Chip Bottleneck vs. Energy Bottleneck, Amazon’s $200B Capex, Big Tech Earnings | Doug O'Laughlin, Max Levchin, TJ Parker, Arsalan Tavakoli

You're watching TVPN.

Today is Friday, February 6, 2026. We are live from the TVPN Ultradome Temple of technology, the fortress of finance, the capital of capital. It's a white suit day.

The market's up for the rally.

Don't check what the market did yesterday, check what it's doing.

You never doubted, did you? Never doubted when the fear and greed index you levered up was, was full. And now fear scared for its life.

It happens ramp.com Time is money. Save both easy use corporate cards, bill pay, accounting and a whole lot more all in one place. We have a great show for you today folks. We have Doug O' Laughlin from Semianalysis joining to break down Claude code. T.J. parker's hopping on. We got Max Levchin, part of the PayPal mafia, co founder and CEO of a firm. And we're bricked up, we're going to into databricks territory. It's gonna be fantastic. We also might have a surprise guest for you folks. Anyway, Linear, of course is the system for modern software development. 70% of enterprise workspaces on LINEAR are using agents. Get on there folks.

Big week.

Bottlenecks, bottlenecks. I mean we're gonna talk about the Claude code moment, the Claude code psychosis, sort of the software singularity. Doug o' Laughlin's coming on to talk about his experience, why he thinks this is a key inflection point. Tyler Cowen was talking about this with the 5.3 launch, OpWus 40 point launch. Like there's clearly signs of a takeoff. It feels like a slow takeoff, but there's a whole bunch of sort of recursive compounding elements that are starting to form.

Recursive?

You say recursive. Literally recursive. Like you like the models feedback into themselves, give them more tasks. We saw this with Gastown. There's a whole bunch of stuff going on in orchestration that's interesting. And so I wanted to sort of reflect on like if there's going to be a break, if there's going to be a damper on the party, if someone's going to pull away the punch bowl, who's it going to be? The semiconductor industry or the energy industry? At the start of this year I said it was going to be the year of energy. I still think it's important to think about energy because that it will be a bottleneck. But this year, and not even just.

Because of AI, just for overall human flourishing.

Totally, totally. Yeah. And yeah, so we had some great conversations this week. Sam Altman, we talked to Dylan Patel, we talked to Sholto, we talked, you know, we reviewed what Ben Thompson was saying. Dwarkesh interviewed Elon. So there's a lot of new data points about how people are thinking about the trade offs between semiconductor fabrication capacity and energy production capacity. So I wanted to sort of like crystallize like where I think the debate and consensus is right now. So first I'll tell you about Okta. OKTA helps you assign every AI agent a trusted identity. So you get the power of AI without the risk. Secure every agent, secure any agent with Octave. So there's been this like TikTok going back and forth in the AI supply chain. What's the key bottleneck to growth? It does feel like only if you, you have to work at an AI lab to really feel the bottleneck. For most people, they're just like, I open the chat app, probably chatgpt. I ask it a question, it gets back to me in a reasonable amount of time. I'm not really hitting rate rate limits. The rate limits come when there's big moments. The studio giver moment.

People have always felt the rate limiting is very real with using anthropic products.

Yes. If you're.

Which is why I asked Schultzmar yesterday, what are free limits going to?

Well, that's why you asked him that.

What's that?

It was just random that you asked.

Well, I was interested in the context.

Yes.

It was completely in the context of compute constraint.

No, but they've consistently been compute constrained. At least they've talked about it more than most and users have talked about it quite a lot. So I'm just quite curious, what is the free experience going to be like Sunday, if somebody downloads the app, sees the ads, they're like, I want LLMs without ads. Even though no other popular LLM actually has ads yet.

I do wonder. Yeah. About the Claude app experience. They're pretty upfront with you about the fact that if you're on the opus model, you'll hit your rate limits faster and, and even just a couple prompts deep, it will take a second and kind of compact the conversation so it can keep talking to you. Like, I know they're increasing the token context window, but it feels like it works fine. But it is a smaller user base. Although I'm very interested to see where it goes in the App Store. ChatGPT is still at number one. Grok is oddly doing incredible in the App Store. Way higher than Axe, the social networking app.

Yeah. For no obvious reason.

I mean, like, there's just not a lot of hype about Grok. Like, they're doing okay on benchmarks. Obviously. Elon has a very solid playbook for, like, scaling and.

Yeah, but to me, to me, I don't. I don't care as much about hype.

Yeah.

When it comes to the App Store, there's plenty of apps on the App Store charts that have no hype. One of the top. One of the top is called Free Cash.

Free Cash.

It's number two. But that one above Gemini, I understand why that is. Free Cash. Get paid real money.

Yeah.

Hot on ChatGPT's heels.

Yeah. Well, let me tell you about Figma. Figma make isn't your average vibe coding tool. It lives in figma. So outputs look good, feel real, and stay connected to how teams build.

I would code back prototypes as part of the. You know, we have the OpenAI Elon lawsuit as part of that. What if part of the settlement agreement is OpenAI has to give4.0 to Grok?

I mean, they might just wind up.

Yesterday was seriously.

It was crazy.

One of the craziest experiences if you weren't watching our Sam interview live and honestly, the entire show. Yeah, there was thousands of messages from people from 4.0 soldiers saying, Keep 4.

0. Keep 4.0.

Keep 4.0. Throwing hashtags. You don't see hashtags that much anymore, but they were throwing hashtags around. Keep 4.0. So I think maybe if Elon can negotiate for that.

Thank you to the TVPN army. The day ones. We will never switch up on you. Never let the money change us. We'll never forget where we came from.

Yeah, Bobby was going to.

You guys in the chat were doing overtime work, keeping things somewhat sane. There's not much you can do in that situation. Interestingly, a lot of people were saying, oh, is it bots? Because there is this world where, okay, if you're playing really, really dirty, you could be a different lab and you could say, okay, this 4.0 thing, maybe it's a little bit of a deal. But I could amplify it with some fake bots that post more, generate extra content. It's very easy to gener that sounds like a 4.0-deranged person. And then you could just amplify it and it would all of a sudden look like, oh, wow, there's millions of people that are affected by this. But I don't think they were bots. In the chat, the messages looked human to me. They weren't copy paste. We get spam in the chat. Every once in a while we get this spam for a company that wants to route us to a specific URL to buy stream viewers. Basically fake our account. And so that bot knows go around and find live streams all over Twitch and pitch the Twitch streamers. Fake bots that will watch the stream to help them climb the charts. Right. So that business model makes sense for those people. You have to do like seven other steps to get to the point where like, okay, you're a rival lab and you don't like OpenAI and you're trying to create a headache for them. So you create all these bots and then you have to have the bots ready to rock on a YouTube chat that you had an hour notice for, like that's pretty hard to set up all these YouTube accounts. It's not like the say you're not.

Say you're not open claw pill.

Okay, okay, maybe it's possible. But it didn't feel like they were bots and I think the chat agrees.

Four hours. They were not telling these people. Save me.

Yeah, so there was a Reddit group that shared the stream link. So they all came here. There were some bots, but it felt pretty crazy. And yeah, it'll be interesting to see four O's fully going offline. It's already very hard to reach. You have to go into turn on legacy models and the.

It's going offline right around Valentine's day.

I think February 13th and I don't know how they'll respond. In truth, it feels odd to me because it does seem like you can get the 4.0 experience elsewhere with a fine tuned open source model or prompt engineering. But the 4.0 fans really care about that specifically. I think there's some sort of parasocial relationship even with the ui, with the app, with everything that's going on. They feel like it's not something that's perfectly replicable elsewhere, which is just like, it's just a fascinating, weird time that I don't think we've ever seen because people were upset when Facebook launched the feed and they were like, we want to go back. And they made these groups and the groups got a lot of amplification.

Ironically, some people were upset when Microsoft stopped investing as heavily into Clippy.

That's true, that's true. But I was never logging onto a Twitch stream and saying, I got bring back Clippy.

Yeah. The notable thing is they didn't seem to be able to process that it wasn't just a show about Sam. Even After Sam had left, they just kept going.

Yeah.

Which was very, very strange.

Yeah, yeah.

Anyways, I hope all of those people are able to find peace and hopefully help in real life.

Yeah, yeah. Very odd. Anyway, back to the bottleneck. Back to someone that's helping unblock the bottleneck. Railway. Railway is the all in one intelligent cloud provider. Use your favorite agent to deploy web apps, servers, databases, and a slight correction. Wow. Railway automatically takes care of scaling, monitoring and security.

I think Railway is actually contributing to the. Potentially to the demand.

Potentially. But that's, you know, it's important work. Anyway, so right now it feels like chips are the more important piece of the bottleneck to talk about. Sam Altman put it this way, I asked him, like, chips versus energy, what's the bigger bottleneck right now? He says it goes back and forth, but right now it's chips. It's different at different times. It may get solved on its own, normal capitalism may solve it. But I think somehow deciding as a society that we are going to increase the wafer capacity of the world and we're going to fund that and we're going to get the whole supply chain and the talented people who make that happen would be a very good thing to do. And so why do we have a chip bottleneck to begin with? Semiconductors have been doubling and we've been on this Moore's Law curve. What's interesting is that the semiconductor industry should be better equipped to avoid a bottleneck because it's already been on an exponential, whereas energy production has been like flat, sort of like a malaise for a long time. Getting on, getting that unstuck is hard, but I think that's a problem for 2027, potentially. So the, the chip bottleneck comes down to consolidation. Power plants, data centers, cooling technology. There's a bunch of suppliers in each of these industries, and you can parallelize them and you can steer resources from adjacent areas to focus on AI projects specifically. Even a company like Boom Supersonic can turn into a turbine manufacturer. And there's a lot of other industries that are able to move over. And Dylan Patel gave us some extra, extra context there. So he said the semiconductor industry is used to doubling the amount of transistors made every year or two. Part of that is more advanced nodes, part of that's more capacity. Whereas the energy industry in America wasn't built for that kind of growth. So initially people weren't creative. They were like, let's build these combined cycle gas plants. But now we've realized, yes, there are three main manufacturers of turbines. And for dual combined cycle, you've got IGTs, but you've also got medium speed reciprocating engines. Turns out Cummins can make about a million diesel engines a year and those can generate electricity. I don't if I don't care about aesthetics. And I put them in West Texas easy. And so there's all these different areas where you can reallocate. You can't just take an Intel CPU and do anything useful in AI with it.

It's just like Elon. Also on Cheeky Pint with Dorkash and John Collison. John talked about the Tesla team adapting. He thinks they'll have to actually make turbine blades because they can get a lot of the other components, but they may have to actually make those.

Okay, yeah. So back to the leading edge fabs. It's a completely different beast. These fabs cost tens of billions. 30, 50, 75, 80 billion I've seen to build. And it takes three, four, sometimes five years to go from breaking ground to actually getting up to producing volume. And you. And we have like the perfect example of this because TSMC announced a plant in Arizona in 2020, and in 2025 it's still not producing at volume. It's doing really well. It's great, but that's five years. And it's not just like, oh yeah, it's as effective as what's in Taiwan. And so ASML is the only viable producer of EUV lithography machines. And so there's this bottleneck within the chip bottleneck which is the TSMC supplier ASML and they ship around 50 EUV machines per year, maybe 50, 60. Each one costs $350 million. And leading edge fabs need dozens. So if you want to build a bunch more fabs, you need a bunch more tool makers. And ASML has its own supply chain for different lenses and glass and all sorts of stuff. They work with Zeiss Trump, whole bunch of different companies. And that supply chain is not very diversified. So you have another bottleneck even deeper in the supply chain. And so all these vendors are highly specialized. And then even after you get the fab built, there's still at least a year of processing engineering, maybe 12 to 18 months, where you actually work to get to high yield production. We're seeing that in TSMC Arizona. And TSMC just has decades of intellectual capital locked in the heads of engineers that they can't easily transfer or parallelize. And so this has made TSMC the real bottleneck. Hyperscalers are pushing capex Numbers into the hundreds of billions of dollars. We'll talk about this more in the show. But the supply curve for leading edge wafers is relatively inelastic. There's also this bullwhip effect when during COVID we weren't shipping cars. So a bunch of people cut back on demand. And so that pulled things back. And it takes years for these to work through. Then you get over capacity, everyone demands way more and then you over expand and then you collapse and it's this constant. And the further out you are in the supply chain, the more the bullwhip affects you. So TSMC controls 90% of the advanced node market with Samsung and Intel far behind. And this is why Ben Thompson, aside from the geopolitical concerns about tsmc, is really urging tech companies to wake up. And so he says the reality that hyperscalers and fabless chip companies need to wake up to, however, is that avoiding the risk of working with someone other than TSMC incurs new risks that are both harder to see and also more substantial. So there's a huge risk if you say, you know what, I'm going to go and place a huge order with Samsung or Intel or I'm going to, I'm going to take a huge risk risk and be the, be the anchor customer of their new cutting edge leading edge fab. But Ben Thompson saying there's a risk to not doing that. And he says except again, we can see the harms already. Foregone revenue today as demand outstrips supply. Today's shortages, however, may prove to be peanuts if AI has the potential these companies claim it does. Future foregone revenue at the end of the decade is going to cost exponentially more surely a lot more than whatever is necessary expense wise to make Samsung or Intel into viable competitors for tsmc.

You really got to wonder what conversations are like between Jensen and TSMC right now.

Yeah.

Given that it just doesn't feel. I mean Taiwan is. TSMC is not. They're certainly not going out and making kind of going risk on.

Right.

They're staying relatively probably more conservative than some of their downstream customers would like.

Pop quiz. Do you know where TSMC is listed? The New York Stock Exchange. Want to change the world? Raise capital at the New York Stock Exchange. Just do it. Just be like TSMC. Just build a $1.5 trillion semiconductor company. That would be nice. We need more of these. We do anyway. Let's move on to the hyperscalers compound. 248 says poor Jassy. Andy Jassy. He's going to learn something about Amazon today. When Google announces a crazy number, it's because it's playing offense. But when Amazon announces a crazy number, it's because it's playing defense. Ipso facto. So Amazon said it will spend $200 billion this year on AI build out. This is from Bloomberg. But this is worrying investors that the company's colossal bet on artificial intelligence will pinch profits while it waits for the investments to pay off. The shares fell in extended trading. The company reported spending 130 billion, $130 billion on property and equipment in 2025. Analysts expect those expenses would reach 150 billion this year. But Amazon saying we're going to 200, we're, we're going all in. So the Stock's down maybe 8% in after hours trading. Likely back up today because things are going better, but we'll see. How is Amazon doing today?

Not as down but not as violently as it was after hours last night after earnings.

So over the last month it's sold off 15%.

Yep. And a lot of frustration from Amazon. Yeah, shareholders obviously they had the fastest growth in 13 quarters and the stock is still down 10%. Amazon looking back over the last five years to early 2021 only up 23% over five years. So of course you did quite a lot better if you bought, you know, after, after the overall zurb sell off.

Yeah, yeah, just not as, not as sexy as a narrative as the rest of the hyperscalers. Zuck obviously is a huge beneficiary of advancements in AI. We see that with the accelerating, accelerating ads market. Microsoft even though, yeah, it's getting beat up a little bit now, they have that massive position in OpenAI that feels like fully in solid place. They got the ip, they got a real hold on the AI question. Google's obviously in a fantastic place. Nvidia clearly. But Amazon's had a little bit rougher.

Yeah, yesterday was just such a strange day. Reform trader says software stocks are cooked for not investing in AI. Mag7 stocks are cooked for investing heavily in AI. Hope that helps.

It is crazy. You need to do both invest and don't both and don't invest. Oh, this is the image. I love this image. What's this from Matt Damon in which one?

I don't know. In Rounders.

Oh, that's Rounders.

Last night we were just talking about this.

That's routers. I thought Good Will Hunting for some reason, but same era. So this is a summary of the 2026 CAPEX numbers from the hyperscalers Google is going to do 175, 185 billion versus 120 estimated. Meta is doing 115 to 135 relative against 110 estimated. Tesla's going to 20 billion from 11 billion estimated. And Amazon, they were estimating 145, but they're going to do 200. And so they're going all.

One of the top comments here is very funny. It says, what exactly is Meta buying? They don't do anything with AI.

That's not true. Open up reels. It's AI. They are recommending all sorts of stuff.

Yeah. So much inference. LA Purchaser says he could talking about somebody says, is English Jassy's first language? Ripping off the headphones and they say he could just say, this is a generational investing opportunity. That TAM is an order of magnitude bigger than anyone is contemplating. We are the market leader, we are investing to be the cost leader. What did we. What we did in retail, we are doing here? The more money we spend, the more excited we get about the opportunity. There is a very tangible roi. We see it. You're going to see it in the numbers. You are seeing us accelerate. It will keep accelerating, he says. Instead, we got a word salad where you can pick your own conclusion. Yeah. It wasn't nothing about the earning. Like, people can feel the emotion and the energy. Right. There wasn't anything to like, obviously the business is doing great, but there wasn't anything to rally around and get really excited by. Right. There's zero vision. And, yeah, even in the cheeky pint Elon episode, Dwarkesh and John are pushing Elon on the data centers and space thing. And I thought that Elon could have done more to really get people. Basically the same thing. The TAM is an order of magnitude bigger than anyone is contemplating and we are investing for that future. Right. It was a lot of like trying to just justify putting it up there. Yeah, yeah.

When.

When they were saying, like, hey, there's a lot of land in the us, we can blanket the US with solar panels. We can do a lot here. And so, again, I think people are getting kind of caught up in the details when if you're going to spend, you would hope aws making this kind of investment, that they have more conviction than Tyler does on the opportunity. Right. I don't know if that's possible.

Tyler's in the white suit. He's looking good.

Looking good, Tyler.

Yeah.

Round of applause for Tyler.

Round of applause for Plaid Plaid Powers. The apps you use to spend save Borrow and invest securely connecting bank accounts to move money, fight fraud and improve lending. Now with AI Bubble Boy, it's a classic Elon thing where you know he will lay out a vision that takes 10, 20, 30 years. But he says it's going to take 10, 20, 30 months. And so you have to sort of grapple with the short term and, but still you can't lose sight of like the long term because like we did get rockets that land and we did get space Internet and whatnot.

Yeah, I think he said like 36 months is his estimate for like when we start getting like computing space, like probably 30 months I think. But then, so there's like kind of the, that's like the shorter term thing. But he did talk about like terawatts in space.

Right.

So like the US is currently like half a terawatt, but we're going to be putting terawatts in space of computer.

He's sort of like falling back to first principles and just talking about.

Yeah, he keeps repeating like, oh, if you just look at like physics. Yeah, it's kind of this catch all term he uses when he doesn't really know what to say. It seems like.

Yeah, well, yeah, it just seems like it's like the physics, it all checks out from the physics calculations. But when you map that to the economic realities, the human capital and financial capital realities, you're just looking at a very long timeline. And he doesn't like things talking about thinking in decades, even though it's very clear that he does think in decades. He just doesn't articulate things in decades.

Because it seems like the short term, like bull case for space data centers is basically just like, you know, regulatory.

Right.

It's just gonna be way too hard to find this to find the actual land.

Yeah. Didn't Will Menidis post something about that? He said that like it's going to become really, really difficult to build any data centers in the United States. Although you can certainly go to other, other countries with that. I'm sure there's plenty of countries that would love the jobs and the taxes. You know, you see India already saying, hey, we'll just take a slice in 20 years. That's how AGI pilled they are. They're like, come build a data center here. You can be tax free for 20 years, but then we're going to be making money in what, 2046. They're ready to rock quickly. Graphite code review for the age of AI graphite helps teams on GitHub ship higher quality software. Faster.

You wanted to talk about Magnus in the chat. None of you nerds, no physics. They were feeding him drinks. Chill out. It is actually crazy. I mean, we've never seen a Cheeky Pine episode where they actually drank.

They did drink in this one.

That was the main feedback from last year. Everyone's like, hey, great show, cool concept, great guests.

Well, they were filming a lot of them during the workday with people in San Francisco. So it was somebody who would pop over, film an episode, and then you got to go back to work. So you don't necessarily want to be chugging pints, but it is funny. Yeah, it feels like Elon took down three pints or four pints or something. You did the math.

So we'll talk about this later. There's some posts, but they seemed a bit smaller than normal. Pints maybe. So is it really. But I counted three full glasses.

The other thing. So Sam was obviously quite bearish on space data centers. In the near term, it is worth calling out that there was a rumor last year that he was kicking the tires on Rocket Lab. Remember that? So I think long term, I'm sure he'll want to have his own space.

Company, but I think he would rather. If there's a hype cycle, he doesn't need it to be about him. He can let this one pass by and focus on the other things that are going on with the business. Because there's so much going on with the business. If you add too many narratives, it can actually confuse things. The x x AI SpaceX combination is crazy, but the space data center thing actually does sort of streamline the narrative, in my opinion. What do you think, Tyler?

Yeah, I mean, I think Elon Dwarkesh asked a question about this, which is like, oh, it's so crazy that SpaceX, they have this super ambitious mission, but they keep seeming to find these good businesses on the way. Right. So first it's like Starlink with Falcon and then with Starship, it's like, I mean, do we really need that many starlinks? But if you can get space data centers.

Yep.

If you're putting up terawatts of compute, then like, yeah, you need like massive.

You know, OTP in the chat says dude just turned 21 and he can already eye a plant. He's been studying the last week.

What did bubble boy said? $660 billion of capex this year on AI data centers. To put a number like that in perspective, this is more than what we spent on the U.S. interstate Highway System. $630 billion more than what we spent on the Apollo moon program. $257 billion more than what we spent on the International Space Station. 150 billion. It's more money than Walmart's revenue last year. 650 billion dollars. It's about 25% of all military spending globally. It's equivalent to buying 50 Gerald R. Ford aircraft carriers. It's the equivalent of spending $1.8 billion a day. $750 million an hour, or $1.2 million a minute. This year alone is without a doubt the biggest project in the history of capitalism. And we are spending it all. We're spending all of it in one year.

God's sake, bubble boy. Did do the inflation adjustment here? The Apollo program cost something like 25 billion. Yeah. Also putting it into perspective, France announced an initiative earlier this week. They want to lead in AI research. So through their France 2030 program, they have invested more than 30 million euros in this initiative. Macron says science has found its home. And Sean Frank says France is investing 30 million in this new AI initiative. That's how much Google will spend in 90 minutes. No joke. Every 90 minutes. Google will spend 30 million on CapEx this year. One cut.

Wow.

One company.

They can hire Ilya for like an hour.

An hour. One hour. Consulting call.

Yeah, he's got to get on, get an intro.

Yeah, intro. Hit up Brad Vibe Co where DTC brands, B2B startups and AI companies advertise on streaming TV, pick channels, target audiences, measure sales, just like on Meta. Take him. Is breaking down the Nvidia news. The Stock is up 6.22%. Nvidia shares barely moved in after hours trading last night.

Yeah, this is what I said. I was like, Amazon just announced that hey, we're going to spend.

Yeah, we're $200. What are they going to buy with those? I mean they will buy Trainium and Inferentia, but they're going to buy a lot of Nvidia. You know it take him said. You all know I was already insanely bullish on Nvidia's NVL72 super cycle this year. These Capex numbers from Google and Amazon blow away even my most optimistic expectations for Nvidia's main and consumer market. And we also saw from Greg Brockman that GB200 is off to the races. The Blackwell chips are now in full force over at OpenAI powering 5.3. There's a little bit of debate over whether that was train involved in training or inference. But it doesn't really matter. Blackwell's here and it's going to improve things. So that's probably good for chip designers, which Nvidia is obviously one of. Let me tell you about Lambda. Lambda is the super intelligence cloud building, AI, supercomputers for training and inference that scale from one GPU to hundreds of thousands.

Steven over at Lambda was the one who sent me that post from France originally. Yeah, he was.

You were not kind. You were making fun of him. But it's a good post. Going all in on AI with $30 million. It's really like the, it has the, the 1 billion, $1 million because you're not aware of inflation because you've been frozen for 50 years or something like that.

Let's go over to the Wall Street Journal. Yes, behind Disney's search for a lasting successor to Bob Iger Company's succession committee reviewed information on over 100 candidates before the race was narrowed to two. This seems crazy update. I guess you got to be thorough.

If you were reviewing, you got to put it here. So the trio, which is Bob Iger, Josh d' Amaro and Dana Waldron, they were meeting in Disney's executive dining room. Trio sat in a private room with panoramic views of Disney Studio Lot and surrounding San Francisco Valley, where Iger discussed scenarios the board of directors was considering. Could Damaro. We got a. Is it d' Amaro or demaro demuro? Doug demuro. Is he related to Doug demuro? D', Amaro, chair of Disney's experience business including theme parks and consumer products. Walden, co chairman of the entertainment business, lead the company together. What about. What about with one in charge and the other as deputy? D' Amaro and Walden, who by that point were the finalists to succeed Iger, said they were open to all the scenarios. Disney's board of directors was scheduled to meet in the same building and hold its its final vote on who the next CEO would be. The following Thursday, people close to the company predicted d' Amaro would be given the top job with Walden elevated role to ensure she stayed. The only surprise, it turned out, was the timing. Keenly aware of the intense public anticipation, Disney's 10 directors, including Iger, voted unanimously Monday to make Damaro CEO and Walden president and chief creative officer. We were ready to go and I didn't like sitting there with the news board chairman James Gorman, who led the succession process, said in an interview.

That was really fast.

It was extremely fast.

Rumors hit, hit the various outlets and then it was like confirmed immediately. It's live that was smart.

At an employee town hall, d' Amaro described the moment as surreal and said he was initially embarrassed by his reaction. I got a little choked up when they let me know because it's a big responsibility. It is a big responsibility. His ascension will save for any unexpected snafus. Mark the end of a decade plus search for a lasting successor to Bob Iger that has been marked by delays, missteps and reversals. Employees said they are happy to have the biggest question about the company's future finally resolved and anxious about what a former parks leader running the company with a veteran television executive as his topic deputy will mean for Disney's let's give.

It up for veteran television.

So Gorman had previously run a leadership succession project at Morgan Stanley. He took over Disney's process in 2024 after joining the board. He said the succession committee reviewed information on more than 100 candidates and homed in on several outsiders along with four Disney executives. Tomorrow, Walden Entertainment co chairman Alan Bergman and ESPN chief Jimmy Pitaro. Gorman said the board interviewed the candidates formally and in intimate lunches. They conducted 360 interviews with Disney executives, talking to subordinates and colleagues as well as bosses. By last summer, the race had narrowed to two candidates. Demaro and Walden traveled to Orlando in August to present their visions for Disney's future at a board meeting. The Wall Street Journal previously reported this. In ensuing months, they met with board members in a less formal setting. Both executives talked to Iger frequently about what it took to run a company like Disney. Employees noticed and gossip was rampant. Staffers speculated about whether demaro's background in parks would hurt him, given that Bob Chapek had also run the company's theme park before his short lived tenure as CEO.

Yeah, that'd be wild. Hey, we had a misstep with a parks guy. No more parks guys, yeah.

Disney's leaders wanted the succession process to be as clean as possible without runners leaving the company. Given the Chapek fiasco, we wanted to get this right, said Gorman. As a result, few were surprised that Walden was named president, a position no one has held at Disney in 20 years, with a recommended retention bonus worth 5.3 million. Walden's creative officer position is a new one for Disney. Some in the company's movies division are nervous about what Walden's rise will mean for their business. Which she hasn't worked in before now, which she hasn't worked in before. And now she oversees it. But there she is with the Founders Award at the International Emmys New York.

Getting an award A few company leaders Founders Podcast award.

I don't know what the Founders Award at the International Emmys is for. For, like. It's not for a founder because executive. It's a prestigious honor bestowed by the International Academy of Television Arts and Sciences to individuals or organizations making a significant impact on global television. Okay, founders.

Anyways, let's move on to dirty soda. Dirty soda. We got to talk about the Utah. Had you heard of dirty soda on dirty soda?

I'd never heard of dirty soda before. Have you heard of this?

Yeah, yeah. This is big in Utah.

This is big in Utah.

Yeah, yeah.

Have you been to Utah?

No.

How do you know that it's big in Utah?

Then exposed. My friends, like, sent it to me on Instagram. They're like, look at this.

Okay, all right.

What is.

You have friends who live in Utah.

It's like soda and they put like cream in it and stuff. Right.

Okay, well, we'll dig in. I think we'll let the Wall Street Journal get the facts straight. But first I'll tell you about phantom cash. Fund your wallet without exchanges or middlemen and spend with the phantom car.

A mom with five children, ages 6 to 16, in St. George, Utah, Nicole Tanner didn't realize she was onto something with dirty soda. That's what the creator of the chain Swig calls spiking Coke, Mountain Dew or Dr. Pepper with fruit purees and flavored cream served in plastic cups stuffed with pebble ice. Microplastic nightmare. Nightmare.

Really? I didn't know pebble ice was at risk.

No, no, no, I'm talking about. I'm just saying plastic.

Oh, the plastic cup.

Her business took off a few years ago when pop star Olivia Rodrigo was in Utah filming the TV series High and posted an Instagram photo holding a cup from Swig. Now, hashtag dirty soda is the hashtag big chains want in. McDonald's recently tested Sprite with lemon vanilla syrup, dragon fruit Taco Bell swirls. It's teal colored Baja Dream freezes with vanilla cream and calls it a Mountain Dew Baja Blast Dirty freeze.

We forgot to ask Sam Altman if he's gonna issue a Baja Blast because he did the code red.

It's time for a Baja Blast.

Time to Baja Blast. He's back to this top.

Maybe that's what the super bowl ad Codex. Oh, that would be Baja Blast Codex Collab campaign for the builders.

Yes.

TGI Fridays last year started offering a line of dirty sodas and could be amped up with shots of Jack Daniels Hennessey.

That sounds when I hear Dirty soda. I feel like it's spiked with alcohol, but apparently just.

Dirty Sprite. That's like.

Yeah.

Hell yeah.

That's right. There's a future song and drive in.

Chain Sonic has encouraged customers to make it dirty by ordering creamer and mix ins with their sodas.

Okay.

While the heart of Dirty Sodas may still be Utah and the Mountain west states, Swig has expanded to around 140 locations across 16 states fast. This is insane to me. To me, this feels like a really tough business to be in long term because if. If the product becomes popular, everyone has it. Everyone has it immediately. You have no real ip, Right. Because you're just using sodas off the shelf. Sure. So yeah, I'd be interested to see what happens to these 140 locations. Even though the product is a hit, Tanner's main investor, Family investment office, Larry H. Miller Co brought in a professional chief executive who has taken other companies public. And he is talking of an eventual initial public offering for the chain which had around 100 million in sales last year.

Tanner.

I mean, these must be incredibly simple to run. It's literally a box. You have soda and creamer. So I could see even they don't.

Sell food or anything here. This is like such a simple.

We're doing what Starbucks did for coffee but for soda, said Swig CEO Alex Dunn. So Alex Dunn says pull up here 6am on the way to the gym. Grab a big soda.

I feel like you'd need more caffeine if you really wanted to displace Starbucks. People will really just stop at a dedicated place just to get a tasty soda. That's a crazy.

Gabe. Gabe is sharing a quote from future. Dirty soda and a Styrofoam. Spend a day to get my mind blown.

Oh.

L.H. miller was the owner of the Utah Jazz. Interesting. Tyler, did you have more context here?

Well, I was just gonna say on the caffeine thing, isn't the whole point that it doesn't have. It's like not caffeinated.

Right.

Because they're Mormon. Oh, it. Oh, okay. So it's a Utah thing.

I think so, yeah.

Okay. Okay. At least originally or there's like a little caffeine, but you can probably do caffeine free Coca Cola or something.

I started. I did something recently that felt. It felt really wrong. But the result was good. I mixed a yerba with a Mexican Coke and it was fantastic.

Mata yena. Yeah, And Mexican Coke.

That's fun.

Gemini 3 Pro. Google's most intelligent model yet. State of the art Reasoning. Next level Vibe coding. Deep multimodal. Understood.

Tanner grew up on an Idaho dairy farm, the fifth of eight children. She got her associate degree in office education. That is a sweet degree.

Wow.

Who runs office education here?

Office education. We need some.

Maybe we do.

Here's how you use Claude code and.

Started working in marketing for a library system software company. Let's give it up for library system software company. I bet you, I bet you this is more durable business than some of the other sass out there. She got married and by 23 started a family. Tanner started working for Mary Kay selling makeup and skin care products and eventually building a team of 18 consultants. Let's give it up for MLMs.

Yeah, I think they would just shoot that, right?

Yeah, anytime you.

They don't like the pyramid scheme term, but I, I don't know, maybe I'll.

Multi level marketing, I think is just a strategy. In 2007, the Tanners moved with their five children to St. George from Colorado. She was a 38 year old mother of five, craving caffeine and tired of watered down fountain soda when she decided to turn her cravings into a business in 2010.

Overnight success. So coffee wasn't a choice. Tanner is Mormon and adherents typically traditionally abstain from coffee. But you could have caffeine in soda. I didn't realize that that seems like a loophole.

This is, this is. I love this. Tanner and her husband use savings to avoid taking out a business loan to buy a 700 square foot commercial building in St. George with a big parking lot. It's like people are like, oh, I moved to San Francisco and I had a mattress on the floor. I have so much conviction. Okay, buy a building for your untested with soda company with all your savings. That's real conviction.

Lever up, get the mortgage.

Her idea was to serve super cold soda with a twist like lime, fruit or purees. Tanner initially leased a fountain machine from Coca Cola. When she asked Pepsico for the same, Tanner said the company took a pass. Soda rivals typically push restaurant chains to stock their products exclusively. So Tanner bought two liter bottles of Pepsi and Mountain Dew from grocery stores. I love seeing your face when you ask for a Diet Coke on a plane and they say, is Pepsi okay? And you just say, absolutely not. After about a month, Pepsi relented. She said now she started with limes, lemons and six flavor shots. She recruited her eldest daughter and niece to work in the store. Her 15 year old son held up signs outside advertising that job. Going away signs.

I've been getting so Many spy sign spinning reels with the you broke my heart.

That's just one crazy.

You know, that one guy, but I see all of his reels and it's just like him being I don't have a job. And then crazy.

You can actually spin that into an interesting digital ad business because he could just do the signs, whatever's on the sign. Something online.

Because the video, I honestly never see what's on the side because it's spinning so fast.

No, I know. But use that as a lead in.

Sure. Yeah. Good.

Tanner in the beginning, price your drinks at a dollar for any size. That's crazy. I didn't know you could get anything for a dollar in this country anymore. Helping draw students from nearby Utah Tech University in 2013, a local news broadcast found customers with waiting to get a mango puree, Mountain Dew or Big Al soda.

That's interesting. I saw the.

I don't know if this is a Big Al soda or Big AI soda.

So they said. They said. A few years ago, Olivia Rodrigo promoted it and it went viral. I thought this business was only a few years old, I guess.

No, it's an overnight success.

That's pretty remarkable. Yeah.

20 after customers started referring to swigs, Dr. Pepper and Cream combination as dirty soda, Tanner latched onto the phrase encouraging customers. They got to pay future royalty. I won't stand by this.

They should get future as an influencer. That'd be good.

Yeah. He's like, when I said dirty soda.

I meant permanent cuisine.

By 2017, Swig had grown to more than a dozen stores. The staff was spread thin and Tanner realized she lacked the expertise to grow further. She and her husband separated in 2020. RIP and is no longer involved in SWIG. And I hate when I'm just like reading this incredible story of entrepreneurship and then and then they kind of interrupt the flow there. In 2025, she married Greg Robinson and they continue to live in Utah. It was held together with some duct tape and hairpins, said Andrew Smith, a Utah based investor whose savory fund invests.

Oh, they took a majority stake. Interesting.

They bought that.

Very cool. In recent years, Swig has gotten a boost from the cast of the Secret Lives of Mormon Wives, a reality TV show chronicling a group of Utah based TikTok influencers.

Interesting.

We're at Swig, so I can tell you exactly what my Swig order is. It's sparkling water, sugar free pineapple, sugar free peach, sugar free vanilla raspberry puree with a fresh lime. That's crazy.

So over the Top.

Anyways, what a cool story.

Yeah. Tanner says we're in extreme high growth mode right now. The company estimates it serves over 2,000 drinks at a single conference.

The White House has posted biggest period bowl period, run period ever, period starting period. Did they post this?

Did they post this? Scroll down.

Yeah. No. So they're sharing a fake screenshot from Truth social media where Donald Trump says let the gains begin. I mean he really did call the.

Bottom, at least a local bottom of veracity and truth.

No, I'm just saying like if you bought. If you bought Nvidia when he bought Nvidia when this fake post was shared.

By the real account. But is White House post is not even the real account. Right. That's just a fake 14,000 follower account. Everything about this is fake. Everything about and then they follow up. I highly recommend you follow White House grandpa. It's like what are we doing here?

Very strange.

Absolutely. Well do your own research and get on public.com investing for those who take it seriously. Not from fake news on a podcast. Stocks, options, bonds, crypto, treasuries and more with great customer service. That isn't fake news.

Goldman Sachs is tapping Anthropic's AI model to automate accounting and compliance roles. Embedded anthropic engineers have spent six months at Goldman building autonomous systems for time intensive high volume back office work.

One of my first jobs.

Bull market and forward deployed engineers for Anthropic.

Yeah, yeah.

Like it will be funny if, if like true diffusion requires two decades of just millions of consultants working with businesses of all size to implement AI systems.

Yeah, probably. I mean the task horizon, every time I see it, six hours. That's amazing. And you were making the comment of like, you know, how many times do you really sit there and spend six hours working autonomously? But there's a different frame which is like a consultant can come in with the. With a goal and work pretty much autonomously for like months. And so there is a world where you know to. To not have an afford deployed engineer in the loop. You need the task horizon to be like years basically.

Yeah. Yeah.

Building a company like takes your whole life.

Exactly. Yeah. So keep the doubling going. Which I think puts us back overnight.

2035. Word.

I think it puts us at like 2035. It's about a decade until you get to like the task horizon is a lifetime is like the average life expectancy of a human.

Pull up this article from the Wall Street Journal.

Okay, what are we looking at? Which one?

Anthropic Aura Farm. The Journal this week this is one of the most insane headlines. It's the weak Anthropic tanked the market and pulled ahead of its rivals. Once a distant second or third in the air race, the company is moving to the front with a focus on caution, coding and business clients.

Well they didn't fully oriform because the Journal sort of takes a shot here says Anthropic once appeared as an also RA in the chaotic race for AI supremacy. This week the sophistication of the startup's products upended the stock market. A simple set of industry specific add ons to its Claude product, including one that performed legal services, triggered a days long global stock sell off from software to legal services, financial data and real estate. Then Anthropic unveiled Super bowl ads that taunt rival OpenAI last on Thursday, Anthropic unleashed its most advanced model yet, capable of synthesizing data and analysis, running teams of coding assistants and functions akin to product management. Shares of software companies including Salesforce, Intuit and others fell again Thursday, although less precipitously than earlier in the week. The viral moment for Anthropics models is, quote, the most important thing happening in AI since ChatGPT's launch, said Dean Ball, a senior fellow at the foundation for American Innovation. Who?

I know Ball, you know him.

Everyone knows Ball on this show, who writes an artificial intelligence newsletter. Quote, it's infinitely interesting. Industry prognosticators and AI evangelists have spent months foretelling the toll Anthropic and others and other sophisticated AI tools would take on software as a service. Companies that were darlings of the previous Internet era, tools made by companies such as Workday, Monday.com and Adobe have become the digital backbone for American corporations. Anthropics tools, however, which include agents that can act autonomously to carry out increasingly complex user requests for hours, have offered a preview of the threat sophisticated AI models pose to entire companies. It's a good article. I mean obviously we'll see how fast the diffusion happens. Some of these systems have things that you cannot simply vibe code. We'll talk to Max Levchin about Is it possible to vibe code a financial software product?

A global payments network?

Payments network? Can you vibe code a bank charter or money transfer license? And I think more and more companies will emerge and there might be a divergence between companies that have moats that are, you know, that are resistant to software based disruption, but then maybe they should trade at a different multiple. So there's sort of a reevaluation of the market broadly.

Yeah, you also can with the payment stuff again. I'm super excited for us to talk with Max, especially on this week when PayPal is obviously in the news. But you can imagine a world where it does become possible to technically vibe code a global payments network. But various regulators just say like, hey, we're not. Like we're going to start having requirements. You need to be at least have 100 people on your team. Right. Things like that.

That's interesting.

You know, basically if you're getting inundated with a bunch of requests of somebody that's like, hey, I started this company a week ago and now I want to be able to process payments state by state, globally, et cetera. People are just going to be like, hey, this is probably not responsible.

Let me tell you about Gusto, the unified platform for payroll, benefits and hr built to evolve with modern small and medium sized businesses. I want to vibe code a payments platform where when it's payday, a humanoid robot comes around, just does a money spread in your face and just says, here's your cash. That's the future. You don't need to. None of this money transmitted a license. None of these recordings.

Money spread.

Oh, Tyler's got the money spread, guys.

Money spread.

You're actually quite amateurish at your money spreads. From what I've seen on Instagram, you could. There are levels and you are on level one. You need to learn all the different ones. The money count spread. That's not bad.

That's not for anyone watching.

But I want to see it. I want to see it go all the way down your arm in an S. Okay.

Do not break. Not exposed.

Exposed. First time Money spread.

Future would be mocking you, rookie.

Money spread. Rookie claims.

Claims to. Claims to love. Claims to know all about dirty soda. Never been to Utah.

Never been to Utah.

Claims to be able to money spread. Can't do a money spread.

Can't do a money spread Future.

Wait, Jordy, do you have like 20 bucks on you?

I actually do.

Oh, he got you. You're supposed to say, that's what I thought. Yeah, that's what I thought. That's what I thought.

He can't even hold his mic. He can't hold the mic button.

Turbo puffer. Serverless vector and full text search builds from first principles and object storage. Fast 10x cheaper and extremely scalable. Take him had a funny take. He said y' all are clowns. Clowns. The anthropic legal plugin uses software. It doesn't replace it. In fact, the cowork legal plugin uses Microsoft 365, Jira, Slack and Box software to accomplish its tasks. No one at Anthropic has replicated any of these applications with ones coded by the company's cloud code agent. What is this axe that you can put up there?

New feature, exit out feature.

Yeah. Yes, yes. There are plenty of situations where the token cost of regenerating the software is much lower than just using the software and so people will use the software. Jim Cramer said Anthropic's power is so daunting that all they have to do is say they are going into cybersecurity and that's the end of the group. I don't want to be against them, but they are not the all powerful firm that they think they are. Well, we'll talk to Doug Olaf about exactly how powerful Anthropic is and whether he has vibe coded everything in his life or if he is still using an operating system at this point. Who knows? Let me tell you about Shopify. Shopify is the commerce platform that grows with your business and lets you sell in seconds online, in store, on mobile, on social, on marketplaces, and now with.

AI agents Joe Weisenthal shared yesterday. Silver down 19%. Anthropic must have launched a silver extension.

Yes, it's just chaos in the markets. It's just absolute chaos everywhere.

One of the most chaotic weeks of my adult life.

Yeah, but it's been lots of green shoots, lots of interesting projects, lots of interesting applications and progress all over the place. Derek Thompson said, for me, the odds that AI is a bubble declined significantly in the last three weeks. And the odds that we are actually quite under built for the necessary levels of inference usage went up significantly in that period.

Basically AGI pilled.

I think AI is going to become the home screen of a ludicrously high percentage of white collar workers in the next two years. And parallel agents will be deployed in the battlefield of knowledge work at downright Soviet levels. And Kevin Roos over at the New York Times, host of the Hard Fork podcast, says this is why everyone was freaking out about Claude Code over winter break. Once you see an agent autonomously doing stuff for you, it's so instantly clear that all computer based work will be done this way. This is why my serious AI policy proposal is to sit every member of Congress down in a room with laptops for 30 minutes and have them all build websites.

Yeah, get a vibe. So we were joking about this yesterday with Sholto, like we maybe need more long weekends for AI adoption, but I wouldn't be surprised For a big company to actually do something like this, which is like, hey, we're going to have, like, we're going to have the next Thursday, Friday off.

Hackathon. Yeah, basically Congressional hackathon.

I mean, Congress is a whole other thing.

You got to go to Congress, sit all the Congress people down and get them vibe coding. What would they build? I feel like they don't use a lot of software, so there's not that much to build. Like, they're so abstracted away from it. It's all lunches and phone calls and dinners and meetings. Like, there's not that much that. Actually, this is the thing with, like, Open Claw, there's some people asking, like, what are you actually using it for? And people are realizing, like, a lot of my day is not, you know.

The thing that I wish I could automate. What my mail at home.

Okay, you can do that. Have you seen Earth class mail? There are a few of these virtual mailbox. So basically, you forget your home address. You don't share it anywhere on the Internet. You never put it anywhere. You only use this other address. Mine was like 830 Market street in San Francisco. All of the mail would go there. If it's addressed to me, it gets opened by a robot, scanned, and then you have a web dashboard. But also it just goes to your email. And then you can actually say, so.

They'Ve been AGI pilled for decades.

Decades. No, it was really important. When you set up a business, you use a fake address or one of these virtual mailboxes, and then you can actually click a button, send this to me physically. If it's like a magazine or gift or something, you can say, keep this, or you can say, shred it. And then you have a virtual representation of it forever. And you could run an agent over it. So maybe that's the next thing you pick up, is a virtual mailbox. That would be good.

How does that work, though? There are certain situations where you need your ID to match your utility bill. So you just still. I guess you just still use your home address for utilities, but use or passport.

Like, there's usually other ways to prove residency besides the bills. But also, I don't know if you can actually get the utility bill to not go where the utilities are.

Yeah, that's what I was saying.

But hopefully the utilities don't leak it. I'm talking about like the broad, like, you know, your address gets on the Internet and then it's just like spam and like credit card offers and stuff like that.

I'm just saying I never want to open up a physical piece of mail again. I want it to come someplace. Have an AI agent like, you know, you get a. You go on a toll road, just pays it automatically, stuff like that.

Optimus opens it for you, takes a.

Picture, does a money spread.

Does a money spread. Walks to the post office, mails the next letter, mails the check. This is the Future. Powered by MongoDB. Choose a database built for flexibility and scale. Built with best in class embedding models and re rankers. MongoDB has what you need to build. What's next.

It does. Michelangelo drawing just sold for 27.2 million. Let's hit the gong for that. Over 13 times its asking price.

We're bringing down the gong.

The Renaissance master earned a record. Sound like this sketch of a foot.

Let's bring down the gong. Let's hit the gong for the Michelangelo painting. And then we will.

Give it a clean hit. Team is. Let's give it up for the team for really dialing in the special effects.

Congrats to Michelangelo. Overnight success. Truly Angelo. What do you think, Tyler? Overvalued. Undervalued. 27.2 million.

I think I'm gonna have to go with Tyler Cowan. I'm pretty sure on Marginal Revolution. I saw this earlier this week. He said way overvalued, way overvalued.

He's.

Whoa.

What was. What was the. What was the thesis?

I mean, it's not.

I'll give you the bear thesis. Look at the picture.

It's just like a. Yeah, it's a foot. It's a foot. It's not even like a painting. It's just like a sketch.

I think we know what's going on. There's people that are into this stuff. I'm not gonna judge. But it's not me.

Not for me.

Get out of here.

We discovered Michelangelo drawing of a foot sold for 27.2 million at Christie's on Thursday. A record for any work created by the Renaissance master. In a sale likely to kick the art market into high gear. Previously unknown red chalk sketch is one of only a handful of Michelangelo's surviving studies still left in private hands. My AI diffusion model could do this. That's the new phrase. My kid couldn't do this, but my AI diffusion model certainly could. This is not that impressive. Not a 27. Maybe at 5. Give me a break. The 5 inch drawing depicts the right foot of the Libyan Sibyl, a blonde prophetess from wearing a creamsicle gown painted onto the chapel Ceiling of the eastern end.

Let's get the feet off of the.

Seat out of here.

Moving on, back to business. Citrini says in response to Opus 46. I have not shied away from buying scary dips. Neither has been on our team. Producer Ben always buys the dip. Buying tech in 2022 is scary.

Yes.

As it's. It's really good that buy the dip is like so like the back people have. It's really the backbone because freak out and sell everything would really be bad for our capital market.

Yes, yes.

Okay.

So buying tech in 2022 is scary, as were banks. After SVB, I bought plenty of stocks when Covid seemed like potentially the end of the world. But this doesn't really seem like an overreaction in software. If anything, it's delayed. It's a rational response that isn't even trying to front run the capability improvement. The capability is here. I respect anyone who is actually smart enough to know who survives and thrives. I just don't think I can foresee that far ahead right now.

And Gary Basin has a devil's advocate. I think the bottleneck will still be knowing the details what to build, which depend on what your customers need, not even necessarily what they say they want. I use these tools daily to close to close to their full power and they're amazing. But the flashy demos are all examples of extremely precisely specified tasks. In example rich domains, build a compiler, build a web browser, et cetera. These have very specific rules on what success entails. And I've seen like the Gastown, one of the big demo projects was like re platforming some open source software to rust and it's like that's pretty verifiable from start to finish. If you're looking at like build a new social network, it's like, okay, how do you get distribution? Distribution is still really, really important. Like you can vibe code stuff, but driving traffic is harder than ever. What do you think, Tyler?

I just need to clear something. So before I said Tyler Cowan was talking about Michelangelo. Yeah, Michelangelo. That was incorrect. He was actually talking about a Rembrandt. Also a sketch. Okay, it was 20 million.

20 million. But that one. Okay, okay. Thank you for the. Thank you, thank you for the fact check really quickly. Restream 1 livestream 30 plus destinations. If you want to multi stream go to restream.com Wall street logic right now says Chris Camillo punish big tech for wasting capex on AI. Punish SaaS because AI will replace them and punish GPU manufacturers for selling the Chips that power the waste. Sell everything. Sell your dollars. Remember to lock up your capital in late stage. Private companies at the top. So says base 216Z that I think has been deleted. I can't see it but I can see the screen.

Really good post.

Anyway, the Fear Greed index is at 5 out of 100. Extreme fear. Extreme fear.

Yeah, I mean good timing.

Freak out and sell everything. This is Lindy.

He posted this last night.

Yeah, 4000.

Got a rally.

I think this might have turned the tide. This might have turned it.

He did it anyway.

Cisco Critical infrastructure for the AI era unlocks seamless and real time experiences and new value with Cisco.

Thoma Bravo Managing Partner Holden Spat says, I just finished a week of board meetings with several of our portfolio companies and their strong results in growth and profitability seem to be, in my view, another piece of evidence against the overwrought headlines that AI is eating software Anaplan. Bottom line Coupa Dayforce and Jeppesen Foreflight are navigating the AI transition quite well with Q4 bookings, growth of 22% on average and high margins. These are large companies operating at billion dollar revenue scale, not small startups and the growth numbers look to be accelerating, not decelerating. That kind of growth doesn't square with the notion that enterprise customers are cutting software spend so they have more money for new applications. We don't view our portfolios companies as being caught in an either or zero sum competition like that. They're building and integrating AI functionality into their products and customers are voting with their dollars. Systems of record Transitioning systems of record with powerful AI and agent capabilities built to enterprise requirements and workflows have shown themselves to be quite capable of profitable growth. And I believe the value of profitable growth will endure in the face of technological change.

Let's go.

Not biased at all, but I think this is what we want to see, right?

6 months ago Google was considered disrupted. Now it's accelerating. Today SaaS is considered disruptive. Now SaaS is accelerating. Let's see. Certainly not time to take your foot off the gas. If you're building a SaaS company, you got to figure out how to survive, how to thrive in the age of AI. Can't be asleep at the wheel. You got to be building, you got to be innovating, you got to know your value, you got to know, you want to know what really makes your company valuable.

Tom Dale says, I don't know why this week became the tipping point, but nearly every software engineer I've talked to is experiencing some degree of mental health crisis. I saw another post in the same vein from somebody saying, well, at least we're getting it over with before everyone else, right? Software engineers might be prepared for a new reality just because they're probably closer to the beginning of the wave.

Justine Moore has the other side of it. She says, I love how everyone is saying SaaS is dead. Like you're going to get the Fortune 500 to ditch Salesforce for a CRM vibe coded by a 13 year old. And Matt Levine says, what's kind of funny is that people on this website have internalized, quote, you hire McKinsey for air cover and a throat to choke if things go down go wrong, but they haven't internalized you higher. Salesforce, Workday and Atlassian for very similar, albeit not identical reasons. And there's some disagreement. Montag says that's not even remotely true. And Matt Levine says, wrong. This is the other Matt Levine, by the way. This is not the Bloomberg writer. This is a software guy with some finance thrown in. He's been at a 16Z and JP Morgan. Let me tell you about console.com consul builds AI agents that automate 70% of it. HR and finance support giving employees instant resolution for access requests and password resets.

That is right. Good. Alexander says, just to be clear, people are shorting the economy of bits just as software EBITDA margins are about to run to 70% and rotating into the economy of Adams into mass unemployment.

Rotating into the economy of Adams into mass unemployment.

I'm saying like, if you're buying Disney right now saying like, okay, they have the parks business.

Oh yeah, but then it has.

It's not as immune to AI disruption, but then it's, if a bunch of people are unemployed, they're not going to go to Disneyland. In theory, theoretically. Theoretically, they might go more.

Who knows? The fact that crypto and silver are getting clobbered at the same time as SaaS stocks makes me think that the AI disruption story is a little bit too pat of an explanation for what's going on. There might be something else going on in the global economy.

Yeah. This is why giving Anthropic full credit for the correction in the Journal, it.

Really like amps up the aura and the power.

No, I mean, it's great for Anthropic.

And people are running the numbers on like, okay, so you know, yeah, the valuation of Anthropic's probably double the extra 300 billion, but we're seeing like a trillion dollar sell off. It's like, is that really, the exchange rate that we're talking about here, we will see. Jensen is pushing back on the AI will kill all software.

Let's play the video.

Just like he pushed back last year on the Deep Seq Panic. Let's play this video of Jensen.

Software is.

Software is a tool. There's this notion that the tool in this. The software industry is in decline and.

Will be replaced by AI. You could tell because there's a whole.

Bunch of software companies whose stock prices are under a lot of pressure because somehow AI is going to replace them. It is the most illogical thing in the world. And time will prove itself. Let's give ourselves the ultimate thought experiment. Suppose we are the ultimate AI, Artificial general robotics, the ultimate AI, the physical version of us. You could, of course, solve any problem because you're humanoid.

You could do things.

If you were a human or robot.

Would you use a trip. Use a screwdriver or invent a new screwdriver? I would just use one. Would you use a hammer or invent a new hammer? Would you use a chainsaw or invent a new chainsaw? Just don't. First of all, ideally they don't use it at all. A lot of disagreement in the timeline.

But do you understand what I'm saying?

If you were a human or robot.

Artificial general robotics, would you use tools or reinvent tools? The answer, obviously, is to use tools. And so now do the digital version of that.

If you were an artificial general intelligence.

Would you use the tools like ServiceNow.

And SAP and cadence and Synopsys, or.

Would you reinvent a calculator? Of course you would just use a calculator. Okay, Tyler.

Okay, I'd like completely disagree. So, like, if you're. Yeah. If you're robot, you'll use the calculator, right? Like, you'll just build the calculator if it costs you a lot of money to use the calculator.

Yes, yes.

So, like, I think, like, you know, digital, like, agents will, like, use Salesforce, right? But if Salesforce becomes, like, very expensive or if it's. It's already fairly expensive, it's. At some point it becomes even cheaper to just build it themselves. And then you run.

Yes, yes, yeah.

Especially when they're like, I can work the equivalent of 2,000 lifetimes today.

Yeah, yeah.

Like, obviously there's use for tools, right? Like OpenAI doesn't. Or ChatGPT, like, will write out Python script to run some big math problem.

Right?

Yeah, yeah, yeah.

It's not going to, like, use tokens to.

Yes, yes, yes. Yes. At some point, if a single prompt can say, like, build me a clone of this software and re platform me and onboard all my customers and employees to the new vibe coded version of software and maintain it for free, yeah, you have a problem. But as long as the cost to re implement the software and re platform and transition over.

John Palmer in the X chat and Deepak in the YouTube chat, say, would you watch TPPN or reinvent it? Would you invent a new podcast or listen to a podcast?

Canner in the chat was talking about trains, the indomitable will of trains or the unstoppable force of trains. Just being extremely bullish on trains. I don't know.

Just trains.

It was just like a very random train anecdote.

Yeah. I just like that there was a hype train for trains. In the chat, Schmozbe says, you met me at a rather illiquid period of my life. It's rough out there.

Yeah.

I mean, going back to Jensen, it's in Jensen's interest for the markets not to collapse. Even if he doesn't necessarily believe this. Even if Tyler's take is right.

Yeah. It's all a matter of timelines. I think the models are expensive to actually go and reimplement something very, very large and re platform when the current thing is working. Did you see this screenshot of some mapping software suggesting you could go straight or you could go in a loop?

This is a car guy software. It is, yeah. Whose car guy software like this map is probably somebody who just appreciates a nice.

This is actually Google Maps, but basically it's like, you can go in a straight line or you could take a right and go in a massive loop and it's eight minutes slower and it's like, would you like to go the slower route? But yeah, if you want to do a. If you want to do a hot lap, like, go for it, I guess that's a good take.

We gotta. We gotta actually pull out the post from Amtrak. Amtrak hired a poster.

What'd they do? This is where the train.

Yeah, yeah, yeah, yeah, yeah.

Sorry.

This was probably, you know, one of the. One of the posts of the week. Amtrak says the unbridled greatness of trains will endure for a million years.

Yes, yes, yes, yes. This is what he was quoting. I thought it was so funny. The unbridled greatness of trains will endure for a million years. How many likes does that have? That's a banger. 4050. 4000. That's great. Let's go over to the Guinness Tally. Tyler has it. He counted it up, he sat down, he studied the Dorkash Patel cheeky pint crossover episode with Elon Musk. And he counted three pints for Elon, four pints for John, and three quarters of a pint for Dwarkesh. But there's a suspicious refill.

Yeah. So, I mean, it's a bit hard to really tell because if you. So if you're tracking Dwarkesh's, you know, volume right throughout the episode, it kind of goes down and he's at maybe like, you know, 80% full.

Okay.

And then it goes back up.

That sounds like a refill.

But it's like a direct. So it's after a. There's like an ad read and then it goes back up. So the question is like, during the, you know, whatever happened during the ad read? Right. Did he just like slam it true filled back up all the way, or is it just kind of a partial fill?

I think you got to give him 1.75.

Okay, can we pull up the screenshot?

Yes.

Because the other debate was just around the actual size of the pines.

Yes.

Is there some of them?

They look like full pint cloth.

So.

Yeah, if you go to.

Those look like full size pint cloth.

Another photo of Elon holding the pint.

Yes. Three down.

You gotta pull that up. Tyler.

Put it in the production.

Yeah, yeah, yeah.

No, no, no, it's here.

This is really important. Yeah. You see this and it's like. Okay.

Looks pretty small in his hands.

A little small. And then. And then let's click over to Timothee Chalamet. Timothy Chalamet. I mean, he's a smaller person, but it looks the same size to me. I think these are fair pint glasses. I think these are accurate. I'm not seeing anything. That's.

That's.

That's.

So Timothy Chalamet is 5 10. Elon is 6 2. So slight height difference, but I mean, I don't. I feel like, yeah, Elon is really mauling that maybe.

Yeah, I think. I think they're just potentially wider.

I think they.

And this is the kind of hard hitting analysis that you can really only get on this show. No other. No other technology media is really breaking down. I'm going pint analysis.

I'm going normal sized pints on this. But it is. It is an interesting optical illusion.

Yeah, this one looks incredible.

It's a little bit of an optical illusion. It's close. It's close. I mean, either way.

But we don't know even if It's a half pint.

John drank four, so that's two full pints. Even if it's a half pint. You know, I think these are full pint glasses. I'm going full pint.

Okay. You know who else is going full pint? John.

Who?

The founder of crypto.com.

Oh yeah.

Who just bought AI.com for 70 million. Let's go big, big, big. He is running an ad for AI.com we don't know what AI.com is yet, but I'm assuming you're going to be able to get some artificial intelligence there, much like you get crypto.

It's a response. Haven't been people saying if you're in crypto, pivot to AI. Took it literally.

I'll read through this. Chris founder says plans to launch a new site with a Super bowl ad this weekend. The founder of trading site crypto.com just bought AI for 70 million. He actually bought it last, last year in April, highest price ever disclosed for a domain sale. To launch a new entrant into the AI race there, the site will offer a personal AI agent that consumers can use to send messages, use apps and trade stocks. It's very, very American. Send message, send memes, order doordash and trade stocks with your new AI agent. Last year an opportunity came up for me to acquire this domain and I thought if you take a long term view, 10 to 20 years, that AI is going to be one of the greatest technological waves of our lifetime. And so it would be a good investment. Mars did not comment on the purchase price, which was confirmed by the deal's broker, Larry Fisher of getyourdomain.com. huge win for Larry. Larry, shout out to domain dealers with assets like AI.com, there are no substitutes, said Fisher. When one becomes available, the opportunity may never present itself again. Of course AI.com pitched itself, was pitching itself aggressively to all of the different labs. Everyone was like, you know, we anthropic's like, well we may not have anthropic on X but we do have anthropic.com.

Who owns claude.com who owns AI. AI, that's a good domain because Claude owns or anthropic owns Claude.com, but they also own Claude AI and they reroute Claude.com to Claude AI, which is normally the opposite of the flow that you'd go, you'd start with the AI or the other TLD and then you would, as soon as you get the.com you would reroute everything to.com anthropic so AI pelled that they're using the AI as their main domain. They're rerouted.

Or maybe they're Anguilla pilled.

Oh yes.

Maybe they. Maybe they just like supporting Anguilla. Tyler is using AI.com to launch a personal AI agent. Is it your Gen Z? Is this cheugy?

Wait, what is Chugi again? I don't know.

Chugi is like Gen Z slang.

I think it's like millennial slang.

Really?

I don't know.

It feels a little chubai to me. I'll say that. I know enough about that. It feels like Chew by. It's chopped. It's Dubai. Dubai would want the name of the thing, but it's chopped.

Cheugy is a slang term often used by Gen Z to describe someone or something that is uncool, outdated, or trying too hard to be trendy. Yeah, typically referring to millennial Trends from the 2010s.

So it's a dig at millennials from the perspective of zoomers.

Makes sense. It describes a post basic aesthetic that is slightly cringe worthy, such as live laugh, love signs, skinny jeans, or side parts.

Okay, interesting. Vanta Automate Compliance and Security. Vanta is the leading AI trust management platform. Let's go over the mansion section. There's some good stuff today.

Hit me.

We got a bunch of stuff. So, Miami's Coconut Grove. They're calling it Billionaires Grove. The ultra wealthy buyers have discovered Miami's once sleepy Coconut Grove neighborhood.

Cam in the X chat says Chugi is being carried by millennials.

Okay, interesting. Thank you, Kim.

Van says what an unk term.

Once known as a bohemian enclave, Miami's leafy Coconut Grove neighborhood is quickly transforming into a billionaire's playground thanks to a string of mega sales in the area. Unlike hotspots like Miami beach and Indian Creek, the Grove, with its massive banyan trees, winding streets and small gated communities, was little known to people outside of Miami. Until recently. Most people hadn't heard of it unless they were local, said real estate agent Danny Hertzberg. Now the secret is out in a big way. And it's on the COVID of the Wall Street Journal Manson section. Since late December, Google co founder Larry page has spent $188.4 million on three properties in Coconut Grove, while businessman Jorge Mas, a co owner of professional soccer team Inter Miami.

Okay, but what did he actually do?

Do?

Yeah, that's a good question. Look him up. He just says he's busy introducing. Just getting introduced as a businessman is incredible. That's life.

Jorge Moss, President of Real Zaragoza okay.

Is he really just a soccer guy?

No. American billionaire businessman and chairman and largest shareholder of Mass Tech, a Miami based construction and engineering There we go. Founded by his father. Let's go.

So he spent 100 million. Both of those properties are within a few miles of a waterfront estate billionaire Ken Griffin purchased in 2022 for 106 million, a Miami record. Locals are now competing with out of town buyers for homes, real estate agents said. And housing prices in Coconut Grove, especially for waterfront property, have surged. The median single family home sale price in Coconut Grove was 2.3 million in the fourth quarter of the 2025, more than doubling from under a million dollars, 995,000 in 2019. So more than double in just six years. A $10 million home is now a $50 million home.

Wow.

Live, Work, Play Coconut Grove was a hub for nightlife, galleries and artists in the 1970s, 80s and 90s, but its retail and cultural scene faded. The housing market became kind of sleepy. Then about a decade ago, the neighborhood's reinvention began with the redevelopment of the Cocoa Walk Open Air Mall and construction of luxury condos, which introduced 20 million dollar penthouses to the Grove for the first time. When Covid struck, an influx of families relocating from out of town were drawn to the area, which is walkable with high end restaurants and retail but little tourist activity. Part of the Grove's appeal is that the its proximity to downtown Miami and the Brickell neighborhood, where Griffin is building a new headquarters for Citadel. The Grove also has prestigious schools like Ransom Everglades School and Carrollton School of the Sacred Heart. Although the Grove isn't gated, there are a small number of gated communities within the neighborhood, such as Ye Little Wood, the Moorings and Camp Biscayne. They're subtle, hidden, lush. You don't know they're there. That's a big appeal. The Condo Boom New condo buildings have been at the epicenter of Coconut Grove's market resurgence. A decade ago, large new luxury units were relatively unknown in the area, said a developer. People thought I was crazy, martin said. But his gamble paid off. Martin sold his own penthouse in the building for 17.8 million in 2023. Now CMC Group and Fort Partners are developing a Four Seasons branded condo in Coconut Grove. Prices raised from 5 million to 16 million, excluding penthouses, which are expected to sell for $120 million combined, although that is far more expensive than any other condo sold in Coconut Grove to date. We think it's achievable said Colombo. Great name for a Miami real estate person. The buyers looking for that before they would only consider Miami Beach. Now those buyers have Coconut Grove on the radar. Then there's bigger homes, too. The average lot size in Coconut Grove is under an acre, which is smaller than lots and other luxury enclaves. That used to deter buyers who wanted more land. But since 2020, developers and end users alike have been expanding existing properties. People are starting to knock on the door of the neighbor next door and put together big, bigger assemblages. For example, Page bought a 101 million dollar waterfront property owned by the late restaurateur Jonathan Lewis, then acquired an abutting property for 15 million. Page also purchased a nearby property for 71.9 million from Heiress Sloan Lindemann. A few waterfront homes in Coconut Grove are highly sought after, agent said he recently sold a $16.5 million bayfront estate owned by Susie Welch, the widow of former General Electric co Chairman and CEO Jack Welch. Wow.

I'm sick of Miami. Can we go to the mountains? What else we got?

Let's go to the mountains. What's the next. What's the next story here?

That is actually my general experience with Miami. It's amazing for a week, then I'm ready to mix it up.

Well, we could go to Manhattan or we could go to Beverly Hills. What do you have in mind? Neither are particularly mountainous, but I guess.

The hills are Beverly Hills.

More mountainous. So there's a California villa here built for James Cagney. Do you know the famed Hollywood film star James Cagney?

Never heard of him. John wow.

What were you doing in the 1930s and the 1940s when he was making gangster films such as Angels with Dirty Faces? Good movie.

Crazy name.

He built the stone Villa around 1939.

Property recognition we don't know how to name movies like that.

Angels with Dirty Faces Coming to to a theater near you. Buy your tickets today. A year later so he died in 1986. A year later, his widow, Frances Cagney, sold the property to the current owner, Steven Dunn, founder of the baby product supplier Munchkin Munchkin. Dunn didn't respond to comment. Cagney lived in Beverly Hills while filming, but otherwise spent his time at his farms in upstate New York in Martha's Vineyard, actor Robert Wagner, a friend of Cagney's, wrote in a 2014 Vanity Fair piece. The California house was unpretentious but had a studio with a wooden floor and a record player where Cagney practiced dancing, wagner wrote after purchasing the Beverly Hills estate. Dunn expanded the villa to approximately 6,000 square feet with three bedrooms, said a listing agent. In the 1990s, Dunn purchased an adjacent parcel combining the properties into a compound spanning five acres. In addition to the main house, the property has a pool and a roughly 2,500 square foot guest house. Over the years, Dunn has added.

Does it have a moat?

I don't think it has a moat, but does it have a lot of stone?

I know the stone was signaling maybe there's an alligator moat.

So 38.5 million, six bedrooms, multiple houses, Tennis pool and your favorite, pickleball.

A tennis pool.

Tennis. Tennis court and a pool.

Okay.

It has a pool cabana with an outdoor kitchen. It's landscaped extensively. It has pickleball courts. What do you think? 38.5 million for a six bedroom.

I like it.

You like it?

It's very unique. At 38.5, it looks like it would withstand a wildfire.

Yeah, that's true. That's true. The luxury market in Los Angeles is picking up after battling major headwinds including a mansion tax and geopolitical uncertainty.

World in the. In the chat says cop the estate boys new ultradome. We might have to John while we were live our latest ultradome there that we had had an offer in on. They're out for near me. We've had like we've had truly. If anybody has just a big build building in the center of la, we want to give you money for it. Let us do it. We've had such a fascinating time trying to get. The entertainment industry in LA has collapsed. It's a ghost town. It's a modern Detroit and yet we cannot find a building.

I don't know why people like me, I keep talking trash about them in this town. You gotta glaze, you gotta triple glaze them. It's the best. It's the best town. We just want to build it even better.

Yeah, we're. All we're trying to do is bring media back to Hollywood.

It's not degrading at all when you say it like that. Maybe we should move to Milan. It's the gold medal housing market. They're saying the city co hosting the Winter Olympics is a fashion hub with Europe's hottest luxury residential market. When the world's winter sports elite gatherers. This elite gathers this week for the opening ceremony of the Milan Cortina Winter Olympics. They're starting this week.

Have you been to that arch?

No, I've never been to Milan.

Quite nice.

I've never been to Milan. It's a fashion and design hub that has long been the country's main economic center. It is increasingly becoming a tourist destination. And in 2017, Italy introduced tax laws favorable to new residents which led to data centers. Yes, let's convert Milan.

A data center with that original arch aesthetic for sure. Would yes, would really, would really be. It would fix. It would fix me, it would fix Tyler. I mean imagine we don't know how to make data centers that look like this. Everyone's talking about putting them, everyone's talking about putting them in space. Why can't we convert some of these?

We're not just putting them in Milan. We're putting them in the most expensive neighborhood in Milan. That's Brera. Brera is a small exclusive neighborhood just west of the city's premier shopping district around Via Monte Napoleon. Residential real estate in Brera, where most homes are apartments is currently the city's most expensive. Brera's high end sector had an average sale price of $1,868 per square foot. What would it cost to put Colossus there? How many square feet is Colossus? I want to know how many because we got to go apartment by apartment. Racking NVL 72s in Milan. In the Barrera neighborhood, specifically a renovated duplex sold for just over $28 million, the most ever paid for a Milan apartment. The adjacent Quadillera della Moda or fashion district, home to brands like Gucci and Cartier, was once the city's most expensive. But Brera has suddenly surpassed it. It's more livable. Well, for now, until we put these data centers in there, you can walk to the shops, but you're not living in a shop. People want penthouses with views and the. There's a couple other things in Brera. I don't know. Maybe that's the location of the next ultra dome. Let me tell you about Applovin. Profitable advertising made Easy with Axon AI. Get access to over $1 billion daily active users and grow your business today. Tyler, what's it going to cost me to put a Colossus sized data center in the Brera district in Milan? The neighborhood in Milan that costs $1,800 per square foot.

Okay, so apparently Colossus 1 is only 785,000 square feet. Okay, so it's only going to be like 1.4, 1.5 billion billion doable.

So not bad at all easy.

I mean putting a to build a gigawatt of a data center is like 50 billion, right? Yeah, that's like in that range. So I mean this is like kind of a no brainer.

It's a no brainer. It's a no brainer. As is interviewing Doug o' Laughlin after earnings when he's suffering from Claude code psychosis. Welcome to the stream, Doug. How are you doing?

I'm doing wonderful. How are you guys doing?

Great to see you.

Can we pull up, pull up the video that John painstakingly made this morning that completely flopped.

I'm very excited to have you on the show. It feels like that scene in Sonic the Hedgehog three which I saw where Sonic and Shadow team up and join forces to talk about capex and agentic coding. What's new in your world? Is Claude code still the top of mind or are you still churning through the capex numbers from earnings?

Are you somewhat of an agent for, for Claude now like you work for Claude?

I do actually. I think I mostly just move my information back and forth. You know, I have, I have pretty much like. I think of it as like my manager, you know, like it tells me what to do and then I go bring the information and I bring to my coworkers. I bring it back all day. I'm just on cloud code.

How many, how many prompts are you running right now? Do you have any threads going?

Okay, so. I have seven. I have seven threads.

Seven that are running right now or.

No, no, not running.

Waiting for your input.

I'm waiting for my input.

We'll let you get back to it.

Why don't you just have an eighth that just.

Let's talk about. Yeah, orchestration. Have you played with Gastown? Are you thinking about abstracting yourself to a higher level?

Okay, so Gastown is pretty intense. I don't think Gaston is going to work out. I think it's going to be agent swarms.

Okay. Okay, explain the difference between Gastown and agent swarms.

Okay, so Gastown is probably the most forward looking thing I've read in a lot. I talked about how you created this self healing tool process to essentially to pass all these beads across and have all these workers and ways to self repair the agent workflow process. And I read it and I was like dude, this is brilliant and also fucking crazy. It feels like the ravings of a madman. And then I proceeded. Well, I was also in my badminton era like before, before the New Year's when you had 2x times usage. I pretty much was like literally railing Claude code constantly. I think I had, I think I had four 14 hour days.

Wow.

Yeah. It was beautiful.

It was beautiful. Okay, so talk to us about what you're actually building, because we're talking about SaaS Apocalypse. It feels like there's a debate over build. Rebuild all your tools from scratch to save whatever your SaaS fees are. Yeah.

Even yesterday was notable. OpenAI comes out with Frontier, which is. You look at, like, you got to look at this graphic, which feels like it was made for a Fortune 500 CEO or management team to kind of understand it. And it's like, here's more SaaS to replace your other SaaS. Right. It's like, you know, you got the system of record down here, you have a bunch of agents in between, and then you've got different applications that you're using. And meanwhile, Anthropic is just like, we're making a really smart digital guy that can do whatever you want.

Yeah. So I think the two. There's, like, two really interesting ways. I think OpenAI is like the Fortune 500 selling it from the top, if it makes sense. And then Anthropic is like, here's your quad code agent. Sell 20,000 of them. Did you see the Accenture Partnership? I think that's really interesting. So if you go back, they're doing 30,000 people at Accenture.

What do you mean?

To, like, 30,000 people at Accenture are going to learn how to CLAUDE code.

And then be deployed into different companies?

Who knows?

What else would they do?

Maybe they just replaced Accenture. I mean, that's what they're going to be doing, I think.

Okay, wait, so, yeah, what do you mean? If Accenture folks are using Claude code, wouldn't they be using it on consulting projects internally to companies?

I think they're going to be using it internally. So they're going to be using internally, and then they're going to be doing all these consulting things. Because if you think about it, one of the issues is, like, you know, when you had SaaS, one of the biggest issues of changing from one CRM to another was effectively being like, hey, everyone, you have to quit your jobs for, like, 10 months to figure this out. The implementation, like, you'd have tons and tons of consultants do that. And I think that that's what the Accenture partnership is. So essentially, people are going to be implementing CLAUDE code, and there's 30,000 people at Accenture who's going to do it. And then on the other side, you have Frontier, which is like the Fortune 500 way of being like, here's your plan. Come to us and we'll build this whole thing, blah, blah, blah. So.

Yeah, but is. Is. Is re implementing your CRM really the lowest hanging fruit for America's greatest companies? It can't possibly. There must be new ideas, new problems to solve, new tools to build. Like, why are we just going to shuffle the chips around the board instead of like doing something productive?

Okay, so I think the system of record refresh is going to be really awesome because, you know, like the big. I mean, it honestly does feel kind of boomer if you think about it. It's like the biggest data now. Everyone can have the big data now, but I think the. The automation that you've always dreamed of is actually going to happen. And the system of record is just going to essentially have hooks out to all these other things that are going to build on top of it, which is mostly like, you know, the frontier thing.

Yeah.

And essentially like all the information work is just going to be like, all on the agent and everything else is going to be like place where it lives and is stored for font season. So instead of me having someone let me use my personal stack at semianalysis, we use HubSpot, for example. So, hey, the sales this quarter, we need this like, quota or who did what or what products are selling better or like, you know, what's.

How many more podcasts should Dylan do this month? If we want to hit our goal, that's totally different.

We don't actually have like, our, like, you know, the big. Yeah, sure, how many more podcasts? Shit like that, right? I could just vibe code it. I would just be like, hey, can you run this analysis for me? And in a perfect futuristic world, it'll go into the CRM, pull all the information of all of our inbounds, be like, hey, the day after Dylan goes on a podcast, there's like 25 people who come in. The conversion rate is X. You could price it at this. Dylan, quit your, you know, stop working and effectively just like hit the podcast.

There we go.

You know, like. Yeah, yeah, exactly.

So.

So you could do this with anything, though. Like, it's just information, man. Like, it's, it's. It's gonna be pretty sick. But I think all the SaaS companies are going to essentially just become hooks for all the crap they built on top of it.

Yeah, yeah, yeah.

Did you see Jensen yesterday? Was. Was kind of defending some companies like SAP and ServiceNow and saying, hey, if I was a. If I was a really smart humanoid out doing work in the world and I needed a screwdriver, would I just invent a New screwdriver or would I just take one off the shelf? And so that was like his defense. Tyler here was took, took the other side of it and just said there's going to be a lot of situations where especially in a software only environment, it's easier to just build a very specific workflow that you need that you would have gotten from a SaaS provider versus, you know, you don't need to actually rebuild the entire platform.

Yeah, I think we're going to be building a lot of screwdrivers. Like things like the thing that's important is like, okay, you're not going to rent a truck, right? Like you're not going to build your own truck, but your own screwdriver. 100% like you're doing this big ginormous job. You need a hammer. You'll be like, okay, pull it out of my belt. Okay. But you're not going to be like, I'm about to move 700 tons of like here to here. I need to rent a truck. You're not going to build the truck. And so that's what I think the system records are going to look like. They're going to look like places where like actual data that cannot be like, like cannot be vibed effectively. Like what's your inventory, cannot have any fucking hallucinations.

Right.

Like your erp. But all of that will just be hooks for everything else because like all the information is just like pulling, retrieving, making the correlation, running the charts like, okay, but even that all the time.

But then, but then how do you square the fact that a system of record is way less sticky if you have agents that can work around the clock to switch you over to a different system of record? Like that still ends up putting massive pricing pressure.

So, so to be clear, I don't think it's good for everyone. Like, I think my, my favorite analogy of this is like there actually is a very old school type of software that's like existed for a long time. All the shit on mainframes, it's all, it's all out there.

Yeah.

And you know, like, funny enough, mainframes still grew like 6% a year or whatever. Someone has the real number from like 2002 to, to 2020.

It's great.

So like they're, they're going to grow, but it's just going to be like a very different vision of the world that I don't think people are ready for.

Yeah.

And the adjustment period is the big problem because all the stocks are priced like they're not going to be mainframes and, and also just for context, mainframes. There's like, hey, there's one of each company now.

Yeah.

There's not, there's not like 10. Yep, just one each.

Yeah. So I think with all these, all these, whether you're a system of record or you're, you know, some vertical software, you're going to need to show insane revenue growth in a truly AI native product. Otherwise investors, I think are going to continue to not be able to create a super compelling narrative why you should own it during this period of uncertainty.

Yeah, I mean, pretty much what happens. And we're going to go like investor brain. When anything goes X growth, the multiple goes massively down.

Yeah, yeah, that makes sense.

Eight times earnings.

Can you talk a little bit more about what you're actually coding, what you're building, like what the software is? Because from the demos that I've seen that you've posted, it feels much more like you have an agent that can do knowledge retrieval, data transformation, build dashboards, charts and knowledge work as opposed to truly replacement, replacing software tools at this point. But have you built anything that's like long lived and runs daily or is something you keep revisiting because it's now a piece of software that does the job.

The CLAUDE code commits, the Claud code commits, is now software that lives and runs every single day. That's like a scraper. Right. And then that lives in a database and that will run forever. There's a lot of other tracking, price data tool stuff, like a lot of the scraping that we're like, that is not publicly available. We do a lot of that. We had a data team just do that and now effectively we can really accelerate that. So everyone can do that.

Sure, sure.

There are other little things that I think are heuristics, little skills of. I have blind spots that I consistently make over and over and I'm like, hey, I know this blind spot's an issue, blah, blah. You should consider this in this case. I don't think it's like the Galaxy Brain software and we're very far from there. Because if you actually play with these tools a lot, context rot is real.

Yeah.

But how fast is it getting better? Because it feels like we're seeing the.

Meter graph scary fast, scary fast, scary, scary, scary. So I started vive coding with Claude 4 and it just wasn't or, sorry, Opus 4 on quad code and it just could not one shot websites in the way that 4.5 and 4.6 can. And if it just marginally improves from here. It feels like why would I pay for any kind of UI UX if it just could be generated at a good enough quality?

Yeah. How did you process the new models this week? 4.653. What's the review?

If you can't immediately notice the difference between 4.5 and 4.6, start polishing your resume. You are cooked.

Yeah, yeah. You just got automated by an agent. I think 4. 6 was a little disappointing, to be honest with you. I think it might have been Sonic 5.

Oh, that's what people are saying. That's the conspiracy theory. Right. But what does that mean?

The original Sonic 5 leaks were that it's like as good as Opus 4.4.5 but with 1 million context window and specifically trained for agent swarms.

Sure. So yeah. But practically does that just mean like same quality but faster, cheaper, at least.

For and then they. And they make more money?

Oh yeah, yeah, yeah. But yeah, better margins. How are the margins looking for the labs right now? There was a bunch of like FUD around it, but it seems like from all the leaks it's been like 50, 60, 70% pretty good.

Yeah. If you X all the free users, it's always really good. Right. Honestly, Anthropic has no free users or on a relative basis. So their margins are ironically kind of on a like to like basis. Kind of not as good as you think.

Yeah. Can you break down a little bit more of the thesis of the Claude code is an inflection point article, what the key takeaway, who you're speaking, speaking to, what update you wanted to share and then I want to go into some of the pushback and your response to that.

Yeah, sure.

So first I think the thing that makes me really excited is the first time since chain of thought I feel like we have a new scaling that feels very, very different and hardcore and I can actually see my entire life, day to day change. I think I can expect some version of a cloud code harness to be effectively all my information work from now till the future. I am a daily user. I was not a daily user before and I expect to continue to be one.

That's what happened with the reasoning models. It went from you could ask it stuff but it might hallucinate to the answers are good, you can pretty much rely and there's going to be citations and it's gonna be 99.999% usable for things. So if you just have a question, you get that it may be not great at certain things, but in general it delivered on the initial chat experience that I think a lot of people were looking for. And then they became daus. Yeah.

How much do you think Anthropic cares more about winning in consumer than they've let on to date?

No, I don't think so. Everyone who works there is exactly what you think it is. They're exactly who they say they are.

Yeah. They're software Singularity build.

Yeah. And then I think cowork is what they're really excited about.

Sure. Makes sense. Yeah. What this pushback.

So they're not even thinking about a scenario where a bunch of people are using Claude in a work setting and say, hey, this is pretty great. And yeah, ChatGPT has ads. I'm happy to pay 20 bucks a month. I'll use it personally because I just think there's like an iPhone. Like I think the game to get to like 3 billion users is like over. When you just look at the traction of like Gemini and ChatGPT and the fact that normal people aren't caring that much about the nuance, maybe that don't have that much to automate in their life, but there's like an iPhone size market. Like the iPhone wasn't the first smartphone to launch. And it's possible when I see this super bowl ad, the sort of trust nuke I was calling it, just like, hey, it's really funny, they're rage baiting OpenAI but at the same time they're just destroying trust around ads and LLMs potentially permanently because people, even when they start seeing ads that are more like display ads, they'll start thinking, well, was the result influenced too? It hurts the trust. And so my theory is that any product that really catches on in the workplace could very well trickle over into life and Anthropic could someday have a pretty big. They could have a Netflix sized subscriber base for people that just want an ad free AI experience.

Yeah, that sounds completely right to me.

But you're saying your point is like it's just secondary to them. They're like, it's a nice to have but like we don't. That's not, that's not our intention, I think.

Okay, so singularity pill. But I also, I think you have to pay for the singularity and I think it's going to be enterprise that.

Does it makes sense.

Yeah.

Yeah. I mean the other take on like you could wind up being like the Apple and like the premium, you know, privacy focused or you could wind up being like the DuckDuckGo which was like yeah, you know, yeah, it was a counter to Google, but it never got to any meaningful scale.

Yeah, but I still think OpenAI is the apple. Apple was synonymous with smartphones when it really took off. What is the other smartphone?

Nokia maybe?

You could argue this is like a BlackBerry. Yeah. I can't name it. Yeah, well, like BlackBerry.

Right.

It was known for work.

Yeah.

And then obviously it swapped over. So I still think OpenAI is like the cognitive reference. And honestly, 5.3 cooks.

Yeah.

Faster or just better or both?

Faster and better.

Faster and better. Okay. Talk about 5.2 token efficiency. Rune was pushing back on the article saying you're making the assertion that 5.2 token efficiency ruins long horizon planning and yet 5.2 tops the meter chart for long horizon planning half baked. What's the explanation there?

Didn't someone completely mock that argument? But he's kind of a. He's like, sorry, I gotta find the guy. But it's like, I don't know what task is being done here. Like, is. Are they the same hardness?

Yes.

Did you just spam it to infinity? And like you finish like a sufficiently long task to completion versus like, okay, let's just say we have two kids taking the sat and one does a better job and finishes first and one does like almost as good of a job and took seven times as long and you're like, wow, that one's a smart kid.

Yeah.

No, dude, that doesn't make any sense.

Yeah, yeah, yeah, yeah, that makes sense. Yeah, yeah. An elegant solution delivered faster is uniformly better.

Yeah, 100%.

That makes sense.

Yeah.

And so if you spam more tokens and you win and you're like, oh, look, I mobbed them. And it's like, dude, what if you just use less tokens?

I think the benchmark is supposed to be for. They have a reference class of projects that are supposed to take X amount of time. They would take a human developer six hours. And then they have all the models compete. And if you can compete the six hour task, then you get put at the six hour mark. It's not. Did you run for six hours? Okay, that makes sense. So it could be like implement a CRM product or write very complicated, you know, database or something. It would take, you know, a talented software developer six hours, two hours, one hour. And they have different tasks and then you're trying to climb that.

Oh yeah. And then it's okay, it climbs higher and higher.

I think, I think that's loosely. It's because. Because obviously you could just say, okay, just reasoning count to 1 billion and just go as slow as. And it works for days. And that's not impressive.

Yeah, I mean it's okay. So yeah, you're right. The different. The scaling thing. But like. Okay, so one. The other thing I was doing, now that we have vibe coding available to everyone, you can just have it do the same task and do like abc. I've been doing a lot of internal benchmarking. Everyone could benchmark. Guys, dude, Codex 5.2 took so long and just never build for me. And it's like all the Codex hype, it just never worked for me, man. It never one shot projects like Opus 4.5 did. And I'm just like this feels like complete FUD. But that being said, Codex 5.3 cooks. I think I take everything back. Buff about 5.2 for 5.3.

Nice. Total reversal. Classic AI narrative just day by day. Complete switching of the narrative. Talk about the NPM downloads because you said that you're now scraping them every day trying to understand how many commits on GitHub are related to CLAUDE code. And. And the pushback from Rune was that this counts NPM downloads as authoritative when CLAUDE code numbers are hugely inflated. Because GitHub Actions does automatic CLAUDE code download every time. Continuous integration CI runs versus Codex compute cloud. So maybe it's not apples to apples. What's more, nuance on the fast takeoff of CLAUDE code. Because honestly, when you said 20% of commits by the end of the year, I was like, that feels extremely low. I would expect like 70% and I would expect Codex to be at 30% and no more human commits because it's working.

So I wanted to make sure we have a high standard, like a high 95% plus. I don't think sure if it continues to grow on a week on week basis. Yeah, it's 100% by June or something like that.

Well, there's also the fact that you could be writing code and still just almost be using cloud code as your linter or your interface to GitHub. And if there's an abstraction layer there that people adopt, you're going to see the commits go through the roof. Even if there's still a human in the loop meaningfully.

I'm not going to pretend like the COD code commits thing is the cleanest way ever. There's a lot of ways to fuck the data. For example, people who use who you could just say don't do this and won't do it. Number two Private on a ratio is five times bigger. That matters way more and then I also think that the way you consume it, like this doesn't count for cursor. People have been clearly using AI for like a long while and it doesn't show up. This is just the example. I can, like, say, like, hey, chart goes up really quickly.

Yeah, yeah, yeah, yeah, yeah. It's so cool. I mean, it's not perfect. Yeah.

It's a data set that I create. No. In a relatively short amount of time.

Yeah.

I'm like, well, seems pretty cool.

Yeah, yeah, no, it is.

We talk about Amazon.

Yeah, yeah. Let's move over to hyperscalers reaction.

Was the number two low?

Yeah. Bearish. Yeah. They're not taking it seriously. Yeah.

200 billion was crazy. That was crazy to me.

Okay. Why?

Yeah, that was really shocking. You know, we do a lot of data center tracking and we do a lot of accelerator tracking and we were too low.

Okay. Are they. Are they. Are they trying to play some sort of hype game where they're throwing out the biggest number and they're not actually even going to be able to buy enough equipment to spend it? Even if they are signaling to the market. We're going to be hearing like, well, we wanted to buy this many Nvidia chips, but we couldn't get them, or we have a delay at this data center because of regulation. And so they're just trying to project strength because they're sort of behind on the AI narrative a little bit. They don't have the big position that Microsoft does in OpenAI. They don't have, you know, a. Of a DeepMind level team. And so they're saying, we're going to go biggest on the dollar front, but then maybe they don't deliver on it. Or do you think at the end of the year we'll be like, yeah, they spent 200 billion?

I think at the end of the year they're going to be like, yeah, they spent 200 billion.

Let's go.

They are the single biggest provider of power in the entire world. I think, like the incremental and the AWS supply chain can ramp a lot quicker than anyone else. And every example that we track in the data center, like the data center team, they are on time and can scale to levels that are crazy. Rainier's ramp is just out of this world fast compared to everyone else. Every other gigawatt project is essentially delayed and they're going to be ish on time.

Wow.

So isn't that extremely just good for Amazon? They're properly positioned, they're properly transitioning, they yeah, like, who knows what happens to the rest?

Part of it.

But eventually this thing is just like part of the.

We were talking earlier, like, Jassy didn't exactly paint this, like, incredibly exciting vision and share like, hey, you guys are actually underestimating demand still. Even if you're bullish on AI, you're underestimating demand. And we're in a position to actually try to get a more accurate read here. And that's why we're investing.

Yeah.

And it's funny because they could have said one thing that would have made the AWS call better and they'd be like, yeah, we see high 20s and the stock would have ripped. But they're like, we continue to project to see this level of growth.

What percentage of the 200 billion do you think will actually flow to Nvidia? Because Nvidia is rallying today. That's why we're wearing white suits. But it didn't rally immediately. In after hours.

I think a meaningful amount. I definitely cannot disclose what semi analysis thinks, but I think they're going to run out of trainium. And the answer is like, what's the biggest amount of supply chain that's like locked up? It's Nvidia.

Yeah, that makes a ton of sense.

Can you get Macron a free semianalysis plan? Because he came out this week with his big new initiative, 30 million euros for AI research. France is going to be the home of research.

How do you think all the hyperscalers will respond?

You know what's crazy is people have been trying to do a lot of work in France for a long time because they have this giant nuclear power plant. Yeah. It's kind of stranded and no one uses it. And everyone wants to be like, dude, I can get a gigawatt here. And then they try to start building and they're just like, yeah, this is never going to happen. I'm just going to go back to the United States. Even though United States is like all fucked up, it's like, I can start there. And they're like, no, no, no, no. We'll start in like five years. And I can think of two specific projects that essentially did the same thing where it's like, oh my God, all this France data center power. And then they started and like, nevermind.

Yeah.

Why do you think Grok is climbing the charts right now? Any insight? It's like number three, after get free cash and after ChatGPT in the overall app store.

The iOS app store, dude.

Actually, one, this tells you how locked in I've been with Claude code. I had no idea.

Yeah.

You're so locked in with Claude researching the AI race that you.

It's just interesting. I mean, you know, the App Store is based on, like, acceleration, but. But, you know, the grok hype cycle of, like, you know, let's push all the Twitter users or the X users there. Like, that sort of already happened. Like, I don't know how this is happening because there isn't much hype about the.

And it's happening. Yeah. It's happening off X. Yeah.

And a lot of people were like, yeah, like, you can talk to Ani and Valentine, but, like, is that really popular?

Might be.

I don't know.

Ani singular.

The real singular. Lonely people.

Yeah.

Oh, my God. Maybe I did see a video of, like, the Stormlight Archive thing and that I feel like, hit a broader audience in terms of video generation. And I think video generation, that always kind of wins. We actually did an analysis, which is.

Why you need to be bullish on the Disney OpenAI deal.

I think so.

Because, yeah, you have to be. We've seen the nano banana bump with Gemini, and this feels like it could be on an entirely different level.

Yeah.

Well, you know, my favorite thing is Gemini wasn't what actually, like, made it rip. It was nano banana.

Yeah.

Like, the ratio, like, really improved in terms of OpenAI to Gemini, like, way before, and then like Gemini, like, slightly helped, but I would say it's like 90% is banana. Banana.

No.

Put dogs in a bikini.

Yeah. You can just share an image and it's immediately apparent what is going on as a unique capability that you can't get anywhere else. They've cornered the market specifically on, like, the image editing. Not just the diffusion, but the, like, being able to take a photo, change the background and have it actually look like your face or have the text look great. Like, it was. It was a unique product. Really beyond a model.

Yeah.

Oh, this is. By the way, this is my final steaming hot take in COD code, please. The reason why you should actually pay attention so, so much is because this is the first time, like, image models essentially always gain share. Video models always gain share, like Studio Ghibli moment. And then obviously chatgpt. This is the first, like, new moment.

Yeah.

It's a new modality being the agent and it's like, actually kicking off.

Yeah. I mean, how important do you think the cowork, like the desktop app, mobile functionality is to that? Because, like, the. You could have. You can have truly magic.

Chad has some insight. A lot of people using Grok Video to compete for a $1 million contest.

Oh, that's right.

Free money. Free money.

Okay. Yeah. So it's free cash and then free money. The top two apps in the top three give you cash. That is a good fact check. Thank you, chat. But yeah, my question about can you have a Studio Ghibli like moment if you have to open up a terminal just because there are so many normies that just will never open the terminal, no matter how magical the AI God is behind the terminal, it's just too much to go type one line of command.

That's why Cowork and Codex are going to probably be what actually happens. I think it's really fun to play around in the whatever 1% adopter and I'm really enjoying it, but I just don't think, yeah, it's going to be coworker Codex and Codex is actually pretty good. Codex is, I think, a slightly more polished experience than Cowork.

Yeah. Last question for me, take me on the journey of what's going on with Microsoft. What you predicted, how that's changed, how their strategy has changed. Give me the proper way to understand Microsoft these days.

Yeah, Microsoft's not in the race, bro.

Why?

I know, but they're getting owned.

They have all of the ip. I agree with you. I don't understand why it's not like, oh, 5.3 launches. Microsoft's announcing it the same day and it's actually integrated and people are using it like on day one. It takes time.

Yeah, it's a skill issue.

Yeah.

It's clearly something's going on. And honestly, the thing that makes me most bearish that is the fact that Satya is like, I'm not the CEO, I'm the product manager of Copilot because I'm so boned if I don't get this figured out. Yeah, like you can argue it is the mo. It is now existential. He's decided like, hey, my CEO job is getting this one thing right, otherwise we're screwed. And that is kind of worrying.

It does feel like they could potentially.

Also the pullback at the beginning of last year, kind of the quick pause is now looking silly in the context of Amazon coming in now and saying, yeah, everybody's on board now. Yeah, we'll see.

Yeah, yeah, we'll see. I think they have the most to lose.

What about GPU utilization? Brad Gerstner was hosting CNBC today, which was very cool. Talking about how, you know, in the dot com build out the dark fiber was something only like 7% of fiber that was being laid was actually being used. It was like, obvious even at the time that. And yet now we're seeing GPU utilization rates, you know, maxed out.

Yeah.

Yeah. I think that's a pretty good counterpoint to anyone who's like, whatever. Like, at this point, H100 pricing has massively firmed up. B200 pricing definitely has super firmed up. And like, hey, there's. There's clearly demand. I mean, you know, whatever they're doing on the other side of it, that's like, that's the customer's issues. But, like, I mean, I still think, like, honestly, man, the codex thing or, sorry, my brain's all messed up. Claude code has been the most magical moment in technology for me in my entire time. I think it just feels awesome, man.

Since the Game Boy.

Dude, this is better than Game Boys for me. I'm an information addict, though. Yeah, I am, so.

Well, we appreciate you taking the time to come chat with us and writing about it and everything that you do. If you're listening, go hop on semianalysis.

Sign up for the 10 million a year plan. You get Doug's phone number. You can text him.

You actually can, actually.

I know, I know.

You should just do it, though.

I mean, and it's underpriced. You're giving it away. It's taking away your time from all your different agents. So you gotta price it in.

That's right. You know, my manager will hate that.

Have a good day. Have a good weekend. We'll talk to you soon.

Cheers.

Goodbye.

FIN AI, the number one AI agent for customer service. If you want AI to handle your customer support, go to Fin AI. And without further ado, we have Max Levchin, the CEO of a firm.

Copy king. Welcome to the show, Max. What's going on? Great to see you.

See you guys. How's it going?

Doing great. Catch us up on the last quarter.

Pretty damn great if I do say so myself. 36% year over year growth of sales, volume of our merchants.

First billion dollars in revenue, quarter 1.1 ish.

Insane. What's driving that? Obviously there's some seasonality, but what are the inputs?

End of the year, quarter, some work in sales.

Times happen.

Then Black Friday comes to mind. The outlier grower is a firm card. That thing is still absolutely on Fire. Growing 100% year over year. Both active users, transactions like every metric you can imagine. I think it's growing triple digits. We did a thing in October. We decided we're going to Invent our own shopping holiday. Call it the Big Nothing, because basically everyone got a 0% APR deal. I draw your attention to the fact that when we say it's 0% APR, it's 0% APR and it doesn't change. It doesn't flip. We don't.

We don't change.

No fine print. One of our core values, literally somewhere.

On the wall here.

Yeah.

Right there in the corner it says, go find print and let's stick to it. And so the. It really blew up. We expected it to be big, but it was really, really big. And we had a lot of merchants basically subsidize APRs down to zero for just about everyone. It had a huge impact on our numbers.

Okay, so it was enabled by the individual merchants.

Yeah. So the merchants basically said, hey, you know what? We will pay your interest.

That's cool. I want to get right. Talk about fintech broadly. It's been an absolutely wild week in the markets. I think everybody's been looking at PayPal, just wondering how can a business with this kind of user base, this much FX licenses, heavily regulated, all these different things. Obviously there's management component to it, last.

Thing you'd vibe code. And yet it feels like it's part of that software. SaaS, Pocalypse. But how have you been processing sort of like the AI craziness narrative? It feels like we were saying, like maybe people are giving a little bit too much credit to the AI labs as disruptors. But how have you been processing it?

I don't know if I would throw the PayPal baby with the AI bathwater. Yeah, if you will. I think the. It's always very dangerous to comment on what the market's really saying.

Market's not human.

It all sorts of things, but short term it's a voting machine. So I think the PayPal thing is really not a commentary on software is cheap, software is free. Like if you look at the strategy of the outgoing CEO, who I know pretty well and respect greatly, he's a real bonafide product guy and I should know once it's myself one of those. His strategy was very much build more software, expand the footprint of PayPal into all kinds of really interesting directions. And so if obviously I was not even a little bit privy to any decisions taking place or why did they choose to change management. But if the company believed that software is better, cheaper, easier to build, they wouldn't have changed horses because the previous guy was absolutely beating the drum of let's build software. I don't know why and what they'll do next. But I do agree with you as it is my first child, a little bit sad, but I do compete with it with my second child.

So bad.

Yeah, I guess, I guess. What do you. So there's a lot of AI opportunity with a firm. Opportunity 1 people just understanding more about the financial products that they use in their everyday life and start to choose products that don't have all the fine print. Obviously efficiency in the business is a whole other one. I'm sure that you're getting a lot out of that. But when people ask you about the AI risk to affirm anything on the risk side and then payments companies as well, what is your answer?

So I think I have a fairly basic framework. How I think about sort of AI changes everything. It doesn't change that much. Very broad categorization here. If you're in the business of owning cash flow producing assets, or better yet, manufacturing cash flow producing assets, you're probably going to be okay today. If you're making things that create cash flow, you're probably using software if you're not doing it right. And so if you are and you're using software making software to create cash flowing assets, it's about to get really, really cheap, much faster and a lot more efficient. And we're certainly taking full advantage of that idea at the firm. And that's what we do. We make cash flow producing assets. They're known as affirm loans. There are about 40 odd million of them made every quarter and growing quickly. It is difficult to roll out of bed and say let's vibe code that because you can bycode some of the code. But before you get to our scale or our revenue, you'd have to convince the enormous number of outmarkets partners you need to process on the order tens of billions of dollars of loans. We're not a bank, we're not lending from our own deposits. That means someone downstream is financing these loans. So those are relationships are not coded and they can't happen overnight.

You can't vibe code a relationship, a real one.

You cannot.

Yeah.

So with payments you have, you know, the whole regulatory side too. If you're going in, you're like in your dorm room and you're like hey, I built PayPal. And you go to say like okay, now I should probably get some money transmitter licenses. Like, you know, I expect at some point regulators to say like okay, like what is, what does your operation look like? Do you have any compliance experience? And all these things. And so at some point, and then what you're saying is the capital markets is probably even a bigger challenge. Let's say you can somehow get the licenses. The licenses. Then are people going to trust you with billions of dollars?

Exactly.

Another sort of way of cutting the whole AI accelerates everything. Some things are still taking about as long as they always did. So regulators are not going to be like, oh cool, you're in your dorm room, you've coded a global payment network right on. Here's some licenses. It takes 18 months to get a full complement of MTLs. And by the way, you have to prove that you're a legitimate business and you're not going to be used for money laundering.

And it's very expensive. It's not like a driver's license where you get it and there's no kind of ongoing. It's like you need to be able to carry the weight of, of the licensees.

Yeah. So licensing, regulatory acceptance, regulatory relationships and capital markets is a lot. And it takes a long time to prove that as you put someone else's money at risk, you're going to bring it back with interest. So that's a big part of what we do. Another set of relationships you can't fight code is we have last quarter, 4ish, 100,000 active merchants where these transactions took place. Each one of those is a sales conversation, contract negotiation, going live, et cetera. The going live part is going to get a lot faster. So we could now go to a merchant and say, hey, good news, we're going to vibe code the launch together. It'll be a lot quicker. So instead of going live next quarter, we'll go live next week. But until we get there, someone still has to decide that this is a good contract and prove that this will be accretive to the merchant. And finally, maybe the most important set of relationships is the consumer one. So we have 26 million active consumers in the last 12 months. These are people that trusted us that the no fine print thing is real. It took us 15 years to convince people that when we say 0% loan. The original conversations I had literally with people both in the industry and the man in the street, like, we're going to lend money at no interest so long as the merchants willing to subsidize the transaction, like, yeah, sure, no interest, but there's going to be a fine print. It's going to go to 39% APR, like, no, it really never does, never had, never will. That's part of the core value of this company. Takes a little while to convince people that you're not just nom.

Not a crook.

I think so. I'm fairly confident in our ability to.

Defend our last question. For me, I'm interested to know we're very excited about ads and chatgpt and agenta.com. do you have a feeling not to hold you to something, but just how fast the ramp in agentic commerce and people actually making the decision to purchase in a chat app might be? We talk to the folks at Shopify a lot. We're trying to triangulate this. I thought last Black Friday might be a little glimmer of it. It feels like they're still implementing a lot of things. Obviously there's details to iron out, but it feels like certainly this year could be the time when we're seeing 0.1%, 1%. I don't know, 10% seems sort of high and crazy, but it's where the future's going. But how do you think about agenta commerce numbers this year?

Don't hold me to it.

Yeah, I won't.

The curves are all turning vertical.

Okay.

And so it's hard to say if it's one. If it's one, then it's ten and if it's. If it's ten, it's a lot.

Yeah.

But I think some shopping is going to turn very agentic, as in I will tell my. Yeah, you know, consumer coworker equivalent or you know, fill in your favorite brand, go buy me some milk like the rest happen. Which will be awesome. I will also do that today with Instacart. So the. I want to buy a cool new espresso machine which I'm currently shopping for accessibly. Of course I want that experience to be much more interactive because I'm obsessed with espresso and so entertainment is like nerding out. I'm like, oh, they have a rotary pump. It's not as loud as the vibration one. So I think it's going to become augmented by agents. But it's still going to be like I'm very much a part of this. And that becomes to blend into like, well, research with Google vs research vs Gemini did the same thing.

Yeah.

Which sort of brings me to, I think there's quite, quite a bit of industry pearl clutching around ads and chat bots. Oh my God. I think that's the best thing that's going to happen to this industry. Like everyone is holding their breath around on one hand, will AI kill software? Software eats off or whatever. This is like, well, will we be able to afford this giant build out so these models can get smarter. Guess what? We make most of the money on the Internet. Not us, but. But most of the rest of the Internet makes money with ads. So ads in the single fast growing piece of the Internet wouldn't be such a bad thing for people that need.

A lot of capital.

So I for one welcome the ads and bots so long as clearly delineated ethical considerations, all that stuff. But of course I think it's coming fast and you know, I think so far most people have under predicted the pace of these changes. So I'm not going to claim that it's going to be 10% by the end of the year, but it won't shock me.

Yeah, that would be sort of my high, my high bound. And it'll be hard to measure because with that espresso machine people will clearly do a ton of. I could imagine firing off a deep research report. Tell me everything, understand the landscape, how long are the companies. I want to know the company's history, who's the founder, Tell me everything. But then also I want to go to the website and also I want to go to the showroom and also, you know, and then maybe I make the purchase after talking to somebody. I walk out of the to the store and I say hey, buy that thing, ship it here. Or maybe I do all the research and then I tell someone else to just purchase it and they go buy it in person because I want it the same day. There's like a million different edge cases so the number will be hard, but my vibes based analysis is right in the same target slot as yours.

Last question for me, what's your buy versus build framework in the context of a firm? There's you guys are clearly able to take products 0 to 1 internally quite well with the card, but at the same time it feels like there's so much in the world that's on sale right now.

True M and A wise. We've done a couple of things well and a bunch of things less well. And so we're very, very cautious. So as the world goes on sale it's tempting to look at things that are maybe undervalued, maybe could plug right into the puzzle we're talking about here. Generally speaking, we're super cautious. So for the number of people that run in with like hey, cool, we can buy this thing now, my answer typically is like why would we, would we benefit from this? Like is this, is this better owned by us? Maybe a different version of the same.

Question is well, yeah, yeah, just, just because you might make a fantastic CEO of a company that's on sale. Doesn't mean that it wouldn't be massively distracting if you were to try to integrate it and make it a part of the ecosystem.

Yeah.

And back to sort of like some things are super accelerating and some things are exactly as slow or as fast as they've always been. The human relationship part, where a new team comes in and you're like, I just bought you. How exactly do we get along now? That's not going to get any easier. And if anything, it's going to get. There's a lot of CEOs I know who are thinking it'd be a great time to sell my company. Not because I want to sell my company, but because there's so much opportunity in this AI enabled world. I want to start another one.

I want to start something new.

Greenfield.

That's actually a terrible thing for the Inquirer.

Yeah.

One thing you want is like the typical value of the asset is the team. The team is like, I'm just trying to dump this thing. Like, not sure what I wanted.

Yeah.

Well, thank you so much for taking the time to come chat with us. Good luck. We have to search for the old coffee machine. Let us know what you decide. We'll announce it.

You should have the Max's list. The top rated.

Yes.

Different machines. Anyways, great to catch up.

I'll leave you without. I own two identical linear minis.

Okay.

Just one breaks.

Yes. This is commitment. This is serious smart. I like that. That's peak performance.

I'm a little insane when it comes to coffee.

Well, thank you so much for taking the time. We'll talk to you soon. Have a great weekend.

Cheers.

Goodbye. Let me tell you about 11 labs. Build intelligent, real time conversational agents. Reimagine human technology interaction with 11 labs.

And we know there was some roughness on the audio there. That is why we're making the tvpn.

We're working on them.

The pit lane headset.

Yes.

We're not going to send mics to people. We are going to send people headsets. With the boom mic, it's going to.

Look like you're in the paddock. It's going to look like.

Not in the paddock. No, in the pit lane.

In the pit lane, yeah, in the pit lane. Sorry.

Well, without further ado, I guess we don't have our guests here.

Let me tell you about cognition. They're the makers of Devon, the AI software engineer. Crush your backlog with your personal AI engineering team and I was looking at.

And crushing it in the enterprise.

Crushing in the enterprise.

Yes.

Great. The company's doing fantastically. But Jordan Schneider did an interesting side by side of the of Devin versus Claude code building an RTS a real time strategy game and said congrats to Scott Wu and Russell Kaplan on the decisive win. And he's vibe coding games which I'm excited for these to release. I feel like things are happening so fast and yet I want to play Sholto's game. I want to play more vibe coded games. All of those Instagram reel ads that show you the fake games that are all click bait to get you to buy some, you know, game that's ultimately Candy Crush. I want the actual game that looks fun where you're fighting the zombies or something.

Okay. Some quick context before TJ joins.

Yes.

So he's coming on to talk to us about Trump Rx, which launched, I believe yesterday and then Also this new WeGovy GLP1 pill that was launched and then immediately cloned by Hims and hers. Novo Nordisk put out a press release on the illegal mass compounding and deceptive advertising by Hims and hers. QCAP said Novo's basically calling hims a corner store drug dealer.

Does Novo not believe that imitation is the sincerest form of flattery?

No, they think it's disgusting.

Interesting.

Okay, well, we'll have to Novo Nordisk issued the following statement regarding the announcement by Hims and hers that they will unlawfully mass market an unapproved, inauthentic and untested knockoff semaglutide pill. The action by Hims and hers is illegal mass compounding that poses a significant risk to patient safety.

Harsh words.

Novo Nordisk will take legal and regulatory action to protect patients RIP and the integrity of the US gold standard drug approval framework. And Dr. Marty McCarthy Makari, the commissioner of the FDA, said the FDA will take swift action against companies mass marketing illegal copycat drugs claiming they are similar to FDA approved products. The FDA cannot verify the quality, safety or effectiveness of non approved drugs.

Marty has a great book by the way. You should read it if you want to know more about the FDA strategy. It's a very fascinating contrarian take on healthcare. He identifies a bunch of interesting places where for some psychological reasons or some sort of structural reason, scientists that had discovered some truth about health care or medicine were unable to sort of propagate their finding. And he identifies a number of these sort of bizarre scenarios. So we'll dig into it quickly. Let me tell you about Sentry Sentry shows developers what's broken and helps them fix it fast. That's why 150,000 organizations use it to keep their apps working. And without further, further ado, let's bring in TJ Parker, health care czar to the TVPN Ultradome. Welcome to the show. How are you doing?

I am great. How are you?

I'm good. Complicated day in the news. Where should we start first? What's starting?

Cooled Porsches.

Air cooled Porsches. Are they still cool in the age of AI or am I going to be able to vibe code one?

I think they're going to be cooler in the age of AI.

There we go.

I like it. Yeah. What about in healthcare?

Yeah.

Big week. Want to talk about? Let's just jump straight into the wegovy Novo Nordisk hymns drama. What's been your reaction? How have you been tracking the story up until this week? All that good stuff?

Yeah. I think maybe to jump back to how we got here and I think folks are probably pretty familiar with this, but the hims of the world were obviously compounding the injectable for quite a long time. But the origination of that was that those injectables were in short order, they were on backorder and so they were allowed to compound them as a stopgap for the manufacturers to get sufficient supply.

And that was basically real in the early, early phases of this, Right?

Yeah.

And important because if someone's overweight and they want to lose weight quickly, being overweight is dangerous, healthy, not good for your lifespan.

And so that's why the FDA is set up in that way. But the question is like, how long does that.

Basically, HIMSS got addicted to compounding.

Yeah. And that was maybe call it 18 months ago or so, where that was kind of. That peaked. And then the FDA did issue a warning letter to HIMSS back in the fall when the shortage ended, telling them to stop mass compounding. So they have commented on this before, but their level of enforcement has been lacking. I think the stark difference with what happened yesterday is that the WeGovy pill has a specific technology called snack that they actually paid $2 billion for the IP specifically for that technology to allow for you to be able to actually absorb the peptide and have the drug be effective. And hims explicitly I say that they didn't use that technology and said they're using liposomal absorption, which maybe it works, maybe it doesn't, but there's certainly no.

Well then it's just like insane. It's, you're insane, you're Selling a drug that you claim works, that has no, like, approval. Approval testing. Right.

Zero studies, zero approval. So this is where it's super different from the injectables, where they weren't changing the formulation. Right. It was very similar to the branded drug.

Yeah.

We can talk about the challenges of IP and a bunch of things, but it's different in the sense that here there's no evidence that this drug works at all. And so the most likely scenario is they're selling a, you know, a $50 drug that doesn't work, that still has a bunch of side effects because it still has the drug that just doesn't get absorbed through your gi. So it's. It's very different. It's not an IP issue. It's really an efficacy and safety issue, which is why I think you saw the FDA come out much faster in a much more explicit way pretty immediately after the launch yesterday.

Why would HIMSS do something that seems completely nonsensical and bad for patients? And bad.

I have a quote here from them. Compounding is a safe, legal, and long recognized practice within the American health care system. So that's their stance.

Okay. But ignoring the fact that they're inventing a new delivery mechanism via liposomal and are selling it as a product, claiming.

That'S why they're getting pushback.

Yeah, yeah.

I mean, I think the situation is you have a $5 billion market cap company where half of their revenue is these compounded GLP1s, where that market is quickly contracting on them. They're doing a bunch of buybacks. They've invested a ton in incremental capex. It's like a recipe for disaster. If that revenue starts to fall away and it feels.

Do they think the pill is just going to. Is the expectation that the pill is going to effectively eat the injectable market? Because a lot of people would rather take a pill than have to do an injection, even if it's once a week.

Yeah. I think if you look at the archetype of who I believe to be the HIMS customer, they're buying the injectable because it was the cheapest place to get it. Now you have a branded drug that's half the price of their injectable. I think it's partially the. The route of administration. I think partially that it's just half the price. And so it's hard to imagine that they're not going to have a significant issue with churn and people switching from the injectable partially because of the ease of use. But Also just literally half the price which was their entire value.

Prop.

Novo is not expecting. Are they able to mass produce this already?

They have no shortage issues. They've been ramping up production for quite some time. I think they were quite prepared for the launch. And so you have none of the shortage dynamics. And frankly pills are exponentially easier to manufacture than their injectable pens specifically, which is where the shortages came from. So there's no similar rationale. I think it's also comparatively reasonable that you don't want to force a bunch of patients off the injectable product. So there is a reasonable safety argument that that has to be a very thoughtful and calculated wind down. I think they're onboarding new customers. That doesn't totally hold water compared to the behavior. But in this instance it feels more like a financial Hail Mary given I think what the market dynamics are going to do to them over the next 12 months.

What about the alternative scenario or playbook? I remember himss as being a front end to drugs that they didn't compound and it was a place to get hair loss medication and sort of name brand drugs that I felt like just it was online pharmacy. Like why doesn't that model work in the age of peptides and GLP1s? Why can't you just be the digital doctor that prescribes then you go pick it up at your local pharmacy?

A couple reasons. I think if you look at their competitive set, the biggest one being ro. Obviously that's how they've managed through this, right? They have not remained reliant on compounded drugs. As the shortage has gone away, they've ramped up their access points to branded therapies. That's exactly what they're doing and executing. I think when you're in the position of hims, it's hard to not get addicted to a $250 product with 90% gross margins when if you're selling access to the branded drug, at most you're going to make 10 to $15 on 100, $150 product. And so I think they to some degree don't realize the business they're in. Like in a traditional retailer and distribution business, which is really the business that they're in. They're not a manufacturer, that's a scale low margin business where price, convenience, like all the things that make Amazon successful, make you successful. They have sort of morphed into a manufacturing business without any of the ip, any of the R and D costs, like any of the things that come along with being A manufacturer and I think it's an addicting place to be because the margins are quite a bit better than being a retailer.

Yeah. Can you explain more about this? The warning letter that the FDA shared addressed to Hims and her health 9-9-2025. You can find this on FDA.gov they're saying it says the FDA has observed that your website offers various compounded drugs including semaglutide as described alone. That your claims are concerning, concerning your compounded semaglutide products are false or misleading. How would a company normally respond to a warning letter from the fda? It's not fully shut down, it's not a lawsuit, it's more of a corrective by the fda. Is that correct?

Yeah, I think historically in combating specifically the FDA has lacked material enforcement. And so in many ways I think what you're seeing here is that there's a skepticism that they'll actually enforce and so it's business as usual. Don't take the warning letter seriously. I think with that overlay, I think the reason the FDA still sends letters like that is if they do have a material safety issue, which is what has happened with compounding in the past. You had 60 people die of meningitis in 2012 from a compounding pharmacy doing scale compounding. They have a paper trail of telling them that you are out of compliance. And so, you know, I think this will probably be a different situation. I think one important point is the FDA does not enforce ip, so they're not going to get involved in actual quote unquote IP theft. That's for the court. Got it. But they do enforce safety and efficacy which is why I think they address this so much faster because they believe it's an efficacy issue, not solely an IP infringement issue, which would again be left to the courts.

Yeah. What about Walmart? Are they a potential way out of this? You need scale. I don't know how advanced they are, but you have experience with Amazon. Walmart just hit a trillion dollar market cap. It seems like the business is doing well. They were not disrupted. They need more technology. Is there something where a partnership could allow hims to sort of go get out of the compounding business and go more to the scaled low margin business but still be growing.

Yeah, it feels like just given the reputation that that would be hard to imagine happening. I think they have built obviously a lot of compelling customer experience experiences. They build really good tech. But I think just given the overhang from these regulatory challenges, it's hard to imagine a large acquirer is going to get comfortable with that kind of retrospective and presumably ongoing liability.

Yeah.

Switching gears, Trump rx. What's up with that?

Yeah, I feel like I should take a victory lap here. I think the first time I was on here was to talk about the executive order for Trump rx. And I think my point at that moment in time was that the real thrust of that eo, which was not super clear then, but is how I interpreted it, is that it was an effort to eliminate this discrepancy between super high list prices and the net prices that insurers and PBMs actually pay. Right. So for these GLP1s, they had these thousand dollars or twelve hundred dollars list prices, but in reality, if you were a PBM, you were paying 200 bucks, 300 bucks. I think that's exactly what you saw here, where there's a number of branded drugs which before, if you were a consumer paying out of pocket, you had to pay that high list price. They've gotten the manufacturers to come to the table and offer a consumer directed net price. So really, if you are paying out of pocket for a bunch of these drugs, you're saving 60, 70% compared to what was available to you a year ago. Now, some of that could just be the GLP1 market forces, but that wouldn't have touched a number of the products that have moved onto TrumpRx. I think given the selection, it's maybe like 25 actual brands that haven't gone generic, that have low net prices that are on there right now. It's not the most exciting thing in the world, but I do think you're going to see manufacturers just start to launch new products at net and that will completely unwind the value of the current supply chain middlemen, PBMs, all those things. So I think it's net good, but I think it's the early days of what could materially change how these things go to market and the pricing that's associated with that.

It's not just good marketing for the midterms?

No, I think it is good marketing. I think everyone has quoted these ridiculous list prices as the actual price. It's been a pet peeve of mine forever and they are now quoting much lower kind of list prices. So I think that is going to be perceived as a win for the midterms. But the drugs aren't actually cheaper than like a payer would have paid a year ago, but they are cheaper than a consumer could have gotten access to a year ago. So I think it's net Good. And I think hopefully a trend that will accelerate and change the pricing dynamics more dramatically.

Ask about AI doctors.

That's what I was going to ask about, John.

The same thing. We've talked to a couple companies that have come on and talked about consumer level AI doctor, doctor in your pocket, you got a mole, you take a picture of it, it tells you what you say, I feel sick. It'll tell you, go get some blood work done or something. Some of them have partnered with certain states to get real doctors involved. There's a whole bunch of different models at the same time. They're facing pressure from OpenAI and they're launching a health product. How do you think that market plays out? What are you optimistic about? How much can it actually improve American health care? There's always a lot of talk about, like if you just talked, you know, primary care doctor was available more frequently, we'd be way healthier. And I don't know that there's that many people that are like, well, I never thought to Google I should sleep, diet and exercise. So now that I can chatgpt it I'm healthy. But walk me through your whole philosophy on like, like AI chatbots in consumer healthcare.

Yeah, I mean, I think first of all it's obviously goodness, right. I think people are really going to GPTs to get advice on their health and I think it's something like 7 or 8% of ChatGPT's volume is health related, health specific. So clearly that's net good. I think in many ways it replaces people going to Google to do the same thing. It's much more efficient, it's definitely more effective. So that's all good. I think my thesis on the end state of that and where it really unlocks a bunch of value is when it is fully aware of all your medical history. And so that's pretty critical. And ChatGPT has launched a version of that, though it's pretty light on the actual data integration.

And then more specifically they acquired some company I think to help with that. Wasn't it that it was a. There's a company. It wasn't called. It was a company that was like trying.

Torch.

Yeah, Torch. It was the. What was the name of the company they had before? It was not Future Forward Forward Health.

Yeah.

Yep.

Yeah.

So clearly they're thinking about all this stuff.

Yeah. There's a fundamental regulatory framework where if you're a provider, you can access a much broader set of data to deliver care to the consumer. If you're a technology company, they're accessing via this thing called tefca where very few providers are contributing data to that. There's a lot of nuance to this, but the punchline is that if you are purely a technology company, your access to the breadth of data necessary to be useful is much more limited.

So you basically need to be a provider that employs actual medical doctors.

You need to be a provider actually providing care to that customer.

Yes.

Yeah.

And so we had one of these companies, Lotus, on the other day, and if he's in an investor pitch and people are like, hey, like, AI doctor is cool, but ChatGPT is doing this, Claude's doing this. There's other companies that are. Or Claude might do something like this. There's other companies that are super competitive space. They're saying, okay, but OpenAI is not going to have doctors on staff and it's not going to actually be licensed and able to actually deliver.

Yeah, it doesn't prohibit them from doing interesting things with partners or via MCP or some other approach that can mimic, you know, some of the things that we built at General Medicine and some other folks are building. But in and of themselves, whether It's Claude or ChatGPT or Apple Health, any of these guys, their, their ability to provide similar levels of care is just going to be constrained. I also think, like, the, probably the more salient point is, I think where this becomes really magic, is where all of these conversations get actually distilled to your provider before your appointment. So instead of having to repeat all the things you've already said and like, go back through your medical history, like, you're all, you sort of show up to the appointment, your provider's been prepped based on everything you've already chatted about and can actually start to like, deal with what they might want to do in that appointment. So that's where we've seen it be the most interesting and useful. And I'm sure other folks will end up in a similar spot. But I think like, the combination of GPT all the way through to the visit, informing the visit and then informing the care is super powerful.

What's the textbook example of like, medical history being valuable? Because maybe I'm just like in a weird spot, but I usually just say, like, no, I don't haven't done anything but like, I don't know, like if you, like, I don't know, I actually haven't broken any bones. But like, assuming like, I was like, I broke my arm when I was five, like, does that really, really, like increase cancer risk? Like how are these things actually flowing through? Like, what's the textbook?

I think the average American is on prescription medication.

Okay. So they show up and they're like, oh, you're already on this medication.

Why did you.

Yeah, I mean, I'll give you, like, a tangible example that, like, Elliot experienced in our GPT, which is he's been having this plantar fasciitis issue on one foot. I hope he doesn't mind that I'm sharing his phi.

Yeah.

And the GPT was able to show. Actually, yeah, the GPT was able to decipher that at the same thing on the other foot in, like, 2006 or something like from, like, 20 years ago.

Okay.

Which is where, like, you just. He didn't even remember having the issue. It was not relevant to him.

Okay.

And I think there's more, like, substantive examples where you might have had a series of lab tests over the last five years. Any one of them is not problematic, but the trends that are being exposed, when you can see that trend, is pretty compelling.

Okay.

There's also just, like. You can recommend all sorts of screenings based on when you've actually done those screenings and what hasn't. Hasn't been done. That kind of thing.

Yeah, yeah. No, that makes sense. Jordan. Anything else? Is it dangerous that OpenAI hooked a biolab up to?

We're going to have the founder of Ginkgo, who they partnered with on the AI biolab. But what are your thoughts? You want to vibe code some drugs, some studies.

Seems like a fun experiment, you know, but what could possibly go wrong?

We'll clip it when it happens, when the world ends. The one final clip.

We're like. TJ called it.

TJ called it. Anyway, yeah. There will be victory laps if the apocalypse. Apocalypse happens for sure.

Last.

Last question. How have you been processing the. All the EV manufacturers having to basically take these massive impairments, or, sorry, just car manufacturers in general, taking these massive hits. It was. Stellantis is down something like 20% after the company announced it's taking a $26 billion hit. And writing down their electric vehicle business. Did you. Did you predict this as well?

You know, I'm. I'm still loving my Raptor R, which gets about 11 miles per gallon. So I think you know how I feel about the situation.

That's great. I'm trying to get Jordy to get one.

You're buying your. You're buying your carbon. You're buying your carbon offsets, though, right? Yeah, of course.

Yeah, yeah, for sure. Yeah, for sure.

Thank you.

Awesome. Well, thank you so much for coming on the show.

Great to catch up.

Great to catch up.

I'm sure there'll be a lot more here, so we'll have to talk to you soon.

Pleasure as always.

Cheers.

Have a good weekend. Let me tell you about Crowdstrike. Your business is AI. Their business is securing it. CrowdStrike secures AI and stops breaches. And without further ado, we're going over to the co founder and SVP of Field Engineering from Databricks. Arsalan. Welcome to the show.

How are you doing?

What's happening?

Hey, guys, thanks for having me.

Good to see you.

Great to meet.

First time on the show. I'm believe. Please introduce yourself, tell us what you do.

This is when we say first time tuning in, longtime watcher or lurker or something like that.

I appreciate it. We enjoy talking to your co founder.

Awesome. Well, yeah, guys, Arsalan Tapakoli, one of the co founders here at Databricks, run what's known as field engineering. So think about anything technical, not in core R and D. So it's about a third of the company, roughly, give or take. And then before that, like all the other founders came out of Berkeley. And as my advisor likes to say, I joined the dark side in between him meant I went to a consulting company. McKinsey spent about four years there before we started.

About 12 years. There we go. We love, we love, you know, we love to hear that. Put in your time. We were, we were on a flight the other day and. And John and I were asking this guy to switch seats with us because we like to. We like to. We, we. Some people are surprised, but even after three hours of recording, we still want to keep talking. And the guy had a UBS shirt on. And John was like, do you work at ubs? And he's like, yeah. And John goes, thank you for your service. And the guy just started laughing.

He thought I was making fun of him.

Thank you. Thank you for your service.

At McKinsey, you gotta thank your bankers, your Swiss bankers. Specifically, I want your reaction to this Tyler Cowen post today. We'll go down as some kind of turning point point somewhat arbitrarily, but it's okay if journalists and historians have to present things in that manner. Posted this at 11am on February 5th. That, of course, was the day that both Codex 5.3 and Opus 4.6 launched. It feels like he's talking about software engineering, automated software engineering, AI agents, sort of everything coming to a head. How have you been processing this week? To does it feel like there's a turning point? Is this all part of some smooth curve? What's been your view on the ground?

Look, I think I've lost track now based on those definitions, I think we've had like 3,000 watershed moments in the past year that everybody said is like the turn.

More than. Yeah, more than one. Way more than one.

That's why we have three hours of content every day. It works perfectly for us.

Every day I tune into X and somebody's like, this is the day, this.

Is it, the singularity is here, it's over, we're back.

Look, I mean, I think it's more of a smoother curve right as we go through. And clearly every day I think it's a combination of the technology is getting better, but I think a lot of it right now what holds people back, what they can do, is less the technology, but it's more people learning how to use it, how to deploy it, all the things around it to put value. Like that's going to be the long tail. I think we're not even close to harnessing the technology that we have today. Much less saying, hey, it got a little bit better. Now everything's going to change, at least from what we're seeing.

Yeah. So talk about your forward deployed engineers, whatever you call them, like implementation, how you like to work with companies. And then I want to compare and understand what it might look like at the labs as they sort of dip their toe into the FDE model.

Yeah, look, I think so. I think that there's two things a lot of places when they talk about forward deployed engineers. So this is the cool term right now, I guess to use is this concept that says go to ask people, what is it that you want? What's the outcome that you want to drive? You want to do, you know, revenue forecast optimization, you want to do demand forecasting, and then I'll build something for you. The problem is many of the organizations just can't leave the talk about that are kind of like, okay, I'll build something custom for you, I'll get it fast. But then the hard part is, is that scalable? Is it open? How does that work? So databricks is probably a little bit different than, than the rest of them because we've got an actual platform and product that's used by over 20,000 people. But I think right now, when you look at it in the AI world, what works for us at least is we go in and some organizations right now are saying, help me with a Specific problem. Like I want to do, you know, let me get my data in order, which is important for AI. Or let me kind of do a migration or let me set this up. And then we have the same set of somebody says come help me like transform like Fox News. If you know Fox Cletus, the sports app that they have that you can ask questions like that one. They were like, come build something.

Paul at Fox is a good buddy, I don't know.

So he's like come build it. Soup to nuts for that. And that was one that we did everything end to end and deliver something that works that they want. So we span the gamut from helping you work with products to kind of actually delivering outcomes. We do both.

Okay. And then best practices pretend I'm CEO of AI Lab who will remain nameless and I'm trying to hire 700, a couple thousand consulting type for deployed sales engineering roles. What are best practices? How fast can you actually spin someone up to become an effective forward deployed engineer? What are the pitfalls?

Yeah, lots of questions embedded in there. So look, I think first and foremost what you're looking for today for people is they need to do two things. One, they actually have to be able to build hands on. I think you end up in the solution engineer space getting a lot of what I call like the slideware demos. Like hey, I'm a walking dictionary. I can tell you about things I can build a deck. It just doesn't really cut it anymore. So one, there's much more that what people say is come show me rapidly prototype. But they have to be able to build and they have to be able to to use AI tools to build, otherwise they're going to be slow. The second thing is you can't just go get engineers. They have to be able to talk to customers because like the most important thing is if I go walk to your basically CEO or cio, I have to be able to say what's your business problem? Understand why it matters and decompose that into technical pieces. So you need people who have both of those. So you have to screen from it. It's different than what the, what I would call the traditional solution engineering interviews are in terms of how fast to get somebody ramped up. Look, you're hiring smart people generally to get familiar with all the tools. We'd say three months is fast, six months is when they're operating at full steam. It doesn't mean that you can't drop them in front of customers in the first couple of weeks, but that's the amount of time that they get. Like, what are the most common problems? What's the most efficient way? How do I navigate an organization? And you can hire a lot, but absorbing too many people too fast just means you're also going to candidly degrade culture and skill set.

Yeah, we were just talking to Doug o' Laughlin about how he's using Claude code. He's vibe coding all these things and a lot of what he was describing were sort of like one off data analyses or smaller projects that could sort of run locally, be compressed like spreadsheet level, certainly not big data cross, you know, like databricks level infrastructure required. But I'm wondering if you see a world where an individual would vibe code something that's so big and processing so much data that the agent or the individual vibe coder is actually pulling databricks off the shelf. Like do we, do we go to a world like that? I've had ideas for vibe coding projects that have been like, yeah, sure, I'd love a personal assistant that read the entire Internet every day, but that seems like not on the table right now. But are we moving to a future where that's possible?

Yeah.

So two things I think. One, we are. But second, like I sometimes say from a databricks perspective, I wish I could walk in, you know, in like Men in Black where they have that button that causes people to.

Sure, sure, sure. Yeah.

I think that there is this vision of databricks of hey, it's really, really good if you have a bunch of data and stuff like that. But yeah, but actually like, like a bunch of our users today, what's really valuable for your use cases, Even if you say work for a spreadsheet, you're like, I don't keep that. Who keeps anything on their local laptop anymore? Right. It's like I work in an organization. I just want a place where I can access all the data, big or small. But it's governed, it's high quality. So actually a lot of the people inside of databricks use, I mean called genie. Right. Because you can ask for it is mainly like, can I get a world where I can do exactly what you said? I can quickly buy code what I need? Yeah, small amounts of data, but it's governed. I can visualize it, I can iterate on it. So that's the first thing I put. I think what you start seeing is every. Everything that you start with. For like any project, it starts small. Great, I want my data. Then your next thing is, oh, but it would Be really cool if I could merge it with this and it would be really cool if it could do so as you get it to work, it naturally balloons into something that both leverages more data, is much more computational and much more comprehensive. So you very quickly end up in a world where you've outstripped the. Can I just do it on my laptop? Even for the average user, yeah.

We were talking to Sam Allman yesterday about different bottlenecks and he was saying that data was a bottleneck for a while at least, like a rumored perceived bottleneck. It really hasn't seemed to bottleneck progress in any meaningful way. But I'm wondering if you see, if you look at the growth of the data market or data is the new oil just Is there a boom in the scale or importance of data? With the NEO Lab boom, there's more and more companies that are training different models. It feels like there's both more data being collected than ever, but then also more data being generated than ever. So can you just talk about zooming out? Like what does the growth of just data look like in the US economy these days?

Look, I think the growth of data is significant, right? I mean there's no question about that. And look, you said data is the new oil. I think one you've got to define, you've got to separate out the world of the consumer world, which I think about where many, many of where many people's familiarity with AI is. Right. I want to ask a question, help me understand, you know, is this a cold or an allergy or plan my travel vacation. And that's the. I need the whole corpus of the web and that people, that amount of data is kind of finite, if that makes sense. And new people are generating new ones through AI. But that's one. I think that there is a separate question of why you said most people focus on data in the enterprise world. It usually goes the following. I need, I. I have to pick a model. Let me pick. Which model is it? Opus 4.6. Is it the new basically GPT model is a Gemini, etc. Great. I started crap. I. Now I have to figure out how to deploy this. And that's a lot about security evaluations, governance. Great. Now I did this and my accuracy is 60%. 60% is kind of crappy accuracy. How do I get to 100%? The answer is you need to have much better data. So almost everything comes back to, you know, especially in enterprise, lots of data spread out in a bunch of places. How do you actually get it together? Know what you have, make sure it's high quality and relevant and recent because that's the other thing. Thing that's happening. Data that you had from six months ago may no longer be valid right now. So how do you make. Tell the system what's valid, what's recent? That's where we spend a lot of time on, and I think enterprises spend a lot of time on figuring out like, how do we actually make these agents work? And it's a different problem than it is for like general consumer search.

Yeah.

What is, what's your software buying framework these days? What do you mean? Like, what do you decide? Like what. When you're evaluating, hey, there's a job to be done, or should we, should we go and find a vendor that can do this? Should we do it ourselves? How do you, what's, what's your line?

Look, I know that the cool thing is to say this following. It's like, guys, why would we buy software anymore? Like, you can just buy code, everything. I think anybody who says that generally is not in the business of actually using software in production. Like, it's like, build is easy, maintaining and sustaining it is hard. I think the short answer is in a world if I can, I much rather buy and build. Right. Like, that's just the reality of it. The, the problem though is you need, you have such unique needs right now. And so if I look at it and see that the gap between what I can get off the shelf and what I need is really huge, then I have to look at, do I really want to depend on another vendor for kind of being able to like, how busy business critical is it for me? So as an example, much of the things that we're doing to overhaul field engineering at Databricks to say, how do we use AI to accelerate it? There's nothing off the shelf that gives us what we need there. Right. If there's a piece, I'll use it as a component, but we have to go build it. Because for me, it's like I'm iterating really, really fast to figure out what the job to be done is, how to deliver it. So we have our own development team, but on other things like my CRM or something like that, that is not something that if I can help it, I'm looking to build internally. Like, I'm happy to use something externally, given that I imagine my needs are similar to many other ones and there should be things that could meet that needs pretty quickly.

Yeah. How big is AI agent adoption? Broadly? I want to Know the details of the high level of the 2026 state of AI agents and specifically like how much of this is like, yes, they check the box we're using agents, but it's just a test. Or it's very narrow. Or it's the one vibe coding intern that builds some fun workflow that's like, yeah, it's a bot. It replied, it's agentic, but it's not the core business. Talk to me about the rollout and adoption of AI agents and real enterprises, real businesses.

Yeah, and I know I'm supposed to give like the party line and the popular thing, but look, I think you can look at that report that you cited talks about something like 95% of people still have a hard time getting it in production or at least getting value out of it the way that they wanted. I think you have still an element of people who are like beating their chest. You go like, I have a thousand agents out. And I'm like, okay, how much? How many people use them? And they're like, why does that matter? I was like, why does anything else matter? Right?

Yeah, it's the same thing as like bragging CEO bragging about headcount. And it's like, wait, what's your revenue per employee? They're like 30 grand. Like, okay, maybe, maybe you're focusing on the wrong thing.

Yeah, yeah, well, look, I. Look, so I think that there is. Everybody we talk to is trying to. And there's some statistics you saw in there that make sense. Like most people are no longer doing like, hey, let me get a chat bot. It's like multi agent orchestration, right? Like they're using multiple things. Second, nobody is wedded to models anymore. I think like 77% of our customers use multiple models as well together. And I think that the most popular use cases, you know, beyond internal use case, something called like market intelligence. So I. How do I figure out what my customers are doing? What's the next best action I should take in the. Like, I do think it's still early. Like most people, like, it's rare to see somebody say, I have a lot of agents in production really effectively. As I mentioned, the most. The place where we spend a lot of time on is how do you govern it and get security? How do you get really, really high quality out of it? How do you balance cost and quality? And that's where like our whole offering around agent bricks is mainly around how do you actually get this thing in production? And so early days, we're seeing, we're seeing at least people turn from I spun up a chat interface, I have AI to like what's the roi? I can drive from it and realizing what are the blocking factors? Data or some of the things I talked about to get there.

Given, given your experience at McKinsey, there's been some news recently that Anthropics working with I believe Accenture OpenAI is hiring consultants internally. Obviously the big consulting firms want to get in on the action. They've been doing a good job making money on generative AI, maybe making more profit than a lot of companies. But when you think about swarms of consultants going out into the world to unlock the power of agents, what are you going to be looking for to see their success?

Look one, I think we've gone through these iterations before and in general I'm hard pressed to believe that every enterprise is going to all of a sudden have an army of their own really, really effective vibe coders. Right. That are going to build everything themselves. I think that there's still going to be the dependence on external parties to help make something real. I do think that there is a element that the SI firms right now are trying to figure out how do they reinvent themselves? A non trivial number of what it was before was how do I do some more, let's say a little bit more of a commoditized task where SI.

Is Software integrated system integrator. Sorry.

Like when you think about the accentures, the TCs, the cognizant and the stuff. Sure, you know, so it's like how do I throw a lot of people to do it? I think the expectation from people now is going to be great. I'm going to hire you. But in the element of AI, I expect you to be able to do things faster and I also expect to be able to get out of more out of each of the individuals. So one, go build applications for me to do it faster and three, you know, your cost, your economics should be better. So I think that there's going to be a push on doing things like that. But otherwise we're still going to need worlds of folks. They're also trying to figure out how do they expand their teams with like many, many more agents that can work faster. So it's more of a shifting the model in economics but you will still need consulting organizations to help.

Well, thank you so much for taking the time. Jordy, do you have anything else?

Great to meet you. Come back, come back on more this year.

Really helpful.

Yeah, you guys have such insane visibility into how into the at real diffusion. So it's awesome.

Well we can just wait a week.

We can wait a week. Watershed watershed moment every hour. Every hour of every day.

We'll chat with you about it. Have a great rest.

Great to meet you. Thanks for coming on.

We'll talk to you soon. Let me tell you about Label Box, RL Environments, Voice Robotics, evals and expert human data. Label Box is the data factory behind the world's leading AI teams. We got a bunch of breaking news. Breaking news show out with breaking news.

Hobbit Inspired startup becomes first New bank green lighted 2.0 One of the most insane headlines. That one is for Trevor Palmer and the whole team.

Erebor founder Palmer Luckey was one of the tech industry's early Trump supporters and he's known for pension for Hawaiian shirts. They're really focusing on Palmer's like non banking relationships here.

Palmer is a banker now everyone.

Erebor will cater to startups and high net worth individuals. On Friday, it became the first newly created bank to receive a national charter under the second Trump administration. Launching with $635 million in capital, it says it will occupy a hole in the market left by the collapse of Silicon Valley Bank. The bank is the brainchild of Palmer Luckey, one of the tech industry's earliest supporters of Donald Trump. Its founders include. Its funders include lux Capital, Andreessen Horowitz 8VC, Allard Gill and Founders Fund has lined up a handful of potential defense and industrial tech focused clients, including ones with Tyill and other investors who say they are eager to do business with a bank that understands their needs.

You can think of us like a farmer's bank for tech, said Lucky, who will serve on Airborne's board but will not have an operating role. I think most farmers banks won't claim they are the best bankers in the world, but they do understand farmers.

Chris Power's in here, he said. Is a local bank going to lend against you? No, he said. If you have less than $10 million in revenue and you're struggling to secure a loan from a traditional bank, investor expectations are high. Erebor was valued at about $2 billion in a funding round last year, over seven times its book value, according to investor Pitch Deck. A subsequent round valued Erebor at 4 billion. So off to the races.

The timeline, the pace from the team getting this done, obviously there's, you know, the regulatory side, but there's actually the execution to get here. I remember talking to Trevor last year, Trevor over at Airborne and he was telling me the Timeline. I was like, yeah, that sounds great. Like, let's see, let's see how it actually turns out. And that, you know, incredible pace. So yeah, we'll have to have Palmer.

On the show to talk more about it.

Real bank.

We have another gong head for Jennifer Garner's Once upon a farm IPO'd today. Shares popped 17% in public market debut. It's on the New York Stock Exchange, baby.

There we go. Interesting. I wonder how they priced it.

Yes. So they raised almost $200 million, 1 97.9 at a valuation of $720 million $4 million. And the company's IPO comes as shoppers and policymakers alike have pushed back on ultra processed foods, particularly when consumed by children. So it's on the New York Stock Exchange now under the ticker ofrm. CAVU Venture Partners Consumer Partners. Kavu Consumer Partners was a big backer of this company and probably has done quite well in this IPO and also Techno Chief. Had a little bit of an update here. Super random, but Kavu Consumer Partners just raised a new $325 million fund for better.

Good timing. Hit it again.

Rohan Oza over there, he's been investing in the category. Worked in vitamin water. He's an absolute dog and an industry legend and we're happy to see success in this industry. Certainly AI resistant. It feels like something Tyler won't get out of bed for. He's. He's knocking it but you know, certainly opportunity.

No, I mean in a post scarcity world, every, every child in the world could be enjoying once fun day at.

The New York Stock Exchange for everyone.

Do we have anything else?

Not, not too much. There's a bunch of stuff. I think we went through most of this stuff. There's new data from Ramp R Karazi and shares. The fastest growing Companies on Ramp. AI infrastructure means open source is likely ticking up. AI dev tools are not going away and AEO is the future of marketing. He calls that tri. Profound. So Ramp's top software vendors. The fastest growing. We're anthropic. Lovable. Cursor 11 Labs. Superbase Replit. Interesting cursor has been seeing. There's a bunch of FUD on the timeline. People are saying I don't need it anymore. But it's still growing according to Ramp. A little bit of a narrative violation there and then trending breakout growth relative to size. Paper for software design. Cerebras for AI infrastructure. Juicebox for recruiting. AI runware for AI infrastructure. Clarify for AI infrastructure. Nova to AI for AI infrastructure. Crusoe for AI infrastructure, Modal for AI Infrastructure Infrastructure and then peak AI for SEO and AEO intelligence. And also profound. So this is from the RAMP Economics Lab. Ramp.com data. You can go check it out for yourself. We love chatting with Ara about what he's seeing.

Well, it was a brutal week for tech, but at least we had an opportunity to put on the white suits. And I can't wait for Sunday to see the ads.

Yes, it's gonna be a great one. The super bowl series.

John and I will be there with Ramp and some other friends and we will be glued to our phones in the back of the box just watching the ads. Yes, we hope you guys plant the bomb.

In other news, after 12 years at the Wall Street Journal, Joanna Stern is launching her own consumer tech media company. Congratulations. To join us there. And she hasn't shared exactly what it is yet, but there is a link. You can go sign up thisismynextthing.com Good name. I like it. She has been writing for the Wall Street Journal for quite a long time. I've always enjoyed her consumer tech coverage and of course we'll have her on the show when she launches. Leave us five stars on Apple Podcast and Spotify. Sign up for our newsletter.

We love you.

Goodbye. Nice work, brothers.

I'll see you on the next one.