I wish in college if you could generate a podcast on any topic. I mean, that is. Totally. That would be amazing. be like I'm going to go to the gym and when I finish this I'm going to know like 90% of what I need to know about this book I'm going to go into the test I'm going to write the essay I used to get the audio books for every book I'd buy the book for the course and then I'd also buy the audio book because there's an auditory learner but we can't assume that just what humans want is efficiency all the time right because I don't want efficiency all the time like you think about going to a concert going to a concert is the least efficient way to hear the music from that artist you go you have to park especially in LA it sucks you got you have to figure out part And there's traffic on the way in. You go in and like, why are you there?

You're there for a collective human experience. You're there for the messiness of how John Mayer is going to play. I'm speaking about a John Mayer concert. So you want to hear him play the guitar differently than he would. If I just want to hear the song, Daughters, I'm just going to go on Spotify and hear the song, daughters. So I think you just play with those two worlds of like, we will always kind of crave efficiency and inefficiency. There might be areas where we want it more. So like, brainstorming is a really interesting example.

example, hey, I want 100 variations of this idea. Presently, an LLM can do that better than five people in a room and faster than five people in a room. If you want to see every variation of an idea, but you know, you want to just watch a movie and see if you can come up with ideas based on that movie. Like it'd be hard for an LLM to match your taste in that way. Yeah, I just think there's like efficiency and inefficiency and they're both going to play together. We really got to book an hour for this. We could just keep going. Yeah, yeah, yeah. this is fantastic we do have another guest no let's do it in person next time yeah oh yeah you're in l a

you can just come by dude i would love to come see it also that's amazing i mean i totally botched my uh my good looking set i'm sorry it's high stakes of the live stuff it's high stage we saw we we saw the potential we saw a glimpse of the potential it was a human it was a humanizing moment for your audit you know yeah you should come in you should come one day and close out the show with us and then we can pick your brain and you can help us get an up great John just said pick your brain. What's that?

Well, I mean, I just asking Samir, I'm sure he gets that message a thousand times a day. I'd love to pick your brain, Samir. Tim Ferriss, Tim Ferriss gave me a hard and fast rule. He said, if anyone says pick your brain, don't go to the meeting. Yeah, don't go to the meeting.

So that's my hard and fast rule. I will not be there. Okay. You have another suggestion. Then I'll come in for that. We'll hang out. We'll have lunch. We'll get the gone.

Awesome. All right, guys. Great to see you. Thanks for having me on. See you guys. Cheers. Talk to you later. Bye.

Up next, we have Flo. Coming in the studio to the TVPN Ultradome. Legend. Welcome to the stream. We kept him waiting. We'll try and move through these next guests quickly, but very excited to catch up with him. How are you doing? Boom.

Welcome. John, we'll be back in a minute. How are you doing, Flo? Good, good, good. How's everyone doing? Great, great. You know, it's been a slow summer. Not much news and AI.

Everyone's just clearly taking a breather. Of course not. I know you guys have been busy. What's going on in your world? Yeah, just announced 3.0. It's been an insane last couple of weeks. I've been working on this update all year. Yeah, I mean, we're just making it way, way, way easier to create way more powerful AI agents.

Incredible. When did you decide to adopt the versioning for Lindy? Because it feels smart to kind of anchor like the team and your customer. around these major updates to the product versus historically people might have a product launch, but they're not necessarily like signaling like, hey, this is a massive advancement. It's an entirely new version. Yeah, no, we've done that from the beginning. Like 2.0 was like in November last year.

Actually, you know, even smarter, and that's what I want to do next time, would have been to name that. I want to, like, Lindy's proponent. You know, like, how Apple does, like, Yosemite and stuff? I want to give them good names moving forward. Yeah, I'm trying to think we'll come up. We'll come up with the right series of names and the right category. Do something. I'm thinking like sci-fi books.

Like what would you think of like a fire up in the deep, you know, or like the metamorphosis of pride? Or Lindy Snow Crash or something like that. Yeah, exactly. That's great. How are you guys using the new Lindys internally? Oh, we're using it all day. Like I think one of the most insane things that it's automated for us is like it's replaying. to QA engineer.

Wow. It's just like, so it's getting very metallic. We are using Lindy to QA Lindy. And so every hour we have like a Lindy agent that wakes up that tries the entire like cool flow like literally signs up with a new account, tries to pay, create a new agent, tests the new agents, you have an agent creating an agent and testing it. And then if anything goes wrong along the way, pings the on-call engineer. So we've been using that for weeks and it worked beautiful.

It's amazing. How do you think about reducing the time? I guess time to value for somebody that's coming in to use. Lindy, I imagine that's key. People are so used to signing up for a SaaS product knowing exactly what they're getting and having there be a lot more kind of controls around what the product does. So it's not sort of like build your own adventure, whereas I feel like with Lindy historically, you do, like the customer needs to really embrace,

like, hey, I have to figure out how to, when you hire a new employee, very few roles can somebody just come in and without any ramp up time just immediately start crushing it and and so I'm curious what that looks like for you guys yeah can I show my screen does that work it does everything's live so whatever you share will be okay what if you're reading to the immutable record of the of the internet oh that looks great well yeah so I mean this is like an excellent question like that's the problem we've always had from day one is like it's almost too powerful. You have to do everything yourself.

And that's why the vision from day one has been like, it's just like an employee. Like it's very easy to set up. You can just talk to it and tell it what you want to do. And so now the new app is just text field. It's like this thing and you tell it what you want to do. It's like, hey, I want an agent that will monitor my support inbox. Then you can like iterate with it like very interactively.

And we have a bunch of templates here for like sales, customer support, personal productivity, engineering right here. Like the QA assistant is one of them. Or, you know, like a software engineer agent, like human. like human resources, like a recruiting agent, or like a company knowledge base. So we give you a bunch of templates, and then you can just create your agent in like a minute and a half.

Are you running into any problems, just linking up private data sources? I mean, I'm just thinking from like a consumer's perspective. You have a bunch of news subscriptions. You want to log into those news websites and kind of collate a digest for you. That's something it would need a password for, it would run into a number of paywalls.

Are you seeing a lot of luck with jumping through login flows or are you getting hit with CAPTCHAs constantly that are intractable? Oh, no, I mean like, so that is a big part of what we're announcing today is like autopilot. So basically your agents now have their own computer. And if you need to give it a login, you can literally take control of the computer

and login for them. So let me just pause the screen to here because I don't want to show anything crazy. Yeah. Is that something that you do like once or because I'm, I, like, when I think of the benefit here, I think of like a really long running like cron job based, uh, uh, agent, essentially something that feels more like a worker on my team. You're still screen sharing by the way.

Yeah, your list of passwords is up now. Just kidding. API keys are leaked. And it's, oh wow, you did a billion in revenue yesterday? No, I mean, yeah, you're looking only once. So it's like this is, I am screen during on topos. So this is an agent that set up. This is actually a little bit of a dirty secret.

Like I don't like LinkedIn, I'm all of a Twitter guy. Sure. And so I post on Twitter. I don't want to post on LinkedIn. But I have an agent that, so it is a crunch up. Wakes up every day at 9 a.m. Cross-busters. And you can see it's logged in from my Twitter account.

Like this is authenticated as me. And so it goes on my Twitter. It's also authenticated as me on my LinkedIn. And it basically looks. said the posts on my Twitter and it's like, hey, this is what I think you should post on LinkedIn. That's very cool. Yeah. Yeah, that makes sense.

That's awesome. And it can obviously, like, hydrate things and add extra context and really, like, transform the post properly as opposed to just, like, a default binding. Flow posts on X. Very excited to announce Lindy 3.0. It's our most powerful Lindy. And then it converts it to the Lindy then converts it to LinkedIn. It's like, what's up, LinkedIn?

What I learned about B2B SaaS? Today we're transforming the power of agents to unlock productivity for everyone in your organization. No, the real clickbait for LinkedIn would be like, you know, three weeks ago, three weeks ago, I hired a software engineer who was working at 25 different companies. Here's what it taught me about building 3.0. Something like that. Anyway. Well, congrats on the launch. Anything else we should know.

Where can people sign up? Yeah. Linda does AI, it's available today. I don't know about y'all, I find it frustrating. There's a lot of vapor wear in AI. It's like, this is coming one day. It's like, no, this is available right now. It works.

It's incredible. What's the pricing model right now? Don't worry about it. But consumption or subscription? It's consumption. It starts at $50 a month, and it's a tenths of the price of a human. Okay. This is going to be one of those things. Am I going to hear horror stories about people that, like,

forgot to, you know, turn off their Lindy, and they ran up. a $10,000 bill? Don't worry about. It's not. The cost of inference is declining rapidly. It is. Everybody's going to make it.

We're all going to be fine. Yeah, we're going to distill the models. These aren't. I mean, look, you know, this has happened before and we're very generous with credit. People reach that to us. They're like, can I get a refund? This happened. Yeah.

Of course. Yeah, that makes sense. I need cool. Cool. Any other, any other tools or custom models or like RL environments that you're tracking or excited for progress to be made? You mentioned computer use, but what else do you want the foundation labs to solve so that you're like a beneficiary riding the wave? Memory.

Memory. Definitely. Memory, I think is the big gap. And I think like there's a lot of attempts to solve this problem across the industry, but I'm noticing this pattern where it's like people and companies like ours that are stuck and cannot change the weights of the models. They think of all of those like convoluted schemes to like hack the model to do something it doesn't when really. to belong to the model. And when you talk to the research shows at the labs and you ask them, it's like, ah, we're working on memory, this is what we're doing.

They tell us, don't do it. Don't do it. Let us do this. We'll do it. We'll do it. This belongs to the model. And the model is going to do it a thousand X better than you do. And I expect a much more advanced memory schemes where it's like basically the model is going to be retrained to come out in like a year or something, like six or 12 months.

What are you expecting from GPT5? Just I'm expecting it to be smarter. Look, we feel sparks of a GI daily at this point. Like we had this moment last week, like we basically accidentally built lovable. Like because we gave a computer to Lindy like someone who was like 10 p.m. at the office is like Flo, like, hey, if you ask me to build a website, she builds it and she deploys it, it works. And so like to me that was like a huge holy shit like like a GI spark of a GI moment. But like look, it's still not as good as a human. Like sometimes it's just just. It just makes dumb mistakes still today, you know.

And so I'm expecting GPT5 to be like a step change and to be a big step to wheel to GI. But I still people don't realize what's happening. And I get it. People are tired of like all the AI hype and the noise. And I tell them like if you're tired, man, like this is only the beginning. It's going to get a lot more noisy out there. Like it's going to get pretty crazy. Yeah.

I mean, I'm just, it should be exciting for everyone because Lindy today feels like the promises is automate. these tasks that are boring that a human does the first 10 times they do it. It's interesting. The next 10,000 times they do it is very boring. And so you just free people up to do these more like higher levered zero to one type type activities that can then be automated and systemized over time. So very, very cool. Congrats on the launch.

I'm sure, I'm sure, you know, people are going to be, you know, making bets on Lindy 4.0. Is Lindy 4.0 going to be able to get a college degree for me, you know, all that good. stuff congratulations that is that is that these models all actually today able to get college degrees yeah as an interim right yeah yeah yeah the top to the bottom so yeah yeah awesome well congratulations we'll talk to you soon thank you have a great day bye up next we have Dave from Rune technologies coming in to the studio get that gong ready jordie hayes and welcome to the stream how you doing Hey, how are you?

I'm good. Welcome to the stream. Thank you. Give us a brief intro. Give us an update and tell us the news. Yeah. Yeah, so Dave Tull, co-founder of CEO of Rune Technologies, founded the company last year with my co-founder, Peter, who I know was on a little while.

We just raised our series A, just announced it about two weeks ago. So excited to have closed that. Didn't even give you the number. Sorry. Sorry. Didn't even get to the number. They didn't even get to the number, but congratulations on the series A. Give us the number.

Give us who led, and then we'll go into more about what the company's building. Yes, we closed $24 million, led by human capital. All of our current investors from Andrescent, 0.72 and XYZ reinvested. And then, you know, new investor is human and, you know, Washington Harbor Partners as well, a new investor here. And we won't announce it yet. There's going to be another one. But we'll save that in a few weeks as a little surprise. Fantastic.

Nice. Give us the brief history on logistics and supply chains in the U.S. military. I mean, what we're using in World War II? What are we using today? And then what does the future look like? Yeah. And I mean, in many ways, the insane part is some of it hasn't changed, right? Like we're tracking things in analog systems. You know, the means of a whiteboard has changed, obviously, from scratch paper back in the day. And, you know, sometimes it's even Excel documents. But great. Like the basic. nature of it is still an analog human-centric tracking of things, whether that supply inventory levels, vehicle maintenance levels, vehicle locations, and those sorts of things. And especially of the last, you know, I came out of uniform. I'm still serving part-time, even in the Army National Guard. And when I think about, like, how do we track what we need and where we need it and when we need it,

that is still a human-centric analog process that has remained generally unchanged, at least in the last 20 years with the advent of, you know, computers at some level and connectivity at some level. So yeah, what's the path to actually getting this in the hands of the warfighter? Obviously, it's a newer company. You've made a lot of progress, but it's never easy getting a program record. It's never easy getting full deployment. We've talked about the history of Palantir took a long time to get that out into the world. What's the plan?

What's the update? Yeah, yeah. I mean, you said it. I mean, it is a long path. We know that. All of us have come from this industry. We've lived in the defense space for years. Roon, you know, from many other companies. I'm excited to say right now, like there's a team down

at an Army base as we speak, deploying Tiroz at an exercise with an Army Corps. So pretty excited about that. That'll be our second deployment within the Army. We haven't announced you, but we also have worked within the Marine Corps as well. So we know we got to spin the flywheel with these types of things. And core to our DNA is how do we get out to these exercises in real world scenarios with warfires, with our sustainer, you know, sustainment customers and do this. And that, you know, build closely with the war fighters in many ways what what palliare pioneered and what andro pioneered right how do we do this with them to then build momentum get those get ahead of the requirements do things very quickly and iteratively to then get to that goal of writing enterprise you

know our production deployment in those types of situations so you know you guys well know that you know seed round is about hey do we have product market fit is the thing that we're building needed right we have proven that and what we thought was going to take 12 to 18 months it we did it in eight months with initial pilot contracts in the military. And now series A is about how do we scale the team and how do we ruthlessly execute on the things we need to do, right, to meet the milestones we have to actually deploy this across the, not just the joint force, but even potentially our allies as well. How much, you know, in the recent conflict in the Middle East, how much are you trying to learn and how much are you guys paying attention to the challenges

they're facing there versus, hey, we already know our systems and processes are so broken that we just need to bring them. And so, you know, ancient, I'm sure, you know, we just need to bring them into the present and then worry about kind of more specific deployments and problems. Yeah, well, I mean, everything's a learning opportunity, right? So whether we're talking about the Russia-Ukraine war and how the speed and the scale of sustainment that has to happen in a peer or near peer adversary, or we're even looking at the recent events in the Middle East, right? There was a not insignificant consumption or expenditure of U.S. surface-to-air missiles, even in the 12 days of conflict in CENTCOM, right?

So how do we think through knowing those expenditures in real-time or near real-time, being able to even predict those now with those expenditures into future needs and being able to position things and reposition things, inventories, right, in this case, munitions in the right place? To me, those are all important lessons learned, right? We have to be able to do it at the scale that is going to be required and at machine speed. And then we have to take a real hard look at like, what are the exquisite things that are very low density that we need to get? Like exquisite munitions in systems that are going to be required in a conflict.

Should a conflict come? Hopefully not, but should it come. And I think, you know, Peter and I founded this company in many ways because I don't question that we're not going to have the most exquisite munitions, the best weapon systems in the world. What keeps me up at night is like, can we get them in the place we need them out? at the scale we need them at the right time, right? Can we actually do that? Like that's something that I think is lacking. And frankly, has been overlooked by the technology,

you know, companies of today, right? And how do we sustain the force and how do we do this with modern technology? Yeah, yeah, that's the entire conversation. Just, yeah, capacity, more than capability. So thank you for everything that you're doing. And thanks for joining the stream. Congratulations to the team too.

Tremendous progress. Love to see you guys making 18 months of progress eight months. That's right. Keep that right. Keep that right in the future. Yeah. Awesome.

Welcome back on the stream anytime. We'll talk to you soon and have a great day. Talk soon. Thank you. See you later. Up next, we have Will Ahmed from Woop coming in the studio. Welcome to the stream. Second time on the show.

We had him on around the last product release and we will invite him back to the stream. There he is. How you doing well? Good to see you. What's up guys? Good to see you. What's up? Great to see you.

Where are you right now? What's in the background? This is Whoop HQ. I'm in my office. And if you look right here, that's Fenway Park. Oh, that's right. We're in Boston, yeah. Boston, very nice. We got some claps from the team in the background. Someone's happy about Boston. Anyway, give us the update. What's new in your world? Well, we recently launched the Whoop 5.0 and the Whoop MG. So two new hardware, three membership tiers, a bunch of new bands and accessories. the market seems to be loving the product. Business has been really good. And yeah, we've launched a lot of new features. We just came out with a health span feature that will tell you your whoop age. We worked with the Buck Institute, which is the leading institute in the country for longevity. So we found that, you know, the top nine biometrics that correlate with all cause mortality. I'm sure you guys have very low whoop ages. And how do you think about when just sorry to interrupt, but a lot of anybody that, anybody that does any type of like health testing diagnostics, a lot of people are adding an age or this kind of like age score. And it feels like there's incentive for some players to just put the age as low as possible so that people share it.

That's obviously not great for the user. They're like, I'm 35, but this score says I'm 25, so I'm going to share it, you know, on social media. I took one of these biological age tests and it said that I had the mind of a four-year-old. The mind specifically, the body of an elderly man, but the mind of a... By the way, is that supposed to be a compliment? But it is funny. People have said I have the mind of a golden retriever. Yes, yes.

It says you're as smart as a four-year-old. It was impressive. But my question is, yeah, it is kind of like a little bit of an optimistic flip because you could flip this around and just say your life expectancy. Like the average life expectancy is 75 and you're at 85 based on how you're aging. So that's good, but that's much darker. It does like the death calendar. Yeah, I remember those apps. And those are much darker and spookier and weird and macab.

But this feels this feels much more optimistic, much more tractable, something you can work towards. But what are the pitfalls? What do you, what did you make sure where you're, without naming competitors? Like, like, what did you want to avoid in the, in the, in how you designed your product? Yeah, it felt like, it feels like some are just a random number generator. Yeah. And you're doing it with the Buck Institute. Yeah.

Sounds. Yeah. Yeah. Look, we, we worked with Dr. Eric Verdon, who's literally one of the best doctors when it comes to longevity and the research institute there. we found nine biometrics that most closely correlate with all-cause mortality. And they were all things that we measure. And so we're able to show in great detail how WOOP members stack up to other people their age in a bunch of different factors, like sleep consistency, hours of sleep, steps,

you know, time spent in heart rate zones, your VO2 max, your weight, a number of, you know, kind of obvious metrics that we can qualify. And we also show all the research that we used in developing the feature. And look, it's not everyone wins at soccer. This is not a participation trophy type product. You know, I think like people generally associate WOOP members as being pretty fit and healthy. And I think it's like 55% of people on Woop are younger. I was going to ask.

That means like 45% are seeing that they've got. some work to do and and that's how we think it should be you know it's it's designed to be really transparent and honest yeah but that's a good balance because if it was like 99% of whoop customers just happen to be you know half their biological age you'd be raising some eyebrows but it seems like you dialed it in appropriately anyway is that is that a health claim do you need to get that approved by the FDA are there other things that you that where the technology has progressed but maybe the legislative framework or the government regulation framework hasn't advanced as fast as you'd like and you feel like the consumer might be getting left behind because of some some

some croft in dc it's a good question because we've had some challenges with the fda recently i think it's worth just stating that the the role of the fda is to regulate products that have a medical diagnosis and uh and the 21st century cures act makes it pretty clear that if a product is intended for wellness, it's not designed to be regulated. And the, you know, the structure of this exists for heart rate, sleep, so on and so forth, where let's take heart rate, for example, that's existed, heart rate monitoring has existed since the 80s with chest straps and other tools. Obviously, WOOP does it today, Apple Watch, so on and so forth. And so you can show heart rate in the context of sleep or exercise or whatever,

and that is wellness and therefore is not regulated. If you want to tell someone that they have AFIB, that all of a sudden is a diagnosis, and so that would need to be regulated by the FDA. Now, just to say it, we also spent the last two and a half years getting regulated with the FDA doing clinical trials, and so now we have a medically cleared product by the FDA to do ECG monitoring from the wrist and also to do AFIB detection.

So that's just one example of the difference between wellness and medical. You know, with sleep, you can say hours of sleep, you can do staging, that's wellness. If you want to tell someone they have sleep apnea, that's medical. And where the FDA is inconsistent on this is with blood pressure. So who spent the last three years and tens of millions of dollars innovating to develop blood pressure insights. And this is a really cutting-edge feature that allows us to measure blood pressure from your wrist and give you a daily estimate, which just to say it is a pretty big innovation in the wearable space. And we brought this to market in May.

And the FDA has come out saying that they think that it should be a regulated product. Now, again, back to this distinction between wellness and medical, we built the feature for wellness intended use. So it has a lot of disclaimers around it that it's not designed to diagnose hypertension. It's not a medical device. And it shows you how your blood pressure, uh, impact. sleep and stress, how it's impacted by nutrition. I had a woman in my office last week, and she was seeing low readings on her blood pressure

insights page. It turned out she was dehydrated. She drinks more water. It goes back to a more normal reading. That's like a classic wellness use case. So we built this wellness feature, and now we're in a debate with the FDA over it. But we're fighting to keep this product in market because it's a... At the end of the day, it's the right thing to do.

Consumers are loving it. It's innovative. And by the way, it's the law because the 21st Century Cures Act says that you can't regulate wellness intended features. What are some of the break points for amount of money to really have a good relationship with the FDA? Like I imagine like a seed funded startup has no chance of getting cleared as a medical device for you know $300,000 investment. I imagine Apple should have no problem because they probably

have an office in Washington DC but you know you're like a pretty big company I feel like you should be able to afford it but like how how tricky is it and how expensive is it to actually just go for the other designation with any of the the wellness features and just kind of shift them over into the medical feature. Would there actually be a product or consumer benefit if you did that? Well, it's a good question. There are a lot of costs associated with being a medical device. You have to go through a clinical trial process. You have to work with the FDA.

Depending on the functionality, you might need to hire a lot of consultants to support that. So it's a hard thing for smaller companies to do. The other thing just to say it, though, is it's time consuming. So, you know, it'll delay your time to market by two, three, four years depending on the type of FDA approval that you're seeking. And then I think the real main question is what's the intended use? Is the intended use around describing wellness or is the intended use around diagnosis? And so our intended use today is using blood pressure insights to help people understand their

wellness and in the long run of course we can imagine having a more diagnostic product but that's not what we've we've shipped yet today nor is it what we're claiming to have yeah and I imagine that even even if you have unlimited money you still will be delayed by a couple years while the FDA reviews these claims and then the problem is is that you're shipping your hardware like every year so you have millions of people that are dehydrated yeah that blood pressure because of it and aren't even able to be aware. Very rough. Well, what's next?

What is the plan? Is there, is there anything on the horizon that you want to get through or, or move over? Or is it just kind of like sort this debate out and then move on? Oh, we're going to keep working with the FDA. I mean, like I said, we've been working with the FDA for years. And I have a lot of respect for the FDA. I think they have an important governing function in this country.

I just think they've missed the mark on this particular case. And so we're going to fight for that and we're going to fight for health data for Americans. In terms of where WOOP is going right now, we just announced a bunch of new features that are coming out before the end of the year. One of which I'm very excited about is advanced labs. So WOOP is going to be coming out with our own blood testing. You're going to be able to upload all of your past blood test. It doesn't matter if it was through a doctor or another product or what. That's all going to be able to live inside whoop.

It'll be grafted over time. It'll become part of your whoop data. And then on a go-forward basis, you'll also be able to take blood tests. And this is a real new capability for the company. But it's going to help bring more and more of this data under one umbrella, one roof, so to speak. That's all HIPAA compliant, I assume? Is that the compliance framework that matters? Yeah, you have to work with the right partners in order to enable it.

And obviously the testing is pretty sophisticated. Sure. We've got, you know, some great partners in the space. But a lot of it's also explaining what the data means. Yep. And I think that's where WOOP historically has been quite good is, you know, presenting information and trying to coach on it. Yep.

And what's powerful about the blood testing is we'll now be able to pair it with your 24-7 physiological data. So, you know, maybe if you're not sleeping well, it has something to do with a biomarker or a blood test, you know, maybe your exercise is improving something here. You want to be able to have all this data under one umbrella so you can understand it. Makes sense. Well, thank you so much for stopping by. Good luck with everything at the FDA and good luck with the rest of your day. Have fun with your friends over there.

I'm sure you guys are spending a lot of time together. The frequent flyer miles as you go to Washington, D.C. regularly. Or maybe you take the train. Who knows? Anyway, thank you so much for stopping by. We'll talk to you, dude. Thanks, Jets. See you.

Have a good one. Bye. Up next, we have a niche from NIT coming into the studio. We are talking about them, and I think we got some good news. I think we might be hitting the gong game. I'll let you do the honors, Sean. Ooh, first gong hit for me for the show for the day. And we're going to have to hit the gong for Palantir.

How are you doing? Would you mind kicking us off with a little bit of introduction on yourself and give us the latest news? Yeah, for sure. Well, first of all, thanks for having me here. My name is Niche, one of the co-founders and the CEO. of NIT and yeah we're excited we announced our 16.1 million dollar series A on Thursday. Let's go. Congratulations.

Backhand. Backhand. Nice. Backhand. You love it. Amazing. Talk to us about the company. How are you pitching it right now?

Who are the customers? Break it down for us. Yeah, for sure. So yeah, NIT, we have built a system of agents that is really building the simplest way for enterprise consumer researchers to go from, hey, I have a business question to here's a story I can share with my stakeholders. So yeah, what we're solving is this age-old problem of any time you want to talk to your customers, especially these large enterprise brands. It's taking them four to six weeks. They're spending anywhere from tens of thousands to hundreds of thousands of dollars. And we're

helping them do that now in days instead of weeks and at a fraction of the cost. So working with the largest enterprise consumer brands today. What is the is the key differentiator just AI? Are there other differentiations on the go-to-market strategy or the way you build the product? Like, is there counter-positioning against the big, you know, 800-pound gorilla in this category that, you know, we know, we know, and it might be at a basketball game, but, you know, he's definitely not taken days off recently. Yeah. So, yeah, I mean, great question there. So basically, at NID, you know, what we've really kind of, what we're putting our hat or hanging our hat on right now is this idea researcher-driven AI, right?

So we were an AI-native company. We started the company a year or two ago when AI was still not as prevalent in our industry as it is today. But what we realize is that, you know, you can run this data through the AI, but really what it's missing is the context that the researcher brings, understanding how this shows up in your business, how do you actually sell it into the organization? So throughout our platform, the way that shows up is one, we always have the researcher at the center of how we run research. So the researcher on our platform will go in and share,

this is more information about my business, here's the context of my business, here are my stakeholders and how they like seeing the data so that we can get the output as close to that stakeholder ready. And then the second is we've invested pretty heavily into an internal research team. So throughout the process, all of our customers are paired with a dedicated expert researcher. And what we've seen is that that gets the AI to a better spot, a place where our research customers can actually then take this. It's a really good first draft. They cleaned it up a little bit, and then they can share with their stakeholders pretty

quickly. How are you closing all these logos? You got Amazon, Mars, ESPN, T-Mobile, Paramount, NASCAR, Moet, Hennessy, overtime. It's an insane lineup. Congratulations. Crazy stacks for Series A. Yeah. Well, start off with some good old hustle.

You know, I flew out, met a bunch of our first. customers that just kind of figured it out. But really what we've invested, going back to this research-driven AI piece, is we've really invested in the human aspect of it. So we primarily meet most of our customers through events. We show up. We go to all of the major events in our industry. We show up to their offices, take them out to, you know, take them out to dinner and invest heavily in business-based meetings. The other aspect is really around our... What about NASCAR? Did you ever cross? a beer at a NASCAR event for research purposes.

I feel like that's a good sign of respect in NASCAR culture. The NASCAR team was kind enough to invite us out to Daytona when we kicked off our partnership. So yes, they know how to have a good time for sure. Just watch they only drink champagne and it's just the biggest narrative violation of all time. It'd be great. So where, I mean, you're talking about the customer, the human centric research.

Where do humans excel in this problem set, in this domain? And then where does AI thrive? Like where do you not want to delegate a piece of the work to an AI and where do you want to spend zero minutes of human time focused on something? Yeah, absolutely. So where we've seen humans really excel is storytelling. And our entire platform is built around,

how can we take all the time intensive kind of grunt work that goes into research? So think, you know, setting up your first draft of the questionnaire, analyzing all this unstructured quality data, whether it's video data, open-end text data, and we've built really good AI that can get you that first draft, whether it's a questionnaire or a final report, and really where the human researcher comes, both our internal team as well as, you know, the external kind of partners that we work with is how do we take those insights and tell a really powerful story here, something

that will actually drive action from our marketing team or product team or strategy team. Yeah. How important is it to build a like liquid, supply of people that are willing to answer surveys. I, you know, when I think about Amazon, Mars, ESPN, T-Mobile, Paramount, NASCAR, Moe Hennessey, JBL, like a lot of those research projects are probably just, I want to know how Americans feel about my product. It's not this hyper-specific talk to just my customers, my 10 people. I can probably call those people if I'm running that company, but if I want to know,

how is my brand perceived national- or globally. Team mobile also can do like easily survey active customers. Their own customers. New cohorts, old cohorts. Yep. Whereas like Moe can't necessarily. And so how important is it over time for you to be able to bring surveyorable people to a platform? Yeah. Yeah. Well, first of all, I would say we're very data agnostic, right? So we work with our partners where they layer us on top of their existing customer list. We work with some of the larger panel companies in our space. But as making that stance that we're data agnostic, one thing that we've invested a lot of energy in and technology in is how do you make sure the quality of that data is really good? So a big problem in our industry today is data quality challenges. There's fraud, there's bots, there's a lot of not so great stuff happening. And so what we do is we've invested a lot in our AI tech too, especially if you think about video, for example.

If you collect video responses, you can actually see the human on the other end. We analyze across seven different plus parameters to make sure that that human is who they say they are and actually speaking on topic. So the data quality is an area we've really stood out for our partners as well. Very cool. Anything else? No, this is great. Congratulations on the progress. We will talk to you soon.

Have a great rest of your day. Cheers. Talk to you soon. Let's rip some timeline. I have been waiting for this. You have some breaking news. Breaking news. Saw Hill Bloom, five types of wealth.

over 300,000 copies of that book sold. Incredible. Not surprised. You know why? Because we reviewed it on the show right as it went live. And remember, interestingly, we tried to guess the five types of wealth. Cars, watches, art, homes. Horses, yes.

And we were wildly wrong. Apparently, it's something about like living some sort of balanced life that's not that materialistic. Money's only one of them. Yeah, did we put it in the true zone? He had a little bit of contrarian take, yeah, but it's done very well and congratulations to Sahel. What a run, 300,000 copies of that book sold, absolutely massive number. So congratulations to him. Well, Palantir earnings came out after the close north of a billion dollars in revenues for the last quarter, up 48% year over year.

beat by 7%. It's going to be 1.08 billion next quarter. Estimate 11%. And then they raised their full year guide 6% from last quarter and they had already raised it 11% from the start of the year. So they have 57% free cash flow margins. Congratulations. Bad and 81% gross.

margins. It doesn't look like a consulting company to me. It does not. Jammin at Altimeter, broke it down. Let's see how it's doing after hours. Benjamin. Jammin. Up another 3% after hours.

This is about to be, I mean. It's a big company. It's going to have to be MAG8. Throw them in there. Mag 8. Nice number. Nice number. Chinese number of wealth, right?

That's right. That's right, John. Of course, let's do some timeline because I love it so much. I love time line. Dumer says, what you got? So Dave Asprey posted yesterday, huge news for Jim Bros. A new study reveals that incline walking burns more fat than running. We've been getting into this.

We've been doing this because of the bodybuilders. When they're on cuts, they do incline walking. And it's so much easier to get yourself into the mood of just walking. at three miles an hour on a 10% incline. Total bro science victory. It's total bro science victory. This is hilarious. Doomer says jog cells absolutely molding and seething over hike chads.

It's remarkable. Yeah. I mean, when you do that incline walk, you throw on a podcast, throw on some TVP ad, listen to us, watch this, and start burning the fat. Mickey Friedman says it's crazy that an L.M can win IMO gold before making a single funny joke. The only thing... It's not that surprise. They do that format where it's the be-me jokes. Yeah, they're pretty good at that.

But in general, yeah, RL is... Funny jokes seem to be RL-resistant, whereas the IMO is easily defined by a benchmark. And therefore, hackable. But still, you know, still valuable. And I like this. I prefer that it focuses on IMO gold because I can't do that.

but I can make funny jokes. Take Kim, author of the NVIDIAW says, here's what I would do if I was CEO of Apple. Quadruple the RAM and iPhones are 32 gigabytes and have Max model at 64 gigabytes. Memory is oxygen for local on device AI. More equal smarter and more powerful. Take the margin hit. Memory isn't even that expensive.

Buy Mistral or Anthropic and invest $100 billion in AI compute annually. I don't agree with this take. How much does it cost to buy Anthropic right now? 500 billion or something? I mean, the bigger issue here is, like, one, I just don't see France selling their national AI champion. Oh, totally. Why would Europe ever allow that? That's the whole point in company.

And then Anthropic is dominating in code gen, and, like, very clearly could be, it seems like they could be a trillion-dollar business someday, just generating code for the world. So I just don't see Apple. You know, also Apple buying a, you know, whatever price it would have to be a pay. Yep. And then like kind of, I don't know, keep the core code gen business running. And then. Well, regardless, shout out to take him. He's sending us some books of his book, the NVIDIA way.

But I do agree. He says Apple has the capital. Apple now knows the AI computing shift is existential. Where is the bold action and urgency there? The only big tech company not in the AI I raise. Completely agree. Take the short-term pain to have long-term relevance. Yeah, to definitely invest.

Might tweak my recommendation, but love the thought experiment and very interesting. And we should switch to the Mark Gurman article, Power On. Apple has a new Answers team working on a stripped-down alternative to chat GPT for world knowledge as it looks to catch up in AI. So Apple has a new answers team developing a strip-down rival to check-chip-to-check. This of course is like distill the model down, focus it, and you're going to have this, there's this idea that you're going to have the really smart genius models, the big models, training smaller models for specific things. You're going to have one that just does code, one that just does, you know, Wikipedia,

one that just does weather really well. And so if you can just kind of route between those, you can be very efficient and you can probably run it on an iPhone. So, sort of- Equally important, Mark Zuckerberg says without AI glasses, you'll be at a cognitive advantage. Hope's revenge says good. I'm used to buy my product or you one of the best posters in history. Hope's revenge is so good. I have them on the show. Pavel Asperuhov former guest and friend of the show says people will congratulate me on raising money and it's like bro I am poor and stressed

out it should be condolences. Did you talk to Pavel in New York? And Will says tell that to your ramp equity, Pavel says the Patec store doesn't take ramp shares. Well, they should. It's better than cash, better than a store of value. Exactly. Sam Altman says entering the fast fashion era of SaaS very soon gets 16K, a banger here. The thing that's interesting here is my immediate thought was look at the performance of Alvia M.H during the fast fashion era.

True. And if you extrapolate this example, it says that, you know, craft driven luxury products will, you know, will endure. Obviously, fast fashion, fashion is much more about like signaling and trends. And SaaS can be very utilitarian. Were you going to hit a gong or something? No, I was looking for the red flag. We have the red flag somewhere, but I think the fast fashion, we have a red flag in the studio.

Ben, you want to come wave that on camera? I'll wave it. Wave the red flag. But the reason I'm waving the red flag is it's more like maybe, maybe red flag is not the quite the right one. But the fast fashion is there's a different, there's a different term here, which is like disposable software. And I like the formulation of like the meme generator era of SaaS. The meme generator era of SaaS where, you know, everyone got, everyone got the ability to make memes on their phone. And that's just a lot less controversial because it's like democratizing, democratizing.

This is, it's like the tea spring of SaaS. It's like I want to be able to print my logo on a T-shirt at the click of a button. I want the Shopify for SaaS soon. I want the, you know, the chat GPT for SaaS. The fast fashion feels like, feels like very sloppy, but all the value and all the creativity being housed within Sheen and Timu. And I don't like that as much. It's way more cooler when you see someone who's wearing like a small merch drop from one

artist that they like. And you're like, oh, that expresses something that's unique. And it's only made possible. by the industrial capacity that is so ramped up to the point that you don't just need to have the Nike shirt and everyone has the same shirt. Everyone can express themselves differently. Like when the screen printing era, I remember like you used to just be able to take spray paint and just spray paint the logo on a shirt. Like that is what I'm looking forward to. We don't know how to do that. But I mean, you walk around, uh, Vass Beach or New York City, Times Square. You'll see like unique

shirts that are very weird cringe usually but I like I like that idea of like democratizing the creation of SaaS so that there can be one-off websites but that doesn't feel fast fashion to me I'm sure there will be fast fashion but I think it's more there's talk about this yeah apps as memes there's a lot of apps that there's apps that should exist that shouldn't that you shouldn't spend time producing them because they don't have like an obvious business application yep there's also apps that should exist but shouldn't have 10 engineers working on them.

They just have one person who loves building the product. My dad showed me an app yesterday that you just set outside. We were having lunch in my backyard. And he put his phone down. And it just was using the microphone to pick up which birds were. No way. Which birds were in our yard. That's super cool.

Within like five minutes, it identified like 10 different birds and was listing them out by name. Yep. That's really cool. He was showing me the profile. and he can see all the birds he's been around. It's like insane. And so there's apps like that that I just think like probably a great small business. Yep.

And way cooler if you don't need to go and raise a bunch of money and then have this like cap table prefer pref stack like hanging over your head. It's just like that might be at its best a $5 million year business. Yeah. You have a small team on it. And so if the fixed cost of developing the first version is $5 million in engineer salaries, it's like you never really clear the pref stack. Whereas if it's just an idea and you can just get it up and and then run it it like can actually be a good business So I I like that. I like that world Well anyway we have a funny we have a funny take in the in the chat from a czar

Talking about Apple should buy perplexity which we've talked about before and I think that one makes more sense than mistral or Anthropic Much more getable at the current valuation 18 billion versus what what's Anthropic at 150 or something? 160 that seemed and then they'd have to pay more obviously So that, but I don't know why he said Sony. Sony is an interesting choice. I don't know what. Sonos? Sonos would be an interesting one for sure because then they could, I mean, they already have

Apple. But they try to do their speakers thing. But Sony, I mean, they make great cameras. Perplexity is a good example where Apple doesn't need to become a foundation model lab. They just need to be able to build good products. Yep. And I think perplexity is consistently built pretty solid products. I think at some point, you know, if you think like, I cloud,

is not hosted on AWS. They did build their own data centers eventually, even though they're not a true hyperscaler. I could see them just being like, we trained our own models so that we know the legal risks, and I think that's a lot of what the hang-up is with these partnerships and stuff. There's a lot of legal risk, a lot of brand risk,

hallucination risk, all the stuff that the foundation model labs are grappling with, whether it's AI psychosis or intellectual property infringement or hallucinations or just bad recommendations, mediocre products, all the rough stuff, it's great when that lives in the founder mode startup world. Apple's not in that world right now. And so it makes,

but eventually you will be able to develop these foundation models in a way that obviates all that risk. Totally. While we're at it, Mark in the chat says this has quickly become my favorite show. Oh, thanks, Mark. Well, your last message has quickly become my favorite message in the last. I appreciate it, Mark.

We have another post here from Matthew Zitlin. Did you see this one for Frank? Sorry. Todd Graves should lead product at perplexity. That's the CEO of Raisin Cain's chicken fingers. Someone's clearly listening to Founders podcast 100%. If you were a pre-Founders podcast, Todd Graves fan, hats off to you.

Apple should acquire Raisin Cains. That'd be amazing. Put them right in the Apple store. You come in for some chicken, you pick up a new iPad. Do you need the iPad? Maybe not, but you enjoy the iPad. You enjoyed the chicken. It's great.

Matthew Zitland says, I only trust reporters that are fully locked in. And this is from a piece on Joe Wisenthall. In the New York Times. In The New York Times. The other amazing profile. Yesterday, he says, Mr. Wisenthall, who said he used nicotine probably more than is optimal

and was going through a full 15 pouch can in a day. Yeah. You know what's better than using a nicotine pouch? Developing a new nicotine pouch with your co-host before you launch your podcast. As a way to kind of test if we have, you know, on-air. Exactly. I was reading this and it's like, how are we the more extreme version of this somehow? It's like extremely odd and rare that we would be.

But we actually did build a nicotine pouch brand for finance bros called Excel. And we launched this product before. launching this podcast. And that was kind of where we got this. We realized that we were on to something because we sat down to record a promotional video that was supposed to be like three minutes long and we wound up yapping for over an hour. Ben was like did you guys prep for that? Yeah. What was that? Ben was that? So we'll end on this post. Lena says a August 1st. Lena. Yeah, we just go about her first name. Talk about her a lot. She really is the biggest Lena in tech. She says a great reminder that letting startups grow into independently

successful businesses rather than be bought up by existing giants can generate enormous value. A win for employees, investors, innovation in the public. And investment beggars. Obviously, talking about Figma and Shia Levy says, Lina Han cuts off the right hand of a genius. Lena Khan cuts off the right hand. He said Lena Hand. No, Lena Khan. Lena Khan cuts off the right hand of a genius pianist who nevertheless perseveres and produces a one

handed masterpiece for which she then takes credit yeah interesting I don't know it's a it's a it's a it's a it's a hot take in tech yeah I mean I think Alex at Andreessen said something which was in a in a in a free society we let you know we should let people do things if they're not creating immediate harm right yeah and so deciding how other people are going to run their business against is against I still think you had the best the best post about the Lena Khan news I'm wondering if I can pull it up but where was it it was you replying to Lena Khan I'll put it in the chat you found it I found it it's in the timeline chat now but if we can pull up this image of Jordy Hayes replying to Lena

con Lena Kahn says a great reminder that letting startup grow into independently successful businesses rather than being bought up by existing giants can generate enormous value. And Jordy Hayes shared a picture of a, what is this, is a soldier who's fought of war and there's a, and there's a dolphin and rainbows in the background. Just like this guy who just lived through, like, he's clear, you know, he's in paradise, right?

There's tropical rainbows and, and it's like, you know, he made it, you made it through a war and he's smiling now, but clearly, you know, been through, been through, through it. So Lena really made the Figma team. She put them in a dark place. They made it. They earned it, though.

And now it's a better story. I don't know. Definitely. Lots of nuance there. Well, thank you for watching. Thank you for listening. It's going to be a massive week in technology. There's some news dropping later in the week that we will be covering live.

We'll be up in San Francisco. You're excited. And in the meantime, I can't wait for tomorrow. Yeah. Leave us a five-star review, Apple Podcasts. If you're listening on Apple Podcasts, I recommend switching to a more video native platform like Spotify or YouTube.

And we appreciate all your support. See tomorrow. We'll see you tomorrow. Bye.