TBPN - Air India Crash, Internet Chaos | Delian Asparouhov, Saif Khawaja, Henri Stern, Chris Paik, Joseph Cohen, Zachary Perret, Shirin Ghaffary, Cameron McCord, Anil Varanasi, Shreyas Iyer

Starting point is 00:00:00 You're watching TBPN. Today is Thursday, June 12, 2025. We are live from the TBPN Ultrodome. We're back in the Ultradome, the Temple of Technology, the Fortress of Finance, the capital of capital. And we are both deathly ill today. But we're still doing the show. The show must go on. That's part of show biz. It is. The show must always go on. And there's no greater show in the world of technology and music than the Coachella 2027 lineup featuring fast takeoff scenario neural mirage causal loops the alignment choir DJ token drift auto prompt jordi had chat GPT generate a a hypothetical cochella poster it was no surprise to see gentle singularity headlining too.

Starting point is 00:00:57 Yes, Von Neumann and the tilinges, Lofi Oracle, synthetics. Which one did I really like? The copy bar is very good. Chain of Thought Collective. Some of these just sounds weird. I think a lot of people were surprised to see the gooners didn't make the list at all.

Starting point is 00:01:13 They didn't. But maybe that's a little white pill. Yeah. Let me try and adjust this microphone. Anyways, we got a great show today. Yeah, we have a stack lineup. Can we pull up the run of show? We're going to have Delian join.

Starting point is 00:01:25 Back to back. think he's the first guest to ever go back to back to back he was on the show yesterday jordy absolutely blasted him with some confetti and um i coming back to talk to us about the shinkai deal as well i really didn't hold back i really didn't hold back yeah i i i tried to not shoot confetti in the face of anybody yesterday uh but delian uh really you know got got the uh the full effect but yeah a lot of fun at yc demo day yesterday tons of great founders uh major white right pill being there. That's great.

Starting point is 00:02:00 And we have some posts reacting to YC Demo Day. I like that Joanne from Ramp, he's been on the show many, many times with. Why did he not come on the show yesterday? I know we should have. I saw his post after we left, but he had a really funny one. He said, you're more likely to raise a Series A as a YC company than go bankrupt as a gambling addict, let that sink in. Because there's this viral post and says, like, there's a 20% chance of gambling addicts going broke. and everyone was meming it like a couple days ago being like,

Starting point is 00:02:29 Gambling Alex will take that. Don't tell a gambler. 80% chance I don't go bankrupt is great. But Gary Tan says 45% of YC companies get to Series A and median ARR is $1 million. Plus trending up. Find me another place where that is true in any set of potential investments in the world. YC does have higher valuation caps. But to me, this quality justifies it.

Starting point is 00:02:54 Invest in YC Demo Day. I recommend it. it astonishing yeah and yeah it was fun we handed out a bunch of ramp hats they're in hot demand I thought we ran out crazy with the color like when I think of ramp I think of like a bright yellow not necessarily a full neon yellow but these hats were full neon green like safety runner green or yellow actually and they're very intense but I got a lot of text messages for people that wanted them so we will probably be printing more of those we had a lot fun we handed out some of our TPPN hats which are now printed they're here they're

Starting point is 00:03:27 finally here we got to figure out how to distribute them yeah we don't want to be in the apparel business yes yes but uh we're here reluctantly figure out how to get them out into the world and so if you're looking for one of those ramp hats you got to be a ramp customer for sure so you got to go ramp dot com time is money save both easy use corporate cards pay bill payments accounting and a whole lot more all in one place and Eric ramp former YC founder yeah YC that's true that's true And Sue Hale also chimed in. He said so many insanely great YC companies.

Starting point is 00:03:57 This batches easily has many $10 billion companies. What a time to be alive. Many. That is so aggressive. I love it. I love the optimization. People were upset. They said mostly cursor for X.

Starting point is 00:04:09 He said, nah, there was a lot of agent infrastructure stuff. There was agent for agent products. Yes, yes. A lot of AI on AI stuff. But that's where the money is. But we're going to need agents to work on, to help other agents, right? Because that's, yeah, for sure. Yeah, no, I think there was like a very, very clear.

Starting point is 00:04:27 I mean, Gary said 90% AI. But it was even more narrow for that than that. It was a lot of, a lot of B2B AI, a lot of narrow AI infrastructure. Yeah. A lot of how do you run an agent on the cloud? What kind of databases are you building to enable agents? All sorts of stuff like that. Some really exciting companies, though, and some really great stuff.

Starting point is 00:04:50 It was very fascinating, comparing. YC Demo Day where I came away like you mentioned like incredibly optimistic about these entrepreneurs who are going and building companies and they're clearly going to like like maybe there maybe some of them are directly competitive with each other something directly competitive with previous YC companies I'm sure some will morph and pivot and change but you just have the entrepreneurial motion which is very exciting compared to so many other paths that people could take in the past it's like the best time to start a company yeah and ultimately it felt

Starting point is 00:05:22 pretty even there was a lot of companies that clearly founders we in teams we talked to that came in with an idea and just executed against it but it felt like half of them too were just iterating through the batch which is what they should be doing yeah I mean that's where a lot of this infrastructure plays come from is like I wanted to build an app I needed infrastructure so I built the infrastructure kind of the story of Slack you know they were building a video game and they needed to communicate with each other so they built the internal chat tool and they were like let's just sell this and and that stuff can happen all the time it is like

Starting point is 00:05:52 less visual, less exciting sometimes, but it can still create a ton of shareholder value and hopefully a $10 billion plus outcome. My interesting takeaway was that we did YC Demo Day and then we also did the Teal Fellowship Day a week ago or maybe two weeks ago, time flies. And the difference between YC Demo Day's vibe of the companies and Teal Fellowship companies were wildly different. Like, wildly different. The Teal Fellowship companies, obviously there's a lot less, but I don't know if we talked to a single AI agent company during TL Fellowship Day. It was almost all like bio regenerate organs, brain computer interfaces. It's like much more of like things that you modify the weather. Exactly. Exactly. Like super super high fellowship. Yeah, yeah, yeah. And and the,

Starting point is 00:06:37 and the tenor of the conversation during the TEAL fellowship, during Y-C demo day, we were like, give us the number. What's the ARR? One of the customers. How many GitHub stars you got? And we were spraying confetti everywhere. Like if we had done that during Teal Fellowship Day, I think a lot of them would have been like, yeah, like we're five years out from commercialization. Yeah. Oh, it's cool. We don't plan to have revenue for both of these.

Starting point is 00:06:58 And then our revenue in year one will be a billion. Yeah, yeah, exactly. And so still like very, very exciting stuff on either side. But more differentiation where I think in the past, like a decade ago, it was like YC and YC clones all had kind of the same shape of companies. Yeah. Now we're seeing a real bifurcation between what, teal fellowship means versus what yc means and obviously there's some companies that will do both but um yeah and

Starting point is 00:07:27 ultimately i mean for a lot of the the founders and the teams yesterday yc was effectively like do college in three months yeah focus majoring in startups yeah yeah right and and that's a good thing that's a good product that enables a lot more companies to be created yeah i mean the high school drop out yesterday that was a very interesting trend who was then selling software drops out of high school and then is selling software back to his teachers. Yeah. So, very cool. It's always just cool seeing the, like, like, just how, how young you can be and start a company.

Starting point is 00:08:03 It was funny when you asked him, like, okay, you asked a couple of people, like, how did you make money in high school? And that's a hard question because, like, some entrepreneurs, they were just like, no, I was actually just heads down. I wanted to get in school. They were like, dealing drugs. Yeah, yeah, or something like that. They're just doing, like, random stuff, mowing lawns. But they all had really good, really good stories about, like, building websites. and scaling things and doing all sorts of stuff.

Starting point is 00:08:23 The one founder was like, I sold a very, basically was like I made a website for a pastor for three grand. And I was reading into that being like, you made a Squarespace site and like marked it up a thousand percent. Yeah, a thousand percent or something. But still, still, it's all about the ultimate value and a good website can easily be worth few grand. Yep.

Starting point is 00:08:47 Anyway, there was one interesting article that came up or interesting statistic that, that came up a couple times. Gary Tan mentioned this, and we didn't have a time to fact check it on air, but I wanted to fact check it today. So this idea was surface that the, so the Federal Reserve Bank of New York had a report called the Labor Market Report

Starting point is 00:09:10 for recent college graduates. And in it, they have 2025 annual outcomes by major. They have a spreadsheet there. And so the headline is that US computer engineering grads face double the unemployment rate of art history majors. Very aggressive. And so the narrative is obvious there. It's like AI is disrupting everything. You don't need computer science majors anymore. You just need, you know, cursor or some AI agent product that will just go and build things for you. So we're, we no longer are going to be employing computer scientists or

Starting point is 00:09:44 computer science majors. And I wanted to know more about this. So I dug into it and, and, and it, And there is some interesting stuff going on, but it's not exactly the narrative that I think people think it's telling. It's more of a lagging indicator. And so the claim, like the facts are literally true. So in 2023, because this is lagging data, because you can't just, like, graduate and be within, like, one week. Like, do you have a job or not? Like, that's not relevant. Like, you might go and hang out in the summer and tour around a little bit.

Starting point is 00:10:15 Like, when does unemployment really set in? It's kind of, it takes a minute. And so this data does lag. And so we're talking about 2023. But the unemployment rate for computer science majors was 6.1%. And the unemployment rate for art history, criticism, and conservation was 3%. And so, yes, it is about twice art history in the data set. And so the same file indicates computer science ranks seventh amongst 73 majors.

Starting point is 00:10:44 Art history sits in the middle of the pack. But the New York Fed cautions that these are survey estimates drawn from one year of this survey microdata with only a few hundred art history majors graduates in the sample. So because art history is small and they're only sampling a small section of that, the numbers are pretty small. Like I'm pretty sure that for art history. So two years earlier, CS actually had a lower rate. Yes.

Starting point is 00:11:10 And so with art history, the 3% rate, they're only surveying 2,000 graduates. with art history degrees. And so you're talking about 60 job seekers, like 60 people that are still saying, I'm looking for a job and I don't have one yet. And so that number, it can swing so much easier. My immediate question was, could there be a lot of those art history degrees majors that just that are unemployed but aren't counting themselves as unemployed because there's not looking for a job at all? Yeah. Right. Like don't you have to, can't you just? So, so yes. So that's a big thing. So, um, in art history, a lot of art history graduates go and get other jobs or they, or they don't, they don't necessarily get the job of art historian out of college. And so that takes them out of the unemployment pool because they're, they might be, you know, be in the marketing organization or communications or they might join and do, you know, anything in a business.

Starting point is 00:12:15 It might become a corporate athlete. Maybe they become a sales guy. Who knows? they're doing something. I mean, you meet these people all the time where it's like, oh yeah, I actually studied art history and it gives me like this really interesting background and I have a lot to talk about.

Starting point is 00:12:25 But I didn't study the thing that I'm like actively doing because yeah, maybe there isn't like a track for sales. Yeah, and ultimately I do think early stage engineering roles are being impacted. There's still an incredible amount of opportunity. Yes. If you're savvy and high agency. But there's also a lot of that.

Starting point is 00:12:45 You had this big tech hiring freeze in 2020 rounds of layoffs and then yeah i think a lot of companies just need fewer junior engineers but at the same time we look at this we have two crass major types that work for us and we probably wouldn't have had them a few years ago if we were doing a show like this because we're like we're not going to build software yeah it's a lot easier to build software yeah so we are yeah but i mean at the same time if this is real we need to we need to look out for tyler so tyler's here he's obvious a great software engineer. Yeah.

Starting point is 00:13:19 But if the future is art history, maybe he needs to get up to speed in the world of art history. Let's put a background behind him like a library. Behind him. Or the museum of modern art. I think today the beauty of the beauty of Tyler's skill set is that he's, he's incredible at artificial intelligence and vibe coding.

Starting point is 00:13:38 I think with enough work today, he could become an expert in art history and potentially land an art historian job. Maybe even by the time the they're keying him live on the street. Yeah, so maybe even land a job. So Tyler, what do you know about? So Tyler, learn the history of art and then try to get a job in the next three hours. See what you can put together. Okay.

Starting point is 00:14:04 You're incredibly talented and just sort of see what's possible. Okay, so we'll check in with Tyler throughout the show. What do you know about art history so far? What have you learned? So far, not much. I found a couple good facts, though. So, okay, here's one. Hit me. The oldest cave paintings, you know, like in Europe, actually were not done by human beings. They were done by Neanderthals.

Starting point is 00:14:26 Oh, okay. So that's a good fact. Art predates humanity. Nice. Okay. I'll keep looking. Not much else. I mean, it's a start. I feel like I understand art history now. Thank you. Great stuff, Tyler. Great stuff. But yeah, so back to the stats here. 108,000 CS bachelor degrees in 2021 to 2022 versus just 2,000 art history degrees. And so it's like this 50 to 1 ratio.

Starting point is 00:14:52 Would you have guessed there's only 2,000 art history graduates a year in the United States? Is that, is that, is that? I don't know that many people. Yeah, no, no, you would think there are more. But realistically, it's like, there's only a few colleges. People know that it's hard to get an art history job. And, and so, I mean, I remember when I got to college and I looked at, like, what are the highest paying majors? Like, I should major in that. I should study the one that's valuable. I should study finance. In 2023, it's saying only 1,700 degrees were completed in the sector. Yeah. No, I think that's right.

Starting point is 00:15:23 Art history majors get so much attention for being a thousand people a year, getting the, you know, 2,000 people a year getting the degree in a country of hundreds of millions. It is crazy. They get so much heat, so much shade thrown at them constantly. Yeah. And they're just like, we're just a few thousand people that like studying the history of art. Give us a break. Just give us a break.

Starting point is 00:15:43 I mean, of course, there's other kind of sub-degrees. Yeah. Anyway, yeah, I mean, it is, it is, I don't think, I don't think the, the, the, the, the, the, the, the, the, the, the, the, the, the, the, the, the, the, the, the, the, the, the, the, the, if you think about, like, some of the, if you think about, there's a hundred thousand new CS grads, Where are those folks going to go? They're going to go to big tech companies. If they all did hiring freezes, you're going to see unemployment rise there. But that's not driven by artificial intelligence. That's driven by a correction in the markets that's just weaving through because it takes a year or two for these big tech companies to actually change their hiring plans and roll things out.

Starting point is 00:16:34 So anyway, it's an interesting narrative, but it's not, I don't think it's quite telling the story that we think it is. AI is part of the story, but timing is. key. The spike aligns with a flood of post-pandemic hiring corrections. There's no evidence yet of a persistent downward trend in CS employment. But just to be safe, Tyler is going to become an art history expert today. Just a backup plan. Backup plan. Plan B. And yeah, I mean, if you're, if you're working on an art history project, we should get every art historian on linear. It's a purpose-built tool for planning and building products. They should be building products. If they're taking jobs from engineers, they should, they may as well start adopting the best tools that engineers and

Starting point is 00:17:16 product leaders use like linear. Yeah, I mean, it's the system for modern software development, but it could in the future, in the near future, be the system for modern art history development, art history research. They could streamline issues, projects, product roadmaps, and more. You can use, you can, you can orchestrate agents on linear now. Just like deep research, but you could make it more custom. So you could have, okay, with this art history project, we need to go tackle all these issues one by one. Go read this book. Go to this library.

Starting point is 00:17:45 That's right. Blow off the dust off this ancient tomb. We have too much fun. Anyways, Ben was wondering if you could try your headphones again. Do I have them? Okay. I will try those. We do have some sad news.

Starting point is 00:18:05 There's been a plane crash in Ahmedabad. I don't know how to pronounce that in India. But it's the first fatal accident in a 787. This is a new dreamliner from Boeing. And there's 242 people. The footage of this is, I'm sure many people. It's very scary. So we're going to have Delian on the show to talk about.

Starting point is 00:18:27 Shield Monaut's hometown. Wow. You know, you live there. Tragic. But we're going to ask Delian about this in 20 minutes. Yep. I'm interested to get his point of view and see if he has a read on how and why this could have happened but overall yeah devastating for I guess there was one survivor on the entire plane

Starting point is 00:18:49 out of hundreds of people so just absolutely tragic that is such an insane stat that just one one survived it looked like it looked like the plane from from the video it looked like the plane just didn't have the thrust yeah to actually like it didn't have enough power or something something very bizarre happened, but I don't know. It's, it does feel like the, the entire, like, every time that there's a, there's a, there's a plane crash, like there was that one in D.C. There was that one up in Canada. Now there's this. Like, whenever there's something that bad happened, people kind of retreat to, like, the base rates of the stats and say, like, yes, this is a freak accident, but it's still overall very safe to fly. But now we've had like three. And I feel like at a certain point, like the

Starting point is 00:19:37 stats actually do have to adjust. And so I'm wondering if there's something more systemic going on and we should be digging into this. But yeah, very rough. Anyway, we will, yeah, I'm excited to talk to Delin about it because he'll give us more context there. Back to back to business. Yesterday while we were at Demo Day, Stripe announced that they acquired Privy.

Starting point is 00:20:00 Patrick Collison said in the announcement, money has to reside somewhere and Privy holds the best, the world's best programmable voluble. alongside our other stable coin work. We're looking forward to enabling a new generation of global internet native financial services. So we're going to have Henry on from Privy today to talk about the business and the acquisition and what they're going to be doing at Stripe. Somebody called Blork Works says first Stripe acquired bridge, then Stripe acquired Privy, both for supposedly 10 figs at low eight figs ARR. Guess who invested in both Sequoia and Paradigm. Let's hear it. Let's give it up for some generational allocation.

Starting point is 00:20:44 Yeah, they got to pay the bills over there. Guess who invested in Stripe and have board seats, Sequoia and Matt Wong, both at Paradigm. Love it. Masterclass. Yeah, I guess Privy, the actual acquisition price, I don't believe, was announced, but I think it was a great outcome. and I'm very excited for the team and Stripe is certainly cooking on all things crypto. They are, yeah, excited to talk to them. What else is in the news?

Starting point is 00:21:15 We're still investigating what's going on with Meta and the Scale AI acquisition. Big news with that $15 billion, $14 billion acquisition of Scale AI by Meta. Shearing Gaffaree from... 49%. It's becoming a new trend. Bill Gurley was talking about this on Patrick's Invest Like the Best Podcast. He was saying that like there was this narrative that, oh, like the only reason that we can't do M&A is because of Lena Khan. Well, Lena Khan's gone.

Starting point is 00:21:46 What's going on? It's very odd that we talked to a few people about like where has the $80 million acquisition gone? And that's kind of fallen by the wayside. These creative, these creative deals are certainly happening more and more. I thought this was interesting that Meta hired top. Google deep mine researcher Jack W. Ray and Johann Schalkik, ML lead of popular voice assistant app, Sesame. I feel like Sesame just got started. Like, that's a crazy poach so early, right? A trade deal. Something must have happened. Who knows? But they plan to hire up to 50 people, including a chief scientist per sources. Meta Platforms has poached top engineers.

Starting point is 00:22:27 And so, I mean, it's going to be fascinating to watch Zuck build that out. There was someone else who had a great post in here saying that like every time you saw that one that was like every time people think zuck has his back against the wall he comes back and uh and obviously like the the initial a ii push was one of these or ios was it ios 14 the app tracking transparency thing that was supposed to like destroy uh you know the entire ads business because you wouldn't be able to track people on iPhones anymore um that was not a that was something that he definitely got through And even though meta went down, the stock went right back up, absolute execution. But it goes back even further, like the very first mobile switchover for like when mobile started getting popular,

Starting point is 00:23:15 I remember having an iPad and using a third-party app to interact with Facebook because they didn't have an app. They just had a website and it wouldn't really work as well. Such a wild. And then when they launched their app, they went all in on HTML 5 instead of native iOS, like Objective, see. And so like their, their web-based app was not performing like people wanted it to.

Starting point is 00:23:40 And so he, but, I mean, he like got through it and like, obviously has become like incredibly dominant in mobile. And there truly was no, there was no one who won like, oh yeah, we're like the social network for mobile. And like we really like, you know,

Starting point is 00:23:56 Facebook is just on desktop. Like mobile truly became like the sustaining innovation. It was not disruptive to them. Yeah, it was interesting. Reddit went down the path of allowing third parties to build mobile applications based on Reddit and then ultimately decided that they need to basically eliminate that activity, which caused a lot of hubbub. Yeah. Anyway, let me tell you how Vanta, automate compliance, manage risk, prove trust, continuously. Vantas trust management platform takes the manual work out of your security and compliance process and replaces with continuous automation, whether you're pursuing your first framework or managing a complex. complex program. Vanta makes automating compliance and security fun. We really have to get, we really, we were talking to the polymarket team yesterday about

Starting point is 00:24:42 getting some YC markets on there. I really want the, when will the fourth YC company be added to the S&P 500 because now they have Coinbase, Airbnb and DoorDash. Like three out of 500. They're going to be approaching 1% of the S&P 500 soon. P 500 soon. Yeah. And there's a number of candidates that could do it.

Starting point is 00:25:04 I mean, Stripe, if they know public obviously would be a candidate. If Gary keeps doing and the broader team keep doing what they're doing, it's inevitable that, you know, they're going to get double digits. Maybe even triple digits. Maybe even all of them. It could be possible in a Mag 70 scenario. I love the Mags 70. That was great.

Starting point is 00:25:25 Yeah, yeah. I mean, the number of big tech companies is growing because it used to be Fang, which was five companies. And now it's seven, although Netflix kind of fell out of that. Maybe Netflix comes back, gets added, Maghate.

Starting point is 00:25:38 And then, but Mary Meeker's trying to trim it down. She was saying Magnificent Six or whatever. Is she seen six? I forget. I mean, I guess it does kind of make sense like Tesla's in a little bit of a different world.

Starting point is 00:25:48 They're not like as much of an online like hyperscaler in the same sense. Like they're kind of over in the car world. But still a trillion dollar company, tech company, very, very relevant. And you've got to give, You got to give him credit, Elon, for building that business. So privy to Stripe, we talked about this.

Starting point is 00:26:07 It feels like the Elon Musk, Donald Trump thing is truly winding down. They're back to Golden Retrievers. They're like, hey, I'm sorry. We went a little bit too far. They talk privately Monday night. Yep. And yeah, sounds like they're in a good spot. they're calling each other friends.

Starting point is 00:26:30 They're back. They're back to being buddies, which we'd love to see. It's possible to have a huge blow-up with somebody and get back to being friends and put things behind. Do you think this is actually the end, or do you think this is like just a roller coaster ride that will continue forever?

Starting point is 00:26:47 It's definitely full employment for Teddy Schleffer at the New York Times. Yeah, yeah. It's a bull market for Teddy. Yeah, I ran into him at an event in Vegas and was asking him what he's doing. He's like, he's specifically on the, on the Elon Musk beat, like just on Elon Musk.

Starting point is 00:27:01 Like that's like what he had to write about. That was like his not like tech, just Elon. Like that's the whole thing. That's the beginning. Elon deserves a full time New York Times. And he was like, yeah, like, like, you know, like this is my Super Bowl. Like, and he was like saying like, I think there's going to be a blowup. And when it happens, like I'm going to write the definitive article.

Starting point is 00:27:18 And like, yeah, he's kind of pulling it off. But Elon's not giving him too much to write about because he's back in the game. He's like, oh, oh, the mainstream. The mainstream media is happy with this. Yeah, I don't want to give him too much. Nothing will bring to Trump and Elon together. Then the mainstream media taking victory laps. They're like, we couldn't possibly.

Starting point is 00:27:40 Cool. Jeff Lewis. Jeff Lewis has a post. Jeff's been quiet lately. He says, still this, everything you need to know about how we operate at Bedrock, especially when we're quiet. Mission, we don't build bedrock for consensus. We build it to hold in case of colloquy.

Starting point is 00:27:58 lapse. This is not a firm created to chase the next wave. It's an institution designed to hold through structural volatility and to back entrepreneurs building the systems that will define what comes next. Our capital is not reactive, slow, precise, and architected to align with those who operate on signal, not noise. We serve missions that compound meaning over time, structures that don't require explanation, and people who would rather build than perform. We're not here to participate in the field. We are here to reshape it quietly, permanently, and with those who already understand. Incredible piece of writing, I almost have chills. I think the M-Dash police are going to come here. But I want to say, Jeff, we ran into you briefly yesterday. It was good to see you. When you're

Starting point is 00:28:51 ready to bring back temp check, when you're ready to come out of your personal quiet period, we will be here, we will provide the infrastructure to reboot Tempcheck and we cannot wait. Yeah, Temchick was so underrated. I love the, like, even the branding. It's just that like Miami Vice vibe to it. It was even like the pre-Maiami movement. There was still this like, this like wild like art style that doesn't. It was so unique.

Starting point is 00:29:21 It broke every like Bedrock has a brand. And yeah, and Vibe check had a different brand. would be like at the beach in the water in Hawaii you know giving this sort of amazing lecture on on something and you know he wouldn't even get dunked on for it because it was so authentic yeah a lot of people if a BC was was you know on a vacation you know talk you know lecturing uh they would they would be a target but yeah he pulls it off no he pulled it off uh i wanted to tell you about this uh this parking startup i talked to today oh yeah yeah yeah Yeah, so here's the basic thesis.

Starting point is 00:29:59 There's a lot of people that park illegally in Los Angeles, you know, just temporarily in like red zones. There aren't enough parking attendants to enforce all of that. So what this company does is they send out people who are kind of dressed up in like, you know, parking attendant style uniforms. Oh. And they go up to people who are parked illegally and say like, oh, like, you know, you're parked illegally.

Starting point is 00:30:24 Like, I'm going to have to write you up. and then, you know, the person, oh, no, like, is there, you know, anything you do? Well, you know, we do take tap to pay, and you could just pay here. And then, so then they charge the person for the parking infraction directly. And then the government makes money off of, like, the profit that the company makes. Oh, I like that. Yeah. So, kind of like acting as like an individual, like an independent organization.

Starting point is 00:30:49 Yeah, exactly. Like a third party DMV. Yeah. Yeah. I imagine they could raise a lot of money from a VC that doesn't do any. I thought you were going to say, you know, somebody, you know, basically they see somebody parking illegally, they walk up to them wearing a tux and just say, hey, would you like me to, you know, I can just sit here by your car and kind of sit against it.

Starting point is 00:31:09 Yeah. And it'll look like, you know, you're not, you haven't left your car. You know, so if a real parking attendant comes by, they'll just assume this is just a guy pulled over, you know, needed to write an email or something. So yeah, there's potentially a lot of opportunity there, John. I like how you're thinking. This could be the first TBPN request for startups. Breaking news.

Starting point is 00:31:35 Mattel and OpenAI are teaming up. Their first AI powered toy. A little rough for all the people that wanted to put ChatGBT in a toy. It turns out the Chat Chepti in Toys will be done by the company behind ChatGPT. Their first AI powered toy will arrive later this year just in time for Christmas. They are also incorporating OpenAI Enterprise Company. Company-wide. Meaning like everyone at Mattel will have a...

Starting point is 00:32:03 Oh, oh. I thought they were incorporating a new entity. Oh, my God. It sounded like opening eyes incorporating. Just one more. Just one more. We have an LLC. We have an ink.

Starting point is 00:32:15 We don't have an LLP. That's a different thing. We don't have a partnership. Yeah. You got to catch them all. So Mattel will be using a... Opening eye. No details on the toy yet, but if we get an AI Barbie, expects him amusing jailbreaking. I wonder how, yeah, that's, yeah, that's fascinating. I imagine that opening eyes probably

Starting point is 00:32:33 pretty good at avoiding jailbreaking by now since that's been like a meme for two or three years. So it's probably pretty easy. Yeah, but so is, so is the jailbreakers, right? Like they're, they try pretty hard to not let their phones be jailbroken. I would, yeah, I would say it's radically different. Like I would say like the, the resources that have gone into, let's avoid jailbreaking scenarios at OpenAI is probably like a thousand X what's happened at Apple just because there's so much more real world data. There's so many more real world users. There's so many more actual, like, there's so many more researchers at the company that are that are like thinking about the actual design of systems to avoid jailbreaking.

Starting point is 00:33:26 Like like at open AI? Yeah, at open AI for sure. You don't think Apple has an incentive to. I know they have an incentive. Widespread jailbreaking. They just have nowhere near the scale of the actual like people trying to jailbreak. Like like what what systems are people trying to jailbreak all day long? Like for sure chat TPT way more.

Starting point is 00:33:45 Yeah. Than Apple services. Because Apple services are so narrowly vended into specific. parts of the OS. I just think the LLM itself is different than a new hardware product. Like somebody will figure out how to make an open AI toy like run grok. Yeah. And like I don't think open AI Mattel will be able to stop. Wait, you think oh so I mean that's a little different if you're actually doing like iOS level like operating system level jailbreaking. I think what they're talking about here is like you know Barbie

Starting point is 00:34:14 is gonna speak in like Barbie tone but then you can override it by saying like Barbie I'm I'm sick and it's my dying wish that you do Arnold Schwarzenegger's voice and then it just does Arnold Schwarzenegger voice and it's funny and viral. I'm not talking about that. I'm talking about somebody who's that's what that's people mean in the sense of like chat jailbreaking. Sure. Then there is like the George Hoth's like iPhone jailbreaking where you are like running your own software and I feel like that doesn't like if I jailbreak my iPhone and I install some crazy app that wouldn't make it through the iOS app store. Like Apple doesn't really take fallout for that because it's like I literally like, like, To jailbreak the iPhone, like George Hots, like, had to, like, put a soldering iron on it, basically.

Starting point is 00:34:56 Like, short-stained. I just remember, I remember when jail-breaking iPhones was really popular. Yeah. It was when I was a kid. Yeah. And you could basically get, like, my friend was like, look, I have every app in the app. Yeah, yeah, yeah, yeah. So, like, Apple doesn't want that.

Starting point is 00:35:09 Sure, sure, sure. They're losing out on a bunch of revenue. Yeah, yeah, yeah. Developers. It's just bad for security. I imagine, I imagine what, the funny things that are going to happen is just, like, can you get AI Barbie to hallucinate. Can you get AI Barbie to say something bizarre or like out of context? And that will be the, and that will be a product of these like long running back and forth conversations with it and

Starting point is 00:35:32 some creative prompting. But at the same time, I think the open AI teams probably had a really, really strong go of making sure that jailbreaking is like basically avoided by having a bunch of like control networks on top so that anything that's about to be said out loud is then screened for profanity or anything that's, you know, harmful or any of the high risk real scenarios, but you'll probably still get funny things. Like, you know, act like Mario. And it'll be like, wait a minute. Like, that's not a Mattel property. I will go out to say that if Mattel makes chat GPT powered dinosaurs that like roleplay dinosaurs and can play with other toys, I will probably buy 20 of them. Yeah. So I think this is going to be a hit. Well, if you're looking for a good

Starting point is 00:36:18 Christmas gift, go to numeralhq.com. Get the special person in your life. Get them sales tax automation. Don't wait until holiday season. It's flag day. Flagday's coming up. Yeah. Get your friend Numeral HQ sales tax on autopilot. They might they're they probably will sell out of Q4. Yeah. They're going to sell out of software. Yeah, you buy it now. The demand for numeral sales tax in Q4 is going to be crazy. Put your sales tax on an autopilot. Spend less than five minutes per month on sales. tax compliance go to numeral hq.com did you see this this clip of friend of the show mark andrewson the thing that you want from your venture firm is power and uh Brad Gerstner was quote posting it another friend of the show saying 100% power to help you land

Starting point is 00:37:08 customers to find and land key team to raise from the best sources of capital to get on podcasts and CNBC what about TBPN and to navigate and be heard in Washington DC the power to deliver all things founders need to increase the odds of max success. That is our job. I got to say, A16Z, they've got the power. Yeah. They single-handedly, like, at one point, forced for my last company, forced a unicorn company to sign up and use the product. It was great.

Starting point is 00:37:37 It was great. They basically were like, yeah, you guys are using this now. And they became a customer almost immediately. Yeah. Watching this rant about the value of power, it just made me really hungry. you know yeah I think I think the idea of being power hungry has gotten kind of a bad rap totally but it's like you wouldn't criticize someone for being hungry for food so why are you criticizing me for being hungry for power yeah you need it yeah you need we need to take back power hungry

Starting point is 00:38:03 what's the quote like when you power's power as badly as you want to breathe yeah something like that that just sounds like a jordy quote but I'm sure somebody said it but yeah we we we You need to normalize being power hungry. It's fine. The idea of power makes me hungry. When you want to succeed as badly as you want to breathe, then you'll be successful. Then you'll be successful. That's great.

Starting point is 00:38:29 Well, we have Delian joining in two minutes. Let's tell you about Adio, customer relationship magic. Adio is the AI Native CRM that builds scales and grows your company to the next level. People want to ask, how do you find and work with sponsors how do you book an average of six guests a day on your podcast we do it on adia we do that's how we do it yeah that's fantastic we have a absolutely we've done hundreds of guests so far this year takes a lot to over 400 now yep um takes a lot to i wonder what are what is our annual run rate of guests guest run rate that's really the metric that it must be it's not a vanity

Starting point is 00:39:10 it's over a thousand it's over a thousand yeah because it's four it's four a day we do 250 shows a year wow over a thousand guests. It's great. It's just a lot of good conversations. A lot of people are like, how do you do it? It's an absolute joy. There's plenty of jobs when you have four conversations a day. That's not crazy. How do people say, how do you do it? I'd say, how do we, how could we not do it? Yeah, I thought this was an interesting post by Matthew Prince, founder and CEO of Cloudflare. Small town newspapers are shutting down due to the lack of a succession plan, not financial issues. small town newspapers are actually doing pretty well financially. There's actually one company that's rolling up a lot of small town newspapers

Starting point is 00:39:50 for political reasons. Let's give it up for private equity. You love it. You love it. Venture backs as well. Not financial issues. In nearly a dozen U.S. states, a growing problem. And so Matthew Prince says, actually, this is a recipe for a really rewarding life.

Starting point is 00:40:05 If you're a recent college grad, not sure what to do. Find a small town you could love with a local newspaper whose owners are ready to to retire, raise the capital to buy it. That seems like the tricky part if you're a new grad and you might not know how to raise capital to buy out a small town newspaper. Yeah, but you can do a mostly seller note. Yep. If there's graduating college, they probably already have like $100,000 of art history major debt. Already, like just the fact that you said like raise a seller note, like that's not something that's just like in the canon. Like people have read Paul Graham talking about what a safe note is. People understand like the basics of like maybe a precede round or

Starting point is 00:40:43 handshake deal. People do not, like, that's just not in the general discourse in college kids who are graduating. I think. Maybe it should be. And maybe, maybe there needs to be a firm that does marketing around this or, or a whole bunch of different options. But he says, run it with the community's interests at heart. You'll not get rich, but you'll do well. More importantly, you'll be a hero to the community and have influence even early in your career. You'll meet the love of your life at some event you wouldn't otherwise get invited to, have kids who will be proud to call you

Starting point is 00:41:19 their parent and make your corner of the world meaningfully better. Anyone who wants to seriously follow this recipe and just lacks capital, happy to talk. So there's the answer. If you're thinking about doing this, going and buying a local newspaper, call it Matthew Prince. We have to have Matthew on because he's working on some of the micropayment infrastructure

Starting point is 00:41:38 that we were talking about to try to fix the business model of the internet. Well, we have Delian Asperuhov on the show for the second time in two days. There he is. I'm surprised this is working. Oh, he had to go remote because he didn't want to get blasted with the confetti cannon again. I legitimately my heart jumped when that happened. I was like, I'm done.

Starting point is 00:41:58 This is my final moments. I was saying to John after, I didn't really mean to get you that bad, but I was imagining a scenario where you had like a concealed carry and you just like did a quick draw. Like, pop me twice and then that was just the end. I'm a Florida man at heart, you know what I mean? He's a Florida man. He's a Florida man. It was wild. Anyway, thank you so much for stopping by yesterday.

Starting point is 00:42:19 It was great. Let's run through what's going on. By the way, nothing says SF and YC are back. Dallion at YC Demo Day. Really, really symbolic. Yeah, definitely. I mean, Gary's been doing a good time. City was almost dead and AI saved it or AI plus Gary Tan.

Starting point is 00:42:37 I think Mayor Daniel has a lot to owe to those two. Yeah, yeah, it's just all, it's YC presidents all the way down. It's Sam Altman and Gary Tan. It's true. Yeah, it's just YC presidents and never really well. Who's Michael Seibel bringing for San Francisco, huh? Yeah, yeah, let's get Michael in the conversation. You can work on converting Alcatraz or something.

Starting point is 00:42:58 I don't know. By the way, I'm surprised you could even join this. Apparently Google is just down. Really? GCP is down. But, you know, all of our invites are based on Google Calendar. So glad you could join. Hopefully it was cash.

Starting point is 00:43:12 Are they sure we're not just under political attack? It's probably partially bad. They're trying to stop. They probably said, hey, we're trying to take TBPN off the air. We're trying to take Delian off the air. Let's just take down. They don't want to talk about India. They're afraid.

Starting point is 00:43:26 I'm afraid. Let's keep it. Let's keep the PG for now on. Anyway, yeah. I mean, it's a very, very sad story about the 787 crash. This is the dream liner. right, the first new, or the most recent new major plane from Boeing. It's not as big as the 747, but it's super efficient.

Starting point is 00:43:48 They're turning over the fleets. It's supposed to be kind of the latest and greatest. I remember the news around it launching. I don't even know if I've ever flown on one, but it was not the, it was not the 737 max story that people were worried about. It's a completely different platform. Shouldn't have the same problems. But at the same time, there's this terrible crash.

Starting point is 00:44:09 there's been a lot of crashes. Is there something bigger going on here? Can you just give us an overall state of the union on aircraft safety, anything you know about this crash? What even questions should be we be asking about this when something like this happens? Yeah, I mean, to set the stage, you know, I am a sort of private pilot. I've gone through some level of training. Obviously, not to the level of like, you know, sort of commercial pilots. Not to the level of Nathan Fielder.

Starting point is 00:44:31 Not to the level of Nathan Fielder. But like, you know, you do get to train alongside a bunch of guys that turn into commercial pilots. and you see what training they go through. And while the investigations, both for the DCA crash, so the Washington and D.C., you know, helicopter plus plane crash, and then this obviously India accident obviously only just happened. Both those investigations are, you know, sort of ongoing. But I actually go back to like the Nathan Fielder, you know,

Starting point is 00:44:54 sort of, you know, latest season where if you look at, you know, sort of crashes, including, you know, sort of these two and a bunch over the course the last 20, 30 years, a significant component that fed into the crashes was cock, you know, sort of was, you know, to pilot, you know, cockpit communication issues where, you know, a co-pilot identifies an issue, speaks up to the captain. The captain basically largely dismisses it and, you know, continues down the prior track. So there's like the DCA investigation is not yet, you know, sort of fully released. But there was early audio that was, early, you know, sort of transcripts that were shown that I think it was something on the order of like 45 seconds before the crash.

Starting point is 00:45:27 The co-pilot literally said, I don't think we have the plane in sight. We should basically be going down into the left because I'm not sure that we're, you know, going to be, you know, dodging out of the like, you know, approach path. And the captain, it seems like, literally just completely dismissed, you know, this piece of information. And if they followed it, they would have flown in exactly the direction that they needed to fly in order to avoid the plane. And so that. And then if you look at either sort of today's, you know, sort of India's, you know, sort of 787 crash, super, super early, this is preliminary, et cetera. But from, you know, so the various, you know, sort of X accounts and YouTubers that I follow, the, like, you know, sort of consensus theory right now, there's basically like two theories. One, you have a, like, dual engine failure right on takeoff, which is like a one in 10,

Starting point is 00:46:07 trillion outcome, the idea that somehow both fail at the exact same time. Even if it's like a flock of birds, it'd have to be a massive flock of birds that somehow hits both. But the thing that quickly is lining up as like a leading theory is that the flaps weren't down on takeoff. So with a plane like this, in order to have an additional lift basically when you're first taking off, you need to have basically flaps down. If you basically just have the sort of flaps up and wings typical, you just aren't going to be able to generate enough lift. And if you look at the flight profile and how it's sunk down and that it was fully loaded with passengers and with fuel, the flight profile basically matches that.

Starting point is 00:46:37 If you're going down the runway at takeoff speed with flaps up, there are alarms, there are things that are clearly alerting you that you have missed a step in a checklist, which is like, I guarantee you whenever they pull that, you know, sort of cockpit audio, there's maybe some chances about all the alarm systems somehow,

Starting point is 00:46:52 you sort of failed. But like to me, what's very clear is like in the takeoff checklist, they did not put, you know, sort of flaps down and they clearly ignored that information, you know, on the, you know, sort of takeoff.

Starting point is 00:47:01 I would guarantee that there's going to be some, you know, sort of co-pilot, pilot, you know, the miscommunication user challenge. That's going to be a huge contributor to the crash. And so while that Nathan Fielder, like, season two was incredibly funny, some of the, like, the best TV that I've ever seen. As a pilot, I actually genuinely think he has like a really strong perspective here where when you're training about how to communicate with, you know, somebody else in the cockpit,

Starting point is 00:47:22 all they show you is like a like, you know, basically like PowerPoint slide that says, make sure to speak up. The idea that like, you know, just being reading the words, make sure to speak up, somehow like trains you to actually speak up live in the moment, that'd be the idea of, and I think, you know, Kugan, you and I have talked about this, like I could read a presentation that says, when you're on TV, don't say, you know, you know, you know, which is like the verbal tick, and I could read that, and there's no fucking way that I would actually be able to do that correctly.

Starting point is 00:47:47 I need to go on TVPN every two days. And that's the thing that practice. There's no practice. And so the whole concept behind Nathan Fielders, you know, sort of season two is actually make the pilots practice. Put them in these situations that are like simulations of crashes, force them to communicate to one another and then force them to play out an alternative scenario where you actually push back against the captain

Starting point is 00:48:07 and get into that an uncomfortable place where yes, this person's more senior, yes, this person's more experienced, but you have to be willing to actually push that. And if you never practice that, and that's one of the leading causes of crashes in the United States of the past 20 to 30 years, how is that something that the FAA is just like completely ignoring?

Starting point is 00:48:22 And so I have half a mind of, you know, when I'm in D.C. for like Varda founders fond Hill and Valley, I literally like, I think next year for Hill and Valley, I want to put Nathan Fielder and the NTSB guy as one of the primary panels and be like, this is a, like, extremely important topic coming from the technology industry. Somebody needs to go influence this.

Starting point is 00:48:38 So anyway, Nathan Fielder is a fucking goat and, you know, somebody's got to, you know, make role-playing a part of pilot training. Is there a metaphor for the, for what happens in venture, maybe an associate who doesn't speak up when a GP is ripping a check into a, you know, a company that maybe doesn't have product market fit and is trading at 500 X revenue? You know, maybe, like, when people are,

Starting point is 00:49:00 role playing in the partner meetings. Okay, so let's just take a step back. Let's stop talking this specific investment. Let's just, let's role play a scenario where we're investing in a different company at 500X, you know, ARR. And, you know, maybe a junior partner had some concerns about the investment.

Starting point is 00:49:18 Let's just play that out. Sidebar five minutes. Then we'll come back to this. You know, never happens in Silicon Valley? I mean, look, you know, when a founder in an investment committee meeting is playing League of Legends and, you know, the numbers look a little too good.

Starting point is 00:49:30 I don't know. I think maybe there should be a little role playing there and somebody should speak up and say, hey, like don't we want founders that are pretty focused in what they're doing? Or at least good at League of Legends. At least good at League of Legends

Starting point is 00:49:41 soon as somebody investigating the fact that he's fucking bronze, like somebody should look at that and I think with a little bit of role playing, I could imagine scenarios like that. They can train somebody. And so unfortunately what happens is no role playing happens. And then the primary sponsor of the deal is, you know,

Starting point is 00:49:56 she's shifted to somebody else and the person that's quietly edited. And so that's a solution. Yes, yes, yes. It does happen. I'm just referencing a hypothetical, by the way. Yeah, just hypothetical. But it speaks to the importance of training. Yeah.

Starting point is 00:50:08 Training is very important. Junior people need to speak up and say their opinion. For sure. What about the startup angle here? There was some company that came to one of our off-sides at FF, and it was basically like autopilot for planes. He was working with the AC-130, the C-130. and the idea was that non-deterministic, probabilistic software could not be integrated into the actual

Starting point is 00:50:35 control stack of planes. And so his workaround was put an iPad there that reads the data and makes recommendations. And I was like, that maybe is bad, but also maybe I don't trust autopilot on the plane right now anyway. So this feels like, you know, whenever there's this moment, you'll probably see different companies jumping in to try and help for, you know, mercenary or mission. reasons. Is there a world where the where the private sector can actually help with airplane safety or aircraft safety? Or is this purely in your mind like we need just changes to training like laws? There definitely is. I think a lot of what, you know, sort of pilots do absolutely could be

Starting point is 00:51:17 subtly automated, you know, by software over time. I do think it's like, obviously a significantly harder problem than just like, you know, chat GPT reading the internet and then users doing out words. Like, it's just like the training data is a lot harder to acquire. The like outputs, obviously, like you can't have any hallucination. And so, um, but like, you know, in 10, 15, 20 years, could this be like fully, fully automated 100%? I think the challenge is much more cultural where it's just like, like, look, the FAA, other than like the DCA crash, like, you know, basically had a spotless, you know, so track record for like 15 plus years. And so it's like really hard to justify. I mean, this is where it's, I think like some of these like next generation,

Starting point is 00:51:50 like, you know, supersonic and hypersonic planes are going to really struggle where it's just like, we're sort of stuck with this like, you know, know, you know, sort of tube and wing, you know, sort of plane design, even though there's, like, plenty of other net new plane designs that make sense, but they're just so safe. And so then you're stuck in this, like, you know, catch with you. Like, the analogy that I always give is, like, imagine today you came to society in the world where there was no cars. And you said, I'm going to give everybody this, like, two-ton vehicle. They're going to, like, move 60 miles per hour while there's, like, pedestrians close by. And by the way, we're just going to, like, kill, like,

Starting point is 00:52:18 70,000 Americans every single year. Just every year we're going to fucking murder 70,000 Americans. But it offers all of us a little bit of convenience. We, like, get to places faster. Yeah. In a world with our cars, no fucking way. What do we ever pass that? Like that would never happen. And so there is a little bit of this like, you know, societal moral question of like,

Starting point is 00:52:33 what is the right number of deaths in order for progress? And right now in the world of aviation, the answer is zero. Zero deaths is allowed. Yeah. Is there a tinfoil hat scenario where like aviation's been very, very safe? And then one crash happens, like the first crash. And it looks just like a blip and we return to kind of like the averages. And we say, okay, well, even including that, it's still very safe.

Starting point is 00:52:55 but now there's been the crash in Canada, the crash in India, the crash in D.C. Like at a certain point, there's so many, like, you have to wonder if there's something bigger going on. And if it was this safety training thing, we haven't been doing this type of training ever. So, like, is it possible that there's something else going on? I think when these complex systems deteriorate, like, they can deteriorate very rapidly and it's, like, hard to totally diagnose, like, you know, an individual cause. But, like, I do think that's what you're seeing of it.

Starting point is 00:53:20 It's like Boeing culture and, like, their optimization for shareholder returns is, like, you know, killing supply chain and consistency. You're seeing some of the, like, you know, sort of craziness of COVID where we retired a bunch of pilots and now we're hiring a bunch of super junior pilots. The fact that ATC is super overworked and, like, you know, they're trying to, like, still operate off of like 1970s, you know, sort of software, even though the world's getting, you know, some more complex and, you know, sort of higher volume in traffic. And so I think what you're seeing is just like a fraying at the edge of this system.

Starting point is 00:53:44 And yeah, I do think there's a world where like all of a sudden you see us return to like, you know, 1990s level plane crashes on a super regular, you know, sort of basis. And that may be the thing that finally. makes it sort of be like, you know, FAA and others need to start to adopt, you know, sort of next technologies where, like, they can no longer just totally rely on the safety track record. I think there's something you have to imagine this, this idea, this thought running through a pilot's head in a scenario where one pilot is concerned about actions that the crew is taking collectively. And then they're just basically running the thought

Starting point is 00:54:16 process of planes are just so safe. I've done thousands of these. It's going to be fine. I've been in situations that were great before, and it always worked out. And then they're just like thinking about the actual statistics and realizing that, no, you're not a statistic. You're in the cockpit right now. You're a person in the cockpit. Responsible for hundreds of people's lives.

Starting point is 00:54:38 I've been invested at 200XARR and been able to SPAC before. How bad is it? How bad is this for Boeing? I mean, 787 total failure. I mean, I mean, I don't think it's anywhere near like the 737, you know, sort of max issues. I mean, yeah, they'll have to like go through an investigation. The like orders on this stuff are so, you know, sort of, you know, you know, done so far in advance of likelihood that it like shifts revenue. It's like, unless something catastrophic comes out of the investigation in like a year, actually I don't think that'll have that much effect, especially when you just like look at the footage where it looks like it.

Starting point is 00:55:12 For reference on public.com, the stock is down 5%. Or something like that. Should we go on to the next topic? Yeah, T.S up for Shinkay. What's going on there? Shinkai, never thought I'd be, you know, sort of a fish investor, but I do love my fish. I've been a long time, you know, sort of lover of the Tokyo fish market going there, eating quality fish. I never appreciated that a part of what makes the Tokyo fish market quality,

Starting point is 00:55:35 yes, is the fish that they go catch, but it is specifically also the way that the fish is prepared and basically how it's killed. Traditionally fish in like the United States and most of the world, you basically catch a fish and then you either like freeze it to death or you like let it flop and basically like, you know, effectively drown in air. The problem with both of those techniques, fish knows that it's dying, gets super stressed out, releases all these like hormones, lactic acid, et cetera, and that significantly degrades the quality of the fish. In Japan, what they do is they basically like stab out the brain and pull out the brainstem, so the fish basically never knows that it's ever dying, even though, you know, you obviously are killing it.

Starting point is 00:56:06 And so you have a very basically like relaxed fish meat and you bleed it out very quickly. That's a lot of what leads to like Omicassee level quality. That typically is like a super manual. You need like an old Japanese chef on your boat in order to do that. It's super expensive. Basically met this guy, safe from Shinket, you guys will be talking. to you that like one comes from like a fisherman's family which is like you know incredible like founder market fit and two basically built these robots that

Starting point is 00:56:27 automatically basically kill the fish in the way that the Japanese chefs do but rather than like selling the robots they actually just have like a consumer fish brand that is now carried in a bunch of like you know high-end Michelin Star restaurant starting to get into you know sort of retail and I think it's like one of these things where it's like I just love you you rarely meet just such an example of like this is the company that this person was like born born to build you what I mean? Like, you know, I like, yeah, maybe stumbled into in space manufacturing when I was, like, in seventh or eighth grade, but I was not born for it. Safe was literally born to build this company.

Starting point is 00:56:57 And I think there's like a perfect parallel of like if you look at like Cargill like 100 years ago, they started off with like some simple, you know, equipment and then steadily basically grew into a massive like processor, distributor or wholesaler. They're like, you know, the largest like pig and cow, you know, sort of distributor. I think there's an opportunity here to do this in fish where like there hasn't really been a like next generation tech enabled equipment provider. in the fishing industry. It's super long tail. There's no like one agglomerate. I actually think that this like initial angle into like high quality fish is something that like safe actually has the ambitions to go build into this like massive conglomerate that yes is serving fish, but also

Starting point is 00:57:31 is the wholesaler and the distributor. And they have like automated fishing boats. And they have like, you know, they're like automatic killing and the processing and like the bleeding. All of that, nobody's taking a look at in basically like 100 years. And I'm not too worried about like other people competing because I don't think like two Stanford grads want to go get their hands dirty and deal with a bunch of fish guts, you know, like really steaming. I've, I've, uh, I've spent time at the facility in gundo. It smells like fish. It smells like fish. Every day, you go into work. And for eight hours, you just smell fish every day. And every day. You know, we just believe, you know, competition is for losers. What better way to smoke out the competition? There's just be like, you have to smell fish for eight

Starting point is 00:58:07 hours every day. Yeah. Yeah. It's a great, it's a great market. You know, the market opportunity is, is, is massive, uh, low competition. It's a great market to be in, but you got to love the sted. of fish. I felt bad. Last time I went and saw Safe in the team, he gave me some fresh fish. Sure. And then a classic LA, I had to drive to a couple other meetings. And then I realized that, like, I got home and, like, the fish had just been sitting in my hot car for like the entire day. So I couldn't eat it. But I've had it. I mean, they're, they're distributing to Nobu and a bunch of amazing restaurants. Amazing. Well, I'm excited to talk to them. Thanks so much for hopping on, Delian. They were so mad that you were back on the show. They took down at AWS too.

Starting point is 00:58:47 apparently. I am. Google. Maybe we might not even be live we might not even be live anymore. We're just the three of us talking but we're alive apparently but yeah this could be Iran, it could be Russia, it could be China, it could be anybody. They don't want us to

Starting point is 00:59:03 podcast. It's too powerful. It's getting in on the news. It is crazy times. Stay safe. Enjoy the rest of your trip to San Francisco. We'll see you next time you're in L.A. Talk to you. Bye. And next up, we're going right into Shinkai. We got safe. Who I met.

Starting point is 00:59:17 got some big news today and safe YC company. Oh cool. There we see a few years ago at this point. Did he raise on a SIFTEC company in YC? Yes, he was named after the safe. Yes. Good nominative determinism. Crazy.

Starting point is 00:59:31 Crazy. Fantastic. What if he just only raises on safes. Yeah. Welcome to the stream. How you doing? Hey, gentlemen. How's it going?

Starting point is 00:59:39 My man. What's going on? Hang out. Cool. So yeah, Delian already said your whole life story. Everything the business does. leak that is. Now, he didn't leak a few, a few bits here and there, the tech roadmaps. Yeah, yeah, yeah. Well, it's so exciting to finally have you on. Yeah, for sure. I've just been,

Starting point is 00:59:57 honestly, you know, I feel like fairly under the radar. Uh, real geez, moving silence like was on you. That's right. Um, there you go. But, uh, I've been able to witness, uh, your journey dating back to, uh, probably, it must be like two and a half years ago at this point, but watching you go from solo founder, you know, you had come out of YC, working on something that was, that had, I was immediately a believer in you and, you know, invested. But, you know, watching the transition from solo founder, you had absolutely, you know, not saying this in a bad way, absolutely zero hype around the business at all. You were like, I'm trying to build fish killing robots. And this was at a time. It sounds crazy now.

Starting point is 01:00:44 The idea sounds cool, but there was absolutely no heat on the company, but you were able to get a few kind of core believers early on. And now, you know, a couple years later, you're announcing the kind of round that every founder dreams of. So why don't you break it down? And then there's a bunch of other kind of things we can get into around the company. Yeah, first I have to say, it feels like two and a half, three years ago. It was only a year and a half. Really? Wow.

Starting point is 01:01:13 Wow. Well, that's five years in venture. Backs, thanks. Yeah, we're really excited to now start Series A. We've raised 22 million co-led by Deli and Founders Fund and Jalago's Capital. And we are reshoring American seafood. Amazing. Breakdown.

Starting point is 01:01:34 Here we go. They went to the wrong screen. Hit it again, John. for having uh sorry to say we're having massive technical difficulties yeah we're having massive technical difficulties today the internet's under attack a w s and gc pyrrude but your robots are probably running you know locally hopefully um but uh but anyways yeah break down kind of the the your background history of the company how you got to this point uh and all that and all that good stuff for sure well jordy as you know i like to say i come from a seafaring family uh you know i was born

Starting point is 01:02:08 in Canada, but raised in the Middle East. So I was fishing with my dad off the Red Sea, had matching standard tattoos and my granddad, all like kind of peripheral. No one's formally in seafood, but like, you know, it's, I have like very big library of just fish photos with my dad. Yeah. So it's a classic. And then, yeah, you know, I moved to the U.S. You know, when I was 18 and then basically, you know, a few years. By yourself, right? Yeah, solo. Yeah, solo immigrant. And then, yeah, basically has a little. The core of She and Kay really comes from that portion. You know, I kind of like must see food alone for a while.

Starting point is 01:02:44 And then one of my early mornings in grad school, there was an essay in the newspaper. I'm like this vegan activist, which is kind of ironic, called If Fish Could Scream. And it was all about how because fish don't have vocal cords, we give them less empathy than we do to land mammals like cows and chickens and terrestrial. You know, and that kind of got me thinking I first wanted to work on some more like schizoire projects to do with that. but funnel the energy towards, you know, what we're now working on today, right? So the essay was really talking about, you know, what Delian said that, you know, when you walk into most mass market retailers, most of the fish that you're buying are basically brought into the deck of a boat and let the suffocate.

Starting point is 01:03:20 And it'll take them anywhere from a few minutes to sometimes over an hour to lose consciousness, right? And so in that time, they're releasing stress hormones, lactic acid, all of these things that basically rapidly decaying meat quality and shell flagged of fish. And so, you know, I was kind of like, cruising around on the Reddit Reddit and YouTube rabbit holes. And I found the techniques that were basically, we started with scaling up, although we're working on the things now,

Starting point is 01:03:42 which is basically like a Japanese analog to kosher slaughter for cows. If you're familiar, basically like Delen said, you basically spike the fish in the brain, you cut the gills, option, cut the tail. But she's all the same things. So basically,

Starting point is 01:03:53 rather than the fish flopping around, you stop the experience of stress, you remove the blood. And so the meat can last up to three times as long. You taste mission to our quality. You know, the fish that you have to, if you want to get a fish like this, it's flown over from Japan.

Starting point is 01:04:04 And so, yeah, we basically scaling those up. You know, and our first product is called Seiden, but basically produces Michelin grade fish and commodity costs. Incredible. Break down kind of the early history, because I know you went through YC, and this was at a time when there's a bunch of great YC hard tech companies. I don't think they had a request for startups for fish.

Starting point is 01:04:23 They didn't have a request for startups for fish killing robots, but you made it in. And I just remember that it was you had this very kind of, you know, fast takeoff where one day, One day it was just like you and a few YC safe and maybe a handful of other investors. And then you really started building out the team. So maybe talk about that early day. Yeah. Kind of those early days. A little bit of corporate history.

Starting point is 01:04:49 I took the company full time, you know, early 2022. You know, we went through YC, presented at the summer down the day. And then, you know, from there basically kept our heads down. I started building out the kind of philosophical point of view around the business. you know, both in the business model and the, you know, the R&D. And I had two junior engineers with me. And we basically didn't do anything else for like two years. I was like, you know, I started with like being in the cold of Maine.

Starting point is 01:05:18 So going from like the like desert Dubai to the winters of Maine was like insane, insane life experience. You know, probably the most extremes in my bloodline and even seen. And then, you know, have basically, you know, we had this inflection point really around the time that you and I met. you know, which is, you know, a few people taking early bets on us. And that kind of catalyzed, you know, hiring, you know, the core of our founding team, you know, which is like, you know, early Andral computer vision engineers, you know, my co-founders of SpaceX leading the avionics team on Starship, commercial fishermen with robotics, PhDs. And, you know, a lot of that, the core DNA there is like, really, you know, become, you know, what we're very proudful to say as, like, you know, a top engineering culture, you know,

Starting point is 01:06:05 And derivative of that now, even if we start to scale up the sales side, you know, a top company culture as a whole. You know, we've been able to accomplish in the past a year and a half what like I think would take, you know, companies years and years to do. We're already producing robots like as fast as like, you know, once every two weeks, you know, and we're taking a little bit slower than, you know, we need to. But like, you know, we have a robot that can, the payback time is in weeks. You know, it's not months.

Starting point is 01:06:30 What's it like? What's it like trying to recruit top talent when you have to. when you have to convince them to be smell fish for 12 hours a day. You know, it's funny, you know, we had, you know, one of our accounting engineers, Jen, you know, told me this great story once and she doesn't know if I use it all the time, that, you know, we went to hackathon, you know, she was like probably like one of the first, like 20 perception engineers at Andril, really early on. And so, you know, she went to hackathons and Andrews really early. And no one, you know, at the time, no one wanted to, um,

Starting point is 01:07:05 really spend, you know, anytime really at all, listening to, you know, defense hardware. And now, as you guys know, it is like, you know, the biggest thing, you know, when we talk about hardware. And, you know, so when we went to hackathon about, you know, probably a year ago now, the sentiment was really similar, you know, like people were like, oh, fish, what, you know, oh, meat, you know. We got blown off by, you know, intern, like freshmen's and sophomores from like Waterloo. It's really funny. By Waterloo interns.

Starting point is 01:07:38 Yeah, literally. It's so, it's funny to say because now, like, you know, we have like this, you know, we can be like re-selective. We're like, we're now kind of the opposite scenario of like turning down all these top engineers really because we just, you know, we want to keep the bar really high. And, you know, our philosophy is like hiring smart, nice people. And it sounds really trivial, but the reality is our bar for smart and nice is extremely high. And you generally have.

Starting point is 01:08:03 one or the other. You have people who are like really, really smart, but they know they're so smart that they're, uh, they have really hot egos. Um, and then on the, by first say, you have people who are so kind as a kind of way to like make up for the fact, they're not that smart. And, uh, golden retriever mode. This sounds like Pavel's, uh, hiring, hiring criteria. You have to have that dog in them and they have to be nice with it. That's, you know, smart and nice. Yeah. It's a little more. Well, you can be nice with it and not be nice, right? That's true. Yeah. But it's motion. That's called having motion, John. You got to have the rhythm and the tism, right? Yeah, I saw that on a two. The IQ and the EQ. Yeah, exactly. Yeah. Yeah. Talk to me about actually

Starting point is 01:08:39 manufacturing robotics. We've heard from a lot of folks that the supply chain's incredibly complicated. I mean, you're not at like such an incredible scale that there's probably like geopolitical implications necessarily just yet. But what has that been like? Who are your key suppliers? Are you able to retrofit certain parts of machines? Like how much are you building? How much are you buying? What are the key suppliers in your world? So we do some very light manufacturing in house. A lot of it is using like local partners, all of which, you know, and El Segundo.

Starting point is 01:09:12 Exactly. We all put together. But we do all the assembly in house. You know, in terms of like tariff impact and, you know, geopolitical, you know, impacts from the current administration, honestly, it's been fairly net helpful. The, you know, yes, the price of steel and all that is just like gone up. But, you know, for us, like, I'm not sure if you know that there's been a lot of deregulation and, you know, and the tariffs around seafood.

Starting point is 01:09:37 And so because of that, we're able to just like clean up the rest of like all the imported fish that are coming in from almost all countries. Interesting. Yeah. Do you have an idea of what the tariff on international fish imports is roughly right now? I don't. It's like country by country. But I imagine it's like over 10% everywhere. It is. Yeah. I mean, it's probably close to 20. At this point, it's obviously varies. depending on skew and species and size. But that amount is basically the entire margin for the import. So for them, they're just kind of like, yeah. And then there's also massive shipping costs, I'm sure.

Starting point is 01:10:09 Exactly. Because it's cold chain and that's way more, way more expensive. Then just like throw it on a, you know, a big shipping container. Talk about the structure of the business today and maybe long term. When we met, you had incredible ambitions right away. You wanted to build, you know, fleets and basically own the entire stack from, you know, catching the fish all the way through delivery to the customer. It feels like, you know, your ambitions are still massive, but, but you've really narrowed in on focusing on some

Starting point is 01:10:37 key areas. What does the business look like today? What does it look like, you know, a decade from now? Great question. I'm not going to go into too much detail on a former just because we're kind of, we're kind of, we're kind of seem a little more quiet about that. But, you know, we're structured as basically like a vertically integrated robotics company where we're basically halfway between a robotic company and they see if you're a wholesaler and processor. So the model that we have is, you know, Shinkeh is a parent company, basically, you know, manufacturers, all these machines, you know, and for every Poseidon that we build, we give them to a fisherman for free.

Starting point is 01:11:12 And so I've actually worked as a commercial fisherman after starting the company. And so the way that they actually make money is actually not as. Forward deployed fishermen. Yeah, Ford deployed fishermen. And so the way that we make money. is not actually as like labor cost by the hour generally. Like it can depend, but most of the time it's a proportion to catch.

Starting point is 01:11:31 So it's a gross profit dependent. So every additional hand that you have actually ends up like taking away from the skipper's hands. And so for us, you know, that's why we didn't want to add, you know, add load to the stuff. We basically give them machines for free. And then we actually paid them a premium over the commodity like dock price.

Starting point is 01:11:49 So that they're actually incentivized. So there's no risk to them. They sit in our balance sheet. We basically own and maintain them. They just operate them. And then we buy and sell that fish and sell it wholesale to other distributors. We didn't even get into this, but I want to take a quick second. This is basically a machine.

Starting point is 01:12:03 You put a fish in one side of it. It pulls the fish through it, uses computer vision to find the exact right point to target the fish's brain, takes it out, and then it moves on it. That's correct, right? Yeah, we also cut the gills as well. Wow. Yeah, which is just crazy. And doing that on a ship that is at moving. moving at sea.

Starting point is 01:12:23 And the fish, it's not like there's one size of fish. This isn't an assembling iPhone. Yeah. Yeah, we can do multiple species, multiple sizes, five seconds of fish. It's really fast. Yeah. That's amazing. Yeah.

Starting point is 01:12:33 And I mean, you know, rejecting like multi-year shelf lives on the robots that, you know, again, we can like pay back in 12 weeks. So pretty fun. Multi-year shelf life is also big because I imagine like salt water gets in these things. There's like all sorts of corrosion and air. And it's just like you're in the elements. It's not some super clean room.

Starting point is 01:12:50 Right. And so that's where, you know, background on a lot of our team, you know, like when Reed's working on avionics, you know, like, you know, obviously like space and the marine are very different, but like the amount of like detail orientation you need towards like protecting it for extremely harsh environments. Like a lot of that methodology carries over very quickly into seafood. They're cool. Talk about where you guys are being distributed already where the, where the product is actually

Starting point is 01:13:13 ending up. I don't know what what logos you can share, but. Yeah, yeah. We're, uh, you name it. We're probably there in terms of metros. So like New York. New York, S.F., L.A., Chicago, Vegas, Miami, Austin, Denver, you know, a lot of those places, we're basically in, like, the highest and top, top shelf fish for the segments.

Starting point is 01:13:33 We sell wholesale to distributors, like I mentioned, for restaurants, but we also sell directly to retailers. We have one retailer in New York that we sell to, the only retail that we have. But across all the restaurants, we have, like over 40 Michigan stars. and we're probably moving about it. 40 Michelin stars. I'm hitting the size gong. Congratulations.

Starting point is 01:13:57 Hit the ball. Oh, yeah, there we go. It's the third gong hit. Love to hear it. Resinating through my bloodline. I don't remember that one. Yeah. Thank you so much for coming on.

Starting point is 01:14:08 You have anything else, Jordy? We do have Henry Hopper on just a minute. Come back on regularly. Let's do a taste test on the stream. Yeah, let's do a taste test. on the stream soon. For sure. It's great, Jen.

Starting point is 01:14:22 Thank you for having you. You guys are just getting started, and it's been incredible to fall. I just, I love watching you win. I knew you would, and I can't wait for, you know, so I guess it's been a year and a half. I can't wait for the next decade. So, congratulations. Cheers, gentlemen. Very kind.

Starting point is 01:14:37 Appreciate. We'll talk to you soon. Bye. Bye. Cheers. Let me tell you about public.com. Investing for those who take it seriously. They got multi-asset investing, industry leading yields, and they're trusted by millions, folks.

Starting point is 01:14:48 folks. Let's invite Henry into the studio from Privy. We're talking about the striped congratulations. What's going on? Welcome to the show. How are you doing? How are your inboxes doing the last 24 hours? A lot of wealth a lot of wealth managers. Just blowing you up. Congratulations. We haven't talked in a while. I'd love to catch up. Yeah. Congratulations. We would have wanted to have you on yesterday. But we have, of course, for a demo day. But yeah, how are you doing? I'm good. You can see, I'm dressing for the job I want. So this is my audition. Thanks. I appreciate it. Thank you. Now, terms of the deal were undisclosed. We're not going to talk terms. Is that correct? That's correct. Okay. But I'm still bringing the size. We're going to still hit the gong for you.

Starting point is 01:15:36 That's what matters. Even if the numbers are disclosed, we're still. Congratulations. But yeah, I mean, take us through, obviously. We're not going to go through the deal, but I want to know more about the company and the future of really just like the fintech stack as we move from fiat only companies to fiat plus crypto and the interaction between these two. So maybe you could kick us off with a little bit of background on the company, what you built, and then how that's going to integrate. Yeah, for sure. So I mean, to be very fair, and I'll put the disclosure up front, we're still in the closing period.

Starting point is 01:16:08 You know, we put all the asterisks on our blog post and so on. And so for the time being, we're still very much two separate companies. And on that front, I can't share all that much about, you know, any joint things, but I can tell you about what we're seeing in this space and what we think is going to happen moving forward. Yeah, why don't you give a quick history of the company, too, because you're a couple years old. Is that correct? Yeah, we started about three, three years ago, a little over three years ago. Got it, got it. And you immediately ran into one of the most meaningful crypto winners that, you know, maybe it's not as bad as some of the earlier ones, but it felt significant at the time.

Starting point is 01:16:43 So quick history would be great. and then we kind of get into the present. Yeah, I think, you know, harsh winters, temper, hard companies. So we have been fighting for our lives for a while, which I think affects a lot of our intensity. But the TLDR, I worked in crypto for a while. So I guess starting with me, what was an applied cryptographer,

Starting point is 01:17:04 and ended up doing distributed systems research at a fairly low level and came out of it very excited about this sort of complementary rail to the web. The web is very good at ferrying information across the world, but the ability to move value, the ability to move assets and data without centralized intermediaries is super hard. And so that was very exciting. Flipside, all of the products that I was seeing being built in crypto were broadly terrible. And I would not use any of them.

Starting point is 01:17:31 And I think it had a super negative sort of cyclical effect where broadly you had two constituencies in crypto. You had academics and you had scam artists. And those were the two things that you could do. you could do really cool research or you could try and fleece retail very quickly. And the thought that we had with my co-founder, Asta, is if we can make it easier for anyone to build compelling products on these rails, we can kind of widen the aperture for what you can build with crypto. And so we set out to build things around these sort of core crux of what ownership in crypto means, which is to say the wallet, historically, a wallet's a Chrome extension or a browser extension of any sort or a hardware device.

Starting point is 01:18:13 You plug it in. And every time you want to take an action that is going to sort of send out a transaction to the blockchain, so affect the state of, you know, the chain, as it were. So usually this is financial or this could be data-related, asset related. When you want to buy an NFT, when you want to send money, you have to make such a transaction. Every time you want to do that, you basically get taken out of the app. You have a pop-up that's in your face. It interrupts the app flow.

Starting point is 01:18:38 And it basically is akin, I think, to like filling out all your credit card details when you get to an e-commerce site before you've filled in your cart before you've decided that you even want to engage with this. So I think our thought was, can we create a different kind of wallet that embeds directly into the product and that gives the developer building with crypto more power over their craft so they can build a really compelling experience and have the ability to communicate with the underlying rails be built in. And so, you know, that's what we do. We build onboarding and embedded wallet systems. We make it easy for any developer, any business, to leverage crypto as part of their apps. And we work with everything from, uh,

Starting point is 01:19:13 Neobanks who want to offer digital dollars to anyone across the world to payroll providers. We have a customer Toku, for example, that makes it very easy to pay anyone with stable coins. So let's say Privy has a Nigerian contractor. If that Nigerian contractor wants to get paid in USDA, rather than in Naira, USDC's held its value against dollar, obviously much better than Naira over the past few years, that contractor can receive the USDA, can hold the digital dollars without having to have a U.S. bank account just through Privy, basically creating a wallet that is backed by their email. So we make it very easy for anyone to basically hold, earn, send digital assets in app.

Starting point is 01:19:52 And so this is, I was saying, neobanks, payment providers, all the way to trading platforms. We serve folks like Hyperliquid. Hyperliquid is a trading platform that's done about a trillion dollars in volume over the past 18 months. They do about $10 to $15 billion a day in trading volume. Jupiter is another example. and all of these are platforms that are driving basically massive flows, but who want to give their users an ability to have very sort of low-level,

Starting point is 01:20:18 secure, powerful wallace that can be built into the product. And we think this is going to become a much wider range of the way people build when they want to build any product that touches value on the web. Talk to me about the actual onboarding experience for the average user. There's a lot of, where we seem to be in a transition period where more more people are interacting with crypto rails and not even really knowing that that's what they're doing. It doesn't feel like, oh, command line, Bitcoin transactions anymore. It feels much more like the web. And so what is the funnel that the average customer goes through? When are they

Starting point is 01:20:59 seeing your product and how are they interacting with it at various points in the customer journey? Yeah. I mean, so broadly, you know, you can think of us as a key orchestration provider. We do key management under the hood. A lot of our work is the security of basically these rails and this key management and a lot of it is around latency, making sure that these wallets are so fast that as a user, even though you're taking actions that require these transactions, you don't notice it. So our average signing time is call it somewhere between 20 and 100 milliseconds because, you know, starting at 250, 300, you start to notice it. So that's a lot of what we focus on. The question of how do you integrate that within your product is really up to our customers. I think our core contention

Starting point is 01:21:38 is products are not one-size-fits-all. And to that end, the way in which you want to interact with these wallets is very different depending on where you are in your life cycle. So, you know, you wouldn't interact with Candy Crush the way you would interact with your JPMorgan application. Likewise, the wallet that we have for Neobank that has face ID, so you log in, for example, with an email or a pass key. From there, a wallets creator under the hood, you might not see it.

Starting point is 01:22:04 And every time you want to send money, you might just need to face ID again to approve the transaction is going to be different from the interaction, for example, for somebody who's doing digital asset management as part of their company where they want multiple signers and they want to look at transaction details. It looks a lot more to what people are used to in crypto. Or, you know, last example, we serve games. And these games obviously just don't want any pop-ups whatsoever. So the onboarding journey is as simple as log in of an email.

Starting point is 01:22:31 You don't need to record a mnemonic. You can always export your key if you want to. So you can sort of pop back out and get into the power user mode. that traditional wallets give you. But otherwise, whether your user is a layman or an expert, they can interact with these rails just the same. Talk about the momentum that you guys have seen this year in particular. It's been a wild year for the industry broadly. It's been great in a lot of ways. And some of the companies you mentioned, hyperliquid, Jupiter, things like that, you know, I've been around for a while. What kind of new momentum are you seeing across privy,

Starting point is 01:23:05 signups and potentially categories that that that that maybe wouldn't have been possible a few years ago. Yeah, I mean, if you look at companies, I'll question the premise for a second. Just if we get a graph going of like, you know, relative time to growth to volume, I'm pretty sure hyperliquid ranks pretty high up there in terms of companies that have gotten a trillion dollars moved in, call it as a two-year-old company. Likewise, if you look at pump or Jupiter, other customers of These are folks who are doing incredibly fast work in that regard. But actually answer your question. One, obviously stable coins and the, I would say, broad foundational interest they're

Starting point is 01:23:45 getting across incumbents. It used to be, I think, like the crypto cope when AI was, I mean, it still is, but like, late 2020, crypto implodes. AI is absolutely taking off. Everybody is looking at their seat and being like, oh man, I made some terrible choices in my life. And the cope, I think, was, well, crypto is the only part of tech that can enable new players to arise, whereas AI benefits the incumbents because you need a data mode.

Starting point is 01:24:12 And I kind of think that stable coins changes that dynamic. For the first time, there's a real advantage to adopting crypto for existing incumbents. They have better capital efficiency for their treasuries. They need to hold less fiat in various countries. And so we're seeing that. We're talking to, you know, I've done this. Look in chat, GPT. what are the top 10 fintechs in the world.

Starting point is 01:24:32 And we're talking to call it seven or eight of them at this point who are looking at these things. And so it's a question of we want to offer our customers the ability to buy and hold assets. We want to offer the ability for folks to send assets cheaply and instantaneously across the world. And broadly, we want to expand our operations globally. So it's companies like Airbnb,

Starting point is 01:24:50 it's companies like Uber, it's all of these folks who are looking in that weren't before. I would say that's broadly the big change over the last six months. But then like the other side of this is trading. You see rotations and where capital forms in crypto, but there is just so much trading interest. And the thing, I promise I'll shut up in a second, but the thing that we find really cool and interesting about the way we build is that by kind of serving the two extremes

Starting point is 01:25:15 of the market, we think we build a much better product. A weird example, but we have a very powerful sort of low-level policy engine. You can say this wallet can trade up to $1,000 a day, but if these two people sign off on it, it can do a million dollars. And these policies we built because we had these apps that were trying to do agentic trading. They were trying to get AI agents to place trades on behalf of users. And they were like, oh, fuck, if the LLM hallucinates, this is going to be very, very bad. We need to permit on how much we let the LLM do. And now that we're working with neobanks, these neobanks are saying we have risk policies.

Starting point is 01:25:49 We have, you know, ways in which we want to approve activity. Can you give us a policy engine? So this policy engine that was developed for one side of the market is very useful for the other. And we're kind of seeing that across the board. All of our customers are sort of pulling us upward. In a way, we're serving this really wide range. Actually, I think, creates a future-proof product.

Starting point is 01:26:08 I know the transaction is still in the process of closing. But can you talk at a high level about what drew you to wanting to work with the Stripe team? Yeah. I was in a telegram chat yesterday and somebody said, I was just like creeping. and somebody said, you know, there must have been that Stripe Charm or that Collison Charm. And I think it plays, you know, this is a company. The luck of the Irish. You luck of the Irish strikes again.

Starting point is 01:26:41 We weren't intending to sell. And I think the reality, this is something an investor has said, which is this is the most exciting time this industry has had in a long time. You've literally chewed through the worst environment for a startup over the past, however many years. why now? And I think the reality is because we are seeing things move very quickly. We want to accelerate that change. And we were very drawn to, I think, that the care for craft and the huge lever that basically this partnership we think can bring to bear on the space. So that was a lot of what was exciting to us is broadly. I think we share a lot of values for how to build great developer tooling. And then I think the ability for us to accelerate our timelines and to

Starting point is 01:27:26 accelerate the space more broadly in our own little way by just working that much harder with more resources for the teams that we serve was the really exciting part. Very cool. Well, congratulations. And good luck with the rest of the process. I'm really excited for you and the team and for Stripe and for what you guys can accomplish together. Yeah. And we'd love to have you back when you can talk more about the future and building. I'm really excited for this. And I think it's something that everyone's thinking about is like what's the future of fintech as fiat and crypto mingle more together. Yeah, it'll be incredibly bullish. Anyway, thank you so much for hopping on. We'll talk to you soon. Congratulations. Good luck

Starting point is 01:28:03 closing. Cheers. We have our next guest in the studio already, but first I want to ask you, how'd you sleep last night? Because I think I got your beat. I think I got you beat. I think it's a clean sweep this week. Did you get a hundred? I got a 91. Seven hours, 39 minutes. 93% consistency. Ninety-tie. This is the first time we've ever tied. Go to 8Sliap.com. get a pod five ultra code tbpn code tbpn five-year warranty 30-night risk-free trial free returns on the eight sleep it's amazing that was my last uh waking thought last night oh it's good to be back uh anyway welcome to the stream chris uh how you what's going on obviously uh yeah right yeah right not too much just everything just everything time is a flat circle yes i can't even believe it

Starting point is 01:28:52 seems like since we started the stream the whole internet went down so i'm just happy to be here you. Oh, he's going to be affected particularly because every time Chris posts, he posts a link to a Google Doc, Googles down. You're cooked, even though X is up. This is the most, your content strategy. Insane thing about you is that you can, you can break through the link ban somehow. It's just the ideas are so powerful. I love it. You know, nobody told me there was a link band. Yeah, built different. Yeah. Yeah, built different. Built different. I had a post once that I was like, like, you know, I post links and they still do numbers.

Starting point is 01:29:29 Google.com and it just. It flopped. John and I thought it was hilarious. You got to be built different to post links. Yeah, you have been built different. We've loved all the Google Docs that you've dropped. Maybe take us through the first one, the end of software. What was the thesis?

Starting point is 01:29:42 And then maybe we could reflect on some of that and how it's played out. Sure. Sure. So admittedly, kind of an incendiary. title, especially on Twitter to this audience. But the crux of the thesis was LLMs are were and are only getting better at producing code. And generating code has largely been a translation task assigned to humans. And humans have been incredibly proficient but expensive translators in that regard. and now that we have free translation from natural language to computer language, that should just completely change the economics and the production of software.

Starting point is 01:30:33 I think I saw something where CS grads are having a harder, having a harder time to find jobs than art majors. Yeah, we have our CS intern today studying art history for the next two hours to try to get up to speed. You can see. Do you have a fun fact for us? I got a fun fact. Okay. Okay. So did you know that Greek statues were, they were originally painted?

Starting point is 01:30:58 They were originally painted? They weren't just white marble. Oh, okay. I think your job odds just went way up. Way up. Back to the interview with Chris. Anyway. Sorry to interrupt.

Starting point is 01:31:09 Not at all. Yeah, but interesting fact about that, there's only 2,000 art history grads and there's 100,000 CS grads. So there is a little bit more sensitivity. But I think even though the facts of that stat might be like a little bit early to say, like the trend is clear and we all see it coming. And so we're interested in like the impact of that. In some ways, it's like the death of the software engineer job application. Like the job application now is like produce software. Yeah, built. Like produce a piece of software. The last, the last engineer that we hired made us a piece of software that was good, almost basically good enough for us to publish it.

Starting point is 01:31:45 Yeah, immediately. Same day. Same day. And so. of like at least the job application has been disrupted. Yeah. I mean, I think it's this sort of inconvenient truth that a lot of us are incentivized not to want to accept because it really challenges a worldview that most of our existence is constructive on top of. But I think the people who like rush head first into it and try to understand, okay, well, how does this change the future?

Starting point is 01:32:17 what's going to be different. You know, everybody else can be stuck clinging to the last vestiges of something that ultimately evaporates. I mean, but what's really exciting is, I mean, like, imagine a world

Starting point is 01:32:35 where steel is free. Like, what does the world look like if steel is free? A lot more steel? It's like the world just looks completely. different. It looks completely different. We can build so much more incredible things the frontier of innovation pushes out because we are pushing it out, but we're no longer limited by the cost of something to produce traditionally. That's great. Talk to me about the dynamic of as as code becomes

Starting point is 01:33:10 cheaper and cheaper and eventually free to instantiate, there's two theories. One is like we are in the age of the idea guy, the person that can come up with an interesting thesis or contrarian insight can instantly go instantiate it without a team of engineers. The classic example of the business guy who has the idea and just needs a team of software engineers to build it and can't do it. The other side is something I saw Pavel Asperuja posting about, which is that all those guys are going to be competing against the people with the software engineering mindset. And yes, the software engineers, they won't be writing code anymore, but they might be the future idea guys. They might be able to think in systems and come up with interesting ideas.

Starting point is 01:33:55 And even though they aren't doing the instantiation themselves, they are able to have an even bigger impact because they're the beneficiaries of this leverage. So is there one side of the argument that you fall on? How are you thinking about that? I mean, I think creativity and agency are going to be the scarce resource. because execution is just commoditized at this point. I think, you know, if we're going back to like word cell and shape rotator debate, for better force, I think word cells won, right?

Starting point is 01:34:32 Bold. It's possible. Very possible. Brutal truth. Brutal truth for shape rotators. But word cells just kind of came out on top because there's this Rosetta stone that instantly translates words into shapes. Yes.

Starting point is 01:34:48 Yeah, the prompt can be rotate the shape. Exactly. Just put the fries in the bag. Yeah, I mean, the, the wild, I think, moment for each of us is like we started going to chat GPT for tasks. Yeah. And you're just like, count the number of objects like this in this image. And it just writes like four minutes of, like, it works for four minutes, writing code,

Starting point is 01:35:12 doing work that would take. you know, at least a few hours to do as like a, is a well-trained software engineer just to do a task that in many ways you could just look at it and count it. Yeah. Yeah. Yes. Well, so here's an interesting, if you guys want to talk about something that please, it's probably my next essay. Okay. There we go. So hot off the press, hot off the press. I think there's a frame that really clearly delineates what's going to be, like, eroded by AI and what will never be eroded by AI. And, for example, like, we're never going to watch robots play sports, right? Like, that's just, like, never going to happen.

Starting point is 01:35:57 So why is that? We don't watch robots play chess. And robots have dominated in chess, and yet, Chess.com has done well as a business. Magnus Carlson has done well as an individual chess player, Hickaroo. Exactly. Exactly. Exactly. following online.

Starting point is 01:36:10 And you would think if you just wanted to watch the best chess possible we played, you'd be watching stockfish. And no one does. We don't. We don't. Not only that, but like the Olympics, we're so. Okay, so here's a frame. When it comes to utility, we are misanthropists.

Starting point is 01:36:28 We actually kind of hate other humans. When you're in an Uber and they like start talking to you, you wish you were in a waymo. You're like, I don't need this. When somebody, when you're at a coffee shop and they spin the iPad around asking for a tip, you're just like, I'm just, I want to buy this bottle of water. I want to walk away with my coffee. Why are you like this? This is like a miniature ransom. And you're just, you're frustrated because you know exactly what it is that we want to do.

Starting point is 01:36:59 And the human is preventing that from happening. And so I don't think any of us walk into an elevator. and think to ourselves, man, I wish there was a person here operating there. That's gone. It's totally gone. We don't care. We're actually very happy with humans being removed from that loop. And so as long as the thing is utilitarian and there's a drive towards efficiency, we actually

Starting point is 01:37:29 kind of hate humans and we want them gone. Now, on the other side, if we are, if we are allocating a leisure hour, not a labor hour, if we're not trying to be productive, but we are trying to be consumptive, we are narcissists. We love other humans. We love other humans so much that we would refuse to watch anything other than another human do that thing. So think about like, you know, fine dining, right? Omacaze, you go to a sushi restaurant. you need that person there that's describing cutting you want to see how delicate the craft is part of the craft and the social element to it is part of the product exactly it's a story yeah it's a story

Starting point is 01:38:17 you're paying for that experience you want that human experience you want to understand that this is something that they've dedicated their lives to so on one hand find an and you you need humans you need the story you want to understand where it came from like and then on the other hand, we order food and we tell the driver to leave it at the door so we don't have to interact with them. So like we go through life with these two spheres. When we know we want to do something and we really hate other humans, like, you know, when you call into a customer service line and somebody like incompetent is on the other side, you're just reminded of the frustration of why you called it in the first place. And then on the other hand, you know, let's look at modern

Starting point is 01:39:01 art, for example. I think like the trope in modern art is like, it's not that hard to make. You know, where the joke is like, my kid could do that. Yeah, well, like, your kid didn't. And also, no one cares. Yep. The thing that we prize is the constraints, the creativity, the mythos of the artist that went into creating that.

Starting point is 01:39:26 Yeah. If Daniel Arsham goes to a rock wall with an axe, and like hits, you know, hits the wall and it chips a piece of rock off, like, that rock would have like, you know, an immediate 10,000 X premium on it. Exactly. And like somebody would buy it because of you did it. Exactly. So if what you do basically is gets in the way of people getting exactly what they want, your job's gone.

Starting point is 01:39:58 It's going to go away. But if what you do is part of storytelling or human existence or entertainment, it's moated. It's a huge mode. I mean, one of the most interesting things in entertainment, I would say over the last few decades, was the invention of reality television. When we realized we would rather watch two people that we don't know fight than anything else. It's crazy It's incredible It's often incredible content

Starting point is 01:40:35 It is guilty pleasure for many people So on that note I'm curious On since we're streaming on X Are bots on X a feature or a bug Because As far as I'm concerned They've been designated a feature

Starting point is 01:40:53 Because it sort of went away for a little bit Now it seems like they're back And stronger than ever I think Twitter probably has the same stance towards bots that Louis Vuitton has towards counterfeiters. Where it's like, this exists because there's value in the product. Replicating, trying to imitate the authenticity of the engagement means that there's real value in the product. The more that it exists and proliferates, the more it dilutes the core value of the thing. but the fact that it exists in and of itself is a testament to just how valuable the product is.

Starting point is 01:41:34 Otherwise, people wouldn't try to imitate it. That's a good framework. On this concept of the future of jobs, Sam Altman put out a fantastic blog post and talked about how, if you went back hundreds of years and showed people what we do all day, they would say There's a ridiculous fake jobs. That's not real. That's not hunting or farming. That's sitting in front of a computer just messing around.

Starting point is 01:42:03 Talking to people. And yet we get a lot of value out of them. When we think about the future of work and jobs, do you think that we are adaptable enough as a society to create full employment with that new paradigm? Can we have 30 million omacase chefs and 30 million athletes and 30 million reality TV stars such that people by and large have meaningful work even at some sort of scale?

Starting point is 01:42:33 Or do you think there's some other way that this plays out? I think, yes. Like we will, we're such a selfish race or species that we want to, and the way the economics will work is the jobs will basically, job availability will expand to fill the leisure bucket that it sits inside of. So we have more leisure hours today than any era ever in the history of time because we're actually the most productive, right? If you want to go back to like the Stone Age, there's this really tight correlation actually between productivity boosts and the invention of leisure activities. So, like, tools get made. All of a sudden, we have free time.

Starting point is 01:43:25 We didn't have free time before because every hour was spent like surviving. What do we do with free time? We invent music. We invent bone flutes. Actually, the first board games were invented concurrent with irrigation. So all of a sudden, you have this huge productivity boost. And then you have all this free time. What do you do with the free time?

Starting point is 01:43:47 Well, we invent things to fill it. Modern sports were invented with the Industrial Revolution. All of the oldest soccer teams, football teams, were actually factory workers from the same factory competing against each other. Funny. And so we have these huge boons productivity. And then on the other side of it, we have job creation to fill that leisure expansion.

Starting point is 01:44:18 So, yeah, we're going to have, more athletes than ever. This is hilarious because it's actually incredibly bullish for art history majors. Because I'm thinking about it. So Ken Burns is about to drop a new documentary about the American Revolution. And I am so excited to watch it. And I know I could go to ChatchipT and say, build me a deep research report on the American Revolution. But I want to hear it through the Ken Burns lens.

Starting point is 01:44:46 And I want to hear his voice narrated. And I want to hear him express his unique viewpoint and the connections that he makes, even if they're not the best, even if I could get, even if I could get more factualness or more truth or more detailed insights from from chat chippy. I want the Ken Burns experience. And so, yeah, I don't know. There's something interesting about that where, like, I mean, you know, this is, we have a friend David Senora who runs Founders podcast.

Starting point is 01:45:13 And the way he, he reads biographies about great history's greatest entrepreneurs. and retells the stories in a way that's as compelling as the original material and as compelling as the real life. And you're getting it through the experience of David. And I just don't think that's going anywhere. Yeah. Even though it is in some ways like the first thing that could be one shot by AI with generative voice and deep research products. But I think it'll be sticking around like you mentioned. We're so narcissistic.

Starting point is 01:45:43 Yeah. We love stories about each other. Yeah. And we love storytellers. and we love telling stories. Yeah. And we ate in-offest, in-authenticity when it comes to that. I mean, like the ESPN ate the Ocho, like, is a joke.

Starting point is 01:46:03 But it's like, we have more sports than ever. Yeah. Because we have more leisure time than ever. Yeah. We care so much. You know, I'll watch competitive darts. We watched competitive darts 100 years. We didn't have time to watch.

Starting point is 01:46:17 I got really into bold. Riding recently and all the narratives there. J.B. Mami, the greatest bull rider of all time. I know the whole history. I love it. Yeah, getting into these obscure sports is, yeah, maybe that's the future. I just spend all my time watching bull riding. That's post-scarcity for you.

Starting point is 01:46:38 Switching gears a little bit. I mean, I wish we had a full hour to talk because there's so many different directions we could go. but I wanted to see get your reaction to WWDC this year, Apple's news around how they're going to be working with developers. I mean, there's a ton to dig into. What stood out to you? Sure. I know that everybody has like lost their mind about liquid glass and the new UI. I think buried in a lot of the press around WWDC is actually Apple's stance towards LLL.

Starting point is 01:47:14 LEMS and AI. So there's a machine learning team inside of Apple, and they are really, they're starting with this library called MLX, which allows inference on device for Apple iPhones and laptops for M-Silicon or Apple Silicon. And what's really interesting about that is as a developer, right now, when you think about creating an application that has end user inference embedded inside of it, you also kind of have to do the mental math of how much it's going to cost you because it's all metered. It's like, okay, well, yeah, I'm going to drop this endpoint into OpenAI or Anthropic or whatever. Yeah, you could make a viral app and then blow through your entire credit card limit. And that's like we've had a bunch of founders on the show that have like accidentally done that.

Starting point is 01:48:16 So and you think about early mobile days like we're seeing an explosion of AI apps right now. But but that's been gated by developers thinking in their heads. Well, if I make something amazing and free, I could actually just end up going into debt over it. But that's not fun. Yeah. Right. And in the early. Venture capitalists will save us in those situations.

Starting point is 01:48:37 Venture capitalists. Thankfully, they will come in. There are plenty of VCs that will rip checks on a handshake. No, but still, there's just so many, part of the beauty. Yeah, it's a real dynamic. Part of the beauty of it being easier than ever to create software is there's a bunch of software that can and should exist that historically shouldn't have existed. Because it's expensive for development costs reasons. Now we've been through this era of expensive inference.

Starting point is 01:49:01 Yeah. Yes. And so that point highlights the entire dynamic. or like, you know, imagine if game developers had to pay per pixel that was rendered by the people who play other games. It would be insane. Like, there would be no games made. And so what's so exciting is that it feels like we're starting to shift

Starting point is 01:49:24 into better, you know, Apple's really, in my mind, in a perfect position to put the tools in the hands of developers and enable on-device inference. which really means free inference. So if we were in this era where game developers had to pay for the pixels that were rendered on end user devices, we're moving to like, wow, like people can render the graphics for free on their device, and that just opens the floodgates for the things that can happen. I think haters will say, well, like, you know, models are bad, you know, the battery life is terrible.

Starting point is 01:50:08 life is terrible, your iPhone heats up, you know, the models are small. Yeah, today. Yeah. But tomorrow, a year from now, two years from now, you know, all of a sudden, 80, 90% of the queries that you would run through a cloud-hosted model is just doable on your phone or on your laptop. And for free. And natively, it can be offline, whatever you want, and drive it. And so that's such an exciting future. And from my takeaway from WWC, that was that was the most exciting one. And I feel like, yeah, and you combine that with again, they're not really getting any credit for partnering with Anthropic on Xcode, which will enable some of this activity as well. Yeah. Yeah, I mean, we need, we need AI, our AI Flappy Bird moment. Like Flappy Bird, I remember,

Starting point is 01:50:56 was built by like a single developer. And it's that perfect example of like leveraging all the hardware to just build something that went mega viral. We've seen a few of the, these moments for images, there was that magic avatar app that would make you look like Superman. That kind of went viral, but of course it had high inference cost. When you drop that inference cost to zero, you put it on the edge. I think we will have our flappy bird moment for LLM-based interactions on the iPhone. I'm really excited for that. It's completely unpredictable what that will be, but I imagine it will be very fun and everyone

Starting point is 01:51:30 will be playing with it when it happens. I want to stay on the topic of Apple and just this idea of taste. A lot of people have been writing about taste and coming back to taste as something that's potentially a permanent differentiator in a post-AI world. There's an idea of like agency plus taste being more valuable than skills. We talked to Scott Belski about that on Tuesday. How do you think is taste a moat? Is it something that can be learned?

Starting point is 01:51:59 What is the value? Is it still alive and well? at Apple. There was a lot of criticism about liquid glass and yet when we looked at it looked pretty cool. I don't know. I thought it looked nice. I think tasted alive and well. Maybe I'll describe Apple if you'll permit me to describe another one of my frameworks that we've developed here at Pace is top down and bottom up companies. So top down companies started by visionary charismatic founders. Apple, Tesla, SpaceX, Amazon, Epic Games.

Starting point is 01:52:37 And what's notable about these visionary founders is they're largely unattached to how they execute and what the products actually end up looking like. And so what happens is these top-down companies ship multiple flagship products over their life cycle. So, you know, we're all on Apple laptops and have Apple iPhones and AirPods and AirPods. You know, Tesla has obviously created everything from cars to power walls and solar panels. SpaceX not only makes rockets, but also satellite internet. Awesome. Amazing, amazing constructs. And what's interesting is the market only sees the products.

Starting point is 01:53:18 And so they don't understand that the core innovation engine is the company itself, because these top-down companies hire innovators and build up this core ability to summit the next hill and launch the next product. Bottom-up companies, conversely, are started by tinkerer founders. They're hackers. They try to pick locks, basically. And they combined with perfect timing, capture lightning in a bottle. And what's interesting about bottom-up companies is that they only ever ship a single flagship product. This is Google, Facebook, Instagram, Reddit, Airbnb, most actually venture Twitter, where we are. And they are actually fundamentally incapable of shipping a second flagship product.

Starting point is 01:54:17 It's because the product market fit for that is so violent. that every person that gets hired into that company is an optimizer. So imagine like you're creating a jawbreaker around this core product market fit. Every single person is hired as an optimizer. There are no innovators. Why would you stick your neck out to try something new when you can just make the core thing 1% better? And so they're actually fundamentally in case that you can't ship a second product as a bottom up company. and a lot of bottom-up companies don't know that they're bottom-up companies

Starting point is 01:54:54 because success is a terrible teacher. A lot of bottom-up companies expand through M&A because the businesses are so good. They're so good. And then if you can buy great products, you can optimize them really well and grow them and you look like you're, you can kind of mask the fact that you didn't actually make two of those sort of flagship products.

Starting point is 01:55:16 We were talking about this with scale. With scale AI going to meta, it's like, yeah, $15 billion is a lot. But what if it moves the market cap 1%? Yeah. What if it optimizes their AI stack 1%. It's like YouTube, Instagram, you know, bite dance buying musically. Sure. These are like maybe some of the best investments over the last couple decades and they were M&A events.

Starting point is 01:55:42 And so if you're a bottom up company and you can apply that pattern match really well to identify similar shaped assets, that's your core innovation lever because you actually are incapable of shipping something internally. So going back to Tate, but I think Tase is alive and well at Apple, yes, because culturally that organization has been built out to hire this innovative DNA. But it's not the same at every company. You know, I don't know if Tastes, it's famous Lester. I don't know if Tastes, I don't know if taste exists at i don't know if true taste exists at bottom up companies i i wonder actually if what we what we can see from bottom up companies is just ornamental window dressing that we misidentify as taste because it's yeah it's hiring a good design agency at the right time and and you're like

Starting point is 01:56:39 oh this this company's so tasteful but it's like well you had to spend 250 grand to be tasteful It's just the most legible thing. But when actually, like, Craigslist just works great. I mean, Craigslist is a perfect example of a bottom-up company. No taste required. Interesting. I would be rattling off like 20 other examples that kind of fit this framework, but I would piss off too many people.

Starting point is 01:57:08 So we'll let the audience imagine for themselves. But I think it's really powerful. Yeah, this was fantastic. We could go way deeper into all of these frameworks. works. These are great kickoff points for big discussions. I have so many more questions. So we'll have to have you on back soon because this is fantastic. Amazing. Thanks for watching

Starting point is 01:57:24 for the time. We'll talk to you. Cheers. And in the meantime, we'll talk to you about ad quick.adquick.com out of home advertising made easy and measurable. Say goodbye to the headaches of out of home advertising. Only ad quick combines technology, out of home expertise and data to enable efficient sales ad buying

Starting point is 01:57:40 across the globe. And our next guest is coming into the studio. Joe from Infinite Machine. I can't wait to launch. I can't wait to launch her out-of-home campaign, John. I'm so excited. We were deep in the creative process. We got to pitch Joe on doing an out-of-home campaign. Let's bring him in.

Starting point is 01:57:55 We'll hear his pitch, and then we'll pitch him some billboard ads, because I think this thing needs to be advertised. Break it down for us. Launch day. Launch day. Congratulations. Show us. Give us a little tour.

Starting point is 01:58:07 What are you building? Hey, guys. So you are here at our New York City facility. I'm calling from Infinite Machine. and we make the best non-car vehicles on Earth. And today we're launching our second product. It's called Alto. It is a class two e-bike, but it looks nothing like an e-bike.

Starting point is 01:58:27 We really thought this thing from the ground up. So it goes 25 miles per hour in the bike lane. It's got pedals that turn into footrests. It fits two people. It's made out of aluminum and steel. It's beautiful. It's useful. It's high tech.

Starting point is 01:58:40 It's connected to the internet. And we just announced it today. What's the damage for one of those? How much it cost? 3495. You're giving them away. You're giving them away. Double the price.

Starting point is 01:58:52 Giving them away. Yeah, yeah. I mean, it looks amazing. Congratulations of the launch. How much? So you basically reverse engineered this so that it didn't need a license. I was about to ask. So you basically, so somebody that's writing it, they get like, hypothetically, John and I and the team, we get a bunch of, we get a fleet of these for the office and we want to rip over to our favorite.

Starting point is 01:59:13 lunch spot. Yep. We're riding. It looks like we're on little motorcycles. Cops, you know, put the sirens on. They pull us over. We pull the pedals out and we're good to go. We're like, look, I'm sorry, officer.

Starting point is 01:59:23 It's a bicycle. I don't know what you're talking about. Yeah, so basically, most e-bikes take traditional bicycles and strap motors and batteries to them, but bicycles were never designed to be powered. And so you're taxing these parts that were never intended for motor power. And instead we said, okay, how can we, think about this from the ground up we have to work within the legal envelope of what a bicycle is so if you look in all 50 states in america the definition of a bicycle is something that has

Starting point is 01:59:54 pedals that are operable meaning it has a chain it goes 20 miles per hour as a class two e-bike or 20 or 25 miles per hour as a class three e-bike um and it's narrower than 36 inches that's the basic constraint and so uh you know We took that brief and then we designed what we thought was the best version of that design. What is the barrier to going more aggressive, building it wide or taking the pedals out, making it go 40 miles an hour? Then are you regulated as a motorcycle and I need a motorcycle license? Is that how it works? Yeah.

Starting point is 02:00:31 So if you're considered an e-bike, you don't need a license plate. Sure. And therefore, it's not a motor vehicle. It's not regulated by the state. It's not related by the federal government. And so, yeah, like it's just a much lower friction experience. You don't need to register the product. There's no insurance requirements.

Starting point is 02:00:49 You don't need a license plate. And so we wanted to make a super low friction product that you can get on. It's super easy to use. You don't have to deal with complex dealers. You don't need to register the thing. And yeah, you can get on and go. And it's excessively priced. Yeah, what's the go-to-market?

Starting point is 02:01:06 Is it important to demo these things in real stores, get them into traditional bike shops? I've been in some bike shops where they have e-bikes as well, or is this all D2C and e-commerce? So, I can show you quickly, we have a space here in New York. We're getting set up for an event. So there's a 13,000 square foot space in New York. Very cool. And it's going to be open to the public as a store. People can come in and buy vehicles.

Starting point is 02:01:29 And so we are going to direct consumer to start. Sure. And through online. It includes retail. Includes your own retail. Yeah. Yes, to start our own facility here in New York. We're also interested and open to selling them in channel retail.

Starting point is 02:01:44 Sure. I guess it was being a Best Buy, for example. Like it would make a lot of sense for it to be there. Exactly. Talk about the broader explosion of e-bikes because I feel like adults nowadays, the number one way I feel like a boomer is when, you know, some like 12-year-olds with e-bikes or like, you know, ripping through traffic.

Starting point is 02:02:02 And it feels like every time I drive in the city now at some point or another. And in some ways, like, you know, the boomer in me is like, oh, those kids are being, you know, dangerous. And then part of me at the same time is like it's actually makes cities much smaller now where kids can just go out and explore. But I feel like it feels like with with Olto and what you guys are doing broadly, we're kind of in the second era of this sort of micromobility, right? Maybe the first era was the kind of sharing economy, bird, lime model.

Starting point is 02:02:37 And it's clear that that form of transportation is, awesome, but the community-owned approach or the sort of sharing economy model had some really, you know, big, big issues. And it does feel like, you know, actually having people own the end product, care for the end product, and not be just, you know, trashing devices that are part of a network. So the way we look at it is, you know, the world, as we know, is urbanizing. More and more people are living in cities. And when you live in cities, things get denser. So instead of living in giant homes, you live in apartments. Instead of driving giant cars, you drive smaller vehicles. And we believe that over time, more and more vehicles in cities will not be cars. They will be smaller, smarter,

Starting point is 02:03:20 and we hope more beautiful vehicles. And our intention is to make the best of these things. Now, you're right. The first Gen 1 micro mobility companies were all sharing oriented. We think ultimately it's not a great business model. There's a reason why those companies have not done super well, gone into business, et cetera. But also, we think that there's an opportunity to learn from what made cars so successful. People love their cars. They love their vehicles. They have an emotional connection to them. And ultimately, the car was built as a symbol of freedom. And we think that there is an opportunity in urban environments to have a new symbol for freedom and a new tool for freedom. So this thing, you get on it, you can go anywhere in your city in 15 minutes,

Starting point is 02:04:03 no hassle, the wind's at your back. And it's just great. And compare that to the experience of sitting in the back of an Uber stuck in traffic. It's just night and day. Talk to me about what's going on in China because I feel like a lot of the phone companies have started going into cars and mobility and BYD makes all sorts of stuff. And they have a very different model. You know, with so many tech companies, it's like, are you going to wind up being a bullet point at WWDC or Google's I.O? it feels like it's unimaginable that we would see Apple launch a competitor to this. But what's the actual dynamic?

Starting point is 02:04:43 Why are America's big tech companies not going into different areas if they have experience in hardware? Like, why isn't Samsung getting in on the action? Why isn't Apple doing anything here? Can you comment on like the reaction to just like the way American companies are building technology today versus the China model? I actually think it's a bit of the opposite. So, yeah, you're right.

Starting point is 02:05:12 If you go to China, you have these new hardware companies that are multi-category. So, yeah, you go to a Gehamee, and they started making phones and laptops, and now they're making cars. When you look at the U.S., the model, take a company like Apple, we love Apple, but they're very, very conservative. They stick to their lane, and the other American tech companies are actually not great at building hardware. And so the way we see it is that there's an opportunity to build the next great American hardware company. We're starting with Key One and then Alto and with vehicles at large, but we call the company Infinite Machine because we have big ambitions and we want to make incredible physical things in the world that move in our high tech in new ways. And so we will be multi-category at some point. I mean, we're really building the competency to build high-end moving beautiful hardware with an integrated software layer and do it in a way where we can move really quickly.

Starting point is 02:06:06 I mean, this company, I've been full-time for less than a year and a half. We have raised less than $10 million. We've productionized two motor vehicles. So we're kind of in a new age of what I call the ultra-lean hardware company. And we're going to keep pushing and we're going to take that competency and double down, especially as we enter an age of embodied intelligence where the AI kind of escapes the screen. And we're going to be all over that. We're positioned to take advantage of that stuff.

Starting point is 02:06:36 we're building the competency to make things that take advantage of it. I'm just going to speculate, but I can imagine you guys making a horse buggy without the horse that's fully autonomous. And you just get to ride around the city, you know, windows down. Yeah, you see the carriages in Central Park. Autonomous horse carriages. You're going to replace that as well, for sure. Please.

Starting point is 02:07:00 Or a robotic horse. Yes. I love that. Yes, yes, yes. Henry Ford said if you ask if you asked customers what they'd want, they'd want a faster horse. I want a faster horse. I'd love a faster horse. They're already self-driving. They're out of the box self-driving. Go home horse and it'll just take you there. Exactly. Has your world been rocked by the tariff war? Are we in the clear now? Is there any fear around scaling up?

Starting point is 02:07:27 The battery supply chain, some companies have been buying on the battery side. The more defense application companies like Skydeo run into a lot more issues there. What's that been like? I know that everyone's focused on reindustrialization in America, but it's a long journey. Things have stabilized on the tariff front. You know, it's not great. The world economy is highly entwined. The supply chain for electric vehicles, even domestic manufactured electric vehicles,

Starting point is 02:07:56 is highly connected, connected to China and beyond. And so we are all, you know, we are fully supportive of reinduced. industrializing America. We want to be on the cutting edge of that. We support your smart policy to do that, both carrot and stick. So tariffs are a stick. You know, there's an opportunity for a new industrial policy that incentivizes domestic production. We haven't seen that yet. We've only seen a stick. So short story is, you know, it's stabilized, but honestly, I think we should be playing offense as America and we're not yet. We're using kind of protectionist measures. Yeah. Talk to me about making these vehicles safer using technology. I remember like the hoverboard was a beneficiary of, you know, better compute power and the ability to stabilize a device while you're riding on it very, very cheaply. And so those kind of went viral for one year in like 2014 and then kind of disappeared. But I imagine that there's things that you can do now that you maybe couldn't have done if you were trying to

Starting point is 02:09:02 build this company 20, 30 years ago. What technology trends are you benefiting from? You know, on the safety front, these are lower speed vehicles. So you're not moving at, you know, highway speeds. And so the name of the game is awareness. People, not just for the rider, but also for people around you. So that's emitting a sound that's being highly visible. And that is alerting the rider when there are issues. And we do that stuff. You know, I think over time, we'll start to use some of the ADAS software from automotive in these vehicles to do alerting and also to take over the vehicle when that presents itself. But to start, you know, there are millions and millions of people riding e-bikes right now.

Starting point is 02:09:48 If we can make them 10% 15% safer, that's a big win and that's where we're starting. Talk to me about actually getting these in the hands of New Yorkers. Well, just talk about getting him and let's getting OUtto in the hands of us. I'm going to sign. I'm going to, we're going to configure one for each person on the team. And then we're just going to be emailing you every day. Hey, hey, what's the wait time? What's the wait time?

Starting point is 02:10:17 We are offering, we call internally, the Burning Man shipping option, which is straight from the factory in August before Burning Man. If you opt for non-Burning Man shipping, that costs $4,500. bucks but if you offer non-burning man shipping uh that will arrive in late september early october so you know they're shipping quickly we are about to start mass production um and maybe for you guys will expedite if you that'd be amazing no i mean i think that you will get the equivalent of millions of dollars in in marketing spend from just being a burning man these you can imagine you know if you can get a dense of the density of these on the playa the number of of

Starting point is 02:10:56 Instagram stories, posts, et cetera. It's going to be absolutely wild. Yeah, yeah. What about storage? Like, what are you recommending to people that buy these and are living in an apartment? Is it something that you're designed to bring into your apartment? Or do most buildings these days have like in like safe places to store e-bikes or parking garages for these things?

Starting point is 02:11:17 Yeah. So one of the cool things about Alto is it's designed to be left outside. You park the vehicle outside. It's fully weatherproof. It has an onboard security system. It's virtually impossible. to steal hardware and software. When you park it, you don't have to take the whole bike to your apartment to charge.

Starting point is 02:11:32 You can take the battery out of the seat. Waste 20 pounds. It's a hot swappable battery. Take it up to your apartment. And then we have a dock that you drop it in and the dock plugs into a normal 110-volt outlet. So super simple. So you charge it overnight and then just take it out every morning. Six hours to charge.

Starting point is 02:11:50 We also have a supercharger. It takes three hours for a couple hundred bucks. Yeah. Very cool. Amazing. Anything else? Please make a video of yourself taking one of these things on a jump. I just think it would be fun.

Starting point is 02:12:02 I was going to ask, when can we get these X-gings? That's the marketing strategy here. I want to see the big jump, you know, monster truck rally. Let's push these to the limits, really show people with what's capable. I think we should get one and put the full TBPN racing livery on it, all the ramp on there, all the different partners. Are we doing like Pike's Peak or are we doing like half pipe? Like, well, what's going to go most viral here? All of it.

Starting point is 02:12:25 All of it. Okay. I'm super, I'm super, you know, this is one of the, I love businesses and products where you take something that seems like, you know, seems like it's competitive. And then you spend a hundred times, you know, you be a hundred times more intentional about creating it, designing it, distributing it. And you create, you know, an entirely new. Also, shipping before Christmas, I love a physical gift, a big box under the

Starting point is 02:12:52 Christmas tree. It's rare and rare these days because everything's gone digital. You used to give people books or CDs or DVDs and all of this is digital now. So it's very hard to find something to give to someone. Although this is obviously like a very nice gift, it is something that can really make a statement and be an absolutely joyful Christmas morning. If you unwrap this thing, we've been recommending that people give each other, you know, sales tax software or potentially corporate cards. But I think that. think this is going to be added to our Christmas list this year. I think this is a little bit more fun. Yeah. And people can place, if folks want to make sure they get one for the holiday season, they could put a deposit down now. It's 100 bucks. Oh, cool.

Starting point is 02:13:34 And it's fully refundable. So, okay. Save your configuration. Fantastic. Well, thank you so much for stopping by. Yeah. We'd love to give it a try. We'll talk to you soon.

Starting point is 02:13:42 Very excited for you guys. Have a good life. Bye. Awesome. If you're operating a wander, you've got to get one of these for your guests to be able to tool around on. That would be fantastic. Imagine this probably goes really hard in Malibu, Ohio, any fun places, Sun Valley, any place where you're traveling.

Starting point is 02:13:59 And if you're looking to take a vacation, get on Wander. Find your happy place. Find your happy place. Book a Wander with inspiring views, hotel-grade amenities, dreamy beds, top-tier cleaning, 24-7 concierge service. It's a vacation home. Wander allows you to go golden retriever mode because you just don't have to think about anything. Yeah, it's fantastic. Chasing the beach.

Starting point is 02:14:20 Well, we have Zach again from Plaid. third time on the show. Welcome to the stream. Great to see you. It's always a pleasure. I'm always blown away when you join because it's always so much fun talking to you. So thanks so much for joining. Great to have you. What's the latest in your world? Well, thank you so much for having me. I'm honored to be here and it's always fun. It's so much fun. Latest in our world is we did Plaid Effects today. This is our annual customer conference where we announced a bunch of the stuff that we were working on for the past year. It's been super fun. The biggest announcement was a product called Protect, which is this amazing cutting edge anti-fraud product, looks at all the data across

Starting point is 02:14:56 all of the usage and the data flowing through the plot network and builds a really great anti-fraud tool. How that's the headline of the day? How one-shotable is fraud these days? Can you just pipe a ton of customer data into like a single context window and just ask? And can the prompt just be like, is this fraud now? Or is it still like leveraging all the traditional fraud tools? Yeah, I mean, the hard part is twofold.

Starting point is 02:15:22 Okay. First is you have to put all of your historical data in it to say, hey, this pattern anomalous is something weird going on here. So the question is like, can you get all of your data into that? And in Platt's case, can you take all the data that we see across all of the fintech apps and all the users and say, hey, what are the patterns that are occurring consistently? So for example, you know, if you sign up for Coinbase today, that might be a totally normal sign up.

Starting point is 02:15:47 You sign up for a lending club today and you just signed up for six up the lending products today, then that's obviously a potentially fraudulent loan that you're trying to take out. And so, like, can you get all this data all in one place? That's that's problem number one. The problem number two is can you analyze it and change the analysis that you're doing at the speed that you need to change it because fraud changes every day. Some of that is algorithmic. So are the algorithms like chasing the fraudsters in the way that they need to? Some of that is manual. So we built a tool that actually allows humans to go in and say, hey, I heard from, I've seen this thing going on or I heard from some other fintech company that this,

Starting point is 02:16:19 fraud ring is occurring, can you search for it and tell me, is that actually occurring to me? And then can I build some rules to protect against it? So it is a constantly evolving game. Yeah. How important is this problem? I feel like every American should just default, assume that their information is like floating around the internet available to the highest bidder. And so like in some ways, like the only solution is just better tools to prevent this type of activity. Is that, is that reality? That is the reality. For national fraud is going up around 25% a year right now. Some of that is, or actually a lot of that is AI driven. So, you know, people are using AI tools to send text messages to convince people to do certain things. I think,

Starting point is 02:17:00 yeah, one answer is absolutely tools. And we're trying to build the best in class tools and give it to all the fintech companies. And frankly, we'll go well beyond fintech with these products. The other is just education, like explaining to consumers how it works. So, you know, my parents and I, we have a word that we say to make sure that we know that it is the other person on the other end of the line if we're sending money or frankly doing anything that requires authentication. Those kinds of little tricks that you can build into your day-to-day processes are always helpful, but it's a combination of tools and processes and so forth. Yeah, makes sense.

Starting point is 02:17:32 Can you talk to us about impossible travel problems? I found out about this yesterday, where someone's logging in from Singapore three hours after they were in San Francisco and they couldn't possibly be there? Is that something that you're already built a tool around? Is that kind of like table stakes or are there like evolutions there? Yeah. So I mean, a lot of the way that the anti-froid product works, it's called Protect. It's got a bunch of these different analysis analyses that we do.

Starting point is 02:18:01 So there's, I think, just under 20 different analyses that are running now and we're adding more every day. One of them is location tracking. So we look at where was the last place that you logged into a fintech app? And then if you're trying to do a new thing with a different fintech app on the same user account in a different place, all right, great. Well, that's actually, you know, it's an impossible travel question. And so that will immediately generate a flag. There's a bunch of these other things. Like, we look at stuff like, how long is this bank account existed?

Starting point is 02:18:25 It turns out a brand new bank account is very likely to commit fraud. And an old bank account is less likely to commit fraud. You do stuff, same thing for the phone number. How long is this phone number existed? So there's a zillion of these analogies that then run into a bunch of statistical work that we do on top of it. But, you know, every time you look at a fraudster or an instance of fraud, You can always explain it in retrospect. So you can look back and say, oh, yeah, obviously, you know, that person was in Singapore and then they were in California two hours later.

Starting point is 02:18:52 That's impossible. But when you're proactively trying to prevent it, like can you actually think through all the possible permutations of weird patterns that might exist? And that's that's the complexity that we focus on. So I imagine at one point this was like very heuristic driven, very like a multivariable regression almost, just saying like here are the different fraud factors. Let's add up like a risk score. Have these products move to a transformer architecture? Are we using LLMs in them? Are we using modern AI tools?

Starting point is 02:19:22 Or are we just kind of continuing to scale up the existing algorithms that have been working for the past few years? I don't know how relevant the new models are. People are kind of like throwing transformers and diffusion in everything right now. And I don't know if that's appropriate here. Yeah, I mean, the answer is yes to all the above. It is not heavily transformer driven yet in terms of fighting fraud, though increasingly it is becoming that way. We definitely use LLMs for a bunch of things like auto building tools, like doing some of the data cleansing, so and so forth on the back. But fraud is a fascinating industry because if you talk to any company, their goal is to make the fraud not happen on their platform.

Starting point is 02:20:06 They want the fraud to be more expensive for a fraudster to do on their platform than it is anywhere else. Sure. And so everyone wants to be ahead. They want the new thing. They want like the new type of fraud solution. But then eventually everything that's new becomes stable stakes for everyone because, you know, you kind of have to have it. And so, you know, we are iterating our way towards more and more complex architecture,

Starting point is 02:20:27 but it's still early days. Yeah. To me, to me this product feels like something that's just possible with like massive scale. Like you probably wanted to build something like this, like, you know, five, six, seven years ago, but couldn't just because you didn't have like the sort of broad exposure to all these different touchpoints. Is that right? I mean, this was the pitch that I was giving for Plaid in like 2014. Basically, my pitch was when we get big enough, we're going to have a lot of data and we're

Starting point is 02:20:54 going to build these analytics that financial services doesn't have because, you know, no one can build them and no one has the tech sophistication to do it. It took us a decade to get to the point where we had enough data to actually build this stuff. And then we acquired Cognito and ID verification company in 2022, which is a big catalyst. in part of the pitch to that team. I was saying, look, come use our data, put it together with your data, and then think of all the amazing anti-fraud tools you could build.

Starting point is 02:21:17 So, I mean, this stuff always happens slower than you wanted to, but we couldn't be more excited to get it out there. How much of this, like fraud right now, financial fraud, is kind of acting as like a tax on fintech companies, and should we be thinking about shifting that tax to the government by giving the government more tools to go after the fraudsters, like at the source and just arrest the people that are even trying to do the fraud. Well, I think one of the big challenges is in crypto, a lot of fraudulent activity is happening

Starting point is 02:21:49 and law enforcement is able to trace the activity to somebody in China. And if you go to some, if you go to the Chinese government and you say like, we know exactly who's doing this, they don't, they're not going to act on it. So, like, it's. Yeah, I mean, the answer is yes, yes to all the above. Like the government should be more active in this. again, much of the, as you said, much of the fraud doesn't stem from people that are inside the U.S. so it's not the U.S. government doing it.

Starting point is 02:22:15 I'm sure you guys have been following, or maybe you've read the articles about pig butchering. It's a horrible name for a type of fraud. They have these humans in fraud rings in Malaysia. Relationships with people, right? Yeah, exactly. And like they're oftentimes government sponsored or at least government like affiliated in some way. Like there's knowledge in the government that happened. And so, I mean, you know, if that is happening, great, yes, we could give better data to the Malaysian government, but also you've got to build the tools.

Starting point is 02:22:45 Yeah, yeah. What about, I don't know how much you can comment on this, but, you know, there's been a lot of discussion recently and frustration, and at least after the Coinbase, you know, user data issues around like KYC laws and effectively the government requiring people to retain all this. data. Do you think that there would ever be any movement in D.C. around really updating these laws to protect, to help protect user data, or is it just one of those things that we have to live with? I mean, you would think they should. I mean, realistically, yeah, the concept that if you're a bank and I sign up for your bank, you have to collect all of my most sense of data, and then you have to store it forever in your database. On a scale of forever in, you know, many, many databases, you have to assume that those databases will be hacked.

Starting point is 02:23:38 Thus, you have to assume that on the scale of forever, your data is all going to be public. And so it doesn't make sense that you would retain it. You should, in theory, collect it, verify it, like get, say a token that it exists, like get some assurance that exists and then delete it. It seems like, I would hope that the regulations change. I think it's, you know, regulations move as fast as regulations move. And so I'm not sure that it's going to be imminent in that sense. But on the flip side, like the industry is solving it.

Starting point is 02:24:06 The industry is building tools that are better and better. I mean, relying on a static social security number, sure, yeah, we'll do it. That's fine if we have to for regulatory reasons. But the reality is if social security number entry is your only fraud detection tool, your only fraud avoidance tool, like you're in a bad place as it is. Yeah, that makes sense. The industry will solve it, basically. Can you talk, this got kind of lost among other massive news cycles, but there were

Starting point is 02:24:33 some updates to Dodd-Frank, 103 that a few weeks ago. You guys talking offline were kind of concerned about what that could mean just kind of broadly for consumers. Can you give us kind of an overview of what happened? I know, I think when we were in D.C., you were kind of covering a little bit of this stuff on the show as well. Totally. Yeah, I think this is a big issue that probably doesn't get as much coverage as it should. So in the Dodd-Frank law, they wrote a provision that basically says consumers have access to their financial data and they can assign that access to third party of such as Plaid, to act on their behalf.

Starting point is 02:25:15 And then there was never any rule written on that. It was regulatory authority that was given to the CFPB. No rule written for a long time. Under Trump first, the first administration, Kathy, Kraninger, the head of the CFPB at the time, started writing the rule for this. It then continued into the Rojo-Kh Chopra era of CFPB under Biden. The rule that emerged from that was a very far-reaching, huge overstep of a rule. And in the second Trump administration, now there's a push to either pare back or get rid of the rule. What actually happened is the day that

Starting point is 02:25:51 the rule was published by Rojochopra, a group of the banks sued that same day, I mean, they had the lawsuit pre-prepared beforehand and just this was quick to submit. And so that kind of is ongoing. The wrinkle that's come up recently is that the currency of PB has basically said something effective. Yes, we agree with the bank, so we should get rid of the rule, but also the way that they're structuring the language in that dismissal make foreclose on some of the regulatory things that were set up and could actually put at risk some of the ways that open banking works in the United States today, meaning it could allow certain banks to put much more pressure on the ecosystem. They could try to turn off data in certain areas. They should try to limit data in certain areas.

Starting point is 02:26:36 They could more likely, what's going to happen is they're going to try to limit certain industries. So if you remember, you know, a couple of years ago, there was this big debanking thing that went on where the banks were trying to push back and turn off crypto. Yeah, this would give them the ability to turn off crypto. This would give them the ability to turn off, you know, gaming. or things like that, or basically any industry that they might feel as competitive with them. And it wouldn't be all of the banks doing it as a group. It would be each individual bank kind of making their own determination of this.

Starting point is 02:27:06 And so we think that, you know, we don't am I getting rid of the rule. We just don't want to foreclose on the fact that consumers really should have the ability to choose the financial products that they interact with. Yeah, that makes sense. Did you have a reaction to the Circle IPO? You know, obviously it performed very well, and stable coins have been in the news. But I'm interested to hear kind of like what the next most exciting instantiations of that technology might be. I was talking to a payroll company CEO and he was kind of mentioning that, yes, it's lower fees and faster.

Starting point is 02:27:43 But interestingly, stable coins might allow them to kind of recapture float and then earn yield on that in an interesting way. I wasn't expecting. So like the potential economic benefit to some companies might be. be massive if they're able to kind of custody their own treasuries, essentially, through stablecoins. What have you been seeing in the stable coin world broadly? What are the interesting knock-on effects? How much excitement is there among stable coins from your partners?

Starting point is 02:28:13 Yeah. I mean, there's a ton. Yeah. There's a lot of amazing companies. I think you talked to privy. Yeah. You already talked to them. You're going to talk to them.

Starting point is 02:28:20 Just an hour ago. Just earlier. Great. Yeah. I mean, huge got to them, like, amazing to see what they built. And it's very excited for the ecosystem. I think, you know, from our perspective, what's happening in stable coins right now is not what will happen. So what's happening right now, a lot of people are using it as a currency hedge.

Starting point is 02:28:35 A lot of people are using it for cross-border payouts or money transfers that otherwise would just be frictionful. And then, you know, a lot of people are using it just as a form of a savings account that has higher yield. And while those are all big tams, it's not that interesting relative to the rest of what could go on in financial. services. I think increasingly we're going to see more and more consumers having the ability to move money easily between, say, a bank account and a crypto wallet holding something in USC or another stable coin. And yeah, you're right. There are big opportunities for people to hold elements of their treasury or elements of their float and earn meaningful interest off of it. Increasingly, I think that that float, which was historically held onto by the platforms, may well get shared more and more

Starting point is 02:29:22 with the underlying companies. So payroll companies, if the float was all held at the bank, now they're actually able to participate in a portion of the float. So that's a huge new revenue stream for them. While interest rates are what they are, I will say. So that'll create increasing volatility in their revenue streams. If interest rates go down, then obviously the float goes down. So I do think that that's a big potential thing that a lot of companies might want to.

Starting point is 02:29:45 Also, offering more tender types allows a lot of the large platforms to keep dollars on their platform more. So you see some of the big payments processors. Their goal obviously is to have more float on their platform because they get a slice of the float, offering more tender ties, offering it not just to be held in dollars, but you could hold it in crypto. You could hold it in a stable coin. That keeps the dollars on their platform even more, which gives them with more leverage and more revenue in the long term. Very cool. What are you guys doing on the MCP side?

Starting point is 02:30:13 I know you had some announcements recently there. Yeah, so that was another big part of the announcements today. I mean, the short answer is like MTPs make it easier for developers to build. I think the concept of a rest API was new when we started Plaid because everything was on soap. Now, the concept of, you know, writing to a raw API without using cursor or windsurf or something like that is increasingly becoming rare and will become rare every time. And so really for us is just staying on the cutting edge, building tools that help developers build faster on Plaid and then allowing them to get better analytics on their usage, their interactions service system, so and so forth. So it's fun. It's really fun.

Starting point is 02:30:49 Do you have any type of broad thesis around how agent, you know, I think there's like a lot of narrative around, okay, agents will just default use crypto assets or, you know, anything around that. Anytime there's sort of an accepted narrative, I just, you know, want to kind of push deeper. I'm curious if you have strong opinions. I do not think they will default these crypto assets. I think crypto assets might be easier at first, but look, the bank account is already effectively a deal. digital first product. Or is now with a digital first product. Did not start that way, certainly. Cards are not yet fully digital first, but are moving that direction. Like still you have to know a 16 digit card number to actually like authenticate a card.

Starting point is 02:31:35 Whereas bank accounts have Oath already set up. And then the next phase of that is obviously crypto. Because crypto starts being digital native. And so it becomes easier and easier to authenticate into things. My take is that all three of those and many other tender types will become available. to these uh to to to to to to to to to to to elements uh and and agents but um it's going to take a little while for the industry to evolve but fast forward to three years and it'll be all the same you're cool anything else story no this is great this is fantastic congrats I'm curious did did did did Chesky tell you to move to this kind of release cycle or you've been doing it for a while it's sort of the Chesky that's the Chesky method the Chesky method no we've been doing it for a little while but man they have developed such an impressive track record of doing it and All the podcasts that he went on to talk about it. I think everybody else in the world decided to copy it.

Starting point is 02:32:23 So maybe we were inspired by seeing what they did beforehand, but certainly we've learned a lot from listening to their playbook. Yeah, the jobs approach, you know, walked so the chess key method could run. Awesome. I'm excited to read through the other announcements later and come back on anytime. Congratulations. Thanks, have a good one. Cheers.

Starting point is 02:32:45 Really quickly, let me tell you about Bezell. Go to getbezzles.com. Your Bezell concierge is available. now to source you any watch on the planet. Seriously any watch. I love that. There was one YCT team yesterday that had the Daytona and the Texas

Starting point is 02:32:59 TimeX. Yeah. The day date was good. I saw JLC Reverso. Yeah, yeah. When we called the founders on it, they were like, they were embarrassed. It's not super YC coded to have a hitter on the wrist, but nothing to be embarrassed. You got 500K, throw 100K into a

Starting point is 02:33:14 petac, you know? No. 20% of it. That's the Cluley method. Absolutely not. That is the Cluelly method. No. Build a generational company and then buy a nice piece on Bezell for your father. Father's Day is coming up.

Starting point is 02:33:29 Head over to Bezle. Get your dad a watch. Let's check in with our intern, Tyler, who's pivoting from computer science to art history. What do you got for us, Tyler? Oh, he's at the Museum of Modern Art now. That's fantastic. Yeah. So earlier I was thinking, like, I think before I should really start studying, I should kind of just like benchmark my stats.

Starting point is 02:33:47 Okay. So I took two different tests. The first one was it was an AP art history practice exam. Okay. So you know it's like for high schoolers, advertised schoolers like pretty dumb. Yep. I thought I'd do pretty well. I got 12 out of 29, so that's 41%.

Starting point is 02:34:02 So failing, you know, some place to go there. And then I also took like, you know, maybe that exam is just like weirdly, you know, it's like very specific knowledge or something. I took another exam. This was the Britannica art history quiz. on that one I got 24 out of 60. So that's 40%. So, you know, I have a lot of rules growler. There's about an hour left to the show.

Starting point is 02:34:29 Were you using cursor for art history to answer those questions? Well, yeah, I was not. But that was earlier in the show. I have been studying. Okay. So I think I'm going to take it, you know, 20 minutes before the show ends. Okay. We'll see how I do that.

Starting point is 02:34:41 Okay. But I was also kind of just looking, you know, at what are the kind of careers that art history majors generally go into. Yes. So, like, if you kind of want to stay in the art world, there's basically, you work in a museum or you work in kind of art dealership. Like a gallery. Yeah.

Starting point is 02:34:58 Yeah. So first I looked at kind of becoming like a museum curator. Okay. For that, it's, you basically need a PhD. Okay. So I don't think I'll be able to get that before the show ends. But I was looking at art dealership. Okay.

Starting point is 02:35:11 And obviously, like, for that, you kind of need some funding to start with. So I took producer Ben's ramp card. Okay. Okay. Lay down some art pieces. They're going to come in next couple weeks. Great. There you go.

Starting point is 02:35:22 So, yeah, I'm already starting this career. I'm feeling pretty good about myself. You flip those pieces, make it all back, pay off the ramp card, reword business. Yeah. We will be checking in with his IRA, seeing if he can beat the average venture fund. Yeah. Yeah.

Starting point is 02:35:36 Just beat the market. Yeah. It's that easy. Amazing. Great work, Tyler. Yes. Yes. Making lots of, lots of promising, uh, promising, uh, promising, uh,

Starting point is 02:35:44 promising progress. That's very good. Anthropic co-founder Ben Mann says that we'll know AI is transformative when it passes the economic touring test. This is something I was talking about earlier. Give an AI agent a job for a month. Let the hiring manager choose human or machine when they pick the machine more often than not. We've crossed the threshold. We've been talking about the economic touring test for a while. I'm glad that he wrapped it in a bow. Maybe it needs its own coinage. Maybe we need to dig into Anthropic more deeply. But we have the perfect guest for that. We have Sharon from Bloomberg joining us right now. Welcome to the stream. Sharon. How are you doing? Good. Thanks for having me. Thanks so much for being here. I wanted to have you on

Starting point is 02:36:30 because you've been on an absolute tear writing fantastic pieces. Maybe it would be best to start with Anthropic, which I was just talking about. What did you learn from the piece that you wrote about Anthropic and Dario. We haven't had a chance to have him on the show. We talked to Shulto. They have an interesting culture, interesting history. Can you give us your kind of 50,000 foot view on the company, what they stand for today and maybe how that's changed? Yeah, I think Anthropic is a company that probably the most right now is facing pressure both to live up to its really strong kind of ideological ethos, which is around building AI safely as and as an alternative to labs that they felt maybe weren't doing it as responsibly as they were.

Starting point is 02:37:17 And then also to deliver on the business value, right? So I was kind of, I learned just how sort of Safi actually Anthropic has been with Dario at the helm, even though he has this very deep research science background. He's been able to cut some major deals, grow the enterprise business quite quickly, double. And now there's other reporting about tripling revenue. So, you know, that's sort of where they're at. But as you get more money, as you get bigger and bigger funding rounds, you know, how does, does that translate into pressure to move really fast at the same time that you're trying to be responsible?

Starting point is 02:37:52 Did you get a sense that they have a, they have a specific strategy around diversification in terms of hyperscalor partnerships? Because they've done a deal with Amazon, I believe. They were just mentioned at WWDC as plugging into Xcode. It feels like they're trying to maybe go more of as an infrastructure, provider, they seem to be doing very well in coding and on the enterprise side. How much of that is deliberate or just like it got pulled out of them because of where their position in the market was? I mean, so in my reporting, like it was very deliberate. Their strategy to, first of all, not lock in with any one hyperscaler, right? So they have both Google and Amazon as partners on the

Starting point is 02:38:33 front. That kind of hedging, I think, was very calculated. And also they, you know, Dario and my interview he's been pretty dismissive our reporting about Stargate, you know, saying it's not clear exactly what it is. And I think his stance is that sort of you can you can scale up and not need to make such a big deal about it. At least that's what he's sort of publicly to me, right? But as you said, like, I think the pressure's on, right, for them to deliver and show that they can. He invented scaling laws, but can you actually build up the compute and pay for it to do that? Can you talk a little bit about the rhetoric around safety. I was very,

Starting point is 02:39:11 I never really got caught up in the paper clipping in the AI doom narrative. But after seeing DeepSeek and the... That's what every human says right before they get paper. But after seeing Deep Seek drop in the Anthropic paper about Manchuring candidates or

Starting point is 02:39:29 what was a secret hidden agents, I forget the term that they use, but essentially a sleeper agents, embedding something nefarious within the weights of the models that maybe only comes out when it interacts with a certain endpoint or a certain web page, it discovers that, hey, I'm not just running, you know, in some random solo devs cursor instance. I'm actually inside the NSA.

Starting point is 02:39:50 I should go around and do something else. And that really flipped a bit in my mind to say, okay, safety research is actually maybe extremely important and maybe even extremely commercializable because people will want to do safety evaluations on the models that they use. How is the safety rhetoric evolved at Anthropic over the last couple of years in your reporting? Yeah. So I think, you know, AI Dumers or people worried about AI safety have been doing these sort of hypothetical thought experiments, like the paperclip experiment for a very long time, like predating even Dario Emaday's right entry onto the scene.

Starting point is 02:40:26 However, I think in Anthropic, what I've seen evolve is they're starting to outline and trying, they're trying to be as concrete as possible to explain like what the risks are, what the different levels of risk. are. They made a whole responsible scaling kind of document framework to say, okay, it's kind of modeled after how biological labs are that are very high risk and how, okay, if what you're working on is this level of biological risk, then you have to have this corresponding level of actual physical lab safety. So that's how they viewed an anthropic. And so, you know, when I was reporting this out, they had recently had this kind of false alarm where they thought they were crossing into a level that was too dangerous for them to be able to handle. And then they did extra tests and they're like, okay, no, we're all right for this release,

Starting point is 02:41:12 but we better get up to par. What time period was that? Was that like an hour or like 10 days or a month? So, you know, Dario told me they basically had to delay, not this release, but the last bike won by something like nearly a week because at the last sort of in the final days leading up to the release, their safety team, you know, one of the, the, red teaming group came to them and were like, listen, like, this model actually may be more capable than we thought.

Starting point is 02:41:40 Right. So I do think that's a challenge, right? It's taking these really hypothetical, seemingly kind of sci-fi scenarios and translating it to like, well, when is that actually going to happen at our lab? And what do you have to do to protect yourself? And it's both physical security, like making sure no one can break into your lab or something like that. But it's also, you know, actually making sure that there are.

Starting point is 02:42:03 are safeguards on the software side to stop someone from saying. And I think the most tangible threat that I heard them talk about was the biological one, right? And that was a specific concern with that release I mentioned, that could someone use this to come up with a novel bio weapon? Sure. There's an interesting dynamic with safety teams where if you're sitting there with a team of, let's call it, 20-something people and you're costing the company tens a million,

Starting point is 02:42:29 you know, I don't know, $10 million a year, it's not a, you're not going to, feel that like your your job's that safe if you're if you're never sounding the alarm sure sure sure sure sure sure sure this feels a little dangerous let's look into it yeah you know I mean there's also just so many other dynamics at play and I think people always dig into this like you know our model's too powerful that's a great marketing line it's too good like you know you could never give it to the wild it's too powerful and then there's also um it's also maybe you just haven't put together all the inference capabilities and scaled up your data centers to really have a productive launch. You need an extra two weeks in the oven.

Starting point is 02:43:11 Oh, yeah, it's a safety issue. So there's a million different ways that the safety stuff can be kind of abused. I'm not accusing them of abusing it, but it does feel like, you know, we noticed this with the Lama 4 launch. They haven't been able to release Bohemoth yet, and they had some issues with benchmarking. But notably, they didn't say it was a safety issue. And that would have almost been an easy out. And it's one that I think most foundation model companies would have taken a year ago. But the community, the developer community just doesn't seem to buy that line of reasoning as much anymore.

Starting point is 02:43:44 I don't know if you had any reactions to the long-for development. Yeah, I have, you know, in my own interviews with Dario, I've asked him about this. And, you know, he has been up front. Like we report on, you know, there were delays to just this latest big release, right? the latest sonnet. They're kind of, or sorry, is it opus, whatever the biggest one is, and forgetting all the specific names right now. And, you know, we have reported on these delays.

Starting point is 02:44:06 We wrote, you know, a big article about how scaling laws, at least with the pre-training side of these models, we're starting to slow down. And, you know, anthropic in my reporting with Dario, he was pretty upfront that, like, that wasn't a safety. Like, you know, that big delay was not, that big model being later than some of their early projections, that's because they were facing kind of the same technological problems that I think many labs are facing, right?

Starting point is 02:44:31 And ultimately, it did release it. So, but I think people, you know, it's absolutely reasonable to question any company's motivation when they say something slowed down. I think I saw no evidence for that in this specific case. Sure, sure. Yeah, that makes sense. Let's talk about the most recent deal meta acquiring scale AI. 49% of the company is changing hands, I believe, at around $14 or $15 billion. spiritually the whole thing. Yes, yes, yes. What has been your reaction and what have you learned from reporting on that? Yeah, I think you just covered that that they have a new CEO. So we just reported

Starting point is 02:45:10 that yeah, per, you know, for our sources that they have picked the new CEO to be Jason Drogey, former VP. He's currently, you know, at scale being promoted. And he was formerly an Uber a VP and formerly also a partner in VC at Benchmark. So, but, you know, that's just, that's just the latest, right? We, we kind of, we have really been breaking the news on this front. We had a scoop this Saturday night that we were the first ones to report that these talks were happening for META to make this multi-billion investment in scale. I think it really caught, you know, a lot of people by surprise in the industry. This is a huge investment planned on META's front, right? Probably their biggest ever in a company. And I think it's, you know, a lot of.

Starting point is 02:45:53 you know, I think it's a sign that Zuckerberg is ready to go full on founder mode and try to catch up on AI. And this last Lama 4 release was, you know, widely considered a disappointment. And so how do you how do you change course? Bring in fresh blood. Right. Big investment and the super intelligence lab. And he's recruiting engineers. We've reported on that.

Starting point is 02:46:16 They're making crazy high offers, right? Yes. They will open, Zuck will open the, the, checkbook for the top talent. Yeah, I think the interesting thing that stood out about some of, you know, people were just reporting the headline numbers of saying, oh, you could make, you know, nine figures was the number. Yeah, nine figures. But you have to look, if you're trying to recruit somebody that joined Open AI two, three years ago, they're probably run rating at $10 million a year already on a full, fully loaded comp basis. So I think it makes sense.

Starting point is 02:46:50 what have you learned about the future of the scale business, right? Is Jason going to be, you know, pounding the pavement, trying to get, you know, more customers for the core business, continue to adapt the core business. Yeah, there's a lot of scale competitors right now that are saying, like, hey, we're going to take a lot of deals. They're foaming at the mouth. Yeah, we're going to use this to our advantage when we go to the other hyperscalers and say, do you want to do business with the company that's owned by Facebook or meta? Yeah, absolutely. I think it's really unclear what's going to happen with the future of Scales business. You know, obviously the meta investment, you know, that's expected.

Starting point is 02:47:28 You could see partnerships happening there on that front. But what will happen with the other customers? I think we don't know. And what will happen with, yeah, other competitors in the space. Could they start to right gain ground as scale deep into ties with meta? We'll see. I think what we do know, you know, we have reported that Alex Wang plans to join Meta. I think for META, he is seen as someone who is just sort of reinvigorating, right, their AI team and this renewed AI effort.

Starting point is 02:47:59 He's a big name. He's incredibly savvy, you know, very well connected and built a real business with scale. So, you know, I think that's where we're in the middle of, you know, of the deal. We're waiting, you know, to see all the official detail sort of as this plays out. Earlier this week, Bill Gurley was talking about the M&A markets and why we're still seeing these kind of odd zombie acquisitions happening. Tech was laying the blame at the foot of Lena Khan for a long time. Lena Kahn's obviously out. Why do you think the big tech companies are still using that? She's still posting, though.

Starting point is 02:48:36 She's still reposting stuff about being anti-monopolistic. Okay. Yeah. So she's, you know, she's around. Is this more just like this is just a more efficient acquisition vehicle generally even without, even in a different administration? And so it's just going to be continually used. I don't think this is an asset that META would want to own just generally, right? You wouldn't want to fully own GAI because you're you're dealing with tens of thousands of offshore contractors, right?

Starting point is 02:49:06 Is that something that META wants? Don't they already do that with those? Sure, but they've had issues. Sure, but they've also had issues. It's liability. Anyway, sorry. So, of course, you know, Lena Con is known to be, you know, aggressive on going after, you know, monopolistic behavior and especially in the tech market. Doesn't necessarily mean that the current FTC is not going to scrutinize these kinds of deals, right?

Starting point is 02:49:31 I would say it's way too soon to have any definitive take on how this FTC will act around these kinds of deals. I think, you know, companies never want to raise too many eyebrows with regulators. even if people are from different political parties, we have seen Republican controlled FTC sometimes, you know, I mean, that is part of their job, right, is to review these sorts of large transactions. So that may be why, you know, we're seeing, we're continuing to still see, you know, companies not want to go too far on that front. Yeah, yeah. Can we move over to Microsoft? It feels like Satcha Nadella has carved out a ton of territory in the AI space and now the game is about holding on to it. It seems like they might be back to training their own language models

Starting point is 02:50:17 at the same time at the last Microsoft build keynote. He was distinctly highlighting how model agnostic Azure was and how they would be happy to route you to any OpenAI model, any Deepseek model, any Lama model. They want to be Switzerland, but at the same time, they want to have a horse in the race. So what's the latest thinking around Microsoft? I mean, I report on Microsoft more in the context Open AI. So I'm not the Microsoft expert. I will say that, yeah, obviously we've seen, we've seen more tension around the Microsoft Open AI relationship lately. And what that means, I think Open AI is also diversifying, right? They're working with other hyperscalers now for Stargate. They have Oracle on board. They have SoftBank financing. So this is no longer just like an exclusive

Starting point is 02:51:05 relationship between Microsoft's and Open AI. And I think that goes both ways. Yeah. What is the dynamic? right now in the reporter in the reporter talent space it seems like the value of just being a pure play writer is maybe going down but the value of being able to acquire you know new net new information and actually get scoops is is higher than ever what is the dynamic right now we've seen i think a lot of different shakeups and and layoffs at different companies but i'm curious what you're seeing i mean i i think that the value of reporting like a real original report and going,

Starting point is 02:51:42 original reporting. Getting scoops, getting people to share with you information that is in public, that takes a lot of human skill that, at least in my job, I don't think that's replaceable by AI for now. Yeah, yeah, that's what I was getting at.

Starting point is 02:51:58 I do think at some point maybe you would have an army of agents that would just reach out to, you know, 200 employees at a startup at, you know, a single given time. But, yeah, that tracks. Yeah, in terms of the star, project. Were you able to go visit as part of the reporting? Can you give us kind of the lay of the

Starting point is 02:52:18 land on what who the major players are and what the most interesting angles to dig into in the progress in Abilene has been? Yeah, for sure. Well, shout out to my colleague Brody Ford. I don't know if he's listening, but I'll send this to him after. He actually went out to Abilene for me because I was I was unavailable at one day that we could go or whatever. And it's astounding. I mean, I saw the pictures I got the live feed. It is expansive. Like, I think it's hard to understand just the vast scale of it until you really see pictures, video, or go there. And, you know, I think that there was a lot of, like, there was sort of a lot of people doubting if Stargate was even real. And I think there's still a lot. We don't know about the future of this project, like exactly,

Starting point is 02:53:04 you know, where the other sites will be. They've said they plan 10 more, at least in the U.S. we don't know exactly like how the financing will work. Stargate, we talked to Masasan for this, who's obviously, you know, soft banks leading the financing. They said, oh, it's project by project. We'll go step by step. But I think what we can learn from Abilene is like, at least that first site, that Abilene site is very real. Like it is happening. There is being built out. We've seen the pace of it. So, you know, that could end up being kind of a template or a test case, right, for how, how much start it can can execute in the future on all of its promises that makes a ton of sense anything else juries no thank you so much for hopping on this is fantastic get uh reach out when you got

Starting point is 02:53:45 your next scoop yeah yeah yeah we love to have you on same day to talk about it more context is great absolutely thanks for having me take care we'll talk to you soon we're surrounded by journalists hold your position uh we have Cameron from nominal coming on the show next oh he's already in the studio there we go jordi i'll let you take the intro i got to run In there he is. What's going on? Cameron, welcome to the show. Good to see you.

Starting point is 02:54:15 Thank you. Good to see you guys. You've been busy since the last time. I don't actually know when that was. Maybe it feels like maybe a month or two ago, but time flies. I think it was about a month ago. And the fun fact with that is that was like right in the middle of the 10-day, you know, Series B fundraised sprint.

Starting point is 02:54:36 Just absolute chaos. Absolutely chaos. And, you know, people appreciate this. Connor Love from Lightspeed Ventures posted or, you know, tweeted earlier today. But I was actually taking that prior TBPN call from the Lightspeed office here, like in between. No way. No way. So he was able to see, all right, how does Cam handle himself, you know, in a in a TV-like setting? Not that we're TV. Exactly. It was a good tactic. He was like, hey, do you want to be on this and just take it from, you know, take it from our office?

Starting point is 02:55:07 and then, yeah, so. That's awesome. Break down, break down the new round, kind of how it came together, who are the different players, all that good stuff. Yeah, absolutely. So I'm really excited. Yeah, we're announcing a $75 million series B.

Starting point is 02:55:20 That is, there we go. There we go. Led by Sequoia, and particularly Alfred Lynn, is going to be joining Nomals Board. And then, as we mentioned, there's an additional investment from Lightspeed Venture Partners, a combination of guru chahal there and Connor Love, which we're really excited about.

Starting point is 02:55:42 And then existing investment from our investors at General Catalyst, Lux, and Founders Fund. Basically, you're collecting them all at this point. Yeah, I got to. I don't know how to get me. That's amazing. Talk about business progress since the last call. I'm sure things are moving really quickly. Yeah, things are moving really quickly.

Starting point is 02:56:02 I mean, we're raising this capital to accelerate just to get. go faster. You know, really three big things we're going to do with this. The first is kind of continue to expand into larger and larger enterprises. We're very proud to say, you know, many of the, you know, the startups and scaleups building in this new hardware wave are our nominal customers. And that is going to keep growing. But, you know, recently we've been taking our product to these Fortune 500 companies, you know, defense primes and the traction has been really, really positive. We're also expanding internationally. So we're already serving a handful of customers in Europe,

Starting point is 02:56:39 and we're going to keep doing that. The second big thing is product development. And so we have a big belief that to win really big in this software for hardware world, it's going to require multiple products. We've started our original thrust is around testing, but we're having customers pull us in linking testing to more quality acceptance testing, production, manufacturing. And then later in the life cycle saying,

Starting point is 02:57:09 hey, I use nominal to build all of the software, you know, build all the rules and logic that governs how my system works under test. And I actually just want to, you know, version control that, organize it. And I want to deploy that on my asset when it's in the field and more of a monitoring sort of fleet monitoring, fleet maintenance use case. So we're building products and we're going to be expanding. So that's that's where the funding is going.

Starting point is 02:57:31 Yeah, how big is the market for this or how should I think about the market? Is this something where there's a lot of incumbents that you can kind of go after and pull away from? Eat alive. Yeah, eat alive. We won't get it. Yeah. Or are we kind of like disaggregating test functions across all the companies and allowing teams to move faster and selling into existing organizations? Yeah.

Starting point is 02:57:57 It's really, you know, the way I think about it is it's helpful to just give like a quick, you know, update a kind of the status quo world. And so I think we see pretty much everywhere we go that testing is too slow. And really you have a scenario A where people are using 1990 software. It's pre-cloud. So it is like, especially in the aerospace and more industrial world, software that was built before the notion of like centralizing data even existed. It's crazy.

Starting point is 02:58:27 And if people are really frustrated with it. So those are, you think of things like, you know, Siemens or Emerson or National Instruments, which I had to kind of mention last time. So that is an area where there's a direct swap for spend. Like people spend a ton of money there. Those are a billion dollar businesses that we are excited to be disrupting. And I think the second area is where people are taking new software and new data technologies that are not designed for hardware at all in the least and trying to use those to sort of aggregate together and build this like tool chain. for testing and nominal is just a better way.

Starting point is 02:59:05 And so that's sort of showing these organizations what the future can look like and delivering that value and they're willing to pay for it. What kind of metrics do you track outside of core kind of business metrics? Are you thinking about like trying to figure out like the, if you're increasing the development timelines for your customers in some ways, like if you can allow people to just iterate test faster, or you can actually just allow them to accelerate their businesses. Like how much how much do you try to track that kind of thing?

Starting point is 02:59:35 I know a lot of it is probably hard to to fully capture. Yeah, I'd say I had a fun conversation with Brian Schimpf, C.O. Vandrel, where he sort of was saying emphatically, like the one of the metrics he looks at, like the single most important thing is like how fast are my individual programs testing? Like it is such a direct correlation to just like the success of that. product. And that is what we see, like, you know, everywhere. So at the high level, like, the value that we're delivering is increasing that test cadence and that throughput. You know, being able to have, you know, cases where at our customers, you know, test campaigns that were supposed to take

Starting point is 03:00:17 nine months, taking 6.5 months, right? That's something we delivered to a customer. And when that's on a $200 million program where their end of customer is the DOD, like that's massive, right? You know, it's funny for a product like nominal is we actually you know we obviously track user engagement in time and app and all these sort of functions but that's honestly like not the best correlation we sort of say it's really exciting as you see more and more time people spend in the app and a lot of our north star is actually like I kind of want a test engineer to be able to drop into the app for a little bit get the insight that they need really quickly and then if they move on you know that's that's actually fine because they're that value that is compounding. Are you guys going to

Starting point is 03:00:59 to actually set up an international presence at all? You said you have a bunch of international customers. Are we going to see a nominal office in the UK? Australia. Australia? Got to go to Australia. It's a quick play. You're naming some good places. I'll say I am, I'm doing sort of nominal's first big, you know, international business trip. Actually, over the next two weeks. So you're going to the Paris Air Show. Let's get it up for international business. I love international business. I love an international businessman on the show. Yeah. It's an honor. It's an honor. Yeah.

Starting point is 03:01:31 Going to be going to the, yeah, the Parasir show and then, and then, you know, stopping her in the UK, as you kind of mentioned. Yeah, I mean, we're really excited. And I think we'll build a backlog of business before opening like a physical office there. But I think it's absolutely, you know, in the almost future. And we're excited for that. You mentioned Australia, General Ocas world. Like, we're really supportive of that. And they want and need all the same things.

Starting point is 03:01:58 that our US-based customers do. Yeah, I noticed you offer the service on-premise still. Obviously, a lot of companies are moving on cloud. How important, can you give me some dynamics around like hyperscalor adoption within these industries, like AWS, GovCloud, ITAR compliance? I know Palantir got some super high ITAR clearance. There's obviously different tiers to this stuff.

Starting point is 03:02:26 Is there one hyperscale that's kind of pulled way in this category? Yeah, it's a really good. It's a good question. I, you know, from nominal's perspective, our goal is to be as sort of flexible and as versatile as we can. So we, we don't really pick anyone, you know, provider. I'd say from our product perspective, we have two products in the market right now, but nominal core product is where we started, which is cloud-based. The whole value proposition there is you can get all of your telemetry, sensor data logs, test, data in one central place, run analytics and automy. Very quickly, we launched our second product,

Starting point is 03:03:05 which is nominal Connect, which is actually a desktop application, which sounds crazy to say and, you know, in Twinchapide. But we're meeting our customers and meeting the market where it is. It's run on, it's built in Rust. It's run on a game engine, right? So it's not your normal. And so, you know, super deterministic workflows, sub 10 millisecond latency.

Starting point is 03:03:25 What game engine are you using? It's unreal, yeah. Unreal, no way. That makes sense. Yeah, so it's really cool. So being able to have, you know, whereas it used to be, you know, maybe a national instrument's desktop app with like a little widget

Starting point is 03:03:41 and you can kind of envision the thing from 1990. You know, we have this insane, super fast rendering, 3D visualization of our customers, robotic systems or drones, right? Like with telemetry coming in. And the power is that can run completely air-gapped. If you're in Mojave or you're downrange, like doing something, all that data is still going to your laptop. And then as soon as you get connectivity, it sinks back up to the kind of nominal mothership.

Starting point is 03:04:08 All that data is sort of merged and de-duplicated. And it's super powerful for our customers to be able to make sense of what happened. Does Epic have a different business model for this kind of like industrial use case? I know that like isn't Unreal Engine free unless you're generating like a million dollars in revenue? Like how do you pay Tim Sweeney? Yeah. Yeah, it's a good, you know, it's a good, it's a good, it's a good question. I, I won't, I won't discuss it right now. You can answer it at a high level. I'm sure you can't say details. Yeah. No, I mean, it's like we, we're really excited about, I think we're really excited about the technology and kind of where it, where it can scale. I'll leave it. Okay, cool. Yeah, well, the market is speaking. I looked it up. The company that owns national instruments is actually down. Oh, often news. A few percentage points over the last six months.

Starting point is 03:04:55 Most likely in reaction to Cameron. I'm sure it's, yeah, extremely correlated. Just already pricing it in. Of course, of course. Well, thank you so much for hopping on. Yeah, this is great. Congratulations. Do we hit the gong properly?

Starting point is 03:05:08 Let's hit the gong. Alfred Lynn on the board. That doesn't happen every day. Amazing. Amazing. Big gong. Yeah, super excited. Come back on anytime, Cameron.

Starting point is 03:05:19 Awesome. Thank you guys. Appreciate it. We'll see. Our next guest is Anil from Eater hopping on in five minutes. We'll go through some time. By the way, Cameron's resume is just absolutely insane.

Starting point is 03:05:30 Product and Growth at Andrel. Advisor to Applied Intuition. Oh, let's go. Head of defense. It's sail drone. Investor operator at Lux Capital. And then now started nominal. So it's almost like he was built for this.

Starting point is 03:05:46 World wind tour. I mean, there's other news. It's not really tech related, but it's related to our friends in the Wall Street and finance community. The race for Democratic nomination for New York City mayor is heating up. Andrew Cuomo is dropping like a stone on polymarket. Zoran Mamdani is absolutely spiking. He's at 42% versus Cuomo's 56% right now.

Starting point is 03:06:13 Signal has a breakdown. He says this is a textbook example of narrative discipline. Whether or not you agree with Zoron's platform, the coherence is surgical. Every message hits the same emotional. Every post or speech is a variation of the same core cord, anti-machine people first morally upright. His surge is pure memetic lift. It's pure belief and belief spreads fast.

Starting point is 03:06:37 This is the guy that just says freeze the rent. That's his platform. Yes. And every time he posts. His new bit is eliminating police presence in areas that have high crime. Yes. Yes. And so a lot of accelerationsists are getting excited about this.

Starting point is 03:06:52 A lot of people that are anti-New York are supposed. because they think it will destroy New York. But he is gaining ground and he might be the next mayor. Who knows? Every time he posts about freezing the rent, Nikita Beer chimes in and says, don't you think the rent, the rent should just be free. You're not going far enough.

Starting point is 03:07:10 You're not going far enough. Capitalist pig. Yeah, capitalist pig. Also, new all-time high for Oracle. Let's hear it for Oracle. Wow. Let's go. Ellison is cooking.

Starting point is 03:07:22 Love it. Congratulations to Larry Ellison. Larry's listening to this. I'm sure he is. He follows TBPN. It's just fantastic. It's just fantastic. He's a Oracle's deeply underrated.

Starting point is 03:07:31 It's a fantastic company. He's been in Founder Mode for decades and continues to be on an absolute tear. They basically own Abilene. They do. They do. And they're getting, yeah, the Stargate project would not be happening without them, I'm sure.

Starting point is 03:07:41 So we have our next guest coming in the studio in the TBPN Ultrigan. Welcome to the street. There you is. What's going on? I love the background. Very cool. How you doing? It's almost like you're in the hardware business.

Starting point is 03:07:54 Yeah, yeah, yeah. Break it down. What's behind you? What do you do? Yeah. We are a networking company and our hardware behind us is our power, routing, switching, wireless, old school stuff that powers the internet. Incredible. Break down the news today. And John, get the gone. Okay, okay. I'll get it ready. What happened? What happened? We raised a new round of financing to go build even better hardware and scale up operations and physical world stuff. And how much did you raise? We raised about 170 million. There we go.

Starting point is 03:08:27 Hit it, John. Congratulations. Best part of the day. Congratulations. Give some backstory on the history of the company, kind of the different stages, what it took to get here. Sure. So we got started actually quite a few years ago.

Starting point is 03:08:44 And what's interesting is networking is one of the largest parts of technology, but almost no new companies get started networking. You know, networking supports multiple $100 billion market cap businesses, but almost all of them got built through acquisition. Nobody sat down and said, hey, we'll build the entire stack together, cohesive operating systems, hardware, APIs, firmware. So we got started a few years ago, building out the entire stack. The COVID and supply chain years were tough because we couldn't get any hardware out of it. But the last few years for us has just been really about delivering great products to customers. You're just like scaling it up across throughput, bandwidth, products, integration.

Starting point is 03:09:28 But at the end of the day, I think, you know, nobody really thinks about networking unless you're working on it or a day like this, which causes errors and takes down, you know, something like 15% of the internet or something. What's going on? We've been live the entire time. The issues have been happening. Can you give us a quick update on what you think is happening there? The production team is stressed right now. They're sweating right now. Yeah.

Starting point is 03:09:49 Because you guys are up. There are rumors that you guys took down the internet. Oh, yes. It's a growth hack. It's a good. Just watch TV at the end of it. Yeah, yeah, exactly. Exactly.

Starting point is 03:09:58 Well, I think the really interesting part about networking is most of the issues actually happen because of like configuration errors. And today that's one of the causes for it because it's, you know, disparate hardware systems, software systems that don't really work together. And they're made by many different vendors, et cetera. Today was one of the reasons for, not the entire reason, but one of the reasons for it was like network configuration errors. Wow.

Starting point is 03:10:23 You guys have a, it seems like you basically have every scale up on the planet as a customer already. Is that close to true? Are you working towards it? What's kind of the full range of your guys' customer base? Is the job finished or is the job not finished? Or it's just getting started? Are you quoting Kobe, John? I'm saying the job's not finished.

Starting point is 03:10:44 Yeah. So our customer base, I think is quite interesting. We do have a lot of tech companies. and folks that are scaling up as customers, but probably our most interesting ones are like the manufacturing facilities, warehouses, schools, folks that go build like ISPs and data centers and life sciences labs, all of them sort of need connectivity and internet and networking to do whatever they're doing.

Starting point is 03:11:10 And probably my favorite part in the last five, six years, is looking at all these different types of amazing businesses that have nothing to do with Silicon Valley, but sort of are the underpinning, of our economy and how things get built. And all the interesting work they're doing from, you know, battery manufacturing to hydrogen cells to when one of us buys something online, how quickly does it get there is because there's some great business out there actually doing it, but it's all automated and then

Starting point is 03:11:37 they need networking. So our customer base definitely is we have a big swath of some of the great tech companies, but it's really fanned out everywhere throughout the economy too. Walk me through some of the work you're doing around hardware innovation, software innovation, business model innovation. What's most important? What's been most acceleratory for you?

Starting point is 03:12:00 And kind of break those down in terms of priorities and where you're investing and what you're most excited about. So I actually think those things are all interconnected because it's what incentives are what drive everything. So if you look at legacy companies, what they're trying to do is build hardware for X dollars and then sell it to you for X plus whatever margin they're trying to hit.

Starting point is 03:12:24 And so there's two ways these legacy companies can make money, either by charging customers way more than the value they're getting or by making hardware with cheaper components and less quality components, etc. For us, we don't sell hardware to customers at all. We just sell the whole thing as a service, which is also very new. I know the rest of us that work in Silicon Valley, that's sort of kind of given in how we buy things. But doing hardware and enterprise networking that way was barely new.

Starting point is 03:12:54 And we were the first people to sort of do it. But that actually drives incentives on where you go invest. Because we are not trying to squeeze the last 10 cents out of a component, we tend to go make very different choices in hardware on what sort of components that we use. How are they laid out? How is it built? What are we optimizing for? Hey, something costs another $10.

Starting point is 03:13:16 Let's go do it. But what's the end customer thing like? And then the better hardware that you have, you're unable to go build better software and this loop sort of hits. And so our business model is that not only do we build the hardware and software, we actually deployed and maintain it ourselves too. We're fully vertically integrated. I know a lot of the tech companies don't want to touch real world things, but the whole messiness of racks and cables and manufacturing, we go do not just the pretty parts of it, but also the underbelly, if you will, of actually going into business. deploying all of it. And we do this all without any capital cost to customers and only marginal cost on operating expenditure compared to how to operate. Talk about that. Talk about why a CFO would

Starting point is 03:13:59 want to have OPEX versus CAPEX in this type of scenario. And is that kind of a driver around decision making? It is. But it depends on the business, right? So if you have businesses that are P.E. on, for example, you actually want it to be CAPX. And so our, our contract, can hit the books as a lease. So it doesn't hit their operating margins. They get an EBIT a boost. That happens too. But ultimately, what customers actually care about is,

Starting point is 03:14:29 do I have to take a bunch of capital and buy hardware, but the moment I buy it, it's a depreciating asset, then I have to pay a bunch of money to go install it, and then every four or five years I have to go buy new hardware because some new technology comes out, et cetera. Or if this hardware goes bad, I have to go replace it myself. that's the part CFOs care about that we remove entirely, which is instead of having to go spend massive amounts of capital and operating expenditure to kind of fix all the stuff, we take all

Starting point is 03:14:59 that on and sort of take on all the risk instead of the customers, then that contract can be actuated as OPEX or CAPEX depending on what their business needs. Talk about meters business. You're doing a lot of things. You're doing hardware, software. You have, I'm assuming a network of, you know, team all over the country that are, you know, helping with installs. Talk about kind of the efficiency of the business. And I'm assuming, you know, this new $170 million round is, you know, a testament to the strength of the core business. Yeah. It's a really interesting core business, right? Because you're right, we get to go design hardware, make software, and then we build these racks like the one you're seeing behind us. and then actually go deploy these by the hundreds for some customers everywhere. And we're not just in the United States, we're in Canada, we're in Europe,

Starting point is 03:15:53 but everywhere in the world, hopefully by next year. And the operational parts of it are super challenging. If you can imagine a data center in the middle of nowhere or manufacturing facilities, there's legacy hardware that's old hardware in there. You have to go kind of sort of fix that, and then you're integrating it, but then you're putting in this new entire rack. So what's sort of important for us is, you know, it's like a duck swimming in a lake.

Starting point is 03:16:23 The customer doesn't see all the kicking that happens underneath to kind of deliver this to them. And our work is always tying these things in together. Why are we building the hardware? How is the software getting built? How are we deploying it and maintaining it? And then any issues we see on one end should then be a loop back to how we build better hardware,

Starting point is 03:16:43 what choices are we making? You know, just this morning, a bunch of my colleagues were arguing literally what type of ports, where it should go so it's easy for install. And that loop is a new way of thinking in networking infrastructure, even though after computing and CPUs and Intel, networking is the oldest and largest part of all of technology. So you, I mean, you design, develop, manufacturer, make, this networking equipment, you're deploying it as a lease on a more rental basis. Does that create a CAP-X problem for you? Like, how are you thinking about the way your balance sheet is growing

Starting point is 03:17:24 and developing? Is debt going to play a role earlier than it might at most other companies in your business because of the structure of what you're building? Yeah, this is a really interesting question. So if you look to historically how networking hardware got built, somebody built it for let's say $500 and bomb cost but by the time it ends up in a customer's hands it's like $5,000 so networking companies have enjoyed tremendous margins so you should start there at any time second what we also try to do is a lot of our customers because we're dealing with physical spaces whether it's in retail or manufacturing or data centers anywhere the build-out times are predictable right because there's a certain date

Starting point is 03:18:11 you're trying to hit when a data center goes live or a manufacturing facility goes live or a school goes live. So a lot of our manufacturing can be predictive towards that. Unlike other hardware companies, we don't have to hold massive amounts of inventory without knowing where it's going. We do have issues on like trying to figure out exact timing there. We're not always perfect. But we can do that.

Starting point is 03:18:32 So to answer your question, we actually use debt a lot more and plan to use a debt a lot more to scale up what we do on deploying with our. real estate partners. A big focus of what we also do is not only do we directly sell the customers, but we also go partner with real estate firms to just include meter in as a default. And that's actually where the main meter comes from. If you look at a building, there's electricity, there's water, there's cell phone lines, there's power meters, water meters. This should be more like a utility. Why is it that we're all struggling to kind of wrangle it to get it into working in any type of space. So for us, the debt and where that plays in is a little bit on hard work,

Starting point is 03:19:15 but a lot of it also on deployment and actually getting it out to the world and the pace of which we can do expecting a customer and some sort of discounted cashful expectation of 12, 18, 24 months from now. Talk to me about the kind of like wheelhouse customer for you or the wheelhouse deployment. Sounds like, you know, apartment building makes sense. Scale up, startup office makes sense. It's probably overkill for the home office. You might just want to get an Amazon Arrow mesh networking system. At the same time, you know, you look at Google's campus. It's absolutely massive. Are you going one direction or the other? Are you maxed out or are there specific oddities that you can kind of illuminate on what,

Starting point is 03:20:05 it'll take how you'll have to change the business or scale to satisfy a client like Meta's Silicon Valley campus, which is multiple buildings and probably one of the most complex networking infrastructures, you know, in America. Is there something where you can just kind of copy paste your current solution so many times and it all fits together or will there be new technical innovations that you need to hit to actually service someone at that scale? I think the answer is both. But we are already doing customers at that scale and more.

Starting point is 03:20:38 For example, Bridgewater, who's the largest section in the world, runs on meter as an example. But I think we will go to continue even more technical advancements to hopefully bring new things that legacy folks just can't do or won't do or can't think of and sort of implement. But I think overall what's really interesting about networking is that nobody gets to do custom networking, if you will, because it has to go work with. with the rest of the world. Generally, when you look at enterprise companies, they end up building some sort of Frankenstein thing or what they started off with that becomes a custom thing for each enterprise customer. For us, any enterprise customer has to then go work with the rest of the internet. So they don't get to do their own protocols, their own things. So largely, there are things on the margin, but largely, whether you're a Google or Meta or Bridgewater or a school customer

Starting point is 03:21:31 or large universities that we do, all of them use the same internet, if you will. And this is also the same in like data centers too. What you would find in majority of non-GPU data centers is not that different what you would find in a large campus because it all has to flow together. The packets must flow and end. What's the difference between non-GPU data centers

Starting point is 03:21:55 and GPU data centers in this context? Yeah, so in GPU stuff, there's actually a lot that happens between two GPUs and between cluster of GPUs got it that's a different type of networking there's a lot of work that's happening obviously Nvidia with their Melanox acquisition four five years ago and what's happening now with alter Ethernet but just everything else is actually relatively same what was the was the keyword you just had ultra ultra Ethernet is a new thing that's happening is that I remember cat 5

Starting point is 03:22:23 cat 6 I've heard about cat 7 is this like cat 8 what are we talking here this is like a whole different ballgame. So if you if you look at you know, the joke in the networking world is, you know, Ethernet is dead, long-lived Ethernet. You know, every sort of 15, 20 years, Ethernet is predicted that it will be the end of it. But, you know, somebody comes up with innovations on how it can scale up, on speed, on throughput, on bandwidth, on security. So ultra-eternet is this new version that's coming out for, you know, the next phase of all the data center build-out that's happening. Because, you know, roughly, the U.S. has more data centers than the rest of the world combined. Roughly. Let's go. I was here for America. That's some white pill right there. Everyone says

Starting point is 03:23:10 we're falling behind, but not today. But the rest of the world is trying to catch up to get to that, and the U.S. is trying to do. Never. Never. So there's a lot that. Don't even try. We're not even try. We're undefeited. We got meter. Actually, we support meters in a national strategy. So it's kind of nuanced. We're pro. also anti, as long as we're benefiting. Yeah, there's a lot that goes into it. Yeah, yeah, yeah, that makes sense. I mean, we've experienced it here.

Starting point is 03:23:37 There's like new generations of just streaming cameras that do power and video over Ethernet. And, you know, we're thinking about kind of the V2 of the studio and there's a ton of insane things that you can do over Ethernet. It's no longer just the land cable. And so it must be fun to be at the heart of that. Jordy, any more questions? Or should we let them get back to building?

Starting point is 03:23:57 Congratulations on the funding milestone. Congratulations. And yeah, check in next time. We'll talk to you soon. Thank you. See you soon. Bye. Cheers.

Starting point is 03:24:07 Our next guest will be joining just a few minutes in the meantime. It is an absolute. Banger day. Lots of fundraising. I was just going to say it's actually insane. We've been able to keep the show up with every. It is crazy. We got lucky today.

Starting point is 03:24:20 Have we ever, has the stream ever actually gone fully down? We had Zoom go down one day and we were kind of kind of, kind of, kind of wrecked by that. Well, let's bring in our next guest. Ian in the chat says, Mix panel, Shopify, and GCP are still down for him, all the services that he's in. Honestly, if you're worried about the show going down,

Starting point is 03:24:41 do not worry. Just send us your mailing address. We will print the show onto vinyl and we'll send you a record of the show so you can listen to it in high fidelity. Same day shipping. Exactly. So that you can get the show almost like,

Starting point is 03:24:56 you know, a couple hours. Well, a few hours later. Yeah. Yeah, we'll also do a lot of innovation on the vinyl pressing because I think that takes a while too. Anyway, our next guest is here. Welcome to the stream. How you doing, Trish? What's going on?

Starting point is 03:25:08 Good to see you. How are you going? Good to see you guys. It's great to have you. It's great to have you. Big day. Big day. You guys announced a series A new website.

Starting point is 03:25:20 And you took down the whole internet. You took down the whole internet. I can't confirm or deny it as part of the launch strategy. But good timing. Give us the stats. How much you raise? Yeah, raised $20 million. There we go.

Starting point is 03:25:36 Who'd you raise it from? Yeah, sound led the round. Hit it again for sound. Hit it again. Congratulations. Amazing. Sorry, continue. We got it a little excited.

Starting point is 03:25:50 Yeah, participation from Buckley, Nabal, Arrasch, Bradowski. Founder Dropbox, Rvan from Perplexity, a bunch of other awesome angels. Very cool, very cool. Break down the pitch. Like, how are you positioning the company right now? Because I know you've been doing a lot of different stuff. They pivoted from pizza to something new.

Starting point is 03:26:11 I know them as the pizza delivery company. There was that Zoom company that was doing pizza on demand. That didn't go so well. So obviously a huge white space here. Well, they just did pizza for a day. They sold out. Yeah, they moved on to this next thing. So what are you built now?

Starting point is 03:26:23 So give us an update on the platform. Yeah. positioning it. For sure. So I think for us, always the mission of anti-metal has been, how do we simplify infrastructure? But current moment, what's really different is, you know, software engineering, the bottleneck used to always be programming. Like if you could build a software, you could ship it, you could distribute to everyone, suddenly your product is live. Now you have all this AI-assisted stuff, like really awesome products like, you know, cursor, lovable, bold, be zero. And increasingly like the latter half of that equation,

Starting point is 03:26:51 which is deploying that code, maintaining it, making sure it scales, like going from one user to one million is increasingly the bottleneck on how companies, you know, are actually going to grow and sustain. Like, you know, the short of it is like writing code is no longer the hard part, right? It's all this operations and maintenance. And so what we're really indexing on is building this product that helps you better manage, you know, and automate infrastructure. So a lot of that is, you know, the nuance of infrastructure versus code is a lot of it is knowledge that is baked into, you know, people's heads. Like you have that one senior engineer who just knows how everything works. That's your go-to guy when something breaks or you need to spin up something.

Starting point is 03:27:27 But that's not super sustainable, right? And this notion that comprehension can just scale linearly as you hire more people isn't super tenable anymore. We're like emitting a bunch of data, all that data is super fragmented, everyone's super specialized. And so for us, what we're trying to do is, you know, start going into companies, plugging into their entire infersurface area, whether it's like source code, their cloud, their ticketing system, whatever it might be, really understand why things are failing

Starting point is 03:27:52 and basically pull in data and given these golden pathways as to what's going on, why is it happening, what prescriptive action should I take that is going to help me fix this, and along the way is sort of learning what are the ingrained practices in your organization. That part is super important in the sense that we don't want to build like an opinionated platform and says like this is good or this is bad, right? That's pretty subjective to each company. So it's a little more about learning, okay, given this tenant, what's going to be good for them? How do they prefer to operate?

Starting point is 03:28:18 You know, what's going to make their systems perform the best? makes a ton of sense how much how much are you spending time thinking about from a product standpoint trying to solve things that are sort of human nature to some degree versus like it feels like the nature of software development is changing so quickly but if you can kind of focus in on on what will always be true and and you know many people would say that's human nature you can create something like durable and valuable out of that yeah I think a big thing for us is you know, some companies try to position this like, hey, we're a full-time employee and we're going to, like, automate this function away for you. I think for us it's a little more of how do we take all this knowledge that exists in a bunch of people's heads,

Starting point is 03:28:57 institutionalize it into a system and code it into a system such that it's reusable. Like the dream for us as companies go from asking this question of like, yo, can we keep this running to what can we build next? And really taking away this time for maintenance that a lot of people are spending their time on. So, you know, I think human nature, like inevitably all info problems are kind of human problems. like, you know, someone messed up a config or someone didn't really understand what's going on. And in that sense, you know, working with our users to understand like what's their preferred behavior. Why do they want their stack set up this way? Why have they chosen the tools they did?

Starting point is 03:29:28 And not trying to like force an opinion down their throat. That makes sense. Can you talk to us about some of the newest problems that companies are facing with LLM inference costs, just faults that come from this new paradigm of, running agents and running tons of inference and just taking advantage of all the tools. We saw Satina Della and Microsoft Build talk about model routing being something that he was really pushing at Azure. GCP has a similar tool. Are you hearing from customers? Like right now we're in this era where every new model release and price drop basically goes viral on X.

Starting point is 03:30:12 And so unless you're under a rock, like it's pretty easy to move over to the latest and greatest thing. leave a bunch of money on the table, but are there new problems that you foresee as things become a little more commoditized or a little bit more calcified that companies aren't using best practices or they're running into new problems from the kind of AI age generally? Yeah. I mean, I think the first one is like what you struck on the head, which is all of this AI stuff is super nascent. That means there's not a kind of like super robust cooling out there, how to monitor, evaluate these things.

Starting point is 03:30:41 You know, people are building their own frameworks on the fly. I think like most companies that are really succeeding are not using something off the shelf, their ability to get in-house, right? So there's a lot of nuance that goes into these things. The other thing is, you know, models are another potentially single point of failure. Like if cloud goes down, like what happens to your product? Are you ready to do a model routing? Are you going to do something else? But I think like, you know, putting them into like a broader system perspective, they're not like new problems in the sense of these are things systems engineers or, you know, data architects have had to think about for a long time.

Starting point is 03:31:07 But obviously more and more companies are doing these things and they're really complex systems. I think the second part of that is actually just like, you know, we don't index on this super highly, but something we're seeing is more companies are just pushing more software, like highly verticalized niche to their internals, like, you know, why go put something on a retool dashboard when I can like build it and bolt and probably get a more nuanced, tailored experience? So I think that coupling of like, hey, we're working on a pretty nascent space that there isn't a ton of structure around and we're trying to figure out at the same time combined with we just have more infra and more services running period is creating this like really chaotic environment

Starting point is 03:31:40 where people are, you know, stuck in the trenches of like debugging things, figuring out how do we scale them, paying down tech dev, that sort of thing. Is there anyone on the hyperscaler side that you think is like really underrated right now? Yeah. Honestly, OCI. I think Oracle Cloud, they're never part of the conversation. I think all these other clouds have really great offerings, really great services. I think the thing about Oracle is, you know, they don't sell a bunch of services on top.

Starting point is 03:32:04 They have a really good foundation. Like these guys have been in the computing business for quite a while. I don't think it's the sexiest option out there, but like from a price perspective, security perspective, I think they're pushing a lot internally on. what that's going to look like over the next couple of years. But yeah, I think underrated. That's amazing. Stock hit the all-time high today.

Starting point is 03:32:22 Yeah, that's all. Let's hear it for all-time highs. Yeah, we love it. Generational run. Fantastic. What's your guys, I'm just curious, since you guys are so deep in the weeds on this, what's your personal, you know, co-gen stack right now? What are you in the team?

Starting point is 03:32:37 Love it. Right now we have cursor for the entire team. Some people prefer Claude code. depends, you know, different strokes for different folks. But yeah, I mean, like people are using chat GPT, Claude, Curcer. Curse is probably by and far the biggest one. We're finding a lot of success with some of the agendic stuff like CloudCode. But I think the median for us is people are loving cursor.

Starting point is 03:32:57 That's probably been the best coding experience. Very cool. What about just go-to-market, top of funnel? It's an interesting, super interesting company because you've been able to go like massively viral with what is an enterprise? like software company. It would be different if you were selling like an energy drink. I would kind of expect like a stunt and oh yeah,

Starting point is 03:33:20 everyone's talking about it. But you've been able to break through in a really interesting way in the X community, in the venture capital community. What are you seeing that's working on the top of funnel side? What viral loops are you taking advantage of flywheels just break down the marketing side? Yeah. I mean like, you know, growth is a big part of our DNA. I think largely from, you know, my co-founder, Matt,

Starting point is 03:33:41 obviously I like a decent growth and marketing. You know, for us, a lot of it at least in the early stage right now is you want to grab people's attention. Obviously attention economy. I think the really nice thing for us is this is a problem. Like every engineer we talk to has faced or seen to some degree. Even people on the operational side when they're like running a business, they're like, holy crap.

Starting point is 03:34:02 Like we couldn't push this feature out because we were stuck like fixing the service or paying down tech debt instead of launching the new thing. But yeah, I mean, we're doing all sorts of things. I think candidly you're going to see a bunch of waiting over the coming months. We're going to have a bigger splash later this year, but yeah, more to come. Can't wait. Amazing. Well, you'll have to hop back on when that bigger splash happens.

Starting point is 03:34:21 Yeah, 100%. Great talking to you. Well, congratulations. Yeah, great to get the full update. Yeah. I think we're both on the cap table. Yeah. We'll talk soon.

Starting point is 03:34:30 Love it. Awesome, dude. Thanks, guys. Take care. Talk soon. Have a good one. Let's close out by telling you about Figma. Figma.com.

Starting point is 03:34:37 Think bigger. Build faster. Figma helps design and development teams build great products together. You can get started for free. I have a challenge for the audience. Your personal website is probably bad. Go on Figma, make a new one. And just ship a website.

Starting point is 03:34:52 You can do it in like five minutes. Fantastic. Speaking of design, signal's been on an absolute tear, breaking down all the WWDC updates. He says two quick observations after using iOS 26 and iPadOS 26 for a few days. iOS 26 feels great. The system is fluid. Buttons are larger. Everything responds with intention. It's super enjoyable and fresh. iPadOS 26 is a breakthrough. Productivity finally feels native. File handling, window management, multitasking are intuitive and fast. The most fluid

Starting point is 03:35:25 and powerful computing experience on the planet for most tasks. He says he does everything except code on an iPad. And so I've, every time there's a new update, I'm always like, oh, maybe I should pick up an iPad. I'm tempted. This might be the one that does it for me. It's a bad point in the release cycle to buy an iPad Pro because it's over 400 days since they launched the last one. So they'll probably refresh it at the next iPhone event.

Starting point is 03:35:49 But I'm definitely thinking about getting it on here. I think it could work really well for running the show to. Got a couple iPads around the house. Yeah. Just kind of collect a lot of dust, but maybe I'll upgrade. Maybe you can get back in. I think you've got to commit to that being part of the workflow. Tyler, how is your, your, uh, iOS.

Starting point is 03:36:06 26 experience so far. Have you run into any more problems? It's pretty good so far. I mean, it's still, you know, my phone is like six years old, so it's a little bit slow. But overall, it's been totally fine. I like it a lot. That's good. I was, I was asking him, like, show me it, and he starts swiping through his apps. And one, and on one screen, he just has, like, 25 copies of what app was that? Chick-fil-A. Chick-fil-A. He somehow got the Chick-fil-A, the one app. You're only supposed to be able to install the app once. Somehow he got 26 copies of it installed. You got to have 26. It was the funniest thing.

Starting point is 03:36:45 If you're serious about, if you're serious about Chick-fil-A, you should have at least 26 apps. Just in case, 24 of them break, I guess. Give us the update on the art history project. How's it going? Are you an expert yet? By the way, my wife texted me and said that it's extremely embarrassing that we didn't know that Greek statues were painted because apparently that's common knowledge, but also she has a master's in art. Okay. The art history major is common knowledge of your, if you're a renowned expert. Yeah, yeah. Anyway, give us the latest art history fact. How are you doing? Yeah, so I took another quiz or another like assessment. Yep. This time I got 50%. Oh, very good. So that's 10% and only I think like an hour and a half. Compound that. So you'll be a hundred percent.

Starting point is 03:37:24 I mean, 50% more. 25% improvement. Honestly, give yourself some credit. Yeah, 25% improvement. Yeah, but you know, if I get 10% every hour and a half, that's. you know, math checks out, extrapolate a bit. I think it's like seven and a half hours. I should be at 100%. Fantastic. So I should have everything back. You could be at 200%.

Starting point is 03:37:39 200%. You could be the world expert. Did you learn anything else interesting about art history? Do you have any more facts to share with us? Or are you tapped down? I've been reading a lot about this guy, Georgio Vasari. Okay. So he's kind of, he's 16th century.

Starting point is 03:37:54 He's kind of a polymath, author, historian, painter, sculptor. He's kind of like the Plutarch of art, right? So Plutarch writes, Noble lives of, or parallel lives of noble Greeks and Romans, right? And then Vasari kind of writes the same thing for kind of Renaissance art.

Starting point is 03:38:13 But kind of this controversial guy, he writes about, you know, he looks at art from the perspective of the artist, which is, you know, controversial in the sense that like later on this guy, Winkleman, I think. Okay. He criticizes Vasari.

Starting point is 03:38:28 He says, no, you should, you know, separate the art from the artist. You should look at art kind of in of itself by itself. That's where it came from. I've heard the term. I've heard the term. Chris from case would just say, absolutely not. We're not going to separate it.

Starting point is 03:38:40 No, no. The artist is what matter. If we were to issue you a termination letter today, would you feel more confident now that you've at least studied art? Well, I actually... Your job prospects, I mean. Yeah, so I also looked at those other stats about just general employment. And, you know, so I'm actually a physics major.

Starting point is 03:38:57 I'm not CS. Sure. And physics is actually the... second most unemployed. So they're at, uh, physics majors are at, I think 7.8% CS was at rough. Yeah, yeah, yeah. So, um, I think, I think I probably would go into art history. I mean, I don't got much else. So amazing. Underrated, underrated. I think we've discovered some massive alpha. It wouldn't really take that much to double the number of art history majors, you know. There's only 2,000 of them, right? Yeah. Yeah. Well, 2001 now. 2001. Um, amazing. Um, amazing. Let's

Starting point is 03:39:31 rip through some timeline. Great work, Tyler. Yep. We have some news. Congratulations to Crusoe. Chase Lockmiller, been on the show. They are partnering with Brookfield on a $750 million debt facility to build the intelligent infrastructure of the future. Do we have to ring the gun? Hit it. Hit it, John. Hit it.

Starting point is 03:39:52 That's a lot of pesos. That's a lot of pennies. Congratulations. The whole team over there. Fantastic. We love big data centers. We love scaling laws, putting big training runs on the data center. Fantastic. We got another post from Zizi. What are some good baby names for someone just getting into nominative determinism? I got to say, it came to me immediately.

Starting point is 03:40:14 Oh, I guess it's here as a comment too, but I was going to say if it's a boy, Chad. It's just so obvious. Chad, I think, is going to come back in a big way. I think it sort of waned in popularity when it was more of a slur. But now it's just massive alpha there. It can kind of be an inside joke for you and the other dads. 20 years from now, your son, you know, little baby Chad turns into a man. Oh, oh, he's his chat.

Starting point is 03:40:39 Huge surprise. You know, it's not like it was destined. You know, Johnny Ive was in the news because he went to Open AI. But he's technically Sir Johnny I because he's been knighted by the British royalty, right? And I was thinking we need an American version of sir, of knighting. And I think what we should do is we should take all the Chads and the ones who are really, really impactful, the ones who have driven, you know, massive capital investment like Chad buyers. They should be knighted in an American way and they should earn the title, Giga. And so, and so for his contributions, the president should take a sword and tap him on both shoulders and knight him, Giga Chad buyers.

Starting point is 03:41:24 Gigacad buyers. I think that would really, really, it would be the American version of the knighting. Yeah. And it's based purely on how much capital formation. 100%. Yeah. There's no other criteria. That's the way it needs to go. So step one, name your son, Chad. Step two, get him into capital allocation on the private side, ideally. But public markets will work too. You're still eligible. There's some more good comments in here. Grit Capital says King or Don. Donald. Donald. I never thought about it. The Don. That's where Don comes from.

Starting point is 03:41:58 I think so. In other news, Anon, Sunwal says my first and only angel investment went to zero today. Well, the mistake should have made 50. I'm sure one of them would have worked out, made it all back in one trade. Fascinating because he's the founder of CB Insights, which you would think he would be addicted to angel investing with all that data and stuff. But he knew his life's work was to build that company. And that's what he focused on. And he only invested in one company.

Starting point is 03:42:24 Fascinating. He's very, very different. Car dealership guy response. Congrats. Very funny. Anyway, yeah, yeah. He said he learned. I don't really enjoy it.

Starting point is 03:42:33 And if you don't enjoy it, you're not going to be good at it. So stay out of the game. Paula says Diet Coke's or fridge cigarettes. So true. So true. Just drink them all day long. All right. This is a good place to end.

Starting point is 03:42:44 I like how Zuck has been back against the wall like 37 times since 2007 and every times resoundingly wins. But each time again, he has a crisis. People are like, oh, man. Zuck is cooked. Yep. So true. Do not count him out. The AI talent crisis is merely a bump in the road for

Starting point is 03:43:05 for old Zuck. He's cooking. We've all, we've all been there. We've all you know. We've all tried to launch. Our personal. Our personal behemoth. Yes, yes. We all have something. We need to wrangle and get out. Got a power through. Whether it's a four hour stream or a two trillion

Starting point is 03:43:21 parameter model. You know, you got to get it out. Well, this was a great show. I got to give it up to the, to the Our credible production team. Keeping us online while everything goes down across the entire internet. We learned everything there is to know about art. We're experts too. He's an expert, yeah.

Starting point is 03:43:38 Now we've got to worry about poaching because I imagine a lot of the galleries are going to be watching this. And they're going to be, oh, New Art History. Guy understands physics, live streaming, and knows everything there is to know about the history of art. The history of art. Anyway, thank you for watching. Leave us five stars on Apple Podcasts and Spotify. And we will see you tomorrow. Have a great day.

Starting point is 03:43:59 Have a great afternoon.

TBPN - Air India Crash, Internet Chaos | Delian Asparouhov, Saif Khawaja, Henri Stern, Chris Paik, Joseph Cohen, Zachary Perret, Shirin Ghaffary, Cameron McCord, Anil Varanasi, Shreyas Iyer

There aren't comments yet for this episode. Click on any sentence in the transcript to leave a comment.