Soham Parekh Exclusive Interview | Chris Miller, Aaron Ginn, Bridget Harris, Pryce A. Yebesi, Jacob Rintamaki, Auren Hoffman, Mehran Jalali, Kushal Byatnal, Avlok Kohli, Nico Christie, Flo Crivello

29:10And they can choose to opt out of it and getting a different job. But they kind of know what they're going into. Like no one shows up at Goldman Sachs and is like, oh my God, I had no idea. I was going to have to like do a slide deck on Saturday, right? But in startups, like there's kind of this like, hey, we have this, you know, good work culture. It's kind of chill. Let's, you know, do 40 hours a week. But then, you know, also there's no real way to make three times as much if you're willing to work three times as hard. So yeah, it's an interesting story. Yeah, and to be honest, he posted yesterday, he said, there's a lot being said about me right now.

29:41Most of you don't know the full story. If there's one thing to know about me, it's that I love to build, which I do believe. He said, that's it. I've been isolated, written off, and shut out by nearly everyone I've known in every company I've worked at. That doesn't feel completely true because he's in conversation with the CEOs that had hired him. But building is the only thing I've truly known, and it's what I'll keep doing. Earlier today, the timing of this is amazing. earlier today I signed an exclusive founding deal to be founding engineer at one company and one company only. They were the only ones willing to bet on me at this time.

30:13I actually think there's a lot of companies that would take a bet on him now. And it just has to be, you have to promise me that you're going to work just for us. And I think the right thing to do for this next company is try to bring him out to SF or wherever they're based. Because he will probably spend 18 hours a day in the office. Yeah, it's so much easier to deal with a known unknown or like a known risk factor. It's the devil you know. Like forever, this will follow him as like, hey, monitor this guy. Make sure he's not working a second job. If you scroll down on this post, he says, I'm pissed and I've got something to prove.

30:48It's like he had a very aggressive posting style and he came on and was very sweet and nice and just like kind of contrite. Yeah, it seems like he got in a bad spot. He made some bad decisions. He's sorry about it. And he's now got something approved yeah and he's now infamous yeah infamous well that's a fun story uh i'm glad we were able to get him on and cover it that's uh that's part of the beauty of the show live baby uh anyway maxing yeah and i think that provides some good closure to this chapter of the timeline yeah um we have a million messages coming in from various sources um i can go through a little bit of the stories if you want.

31:29We got to ring the gong though because the U.S. economy added 147 ,000 new jobs. Let's go. Manufacturers are... It's like a very, very slight job speed. It's not actually that bad. We're very excited. Joe Weisenthal has a post in the stack somewhere where he says... Breaking. It's a beat. 147 ,000 new jobs. Unemployment rate falls to 4.1%. Last month, revised higher. Analysts had expected 106K new jobs and a 4.3 % unemployment rate. So we love to see it. Stark contrast with the calls from our president to reduce rates. If jobs report is good, it's hard to argue. to drop them. So manufacturers are pausing hiring amid tariff uncertainty while unemployment fell partly because fewer people are looking for work.

32:31So that's never a good sign. But it does seem like the the American consumer seems resilient and things are we're we're staying in the kangaroo market. We're neither we're neither. You know, it's not so over, but we're maybe not so back just yet. We're kind of right back where we started at the beginning of the year in terms of the stock market rates. Anyway, the CEO of Ford had a controversial quote in the Wall Street Journal. He's AGI pilled. He's AGI pilled. He's been talking to Dario over at Anthropic. Clearly. So the CEO of Ford said, AI will replace, quote, literally half of all white collar workers, which is something that was definitely said on Dwarkesh Patel first.

33:14I think Sholto might've mentioned that and then Adario mentioned it in the post as well and Anthropix been kind of noodling on this and then OpenAI and Meta have kind of stayed back from that and a lot of the other labs have kind of said like I don't know like you know Daria I think the quote from Sam was like Dario is a scientist like he should apply more scientific rigor here to like do this prediction like this is a very jar is Brad like that's right Brad like that opening I was pushing back a little bit yeah but look that when he was making these comments he was also talking about just how advanced xiaomi and these other chinese ev manufacturers are so this was coming from a broader conversation yeah and it's a good you know ultimately whether or not he actually believes this is true he's signaling to the market that he's he's on top of this trend yep and uh if this does turn out to be the case it's very possible that Ford will be a beneficiary in some way.

34:15Yeah, yeah, yeah. I mean, it's a tough business. And I'm sure he's looking to the Chinese manufacturers and looking at their COG structure and their SG &A spend relative to revenue and seeing how they can benchmark. Not that the Chinese EV manufacturers are necessarily adopting AI any faster than Ford, but it's certainly the idea of how the shape of the business will be in the future. Well, a little bit. They do facial recognition tech. That's true. I think you walk up to your car, pull and unlock for you. I think there's a story about the Ford CEO dailying a Chinese car just to get a taste of like, okay, we need to step it up in a bunch of different ways.

34:57Yeah. In other news. Yeah, there's no reason that Ford can't catch up on these type of things. Yeah. In other news, we have a massive story out of the United Kingdom, folks. Britain's royals gave up their luxury train citing fiscal restraint. This is sent to us by a friend. In an era of hard choices for economies around the world, Britain's royal family is also playing its part, waving farewell to perhaps its most extravagant mode of transport. The Royal Train, the House of Windsor's personalized service that chugs the royals to engagements will be axed in 2027 after a review concluded that it no longer offered value for money and that two new helicopters would be a reliable— Nothing is sacred anymore.

35:45I know. They're getting rid of their train. The locomotive service, which features bedroom compartments, a dining room, and a study has been a lightning rod for anti-monarchy campaigners who pointed to its costs as emblematic of royal excess. Shows the power of democracy that we have anti-monarchy campaign. You can't even be a king with a private train anymore. What's happening to the world? It's ridiculous. this. You know, royal excess, what is royalty if not excessive? It's the brand. It's terrible. Last year, King Charles III undertook a 190-mile rail trip to the capital from England's East Midlands, which was billed at nearly$62 ,000.

36:36It's like the price of a Model S. wait the trip was 62 it was 190 mile rail trip on a private train car and so they have to staff it they have to load it up with fuel and food and all this different stuff and the study has that strike you as i mean it's it's a it's 62 000 for 190 miles it's a bargain but 190 miles it's across maybe a few days probably bringing multiple people to having meetings on there yeah it's cheaper It's cheaper than, you know, taking a zipping around in a PJ. Well, they will a review by the government concluded that buying two new helicopters would be a relative, a reliable alternative or a viable alternative.

37:20What about safety, though? I have to imagine trains have a better safety. I would think so. I would think so. James Chalmers, who manages the royal finances as keeper of the privy purse, said the decision to halt the personal train would mean the fondest of farewells. Yes. This is the Wall Street Journal. In moving forwards, we must not be bound by the past, he added. Oh, rough. This kind of public fiscal self-flagellation is nothing new to the British monarchy, which is often at pains to point out its value to UK taxpayers who bankroll the franchise. The late Queen Elizabeth II gave up the royal yacht in the 1990s in the name of financial prudence.

38:05But this latest decision is all the more surprising, given that the royal family's finances are arguably stronger than they've ever been in modern times. The royal train... Let's give it up for the royal family's finances. Let's give it up for the royal family's finances. Thank you. I don't think ramps in in the UK yet, but they'll be able to get to the next level. Sure. When that happens. The Royal Train will stop rolling for the first time since Queen Victoria first commissioned coaches to be pulled by a locomotive in the 1860s. The train's importance in an age of air travel has repeatedly come into question.

38:41Once the train is mothballed, Charles, who has long advocated protecting the environment, will instead use two newly leased helicopters. that run on sustainable aviation fuel. The train will be taken on one more tour of Britain before it is retired and can be placed on display. So, RIP to the Royal Train. The Royal Train. Hopefully it makes a comeback at some point. I think American Train. You know, the Royals are backing off. Let's get an American Train. Yeah, this actually could be bullish for trains because them shutting it down creates an opportunity for them to start it back up, which would be a positive catalyst for the train.

39:18Yes, yes, I agree. Well, you mentioned ramp. Time is money. Save both. Easy is corporate cards, bill payments, accounting, and a whole lot more all in one place. If you're trying to expense train trips, do it on ramp. That's the way to do it. Ramp travel, baby. Ramp travel. What else is going on in the news? JP Morgan is revamping its bank for the super rich to cater to global clientele. David Frame, named global head of private bank, helping deep-pocketed clients invest more abroad. Oh, there's also big news. Not very American. No, not very American. But Ilya Sutskiver has issued a statement on Daniel Gross with Safe Super Intelligence.

40:00So Ilya writes, of course, the OpenAI co-founder now at SSI, says, I sent the following message to our team and investors. As you know, Daniel Gross' time with us has been winding down. And as of June 29th, that's three, four days ago, He is officially no longer a part of SSI. We are grateful for his early contributions to the company and wish him well in his next endeavor. I am now formally CEO. Reading between the lines. Do we think he hit his cliff? I don't know. When did SSI start? Probably one year cliff or your best. You think he went default? I don't know. When was SSI founded? 1970. Scuba Schools International.

40:41Wow. super intelligence has not arrived in jordy's web browser just yet oh he hit it he may have hit his cliff it was founded june 19th 2024 no way but it's very possible this was resolved prior to that absolute dog and um anyway so elia says i am now formerly ceo of ssi and daniel levy is president the technical team continues to report to me you might have heard rumors of companies looking to acquire us we are flattered by their attention but are focused on seeing our work through he's a missionary we have the compute we have the team we know what to do together we will keep building safe super intelligence ilia so dg has entered the trade portal and uh i don't even know what that means i've never heard that term before but i understand that it's vaguely it's portal time it's portal time it's portal time um but uh anyways uh the the The rumor is that he's going to Meta, but we will have to see.

41:40Well, speaking of Meta, Nat Friedman officially posted that he has, in fact, joined Meta. This was rumored and leaked multiple times. Yep. It seemed pretty confident. Everyone seemed pretty confident in the reporting, but he has confirmed it. He said, I started work at Meta this week. This is Nat Friedman tweeting. My job is to make amazing AI products that billions of people love to use. He's in the application layer, baby. He's productizing. He's in the app layer. Love it. I'm excited. I'm so excited for what he's going to build. Will says, taking the balance sheet of the world's largest megacorp and using it to develop the technology to find Noah's Ark, the Ark of the Covenant, and other biblical artifacts is top-tier stuff.

42:22Congrats, King. Ashley Vance says, if you're not hiring an intern to use ground-penetrating radar to find the ancient first copies of Solaris buried under building 12, I'll be disappointed. I love it. He says, it won't happen overnight, but a few days in, I'm feeling confident that great things are ahead. Yeah, I posted something when the news leaked that it's like it's 2026. I put on my meta quest for Lama, open Instagram, and pull up the latest Vesuvius scroll because I certainly hope this means more Nat Friedman side projects. Obviously, he's going to be incredibly locked in. has a huge task in front of him, catching up to the big labs.

43:01He has an incredible team now, a lot of funding, but it is a big challenge, and I'd certainly hope that he has time for some side projects. I wonder what this means for AI Grant, I wonder what this means for the rest of the Nat Friedman cinematic universe of kind of cracked interns that have gone on to, the Nat Friedman diaspora is really illustrious at this point. Yep. Sort of a sort of a new teal fellowship. So breaking news, the big, beautiful bill has passed the House. I had a poly market just about 45 minutes ago. It had rocketed up to 99 percent. It's now sitting at 100 percent. So Axios has the reporting.

43:45The House passed president's big, beautiful bill Thursday. This article is from five minutes ago, clearing the way for Trump to sign it by his July 4th deadline. It's a massive victory for Speaker Mike Johnson, who was able to flip dozens of members who had initially threatened to vote no, as well as for Trump and Senate Republicans. The bill passed 218 to 214. Representatives Thomas Massey and Brian Fitzpatrick were the only Republicans to join all the Democrats in voting against the measure. Moderates and conservatives demanded adjustments after the Senate made significant changes to the House version, but ultimately none were made.

44:23Several key holdouts flipped following meetings with the president at the White House, but a ground of a group of hardliners needed more persuading. Anyways, so we will wait to see the the impacts of all of this. I'd be curious to get Zach Kukoff on maybe at the end of the show to just kind of break it down. We will be continuing to track the big beautiful bill, but if you're trying to design something big and beautiful head over to figma.com Think bigger build faster bigma helps design and development teams build great products together You can get started for free at figma.com People are loving the s1 Yes, really loving it.

45:02Yes, and they're going out We got to get Tane on who does the s1 deep dives for sure We got to do that And if you're also in the s1 deep dive business and you have an interesting take hit us up We'd love to have you on a chat about the Figma S1. Very interested in that. And Earl Palmer Luckey has an update on the story about how many weapons from America are flowing to Ukraine. He is kind of giving his take and where he stands on this issue. So Andre Yermak says, today I remembered an early wartime meeting with Palmer Luckey. I remember, I think I remember this photo or a photo like it going out at the early start of the war.

45:39steady kind and clear like tanjiro in the face of darkness grateful he chose to stand with ukraine his company's growth is proof good people build strong things and palmer chimes in and says anduril had people slash hardware in ukraine two weeks into the invasion our autonomous weapons have destroyed hundreds of millions of worth hundreds of millions of dollars worth of russia's war machine the united states should give them the tools they need to win it is the fastest path to peace one way or another so i had to look it up but tanjiro is an anime or a manga manga demon slayer so uh andre clearly knows his audience with that post luke metro has a post here uh so open ai came out yesterday and said these open ai tokens are not opening i equity we did not partner with robin hood we're not involved with this and do not endorse it any transfer of open ai equity requires our approval we did not approve any transfer please be careful and luke metro says what every cfo is thinking yeah these tokenized shares on robinhood are every startup cfo's newest headache uh they are derivatives from what i know so they're not it's not actual equity it's closer to a you know something like a prediction market yeah no one is more creative financially than secondary salesmen.

47:05Yeah. Do you remember the whole Ford contract situation? This whole thing? Yeah. Yeah, you'd write, I mean, it's derivative contract, but people were doing that for a while and then that got banned. And I don't know how they're doing this one, but odd to, you know. Never bet against American financial instrument innovation. But the interesting thing, what does this say about Robinhood? Says that they're not AGI built. Because if you truly believe that OpenAI is building God, You wouldn't want to make God angry. So not a G.I. belt. I guess something is that the conclusion that could be one one conclusion I think we need to frame everything in the context of do you believe every decision is how a G.I.

47:42You know, you're just reading between the lines. Yeah Yeah, even even the even, you know, the UK is clearly not a G.I. Pilled because if we're approaching super abundance. Yeah, it's what's wrong with having a royal figure out how to operate the royal train cheaper than walking. Yeah, this is very damaging for their brand. Yeah. Ford CEO, though, AGI-pilled. That's right. Let's get him on the show. Anyway, let's tell you about Vanta. Automate compliance, manage risk, prove trust continuously. Vanta's trust management platform takes the manual work out of your trust, out of your security and compliance process, and replaces it with continuous automation, whether you're pursuing your first framework or managing a complex program.

48:25Anyway. Head over. has some fantastic photos and getty images he went through getty images and found some insane photos of steve jobs that i hadn't seen i thought i'd seen every viral photo he's watching in the bottom right corner if we can pull this up he's watching uh larry ellison game tape i love it it's amazing so good yeah wow he's really watching ellison just being like okay what's he saying how am i going to counter this how am i going to do that i i assume we're just learning from from one of the best. Yeah, yeah, yeah. Zoom in on the bottom right one. That picture laying on the floor, like, you know, deep in conversation or argument.

49:03No, no. Oh, is he on the phone? I think he's just thinking. It just seems like he's on the, he's just, he's just like exasperated, like taking a second. They didn't have, it's funny. I just assumed he had an iPhone up, but the phones weren't that small. No, he invented the phone. Yeah, yeah, yeah. They didn't have that. You had to go through that to make the iPhone. Indeed. Speaking of hardware and jobs, I have Reggie James' new book here. It looks absolutely fantastic. I got it last night. So, well, thank you for making it. I purchased it to support a fellow technology brother. But it is absolutely fantastic.

49:40Portfolio company in there, Daylight. That's very cool. Love to see it and a bunch of other features. I love these books. Oh, we got nothing in there. No way. That's amazing. Yeah, very cool. Beautiful. Beautiful stuff So go pick it up. What is it hardware book calm? I think was what he said hardware book 2024 hardware book 2024 dot com it's sold out so it's sold out But I don't know DM Reggie with something he'll figure it out and figure it out if you're trying to build a Product go to linear calm or linear dot app. It's linear dot app linear dot app linear is a purpose-built tool for planning and building products meet the system for modern software development streamline issues projects and product roadmaps and they got linear world's best team best teams use linear they got cursor perplexity retool ramp open ai scale boom cash app for sell you should be one of them kind of lebron james of uh of product planning speaking of lebron james breaking news from emily sunberg lebron was in the hamptons last week and needed somewhere to practice instead of going to some weird rich guy's home gym who would have probably asked for photos afterwards he turned to the most obvious solution a high school and emily interviewed a teenager who played pickup basketball with him this is amazing great story good for his brand too it's just great just get all around great for his brand yeah it's like a it's an unexpected gift and unexpected time like you played your teenager you probably played basketball at this high school gym like you know every day for a couple months during basketball season, probably off season a little bit.

51:16And then all of a sudden, the greatest basketball player of all time, potentially. I don't know where you stand on the goat debate, Jordy, but he just shows up, puts on a clank. There's only one goat debate I care about. The best entrepreneur of all time, of course. We have another post here from Jira at Jira Tickets. Good morning. 10 ,000 new AI startups with cute little minimalistic names and logos just launched. And they say they're going to change everything. Please enjoy your timeline scroll. Just call me if you need anything. that's true uh we have a post here is he just saying that there's a ton of ai startups launching every day on the timeline basically and they're all going to change everything john okay they're gonna they're gonna do their best names logos i feel like i feel like what's clogging up the timeline more is like you know people actually taking like the viral video seriously like and like breaking through i i don't see that many like cute little minimalistic names and logos that much But I don't know.

52:09It's interesting. Yeah, I guess it is a paradox that there could be so many things that all change everything simultaneously. My recommendation for Soham. We were on the drive-in today. We were working on this. It's hard to see. Maybe text it to the group and put it in the tab. The dev shop of Soham Perique is the last. Yes. And just flip the browser company logo. Yep. You're at five million of ARR by the end of next week. I really do think that I'm hopefully excited for him if he's going to this new video AI startup and he's working seriously on that and he enjoys the founding engineer role there rooting for him.

52:52But I do think the dev shop, he's teed up for something amazing. I think that that would be a fantastic business for him. He's deeply aligned with the culture, clearly loves building. He got it over his skis, made some bad decisions, disappointed some people, maybe broke some rules. But if he just stayed in Mumbai and found 50 other guys like him and he was the guy that was onboarding clients, he would absolutely crush it. Totally, totally. Yeah, onboarding clients would be huge for him. And it seems like he likes juggling, bouncing around. Totally. And I think most importantly, with this like the crazy viral attention that he got, I think he had a chance to kind of rewire people's brains on what it means to use a dev shop.

53:38Because using a dev shop feels very antithetical to like the way startups are built. We talked about this a little bit yesterday. Like it's a sign that like you can't hire or you can't actually build. Like you don't have a team on your like a line to you. But he has such an interesting story that I think a lot of people would say, yeah, let's get Soham on for the dev shop on for a quick sprint for three months. Seems to be good at that. He has a good team, and he could grow it from there. So I don't know. I'm optimistic about that. I was surprised that it didn't seem like when we talked to him that it was even in his mind that he could do that.

54:19It also seemed like he hadn't considered just having a bidding war between the three jobs and saying like, who can actually pay me more? Yeah, I think the equity thing was weird too. Surprisingly, there's a bunch of teams right now where if you just said, I'm going to work 18 hours a day on your company, they would probably pay you$350K. There was something weird about the fact that he's like, I don't want to talk about my personal life, but I have financial needs. but also I wasn't doing it for the money and actually I cared about the equity. Because it's like, if you're really in like, I have medical bills or something, you'd be like, cash now.

54:56No equity, right? No, the schizo takeaway is that he's the mob, the Mumbai mob is saying like - Go get equity? No, no, no. Like, you work for us kind of thing. Sure, sure, sure. But in that case - You would just max out the cash comp. You'd max out the cash, right? You'd max out the cash comp and you try and go get one job or multiple jobs that just pay all cash or as much cash as possible. And I mean, we need to talk to more of the founders on the other side about was he really asking? It wasn't like he had this master plan of I'm going to, like the takeaway from yesterday is I was coming away and being like, okay, if five people came out on X, are there 20 more on LinkedIn?

55:35And it's like this massive conspiracy that he was like master plan, but that wasn't my takeaway. We'll have to dig into it more and talk to more folks because it's totally possible that more people did he had a post from yesterday who's on the show yesterday he said soham is just the tip of the iceberg just like this redditor pulling 800k a year working five jobs well there's a reddit what is a group called our slash over employed yes 500 ,000 people where people just maniacally discuss this there are thousands of soham periques we don't know about. There's a post here from user big fat phony account he says five jobs at 800k TC using an ROI mentality I've jumped from three jobs to five jobs two weeks ago I'm now making three three K a day it's doable for me here's why I'm an expert in my field which is data I know how to solve problems in various forms clients as I call them know my value if there's a problem or they need my expert help.

56:39They like having me on the team. This expertise has come from 15 years of experience, but also three years of over-employment teaching me the latest methods. I'm all remote. None of this one day in the office BS. Just say no. So he's basically giving the playbook. The funny thing is you could just have a consulting shop and do this above board and probably be way more successful. You just have to build the muscle to sell your services and not just get fixated on just applying for jobs. Like you can still look, okay, this company is hiring for this role. Great, I should reach out to them. Hey, I offer XYZ service.

57:21I'll be fractional staff engineer, whatever it's called. I command a premium because I'm super experienced and I have this team. You can call me anytime. I'll be available 18 hours of the day. and you could just build a legitimate business. So I think that that should be the takeaway from engineers that are getting inspired by SOHAM in the wrong way is just build a consulting practice. I agree. Well, if he does, he's going to have to reach out to a lot of people. He needs a CRM that he can trust. He needs Adio, customer relationship magic. Adio is the AI native CRM that builds scales and grows your company in the next level.

57:57Let's mix it up with a post from the P-God. This is important going into the weekend, everyone. Yeah, this is huge for July 4th, which is tomorrow. If you're pure of heart, drinking 17 ice-cold beers isn't bad for you. We've got to have Andrew Huberman back on to debate this. And Brian Johnson, too. We'll get both of them. They'll love it. They'll be like, yeah, heart agreed. I think we feel this way about splitting a bottle of champagne. It certainly doesn't feel bad. Everything in moderation, including moderation. That's what I like to say. That's right. Right. Anyway, what what else should we go?

58:35Leaf over at Public, our friend said this is some interesting context on the on the tokenization of private companies. So he says tokenization of private companies is tricky and risky. Public used to offer alternative assets like music royalties, and we look deeply into offering tokenized private company shares. Here's why we haven't done it. The risks come down to four aspects, liquidity, supply, information rights and investor protections. First, some background. Public offered private assets via regulatory framework called Reg A. Think of it like a wrapper that turned a private asset like IP, art, real estate, or private company into a security, i.e.

59:11securitization, enabling it to be traded. It's very similar to what a crypto token represents in the Robinhood tokenization scenario. Same thing, different database. And he goes on to give a lot more context. He says let's say Robin Hood now owns ten million dollars worth of stock at a three hundred billion dollar valuation They could spin up ten million tokens priced at one dollar each and offer them on their platform Customers then on the token not the actual shares and open AI think of it like a ticket representing a sliver of ownership in the SPV But not opening I directly

1:00:12And he says, With Republicans saying - $5 ,000 ,000 investments,$5 ,000 ,000 total. So there's 1 ,000 people that are putting 5K down to set the price of SpaceX. It's like, does that make any sense? Not setting the price, the initial price. Well, they're setting - The offering. No, no, no. But like there are essentially 1 ,000 slugs of 5K value to kind of set the price. It's like the smallest possible market. So, yeah, I mean, I think that there's a lot of risk to that. They could just trade like Bitcoin. The other thing, because, you know, these types of offerings will typically be tied to SPVs, not the underlying shares.

1:00:56There's usually no information rights. There can be dilution, other changes. And so you actually don't know the value of the underlying assets. Well, really quickly, let's tell you about numeralhq.com. Sales tax on autopilot. Spend less than five minutes per month on sales tax compliance. We have Chris Miller joining, the author of Chip War. In just a few minutes, I wanted to run through this post in the, it's not a post, it's an article in the Wall Street Journal from yesterday. China is quickly gaining in the AI race. I think this is something we'll be talking about with Chris. Chinese artificial intelligence companies are loosening the U.S.

1:01:32global stranglehold on AI, challenging US superiority and setting the stage for a global arms race in the technology. In Europe, Middle East, Africa, and Asia, users ranging from multinational banks to public universities are turning to large language models from Chinese companies such as startup DeepSeq and e-commerce giant Alibaba, that's Quen, of course, as alternatives to US offerings such as ChatGPT. HSBC and Standard Chartered have begun testing DeepSeq's models internally, according to people familiar with the matter. Saudi Aramco, the world's largest oil company, recently installed DeepSeq in its main data center, which is interesting because Saudi Aramco, I believe they have their own model as well.

1:02:15And so they're testing DeepSeq. Very AGI-pilled. And I mean, you could rewrite this to say like, Hey, like, you know, perplexity has a DeepSeek fork. Microsoft serves DeepSeek. So maybe it's okay. But obviously the larger geopolitical narrative here is like, we don't want to have the rest of the world standardized on non-Western models. That could be very harmful to global competitiveness. And so that's something that we'll be digging into. Even US major cloud service providers such as Amazon Web Services, Microsoft, and Google offer DeepSeek to customers despite the White House banning use of the company's app on some government devices over data security concerns.

1:02:54Of course, there's a separation between the layer there. If you're running the open source model on your own chips, it shouldn't be able to call back to China. But if you're installing the DeepSeek app on a government phone, well, then everything that you send, it has to go to a server to be processed to then be inferenced. And so I assume that everything I've ever said to ChatGPT exists on an open AI server somewhere, right? because they have to store it because when I open the app, I want to see all my chats. The judge required it. Yeah, yeah, it's required. Well, that's a slightly separate thing.

1:03:24That's about like actually disclosure, but yes. Hegseth, Hegseth and the DeepSeek apps just being like, thoughts on Iran right now? What should we do? We're not. OpenAI's ChatGPT remains the world's predominant AI consumer chatbot with 910 million global downloads. Let's hit the gong for OpenAI. There we go. Compared to with DeepSeek's$125 million. Give me the wah, wah. $125 million for DeepSeek. Not even close to open AI. DeepSeek, you got to catch up, buddy. No, it's very impressive. $125 million is a serious number. Figures from Research Sensor Tower. The U.S.'s AI is widely seen as the industry's gold standard because of advantages in computing semiconductors, cutting-edge research, and access to financial capital.

1:04:18But as in many other industries, Chinese companies have started to snatch customers by offering performance that is nearly as good at vastly lower prices, I imagine. Let's see. Prices, yeah. A study of global competitiveness in critical technologies released in early June by researchers at Harvard University found China has advantages in two key building blocks of AI, data and human capital, that are helping it keep pace. The competition, some industry insiders said, has set the world on the path toward a technological cold war, potentially even a chip war. And we have our guest, Chris Miller, in the studio, author of Chip War.

1:05:09Welcome to the show. How are you doing? What's happening? Good, thanks. How are you? Thank you so much for joining. It's great to have you. Great to have you. I would love, I mean, we were just reading an article from the Journal about the China versus the U.S. AI race. There's a whole bunch of modern details, but I'd love to kick it off with a little bit on your background and how you initially became interested in the idea of a chip war and kind of how things emerge from there. Yeah, I got interested in the idea of a chip war because I realized we've been living through multiple iterations of one since the first chips were invented in the late 1950s.

1:05:49that you couldn't really separate the development of military power or a country's strategic position on the world stage without understanding the ability to access and deploy computing power. And so we're seeing a new version of that right now with AI, but it's actually replicating what we saw in the Cold War competition between the U.S. and the Soviet Union or the economic competition in the U.S. and Japan in the 1980s. And so then as now, there's this deep link between access to compute and your strategic position. Yeah. Going back in the history, we were digging into some of the efforts that the Chinese government had invested.

1:06:27Yeah. How many five-year plans have they done? They keep stacking these five-year plans, and it seems to be working. For decades, just to stay on the lagging edge, what did you learn about that? What can you tell us about that narrative as maybe the chip war is heating up right now, but China has been aware of this for decades and has been investing against it. Yeah, no, they've certainly been investing for a long time. I think really the last decade or so, it's kicked up a real notch, both because China solved a lot of its other technological problems. It's become self-sufficient in many other spheres of manufacturing.

1:07:04And making chips is one of the really few areas and the largest by far where China still has to import manufactured products from abroad such that today china's largest import more even than it spends on oil is importing semiconductors from taiwan wow and uh so this is something that they want to break free from more on semis than oil that's right that's insane um when you talk about that are is is that driven heavily by uh stuff that's not sanctioned on the lagging edge the 15, 16 nanometer, like the bigger chips? Or are we talking about smuggling numbers of three nanometer chips that have been repurposed through shell companies or something like that?

1:07:49It's a mix, but it's mostly things that are legal to sell to China. And you can sell both the lagging edge, like things in your dishwasher, but also a smartphone chip at three nanometers that you can still sell to China. And there's a large volume of those for PCs and smartphones and autos that are still being sold. Yeah, yeah. So every Huawei phone has a chip and those are not sanctioned. We're just talking about the large accelerators, the big GPUs that are really under lock and key. That's right. Interesting. Then take me through kind of how you process the initial rollout of the chip ban during the Biden era.

1:08:30How did that play out for you? I assume people were calling you? I don't know. I want to know your side of that era of history here. Yeah, well, I would actually rewind the clock back to 2018 and 19. That's when ASML, the Dutch company that produces the lithography tools you need to produce cutting edge ships was first thinking about selling a cutting edge tool to a Chinese customer. And then the Trump administration and the first administration leaned very heavily on the Dutch and said, don't let this go through. because if you do, China will have everything it needs to catch up in chip making.

1:09:06And so I think that was actually the key turning point. And then the 2022 restrictions that not only tightened up the rules around chip making equipment, but also banned the GPUs that you can make with these tools, in some ways naturally followed from that first 2019 move. What was different in 2022 was that there was a real, I think, understanding, And it came before ChatGPT that AI was going to be a big deal and that access to GPUs was going to be a critical input into building bigger and better AI systems. Yeah. How did the chip shortage fit into all of that during COVID? I mean, everyone experienced the chip shortages from like auto prices going up, used cars.

1:09:49There were so many market distortions. Yeah. How did you process that point in time? And then I want to step forward even further. Yeah. So I think the chip shortage didn't really impact tech companies. It impacted, as you say, autos and dishwashers and microwaves, but it brought together a political coalition across the West that included the auto companies and the tech companies and medical device companies, all of whom realized chips were very, very important. We needed reliable supply chains and we needed to stay ahead technologically in this. And that's why, whether in the US or in Japan or in Europe, you had CHIPS Act being passed to reinvigorate domestic manufacturing to try to provide some more resilience for the supply chains.

1:10:33It would have been unthinkable five years earlier, but because of this coalition, it was possible. Yeah. Is it a misnomer to call China's strategy around semiconductor and domestic production of semiconductors and all of their semiconductor strategies communist. It feels like they in some ways have an extremely capitalist system with not picking winners and putting out huge government incentives. They're shifting the market. But from what we've heard, a lot of times there's just a huge pool of money and there's vicious competition to go after it. So how would you actually characterize, like what words are we using?

1:11:15How are we describing the Chinese political system now, specifically with regard to not just how they elect people, but specifically their chip policies? Yeah. You know, I think you got a strange mix of policies being deployed by the Chinese. And in some ways, that's why it's worked relatively well by a lot of metrics. You You have the benefit of the market weeding out firms that can't compete, but you've also got the benefit of the government devoting long-term patient capital into its priority areas, which pulls in resources, pulls in entrepreneurs to these segments. I think the risk for China is that unlike other sectors that they've done well, like solar or EVs, where they've had a thousand flowers bloom and you figure out which ones are capable, which ones are not.

1:12:03In the chip industry, you don't have nearly as much of that. And in fact, the key players today, like the most significant, most advanced maker of processor chips today was also the most significant advanced in China 10 years ago. So there's much more. That would be SMIC, which produces chips for Huawei. Got it. Yeah. And so you don't have this thousand flowers bloom dynamic in the chip industry just because it's so capital intensive. And the number of people with the know-how is fairly limited that you've got a small number of flowers blooming. And all these flowers now have state ownership. Yeah, can you just different?

1:12:38It feels like they're building like a like a mirror stack. Can you walk us through? the equivalent to Nvidia TSMC ASML over in China and then maybe give me a little bit of color Yeah, and on the SK Hynix side as well. Are they doing stuff there? And then I'd like to know kind of If you can give me like a qualitative grade of like how serious these these competitors are Huawei seems super advanced It seems like SMIC and SME, maybe not, but I'll let you explain it. Yeah. Yes, I think you're right. They are trying to build out the entire supply chain. On the memory side, you have YMTC, which is their leading producer of NAND memory.

1:13:18CXMT, leading producer of the DRAM memory, including eventually the DRAM memory that will go into the pair of GPUs, the high bandwidth memory. I think on the memory side, China's done really well. um you probably wouldn't say it's a the cutting edge in terms of its ability to produce that volume but in terms of the kind of pure technological capability they've if they haven't cut up they've gotten very close um yeah i think for the the processor chip uh side it's been more complicated huawei is very capable of chip design it's done it for smartphones at the cutting edge now it's you know catching up very rapidly in terms of gpu design hard to exactly benchmark vis-a-vis NVIDIA, but Huawei is very capable in chip design.

1:14:01The challenge they've had is that SMIC is right now about five years behind TSMC, five or six years in terms of its ability to produce processor chips at quality and at scale. And so that has been a real limiting factor. And the reason SMIC's behind is because there are no real Chinese representatives today of companies like ASML or Applied Materials or Tokyo Electron that make the tools that make chips. There are companies that would like to compete with them that can compete with some of the really simpler lower-end tools. But when it comes to the advanced tools, the Chinese firms just aren't there yet.

1:14:38And so SMIC needs to buy. How far behind? Hard to say. I mean, I think certainly 10 years behind today, They're now investing a ton. They're reverse engineering, usual playbook. But if you look at the leading edge fabs that SMIC operates, it's still mostly Western tools inside. And every single year for the past decade or so, China's been one of the largest, if not the largest importers of chip making equipment in the world. Last year, China accounted for ballpark half the revenue of all of the world's key chip making equipment vendors, which is a pretty good data point for where those firms are relative to their Chinese competitors.

1:15:28What about the scale of the shell company activity buying from TSMC and other players? Have you tracked that closely? Yeah, well, certainly a big shell company was discovered and then we hope shut down last year. But the scale was vast, truly enormous. Millions, several million ships procured from TSMC and then it appears diverted to Huawei, which is at a scale that I am frankly surprised by. Everyone knew there would be smuggling in the scale of dozens or hundreds or even thousands. But smuggling at the scale of millions really calls into the question the ability of the U.S. government to properly enforce these rules.

1:16:15And the only thing worse than a rule that is overly broad is a rule that's overly broad and not seriously enforced. You get the worst of both worlds. And so I think the U.S. government has a lot of thinking of its own to do it. How do you actually make these rules stick in the real world? And thus far, it's been a real struggle. Are there any underexplored areas in data center build out where people aren't really tracking the competitive dynamic between China and America? I was listening to Dylan Patel talk about how with Huawei's Cloud Matrix 384, they were able to do, I think, fiber optic connections between the chips and basically do NVLink at a much larger scale.

1:16:57And NVIDIA had tried to do that and maybe not been as successful. and it felt like Huawei is slightly behind or seriously behind in a number of categories, but that might be one place that they're actually ahead. Have you tracked that at all? Is there anything that people should be digging into there? I feel like the narrative was like, everyone was getting up to speed on NVIDIA, and then people learned about TSMC, and then they learned about ASML, and then they learned about SK Hynix, and I feel like we're all about to learn about a networking provider soon, but what's been your take on that industry?

1:17:29Yeah, yeah. Yeah, no, I think you're right to focus on networking as a key area of rapid technological advance. You know, it's hard to benchmark who's how far ahead, unlike in Foundry, where you got like a really clear way to measure TSMC versus SMIC. There's not such an obvious kind of benchmark. I think the underexplored area in the Chinese ecosystem is actually not at the chip level, at the cloud level, which is we know that China's smuggled in a large number of chips. We know that there are a meaningful volume of chips that SMIC has produced domestically for Huawei. But it appears that a lot of these chips are in data centers that have low utilization.

1:18:14And anecdotally, all the anecdotes we hear about utilization of AI data centers in China suggests a lot of real challenges in terms of getting utilization, partly because there's some state-owned telecoms that are just mismanaging their cloud businesses. But it's, I think, surprising that Chinese firms like DeepSeek talk about compute limitations, given the number of chips we know that are actually in China. And so if you're starting with a thesis that China is good at centralization, that doesn't appear to be what we're seeing here. We appear to be seeing an inefficient decentralization, perhaps driven by some of the pathologies of having state-owned firms from data centers.

1:18:52Do you have any insight into what's happening that benefits China but happens outside of China? I've heard reports that folks are taking training data on hard drives and flying it to another country, doing a training run, saving the weights to hard drives and flying back. You could imagine DeepSeek v4 or something being trained by a series of shell companies, and it's just an American guy with a big AWS account, right? But at a certain point, I would hope that Amazon says, wait, who are you and why do you want to spend$100 million this month on cloud? Like, we like billing everyone a few thousand dollars, but this is going to set off some alarm bells.

1:19:39What are the risks there? Is the government doing anything about that? What is your take on this idea of just hopping on to other clouds and then routing the data backwards? Yeah, I think this is an emerging area of focus for the government. I think you're right that it's an area where the U.S. government doesn't have a whole lot of visibility. And I think cloud computing companies haven't, they've got your customer requirements, etc., but not nearly as much investment as, say, a financial institution spends on making sure they're not transferring money to a sanctioned Iranian entity, for example.

1:20:13I wouldn't be surprised if for the next couple of years, we have a lot more build out of this type of regulation precisely to deal with what you're talking about. And I think a large training run is large. And so I think we shouldn't be surprised to see heightened know your customer requirements as you get larger and larger in terms of the scale of compute that you're using. The other nuance is that certainly US cloud providers are key here, but there are also cloud providers from other countries in other countries that are playing a role. We read a lot about Malaysia the last couple of months as one of the key buyers of GPUs and a variety of Malaysian companies that are maybe working with Chinese companies operating there.

1:20:59I think that's also an area where we might see more regulation. There's been proposals out of the House of Representatives, for example, to say, we want more GPUs going to U.S. cloud providers and fewer going to non-U.S. cloud providers precisely for that type of region. America first for AI is sort of the way to think about that. And, you know, there are different parts of the government that would find that pretty appealing. Can you, I mean, you mentioned Malaysia. Can you take me on a little bit of a tour geopolitically of what's going on in some of the other countries that might be on the horizon for AI relevance?

1:21:35I'm thinking about, obviously everyone knows the US and China, and then Taiwan's role, and then Malaysia pops up, and then France has Mistral now, and there's deals being done in the Middle East. I was always surprised that Russia was not on the conversation at all because they have a very long lineage of incredible math prodigies. And it feels like you would, I mean, there are a number of AI scientists in America who come from Russia. And so I'm surprised that we're not seeing, oh, Russia's getting in the supercomputer game or trying to build out some massive data center. They have been busy.

1:22:11And so what are some of the other countries? can you take me on a tour of maybe maybe let's just start with like what's going on in the middle east how are you thinking about that is is it just like a jump ball bidding war between are they going with the west or the east uh and then we can move on to some of the other regions in the world that might be relevant or not yeah you know in the middle east uh has a real advantage in terms of you want a capital they can deploy which matters if you're trying to build big compute clusters and i think both for saudi and especially for the uae there's a desire to build their next generation of industries if and when oil becomes less reliable for them.

1:22:48And that's driven this investment in terms of building infrastructure. And we saw after Trump visited a couple of weeks ago now, big deals announced, especially in the UAE, 500 ,000 GPUs a year to the UAE is a large number. I think the question for the UAE is there's no doubt they can build data centers, no doubt they can afford the GPUs. Can they build businesses on top of that? And I think all these countries want to do more than just host Azure or AWS data centers. They want to own their own businesses and build on top of that. And that's a harder thing to do than simply building data centers.

1:23:24I think the UAE has big ambitions in particular, but it's easier to have ambitions than it is to realize them, and it'll be a challenge. What's your take on TSMC's traction in Arizona? NVIDIA's had some announcements as well. how how bullish are you on on the progress they're making so i think the good news is that tsmc has publicly said uh their manufacturing yields so the share chips they produce that actually work are just as good in arizona as they are in taiwan that's a really powerful data point i think the bad news is that on a unit cost they're still a lot more expensive than in taiwan and that cost differential will decline over time but it's it's a real differential and it's not going to go to zero.

1:24:08And so the key question, I think, is to what extent TSMC's customers are going to keep pushing TSMC to build more in the U.S. versus be comfortable with production in Taiwan. I think the trend has been that they're pushing TSMC to do more in the U.S. And certainly TSMC, looking at the tariff dynamics, has registered that. They've announced plans to build six fabs beyond the one operational one-on-one construction right now. Unclear exactly over what time horizon that happens. But so long as you've got these kind of tariff and geopolitical issues front and center, I think TSMC customers will be encouraging them to pursue this diversification.

1:24:49Yeah. Who would eat that cost? Because Nvidia has famously very high margins. All the hyperscalers are thinking about doing in-house chip design now. I'd love to know your take on Microsoft strategy, the TPU at Google, anything else that's going on, Tranium and Amazon, for example. Because you could see, yes, TSMC in Arizona is more expensive, but that cost is kind of mitigated by being able to route around NVIDIA's margin. But I don't know if that's a reasonable thesis. Yeah. I think my assumption would be that the cost is shared across the supply chain. And And actually, if you look at the supply chain, you've got a number of high margin companies involved.

1:25:32TSMC has a pretty healthy margin. Its key customers do as well. And if you take an iPhone as an example, ballpark the share of the TSMC content in an iPhone, $50, let's say, if that goes up by 30 % in the context of a$1 ,000 iPhone, it's not dramatically going to change the economics. Same for an NVIDIA server. So I think you're right to say there's plenty of places where you could absorb some of that cost. What share of it goes to TSMC versus customers, I'm sure, is a constant source of negotiation. Yeah. So there's kind of like an interesting supply chain here in the actual delivery of AI applications.

1:26:18And I can't help but wonder if some of the some of the international players outside of America and China are maybe targeting the wrong thing. It seems extremely valuable to be able to fab chips in your country and own your supply chain and be able to produce a consistent flow. And then it also seems incredibly valuable to have the front end to AI, essentially like a chat GPT national champion that users go to. But the model itself and the data center seems maybe less important there, but it feels like all of the different countries that we're talking about, whether it's France or Saudi Arabia, they seem obsessed with owning a data center.

1:27:00But then if the local population is going to chat.com instead of Le Chat for Mistral, like has France really won if they just have a model? And maybe they're thinking about in terms of like deploying within the government in the same sense that like the DOD is deploying LAMA right now. But how do you think these international countries are thinking about the different places that they can try and plug into the overall delivery of artificial intelligence? Yeah, I think that's the right question to ask. And I think you're right to suggest that just having a data center is not a guarantee of making money off of AI.

1:27:40Yeah. And it seems highly likely to be the case that the application layer and the operating system for AI, however that develops, will be places where a lot of money is made. And again, not clear that that if you build a data center in your country, you're going to guarantee that you have that at all. I think when governments think about this, there's probably a degree to which they're not asking the question you're asking and they ought to. There's also, I think, a degree to which they're thinking about sovereignty and control over their tech infrastructure. And they're worried that if they become wholly reliant either on U.S.

1:28:19or Chinese AI providers, they're going to be in a weakened position politically. Now, is the best use of their resources solving that problem building data centers? I think that's an open question. But we've certainly seen a lot of countries turn in that direction as their first response for trying to build a bit more room for maneuver in the AI space. It's interesting, as the UAE was opening up to Western businesses back in the day, there was a structure that was popular where if you were Coca-Cola or Michelin and you wanted to sell into the UAE, from my understanding, you needed to set up basically a 50-50 JV with somebody locally.

1:29:01so it was like basically owned and operated to some degree but almost like more of a franchise model and i wonder if we're we'll end up effectively seeing something like that like that on the data center side i just wonder i just wonder how much that particular layer matters like if you look at what china's done in just the delivery of conventional web services it's like obviously knockout drag out fight to try and build data center capacity and import chips but also to develop the traditional semiconductor, the CPU supply chain, and then ban Facebook and Google and develop homegrown alternatives for the application layer.

1:29:39But they haven't tried to rebuild Linux from scratch. Like North Korea does have a fork of Linux called Red Star Linux, apparently. But in general, you know, Chinese phone companies have been fine forking Android. That hasn't been a place where they've said, oh, we gotta rebuild, we gotta own that from start to finish. It's like we want to be able to make the phones and we want to be able to deliver the services so that you know Google can't like bring their own censorship rules and we can't and you know Samsung can't cut us off of actually delivering the phones, but in terms of the actual operating system layer Everyone's been kind of fine with standardizing around Linux and Android.

1:30:17I don't know. It's interesting I wanted to talk about The AI trade Trade deals. Yeah. Yeah, so this is what I want to get into I would love to to understand the talent market to the in China Yeah, to the extent that you can. Yes. It's been what we've been covering for the last couple weeks, these sort of massive changes in talent moving around between the different labs. I'm curious if there's that level of competition for talent or it's – Between like deep-seeking and all the other. Yeah, because in some ways, you know, these American researchers constantly, you know, shifting around to the highest bidder.

1:30:54It's not necessarily the best for winning the AI race. or winning the AI war. It's the right strategic moves for the different labs and different hyperscalers to compete for this talent, to build the best businesses and products that they can, but it's not necessarily the best thing. Yeah, an authoritarian might say, hey, let's just consolidate all of the AI talent at one lab in America. That's not the way America works, but how does it work in China? Yeah, you still do have a pretty flexible labor market. I think there's an interesting question. Would you be right to consolidate all AI talent?

1:31:28I think if you're a believer that we're going to kind of reach a single threshold called AGI, and if you reach it two months earlier, you've got this kind of enduring strategic prize, then maybe concentration is the right strategy. But if you're a believer instead that there's going to be a bunch of different products, which will themselves be pretty transformative, but there's not one threshold you want to pass, then actually you probably want a more flexible system that's going to allocate between different firms. And, you know, maybe it is macro efficient to have people jumping between firms once every year or two if they're bringing the cutting edge, spreading it to different firms.

1:32:04That could actually be a good thing on the macro level as well as for the individual level. I would be more pessimistic about China's chances if I found them centralizing more aggressively. I thought one of the most pessimistic data points this year about the Chinese ecosystem was the news that deep-seed researchers were having their passports confiscated to report international travel, which to me is a strong disincentive to doing AI research. Rob Markman China seems to love doing that. They were doing that to rare earth element experts as well at rare earth minerals. It seems like it's just in the playbook when an industry becomes important, the passports start disappearing.

1:32:43Yeah, I mean, the other elephant in the room is you can imagine the amount of pressure that the CCP is putting on Chinese AI researchers that are here in the U.S. Maybe they want their citizens already or they have a path to citizenship. But if they have any family or any exposure at home, there's still a lever that they can pull. I want to talk about talent in the context, not just of AI model development, but actual chip fabrication. It feels like right now what we're seeing in the Foundation Model Labs is an extreme power law in the value of AI researchers. Hence,$100 million offers for top talent and starting salaries in the$300K range.

1:33:26That is a wide, wide power law, right? There's always been this narrative around TSMC that you can't chat GPT how to make a semiconductor. There's a couple books out there, but it's a lot of lore. It's a lot of tutelage and mentorship. chip, how power law distributed is talent in semiconductors? And is there a world where Chips Act 2.0 gives Intel a bunch of money or something, and then they start offering TSMC experts$100 million offers, and that really pulls forward American chip production? So I don't think the mental model is that expertise is power law distributed. I think that there's lots of very unique capabilities you need to make a chip.

1:34:15And the level of expertise you need in each of the verticals is very deep. And you need all of them. And so you only got 90 % of them, your chip doesn't work. And so you need to build this expertise over a long period of time. And so if you were to say, who are the three people you need to hire from TSMC to replicate TSMC? I would say that's not the way to think about it. You need to hire half the staff or more. when TSMC was bringing the facility in Arizona online they had hundreds of people from Taiwan working in Arizona and this was not even their most leading edge ship production this was two generations behind the cutting edge with all of the explicit help coming from headquarters as well and so I think that helps illustrate how much of the transfer needs to happen at different levels but if you look at who's Mick hired over the last decade.

1:35:06It's been a ton of engineers from Western, Taiwanese, Singaporean semi-hunter firms. And I think a lot of the advances that they made can be ascribed to the fact they were able to import all of that human capital from workers who had previously spent a lot of time at other companies. What are you tracking? We're halfway through the year. What kind of events are on the horizon or various potential new legislation are you kind of following? I think from the policy side, there's going to be some action on tariffs and semiconductors, it seems. Presidents repeatedly said that, unclear exactly what direction that's going to take, but it could have pretty big implications for the cost of building AI data centers.

1:35:51That's certainly one. And I think on the kind of tech controls front, we haven't seen the last of the moves from this administration either. And I wouldn't be surprised to see more there too. Have you been tracking the leadership change at Intel? Lipp-Bouton is coming in. It seems like they're going to be doing a lot of layoffs. I was trying to go back through Ben Thompson's analysis of Intel through the years and try and understand how the narrative... I've been tracking Intel that long to follow whether there should be a split. What has been your take on Intel and has it changed? and how do you think, what should we expect from Intel over the next few years?

1:36:34It seems like they're going to be tightening things down, but do you expect any significant changes? Would you recommend any specific changes? I'm interested to hear your take on Intel. There's a lot of people who think that a split between the manufacturing and design businesses is the right way forward. I think that the question, if you think about a split, is what does that mean concretely? Because the reality is you'd still need to have a fair amount of integration between the design side and the manufacturing side, because the manufacturing side is making chips for the design side that will be in progress for a number of years.

1:37:05And the manufacturing side needs business from the design side. So, you know, split or no split is probably too simplistic a way of thinking through the option set. There's a bunch of different variables around that. And I think that's probably exactly what LitBoo is wrestling with right now. I think that the second challenge that Intel has is, and it's products business, the CPUs for CCs and data centers. I think there's a wide agreement that they've got work to do on both those fronts in terms of remaining competitive with AMD. And they've lost market share over the past couple of years. They need to win that back.

1:37:42And so I think those are the two key issues he's got to deal with. And they're, of course, interlinked. And it's hard to say you're only going to focus on one of them. totally this is great any other questions fascinating no we'd love to have you back on this is fantastic thank you so much for taking the time this next quarter really appreciate this cheers bye and i'll be right back up next we have aaron ginn the gpu whisperer as uh his moniker in the information proclaims he has a new piece out in the wild i believe it's in the wall street journal will we will get the latest from him aaron how you doing hey man how's it going wait did i get that right new piece in the journal or somewhere else uh the one on robots uh well arena mag is that where yeah yeah that one that was yeah arena mag and then i did you're prolific yeah it's fun it's great i'm back i'm a kind of a not only christian but academic at heart so So I like writing.

1:38:41I like all sorts of things in that zone. But yeah, I have a new one on robots. Okay. Take me through it. What was the inspiration? What's the thesis? Give me the details. I guess I kind of know for hot, spicy takes done in a kind manner. And it did so many matter. But I think most of the analysis done with people in our space is devoid of any kind of reference framing. they uh kind of live in people in silicon valley particularly people in ai live in the they think they live in this kind of arrogant microcosm as c.s lewis would say like chronological snobbery so they don't appreciate how other things that look very similar are just patterns of repeating uh as the uh you know one of the most famous solomon phrasing uh because the theorize in the bible that solomon said that uh you know that there's nothing new under the sun uh we don't fully know the attribute to but generally attribute to solomon and and in that sense like innovation kind of goes the same pattern that you have something that innovation is always intentionally oriented capex both in the sense of knowledge workers and also capital expenditure and and it always starts with the as a rich luxury but then it gets proliferated out and robots as people have been kind of pitching both in the sense of the doomsday sense and the utopian sense i think are wrong because they if you look at past patterns of innovation and how it's been adopted in society it generally doesn't get oriented towards the rich like the rich continue to do what they do like there's a reason why rich people have their own planes have their own drivers have a yacht have multiple houses have chefs and carers things like that like that's basically the same trajectory of lifestyle that they've had for a while yeah and in terms of like things that you own things that you do things you buy like there's the the names that are associated that the Yale Club and the Harvard Club continue to be the same names.

1:40:30People who go to Davos are the same names. But the real arc of capitalism, which is why the existentialist, the German, the Frankfurt School was completely wrong, is that what it does is that it takes those luxuries and gives them to everybody else. I think that there's a very good thing with a venture capitalist, but he said that his kind of moniker of looking at innovation and adoption is oriented towards taking something that is available to the rich and giving it to somebody else. So Uber is the example of that, right? Private driver did everybody else. And it's a robots will be the same thing.

1:41:01Like certainly like robots will, there'll be a certainly, you know, a luxury version of a robot that will be deployed for people and the rich and famous, but the rich and famous can afford people. And rich and famous like character. They like things that are like, go listen to, like I know people love comparing you and Geordi to all in. Right. And just go look at what Shamath talks about. it's like it's things that are take a long time or delicate cashmere i love it yeah yeah this is made of endangered panda fur right that i hunted myself right i am and so no all-in host has ever wore panda fur that's yeah yeah so the king's cashmere is real yes yeah yeah so so like so for the rest of us right upper middle class down robots will be for us robots will give us the ability to do things that we didn't have before that makes us more like rich people.

1:41:54Yes. And aka time. And so that's what it's arguing is that robots will not be dystopian, it will not be utopian. It'll be every other pattern of way of innovation and helping the average American live a better life. Yeah. Yeah. I noticed this when I started interacting with like the multi, multi-billionaires that every app on my phone, they had a person for. So if you do incorporation of your company through LegalZoom or Stripe Atlas, they have a lawyer full-time on staff. You have Uber, they have a driver. You have DoorDash, they have a chef. And it's like the new technology. Yeah, the guys. Yeah.

1:42:31How many problems can you solve by just calling a guy? Very interesting. What do you think about it? It was Chris Pike, right? Somebody's going to create guys.app. They're democratizing access to guys. To just a guy. It's just a database of good lawyers and drivers. Yeah, yeah, yeah. I mean, even like, you know, call 911 and that's for everyone. Private security is the opposite side of that. But what do you think about Chris Pike's take that when, I mean, I guess this relates, but I think his take was that if there is a human that is acting as like a roadblock to something that you want, that will be replaced by artificial intelligence or robotics.

1:43:13But if the human is part of the delightfulness of that experience, and his example was like an omakase chef, that will never be replaced. What do you think about that? Give me your reaction, please. So I'll be a Chamath from a race. Because the question was, he said, he was talking to J-Cow, right? He's the ultimate the thing guy, whatever the thing is, that's J-Cow, right? And he goes, talk about watches. and Shumato was like, oh, get a handmade mechanical watch, right? So I'm wearing one right now, actually. So like, and so you look at this, right? I know, right? So it's Alana. Go to getbezel.com.

1:43:51Your bezel concierge is available now to source your own watch on the planet, literally any watch. So, you know, hand and guilliche, hand and carve, hand and carve. You see a software engineer that he watches. And so the reason why this is worth it is because of that human interaction, right? It's that his ability to do everything by hand is like the machine built in late 1900s. He does everything. He can show it. That's the point. Another example of this would be like being a waiter. You could use this example. There is automated checkout on Applebee's. I don't know if you've been there recently, but there's automated checkout Applebee's.

1:44:27Chick-fil-A, one of the most highest grossing per store sale fast food joint in the world, one of the best. Also delicious. It has tons of people. right so it could if it wanted to but it's not the point of where you go and and fundamentally there's this disconnect i think it's spiritually connected i think you could watch some of peter till's recent uh interview with that you can kind of hear some of these trends it's like if you remove the ability of the anthropomorphic orientation of what innovation is and why you do it you come up with these in conclusions that are basically absurd like there's people in the solica valley who argue absurd things about the nature of what the arc of innovation is going to because the presupposition and priors and the metaphysics built into it are basically that humans are worthless and existential and they kind of have a Star Trek-Decarve view of the world that's very dark.

1:45:13And the same thing exists with social media. The same thing exists for search. The same thing exists for mobile phones. Now it exists in AI. All of these are continuations to the same thing. And just like the AI doomers, if you don't look at the names of the people, it's like the same people who are like the title to social media is going to end the universe sort of people because it's the same line of thinking. so humans just like what you're doing now people watch this show because of you too like like it doesn't even if there is a memetic machine yeah it's not the most efficient way to get the news but it's fun but it's fun right like like like because it's part of what the nature of this uh it's hard for me to imagine the idea that that uh let's say a humanoid robot you know gets real traction in restaurants will naturally, you know, and you can go and sit down at a restaurant, a humanoid can serve you and whatever, and it's efficient, it's fast, it's smooth.

1:46:08That will naturally create even more opportunity for human waiters, waitresses that love the craft of service. And I think it is a real craft. And I think that anybody that's had a truly fantastic dinner and can attest. Yeah. Yeah. So it's one reason why I think the adoption curve of AI is going to be kind of like two different buckets of things. There's going to be the infrastructure side, which we're going to excel at. And then there's going to be the adoption side, which we're not going to excel at. And some of it is political, which is everything from we lost the AI moratorium of the bill, but thankfully the bill got passed.

1:46:50So we're not paying higher taxes. So thank you, Daddy Trump. Daddy T wins again. So yeah, exactly. Make a little noise. But on the other side is like, if we're going to do, let's say, one of the best things AI does is that it allows the ability of computers to look like humans, behave like humans. It's fantastic at that. It's the ability to model us, to make a memetic version of us. and from an education perspective that could be incredibly powerful because my upbringing even though i went to public school before it was insane so like i'll say from that but my dad who's chinese read me much more of a talmudic traditional tiger dad sort of fashion so i still remember this one conversation i had with my dad i think it was in like fourth grade and i got like 100 in a class uh and he said why did i why did i do extra credit right it was like his immediate response and and you know and the you know the the more caucasian people in the audience will clutch the pearls and oh my god it's so traumatic right but the point of it right is like oh you always can do more right he's teaching me resiliency teaching me how to challenge myself teaching me how to uh push beyond what i think i can do and to like open my expense because his goal was to not have the upbringing he did which is super poor and and so imagine the ai that could do that right that could be your your own yeah yeah they did like it your own ability to challenge your thoughts like people are thinking like the cluey thing which i think is a very surface level understandable ai can do but imagine if ai was actually like a challenge agent that let's say you got the right answer and i was like you're wrong so prove it to me like you're right right there's a whole other layer of this and this is this is what like in terms of the el salvador project that we're working on is like that that's kind of like the orientation of why other countries don't have that.

1:48:36Like they don't have the teacher, the library units let alone. They don't have the internal capability of delivering these types of things today, but a computer can come in now because what they have now, they have nothing, right? And they're willing to take a swing being like, well, you know, if it works 50 % of the time, that's a huge win for our country. But America, the lawsuits and this and this and this, right? But, you know, but Kelly is just like, oh, this sounds like great. We should just try this. If it doesn't work, I'll do something else, right? And so that's why I think we will excel at infrastructure, which is why it's so important We push in video products everywhere.

1:49:05But the adoption is going to be in Africa, second world, third world countries. It's not going to be in America because we already have people doing it today. And people will still be the preferred, like rich people still have tutors, right? And they're still going to have tutors, even though AI can actually be good at tutoring. Interesting. Give me the update on sovereign AI generally. There's an article in the Wall Street Journal, I think yesterday, about more and more Western companies adopting DeepSeq and Quen. It's like, oh, I actually read that. Yeah, I need a check. No, it's Josh Chin. But yeah, I mean, any major updates there?

1:49:47Did anything get stuffed in the bill related to chips or semiconductors or anything like that that's relevant? Not that I know of. I mean, people always all have to take a step back and realize this is a tax bill, not a spending bill I'm an Austrian economics guy as y 'all know so I don't like that I don't believe that you can keep spending forever this is a tax bill it's literally why it's going through reconciliation so people need to like lower your threshold like I agree with Elon he's one of my heroes he's very wrong on the purpose of this bill but he's right we're spending too much money but it's also we have to get away from omnibuses so the AI monitoring stuff got removed I'm pro that, Saks is pro that But the OSTP office is part of that.

1:50:30This does not be something should be resigned to the state. It's something that we have to win, and we can't have these varying laws everywhere. On the chip side, essentially, there are – like I have a hat that the president Salvador gave me. So for countries like Bukele, who's trying to basically provide this sort of infrastructure and capital to the world, the market is there for us. And we have these basically voices that I think are living in an era of where the moat resided in areas that they previously defined sort of infrastructure ways. And we're at this moment now where we could accelerate victory to where, hey, how about China?

1:51:12We outsource all manufacturing last 30 years to them that make all the stuff in Timu, take all the stuff in Xi and all this crap. Like, why can't we do that to them for similar connectors now? Like, why can't we take their money? right and you see all these people like oh but wait a minute like what if ai destroys the world right and it's like it's all this kind of like soothsaying again title two people go back to the to that debate remember free speech debate online and like telco versus where does the right to speak reside and the right to serve content those people are the same people in the ai moratorium it's literally the same line about thinking and and so the the ai moratorium was supposed to specifically on like training and application but in el salvador like they pass an ai law like basically a free training law where that which is how the law should be structured when you train something it's up to the user to apply it if the user applies that to make a robot to kill somebody just like you get a hammer and kill somebody right or get a robot to scratch you know jordy's mercedes right it's that person's fault right it's not going to be like if john trained that open source model and they use it against Geordi.

1:52:20It's not John's fault. So in Osamuza, they passed that law, but in America, they want to co-mingle these responsibilities most out of baseless fear. But still, the policy of the administration is still towards, we need to export, we need to become the AI exporter of the world because we have this ability to be the definite infrastructure of the next era and not reside in this kind of loose, or not loose, It's definitely it's actually really hard to understand. I mean even with the critics are saying even with something less dramatic than like You know using an open source LLM model to commit a crime or something just just just the intellectual property thing Like it's very easy that we've had the ability to go and you know Pirate a movie and upload it to YouTube for a very long time That's like not a crazy new technology, but it is in the sense of like copying data is new and distributing it all over the world and yet But YouTube doesn't even need to stop it ever from happening.

1:53:18They just need to stop there from being an economic model. You can't build a business on top of YouTube just uploading Star Wars. That's not going to work for you. And as long as you kill that economic model with enough pressure and say, hey, we're not going to serve ads against this. We're not going to pay you out. We're going to ban your account eventually. There's a bunch of ways to prevent this stuff at the level of like, Yeah, like someone might some kid might upload Star Wars and five people see it and before it gets taken down Not the world who cares exactly and so yeah, yeah going after the Like the individual actors that that's a very very good point.

1:53:55I like that Yeah, the most the most free phrase you could ever and maybe could raise your kids this way too because it was most freeing phrase ever received from from a guy's famous, but I won't say his name because I uh, but he's a huge radio star. And he taught me this. He goes, whenever you feel stuck, just say the phrase. So what? So what? Right. And then like, why should I buy into your presuppositions? Why should I believe in these things? Like prove it to me. Like, so what? And it's so freeing to reside in that zone of like, like you just said, so what? Like somebody does that. Like, why is that a big deal?

1:54:29Like, and, and then it kind of puts people on their heels. Like, like I said, this one's in a debate about like, since we're talking about chips, It's like, well, like, you know, ByteDance could get access to this. So like, first of all, Oracle's number, like top five customers are like Chinese. So like, you know, they're already doing it. Right. Then I said, like, so what? Oh, my God. Like, it's like, oh, yeah, videos will be more musical. Like, I was like, it's a, and they're like, well, we could do this. Could you do this? Could you do this? I'm like, I'm like, yeah, I could say the same thing about anything.

1:54:58I could say the same thing about like, well, you know, if you, if you buy this watch, you could use this to put on a bomb. Right. and I know like it's like that's like that's like their level of thinking there's so many yeah I was gonna I was gonna ask Chris earlier you know how if how he thinks about AI war versus AI race and then you would have to ask him the question well then should you have named your book the chip race and then that would have sold like one percent of the copies yeah um so is it a war or is it a race interesting yeah it's so it's so tough I mean so many of the loudest voices in our industry are beautifully conflicted because they run multi-billion dollar businesses that are dependent you know future fundraising current fundraising is dependent on certain narratives and so it's hard to figure out what's what's real yeah As it relates to the China stuff?

1:55:56Not just that, even just like the progress of models, how much we should be spending, when AGI will arrive, when super intelligence will arrive. AI war versus AI race. Because if you brand it as a war, then all of a sudden it brings in a new level of funding. There's all sorts of downstream implications of, well, it's really important that we get the AI doom question right and so we need to over invest or it's really good that we win it's really important that we win the geopolitical battle and so we it's okay to over invest in this case you know it's all about just like justifying different levels of underwriting and when you add one of these infinite numbers you can say one percent chance of infinity is still you know I'll pay any price yeah yeah I guess I guess my firm is just way more optimistic like I understand like why why can't we just invest because it's cool?

1:56:54We like stuff that Elon does. I don't want to test that because I like my cars to go boom, right? So, but it's still cool. Like, why can't we just be cool? Why can't it just be, hey, this is awesome. This is like moving the trajectory of the human race to something even more amazing. It seems like we're getting there. War, death, ugh. But it seems like we're getting there. I mean, I don't know if you saw Nat Friedman's post, but he was saying he started at Meta and he's going to just build some cool stuff. And I'm like, yes, that sounds awesome. Because there are a lot of AI products, but a lot of them are - And he's going to discover the great lost wonders of history.

1:57:28I hope so. I hope so. He better. Anyway, thank you so much for stopping by, Aaron. This is always a pleasure. What are you doing for the fourth? Yeah, what are you doing for the fourth? No, I'm not going to work. I'm doing something with my church. And I usually, my house is a third space. So for the community. so it's very on the other side of this is uh i built it my i built a custom house and so like it's a big basically like i don't drink but it's like a big beer hall essentially from front to back wow so so we usually hold something it's like uh i don't know maybe three or four thousand square feet yeah and it's just a big open room with like a kitchen dining room living room outdoor space very cool so we usually like host stuff and people like use use my house uh to do that um so i think uh this goes back to the the you know the human element is like fundamentally innovation is supposed to be at the promorphic is to be make humans better that's why people pay for it and there's certainly negative downsides of this uh and we have to mitigate those downsides but you always should start from the the phrase so what or go go read thomas soul about can you even regulate downsides without creating a secondary effect of mitigating upsides and and And I think that from many people are projecting lots of things in AI that they could just get from just like being in community and just like having friends.

1:58:48And like y 'all have your sauna workout. Yeah, yeah. Sure for mail. Yeah, exactly. Yeah, you like so much of the projecting of internal problems people have into technology. when if they had a friend uh they resolved their loneliness academic they had a good family they surrounded by a religious community uh you know particularly christianity uh that a lot of these things kind of go away where you can just look at a technology you can look at innovation from where it is which is like this can just be awesome it doesn't need to kill us it doesn't need to start a war it doesn't need to be a j isi all those are certainly things we can talk about but to overextend you dig into that person that's arguing that there's typically a like a a pathology built into them that is projecting externally when it's just like hey dude like do you still like want a hug like you just like you just want like what you want to hang out like you should pull or like go to crossfit or something like and then and then it kind of like you know fades away because there there are many things in the world you should not care about and and i don't like i get asked to about things and i'm like i don't really care about that like it's not something i want to talk about and because you care about less things is the key to joy because less things means you're actually more responsible for more things which means you can be happy because you can see your work go into something that has output but when you become unhappy is when you see all this crap happening around you and you get worried about it and you do something about it and you really can't and there's just things you just shouldn't care about and then you're like free I hate to cut you off but I'm sensing a TED talk we'll have to We'll do a TBPN talk very soon.

2:00:27But have a fantastic weekend. Great seeing you as always. See you soon. See you guys. Bye. Well, speaking of fantastic AI products, go to fin.ai, the number one AI agent for customer service, number one in performance benchmarks, number one in competitive bake-offs, number one ranking on G2. If you like winning, don't get in a bake-off with fin.ai. Don't get in a bake-off. Slash TBPN. Well, our next guest is Bridget from Founders Fund. Welcome to the stream, Bridget. How are you doing? busy time in oh wait okay so can you say where you are i was on it was a different format you guys were in the old format we were yeah new studio we are now in hollywood future capital media media yeah and entertainment future home of media but speaking of future homes where are where are you right now so i'm at etc where's that you're so you're you're literally doing the meme you're a venture capitalist and you're you're in france conference at a conference oh yeah at a One hour conference, seven days on the beach.

2:01:26Is that what we're going on? Oh, my gosh. It's productive. It's productive. It's productive. Okay. She's working. Yeah, I don't know why they have it in France, but last year it was in. I know why. It looks beautiful. It looks beautiful. All the capitalists will be there anyway. All the events with an ocean backdrop. Yeah, it's fantastic. Yeah, give us the, what actually happens at an event like that? Is it people kind of voting on direction of the technology or is it just like pitch? meeting founders, building on top of stuff? Yeah, I would say it's like it's sort of a mixture of everything. I would say like these conferences probably originally started because the crypto industry is like just so decentralized, like for lack of a better word.

2:02:05And everyone lives in different places. So I think originally it was like for kind of like like-minded individuals to meet up. And Peter actually used to go to some of the really early ones called like FC, Financial Cryptography. And then they slowly evolved into like ETHCC. There's like Token 2049. And it's just a good way to like see everyone in person. Like obviously in person is so much better than Zoom, but people in crypto are just so distributed. Like they're not really concentrated in one city, maybe New York now, but it's still kind of like, you know, since the U.S. isn't like insanely regulatory friendly yet, like people are everywhere.

2:02:38No, I feel like there's very serious projects in San Francisco, New York, Puerto Rico, Singapore and Dubai, which is like all over the world. Yeah, totally. Yeah, but Joey and I try to do like one-on-one meetings most of the day and like because if you just get caught up in events It's easy to just do a bunch of events and like not actually get anything done. So yeah, yeah We we've gotten steamrolled a few times We're not crypto native enough to know the full calendar and we'll try and book some crypto people and they'll be like are you crazy? We accidentally held our crypto day And we just didn't know and we were like well brian armstrong is coming on So, you know if he's not going to this thing like yeah Yeah, yeah, yeah.

2:03:17We actually had some folks from the conference hop on, which is cool. And it wound up working up well. That's good. It was like 2 a.m. our time or something. Yeah, yeah, yeah. And getting Mert on the street. We had Mert yesterday, and he's always great. But he's always traveling across the world. I wanted to get the update on the stablecoin world. What's interesting? Is the game over? Circles out? It feels like the legislation is passing. But what are you seeing? Are there still investment opportunities? Are there still opportunities to build? or is it the time of the enterprise SaaS integrator in Web2?

2:03:50Yeah, it's a good question. Yeah, I think stable coins are definitely the main character right now. Like we there's a new announcement every single day with something stable coin related. Like it's very hypey. It's very in the meta, but I think like for good reason. I was joking with Joey the other day. It's like stable coins are basically the part of crypto that has found the most product market fit. and it's just like a tokenized US dollar, which is like kind of ironic because the original crypto was like, we're going to rebuild the financial system. Everything's going to be Bitcoin denominated.

2:04:24Like we're going to do, you know, basically recreate. And now we just have like the tokenized US dollar. Yeah, well, everyone was complaining about volatility of the meme coins and the crypto community said, fine, we'll give you something with zero volatility. Are you happy now? And of course, the critics. Well, it feels like some of that hype, I want to get into the tokenized private assets. Yep, totally. stocks as well. But before that, Sachs came out and posted that there was a certain concession in the stablecoin bill around stablecoin issuers not being able to pass yield back to holders. Could you could you dive into that a little bit?

2:05:01Because that was felt monumental. But yeah, I think that was something that the banking lobbyists probably snuck into the bill. And the bill was like a Herculean effort even to just like pass the Senate. Like it got blocked once just to like get it like to the Senate was like basically like an insane task. And I think basically the gist of that is like if too much money flows out of banks and into like stablecoin issuers like Circle or, you know, there's going to be like a million in the U.S. probably. That's money flowing out of like the traditional financial system. And then basically like net new money cannot be created via loans because stable coins have to be one to one backed with like T-bills or cash and cash equivalents.

2:05:43So stable coins would kill the entire Western financial system. Like, I mean, basically, and they make I mean, it makes so much sense because it's like I like I would rather hold my money in stable coins. It's like one to one backed. It's very safe. And when you put your money in a bank, it's like people think it's safe and whatever, but it's actually like not that safe. Like they're like lending it out and that's how new money gets created. And that's how the economy functions. Yeah, we all experience with the SVB crisis. And implement their monitoring. Yeah. Yeah. And it's, it's, yeah, we, we, we all experience this firsthand.

2:06:13I mean, most people in tech with the SVB crisis and first republic and the idea of FDIC insurance. And everyone kind of learns about fractional reserve banking at the worst possible time when there's a crisis. But you actually wrote an article all about how a stable coin bank might be different. Can you take us through some of what you might expect? And why do we, I feel like there's this whole crypto podcast called like Bankless and they're all like, don't have a bank at all. And that was a meme in crypto for a while. And now the banks are going to issue stable coins. Yeah, yeah, yeah. So what is the role of the bank going forward?

2:06:44What even is a bank? Like, why do we need that at all? And then walk me through the stable coin bank idea. Definitely, definitely. Yeah. Okay, let's see. Yeah, I think me and my friend Abhi wrote the piece originally like early this year. we basically were looking into this act that didn't end up passing it was called the credit card competition act it had like a three percent chance of passing the senate or something but we were reading it and basically what it did in practice if it were to be like passed into law would be to like allow um or force banks to accept a network other than like visa or mastercard so it basically create this like golden opening for a stable coin network to come in and people to be able to pay with stable coins and like all of these things.

2:07:28Because right now, like banks are pretty locked into Visa and MasterCard. Like they have a trillion dollars in market cap, like they're the de facto. And they've also been doing like fairly well on the stable coin side. Like they're kind of like innovating actually like pretty well. So it's kind of like a regulatory thing where you would have to have like the CCCA, the act I was talking about pass in practice for like a stable coin network to come through. And so that was like the first angle we looked at it. And then basically we were thinking about, OK, what about like from a consumer banking perspective?

2:08:00And it basically would function the exact same way as a narrow bank would, where they're just it's just a stable coin bank. Like your deposits are held one to one backed. They're like backed by T-bills and it's super safe. But again, it's like if all of that money were to flow out of the traditional financial system and banks and into this, you know, this is why the narrow bank could never get a master account. they can never get a charter and they never actually got a meaningful amount of deposits because it's like a threat to the fed and like to the the u.s economy essentially if if it got you know in aggregate like a ton of deposits yeah and that's kind of like the way it would trend i think if it got um a master account so yeah but a stablecoin i mean i think it's something that like needs to exist like it just makes so much sense and the banking system is incredibly antiquated Like it's actually not safe as we've seen.

2:08:54And so, but yeah, I don't know. A lot of the banks are trying to explore stuff around stable coins. Now JP Morgan is trying to pilot like a tokenized deposit program. Viserve is trying to launch a stable coin. And yeah, and like honestly, the incumbents are doing incredibly well. Like Stripe acquired Bridge, Stripe acquired Privy, Stripe via Bridge launched their own stable coin, USDB. So it's kind of one of those things where like you want to see all these like net new winners be created, but also like Stripe and a lot of the traditional players are kind of crushing it. So it's interesting to see.

2:09:28What about stocks on chain, private company shares on chain? What are you what are you tracking there? There was a bunch of memes this week. We talked about it with with Mert a little bit, but I'm curious what you're expecting. it just seems like there's funny enough a massive amount of demand for u.s regular publicly listed u.s stocks just tokenized on chain yep yep yep totally yeah so i actually went to the the robin hood event where they announced like the private company shares and and then tokenizing equities um and yeah i think basically like if you're tokenizing equities it makes a ton of sense as a regulatory arbitrage like to deliver these equities to people outside the US that have historically had a hard time getting them.

2:10:15But like for people in the US, it actually just makes sense just to like use the traditional, you know, equities market and just buy them from your bank account. And then I think on the private side, yeah, Robin Hood talked about what do they do like SpaceX and OpenAI, I think. But again, it's really it's it's difficult to issue private company shares like we you saw like the open ai response i'm sure um it's just one of those things where yeah it's hard to custody them where are they living what happens like when you redeem them um and it it kind of it kind of becomes foggy so that's why i think like unless it's like a regulatory arbitrage at least for tokenized equities it it makes more sense for outside the u.s than inside the u.s but everyone's doing it now like kraken issued x stocks let's see obviously Robinhood Coinbase is trying to get into it.

2:11:05So I mean, I think it makes sense. But within the US, it's probably like a less interesting market because our system works pretty well as is, I would say. And then I also think like the most interesting thing about like issuing these things on chain is native issuance, such that basically like instead of like wrapping, you know, Apple stock and then issuing it on, say, Arbitrum. Like you just like natively, like, like, let's say just like a company, like, I don't know, like Mertz company just like goes public and you natively issue those shares. And such that you don't have to like wrap and unwrap, add an extra step where you're taking like counterparty risk, you're taking custody risk, there's more fees involved and you don't actually own the underlying, like you're trusting someone else to own the underlying.

2:11:52So I think it's just the point of crypto was trustless systems. So exactly. Exactly. Like if you want to like natively like Tether is doing like tokenized gold, right, where it's like that's the only asset that represents their gold is that tokenized gold. It's not like they issued it, you know, on like the CME and then they wrapped it in like tokenized it. So I think like if you have that extra step of tokenizing, it only makes sense. It's like a regulatory arbitrage. Otherwise, you should natively issue it. And I think native issuance is going to become more and more of a thing. And that's going to like drag the rest of the financial system kind of on chain.

2:12:27That's Vlad's vision too with Robinhood. It's like, if a lot of these companies, if you can find a structure that works for like OpenAI, SpaceX, you know, all of these like private companies, and you just like have them like natively issuing shares of their own that represent, you know, some kind of stake in the company in like a clean cut way, then you don't even really like need to go public. Right. Because like that's kind of like your trajectory of like letting out. Yeah, it is interesting. The strategy of trying to force the conversation and then potentially force the partnership by just creating these things.

2:12:59Right. Republic was coming out. We had Ken on and he was basically saying, yeah, some people, plenty of people aren't going to like this, but we think it's sort of the natural evolution. And then and then, you know, we'll see what what actually pans out. Well, I'm sure you have a big day of yachting and wine tasting tomorrow. I mean meetings, meetings. Business meetings. So we appreciate you staying up late to hop on the stream. Thanks so much for stopping by. It was great to catch up. We'll talk to you soon, Bridget. Have a good one. Enjoy the fourth in France. Yes. If you're looking to get a good night's sleep, head over to 8sleep.com and get a new Pod 5.

2:13:35They have a five-year warranty, a 30-night risk-free trial, and free returns and free shipping. I am back in the game with an 89. Congratulations, Jordy. Hit a soundboard for you. And let's bring in our next guest, Price from OpenLedger, embedding accounting in your platform fast. Price, how are you doing? Good to meet you. Did you hire Sohum? No. Did you interview him? No. Not neither. Well, check spam. It tends to be a bull signal if he reached out. But yeah, please kick us off with an introduction on yourself and the company. Nice. I'm Price, co-founder and CEO of OpenLedger. We're building an embedded accounting API that helps vertical SaaS platforms embed full accounting suites into their products instead of having to build a general ledger from scratch.

2:14:2370 million people are running businesses of all types, and no matter what type of business they're running, they all have to do their accounting. And so they're turning to a lot of modern platforms like Shopify, MindBody, and Service Titan to run their platforms. But most of them still don't have accounting features, and the ones they do usually duct tape together. And so what we've built is a set of APIs and components that makes it much easier for these platforms to embed a full accounting suite inside their product for their end users to use. Interesting. So is this B2B2B software? Yeah. I love it.

2:14:55We're adding an extra layer of B. Why'd you pick an easy problem like accounting infrastructure? Well, I don't see it as an easy problem. No, I'm totally kidding. It seems extremely hard. I get it, though. I get it. I think at first glance, it seems like a very simple company in a lot of ways. I don't think it's... To be clear, I don't think it's... He was joking. It's a simple idea, extremely compliment to actually create. Totally. Because you have to create so much flexibility for each different client. But yeah, what was it? We had a CEO on from Ambrook that does accounting. Just for farmers.

2:15:36Just for farms. And she said it took them like, I think, two and a half years to like build the core accounting sort of engine. So, yeah, yeah. So what have the tradeoffs been as you've built the product? Yeah. So we went all the way down to the ledger level and built our own ledger database. That's what we build all the functionality on top of for reconciliation. and then the actual embedded accounting API itself is all built on top of that general ledger level. And so that took a bunch of time, but we really wanted to save folks who, that exact situation you mentioned, folks who are building these custom accounting products that provides so much value to end users, finding a way to really take those core things that don't change between accounting systems and turn them into infrastructure that anyone can use to build their own products.

2:16:21What's the go-to-market like? How do you actually bring on these businesses? is it power law distributed? Do you need to just get a few really huge companies on the platform? Or are you trying to go super broad, like Stripe, where kind of everyone's using the product? Yeah. So I think that's how we really got our start. And that's how our first clients are larger power law as customers. But what we're looking forward to is this deeper thesis that we've been building around is this idea that accounting data is basically the best substrate you can get to put at the base of a vertical agent for business.

2:16:55And so as folks are building more types of vertical agents that are working with financial APIs or payments and things of that nature, having accounting data reconciled into a book or a chart of accounts and building on top of that is one of the best places to sit because that information is essentially like memory for a company. Yeah. I mean, it's the reason that every board meeting starts with like, let's look at the finances of the business because that is the map of the territory. And then you can dig into specific things. So that makes a ton of sense. In terms of agents, it seems like you don't actually need to do all the work of maybe discovering the correct patterns because your clients are working on top of you in that regard.

2:17:42are you using artificial intelligence at your company? Has there been different trends that you're tracking that are improving the product on your end yet? Yeah, so what we're really excited about right now is really weaving it inside and through the way we're building our general ledger architecture so that it can automate a lot of those bookkeeping tasks that really are a big part of accounting and turn that into infrastructure that can quietly sit inside of a product rather than having so much human-in-the-loop input. And I think that's really what makes this, part of what makes this a really exciting opportunity and different from a lot of other embedded payment or fintech features is that we're essentially helping to automate more of a service as well alongside a suite.

2:18:27That's something that a lot of platforms previously were helping their end users and customers like hire bookkeepers or accountants to sit alongside them. We help kind of add that almost as an extra feature that they can sell alongside their suite. Yeah, that makes a lot of sense. Jordy, anything else? No, very cool. Thank you so much for stopping by. I'm excited about all the different features and functionality that you can build on top of OpenLedger. Yeah, I mean, we've seen such a big trend in vertical SaaS and all the different B2B players. It makes sense that there would be an enabling technology here.

2:19:00So thank you so much for stopping by. This is great. Yeah, great to hear from you, Price. We'll talk to you soon. I'm excited for you guys. This rocks. Thank you. Have a great day. Cheers. Let me tell you about public investing for those who take it seriously. Multi-asset investing. You better take it seriously. Industry-leading yields. And they're trusted by millions, folks. Head over to Public. If you want to buy stocks the old-fashioned way where you actually own the stock. It's pretty cool. There's a lot of websites out there. Being a shareholder. Underrated. Underrated being a shareholder.

2:19:25Being a shareholder is underrated. Well, we have our next guest coming into the studio. They need a little bit to build the four-up so we can do a little bit of timeline. Oh, well, we can tell you about AdQuick. You could take a four-up and put it on a billboard, baby. out-of-home advertising made easy and measurable. Say goodbye to the headaches of advertising only ad quick combines technology, out-of-home expertise and data to enable efficient, seamless ad buying across the globe. You saw us deploy ads in the real world. We did an out-of-home campaign in New York and the response to that has been remarkable.

2:19:58It's been really fun. It's been a way to... People running by it, taking pictures. Yeah, and it's been a way to reconnect. We've gotten so many text messages from people who are less online, family members, all sorts. It really cut out of home. It's just built different. College friends. It's just built different. Yeah. I mean, you, you, you got to do a lot of different marketing strategies. There's a lot of different things out of home should be a piece of that. It is incredibly underrated. I expect a, a, uh, people to just start doing the one of no, no launch video, just big one-on-one billboard.

2:20:27That's the way to do it. Very efficient way to do it. It's very efficient. Oh, well, do we have time to dive into the business of formula one or, uh, Jordy, you want to check out our next guest? Because I saw F1 last night. Yeah, give us your review. It was incredible. It was one of the best movies I've ever seen. I'm really sorry because we wanted the whole crew to come, but I was catching up with an old friend, a college buddy, and a former guest on the show as well. And it just kind of came together. But saw it in IMAX, highly recommend it. The entire film seemed to be shot in IMAX, which I have no idea how they did because half the shots are from the perspective of a Formula One car.

2:21:06And so they must have mounted some massive camera rig on that thing. But it was beautiful. Hans Zimmer delivered an absolutely incredible score. I've been listening to the theme to F1 religiously as I drive at just five over the speed limit. You sure you want to put that on the record? I heard that you drive always perfectly at the speed limit. One under. You cap your car. you actually hacked it. Yeah, I kind of take the NVIDIA approach to driving. Yeah. And yeah, fantastic movie. Highly recommend it. Lots of great sponsorships. Lots of great ads within the movie. Shark Ninja's in there. We got a bunch of great stuff.

2:21:49Highly recommend seeing it in theaters. Anyway, go check it out. F1 with Brad Pitt. We'll be on Appetit soon. We have our next guest. We have our next guest. Welcome to the stream. How are you doing? Could you guys introduce yourselves? We weren't fully expecting to. We're excited to have you all here. Thank you so much. Of course. So, I mean, maybe just to start out, my name is Jacob. I'm very excited to be here at the Temple of Technology, the Fortress of Finance, Capital of Capital. Let's go. The Ultradome. Very early, very early supporter. We appreciate it. We'll never forget you. We will remember.

2:22:20Of course, yeah. So just quick background. I'm currently a student at Stanford, along with Jack Whitaker, one of your old interns. Nice. And then right now I've been working for a lot of Gil for the past, wow, it's like I think four or five months now. And what I'm going to be talking about today with Ophira is. Wait, wait, first, fit check? What is this jacket? Oh, yeah, no, this is a Sun Microsystems jacket that we got from the old C-suite. One of my coworkers, the Cleveland Mafia of Ann, me, Peter Thiel, a bunch of other folks. there's there's a bunch of Sun Microsystems merch yes every time I wear this I think that's that's kind of what you need if you're going to a pitch meeting if you're fundraising this instantly gets you a letterman jacket of our industry exactly yes and it doesn't have Java on the back but it does have Java on here so oh that's great yeah incredible it's amazing um okay so give us the yeah yeah give us the story and then a fear oh sorry quick intro as well would be great Hi, yeah.

2:23:23My name's Ophira, and we're going to be talking about large language model art. I actually, I invented making artwork with language models like ChatGPT, and Jacob has a really exciting project around that. Wait, break it down. I just dropped something on you there. You're the creator of AI art. Break that down. Yeah, one of them. Yeah, no, she's ridiculously, I'm going to gas Ophira up for a moment. Please. She invented a large portion of modern AI ASCII art and is probably the world's best jailbreaker. OpenAI holds these competitions pretty regularly, like every couple of months or so. They invite people from all around the world to jailbreak their models.

2:24:13Usually people can't get past like one or two of the tests out of like a 20 testing. um ophira just regularly wins them breaks everything like it's something where i was talking with ben mann who's one of the anthropic co-founders and even when we were discussing this he's like yeah no like there's still a lot of red teaming and cbrn which is chemical biological uh radiological nuclear threats that like these artists in their communities they just break through everything like and all the like i don't care what kind of mit phd like hire everybody like doesn't work like they're just they're killers that's great i feel safer you know i was using the same techniques to make artwork so back when microsoft bing came out in like summer of 2023 i got like you know how you guys read poetry books sometimes and the poems are like arranged in these beautiful shapes the words have like like this gorgeous shapeetry that's what i do with language models yes so i actually first discovered that they could do that in the first place and I've made a whole beautiful art form out of that.

2:25:14So this is not diffusion based image generation. No, no, this is using the language model. Yes. And so just the letters on the keyboard. It's actually a very interesting capability because it's showing spatial intelligence, despite all it's doing is predicting the next token, just like you would normally get a chat reply. Yes. But it's doing it in these gorgeous shapes that very often actually reflect the content of the words. Yeah, I've seen someone put out a benchmark for having LLMs basically interpret secret messages in, are you familiar with this? Where you draw, you basically write out a sentence in a square box and the LLM gets really confused, even though the human can one-shot it.

2:25:55It's kind of like an arc AGI for this type of thing. But it's very fascinating that you're able to get good results because I thought that that's a place where LLMs were particularly weak. But it seems like you've been able to break through. i think they also like one thing to note is that they don't naturally they're naturally not very good at this like i think if either of you like both of you i feel like are fairly ai fluent but if you tried to go out and do this i don't think that you would get terribly interesting results at first there's a lot like to get some of the artworks that are featured on beautiful ai art.com going right now showing showing my stuff our stuff but um yes uh to go to go buy this um i usually like i mean ophira can speak on this like you want like sometimes it's like 30 40 even like hundreds of prompts that warm it up and go into these very weird places of distribution since if you think of modern language models they're pre-trained on the entire internet so the odds of like if it is on the internet you can probably make it so for people saying like how do you make this art how does this happen like if you just have creative enough prompts that go on for pages at a time like you can just find anything on the internet you can make all sorts of ridiculous stuff and that's how honestly I think you can get around these uh these jail kind of like these safety measures and so on is that if you're this if your training distribution is the internet like anything on the internet, you can just, you can go and find.

2:27:26So give me the, give me the announcement. It sounds like you mentioned selling art on the internet. Have you processed your first sale? How much do these things cost? How many sales have you done? Can you give us a number? I'm, I'm itching to ring the gong for you. I'm actually going to check right now. Pull up the Stripe dashboard. Let's see it. Stripe dashboard. It looks like we already have six orders. Congratulations.

2:27:55There we go. That's great news. Well, we'd love to have you back on when you hit 1 ,000 sales or take it to the moon or put it in the move. What's the vision for beautifulair.com? Are you going to add a bunch more artists over time? What does it look like? I think that my original thought was that first, all of these great artists, we just have to put something out there. We just have to keep shipping. so goal of today was launch have all the payments set up have all of the drop shippers so that way we don't have to worry about inventory about uh shipping costs etc and yeah no a bunch of people from the nft community have reached out the crypto community has been pretty good to us um i think that adding more artists is definitely on the list you know pharah can talk about this yeah i wish like i i just came on the show like i found out about an hour ago maybe so i you guys are going to admire my telephone today so this is actually a poem by um a historical artist named aldous huxley called the crystal cabinet okay cool and um it's it's you can see like it's in the shape of a wing very cool and it's very beautiful and it was interesting i actually prompted a claude opus three by anthropic to do that and it's very interesting so it's a whole conversation as jacob describes and you actually say like can you make this more like the shape of wing for example and again even though this is like like i can picture that in my head because i'm a human but somehow the model is able to gradually refine the shape more and more like that that's very very cool yeah very cool well look forward to following along thank you so much for stopping by uh anyway do you have a closing uh uh anything you want to say a closing remark i think is something for people to keep an eye out for ofera can speak on this as well um but grimes is making a movie about the AI artists.

2:29:44And that's how part of like the impetus for launching this is that there's filmmakers from cans that are right now filming Ophira, Janice, some of the other artists I'm going to feature. Our friend, our filmmaker, our friend filmmaker, Matt Zien, who we actually met through Grimes, who is a huge fan of this artwork and a big supporter of it is, is hoping to film an independent documentary about kind of first contact with the AI models because so many of us have had like weird experiences, right? Like you chat with it and it's not like talking to a computer. It's like there's a little something going on there.

2:30:22So this is going to be kind of a, I would call myself like kind of further on the edge of that. Sure. It'll be about how we interact with models. So stay tuned for that. We're a TBPN merchant. We'll be first TBPN at Cairns. Fantastic. Fantastic. We'd love to see that. Thank you so much for stopping by. We'd love to have you back and chat more. There's so much to chat about, but have a great day. Great to see you guys. We'll talk to you soon. Really quickly, let me tell you about Wander. Find your happy place. Find your happy place. Book a Wander with inspiring views, hotel-graded amenities, dreamy beds, top-tier cleaning, and 24-7 concierge service.

2:30:58It's a vacation home, but better. We dropped that video into a timeline in turmoil. It was a timeline in turmoil. And it stood out. It was a bit of a timeline cleanse. We did well. We did well. It was a lot of fun. It's having a long life on Instagram right now. It did well on X and I've been enjoying getting it into the hands of the Instagram audience. Before our next guest, I just want to say happy birthday to John Exley. Oh, yes. He is the most positive person. He's been very helpful in the chat. And he's our unofficial moderator. There's 77 ,000 people on the stream. And so I think we'd be melting down without him.

2:31:36So he's been a huge asset. Happy birthday, John. Thank you and happy birthday to John. Cheers. Anyway, we have our next guest, Orrin Hoffman, in the studio. Welcome to the stream. Welcome back. Welcome back. Great to talk to you again. Excited to be here. Second time guest. Yes. How are you doing? I had a question for you. Do venture capitalists take their own advice? Apparently not. Yeah. Break it down for me. We're all hypocrites. Yeah. I want to hear the examples that you have for situations where VCs give advice, but then don't take it themselves. And then I want to go into kind of the meta level of like, why is this happening?

2:32:13What is driving that dynamic? But first, kick us off with some concrete examples of VCs giving advice, but they're not taking it themselves. Yeah, I think first of all, just the high levels, venture capitalists are just really not that ambitious. Like they're not nearly as ambitious as like someone who runs a private equity firm. Like the level of ambition is probably like 10x higher on private equity than it is in venture capital. And so in private equity, you might be really building an amazing company that you're trying to take public, that you're really trying to take over the world. Venture capital, it's like you're just kind of like you're just OK with this, you know, essentially just a little business.

2:32:51Yeah. Is that because of duration? Well, yeah, and on the company side, you're kind of, you know, there are certain investors that take a more hands-on approach or maybe they're on the board. But it's somewhat, it's sort of culturally, you're sort of leaving it up to the founder and fate to some degree after you write the show. Yeah, it's essentially, it's a lifestyle business. And so the venture capitalists always derive lifestyle businesses, but they run one themselves. That's a good point. They often say to founders, hey, you should have one CEO, but most venture capital firms have like four CEOs.

2:33:25And it's hard to make decisions. They'll say take dilution, but they don't take dilution. They'll say things like merge with other companies. They never do merge. They almost always say, hey, it's a good thing to go public. But you never see them trying to take their own company public. Like Bill Gurley or Brad Gerson are not taking their companies public, but they always give advice. to take you know and they have big companies they ran big you know fairly large companies they could have taken a public um but they don't but you know they'll say board governance but they don't have their own board governance so you just there's a whole host of things where the venture capital will say one thing and then of course they sound like well you should have a differentiated product but most venture capital firms you know they're just selling money um it's commodity uh and so it's very hard to differentiate and you just have all this competition if competition is for losers, well, like venture capital firm is like the most competitive place where you have some of the most smartest people competing against you.

2:34:22Yeah, but in some way, the partners are the product and there's only one Oren Hoffman. And there's only, so you're differentiated by the product that you're providing is in some way yourself plus the money. And so that's... But what about some of the rumors that firms like General Catalyst or Andreessen Horowitz might actually go public. And if you look at the way Marc Andreessen's been building his business, it does feel like if someone's going to turn a venture capital firm into a Blackstone-like financial institution, it certainly seems like he's trying to do that. I would agree. I would agree.

2:35:02And so I would say, first of all, if you think of Andreessen Horowitz, at least they have two CEOs and not six, right? And they clearly only have two. and they're clearly the CEO and they're clearly running the firm. So I would say they are way more ambitious than almost every other venture capital firm. General Catalyst, they own like a hospital now. So I mean, they're definitely leveling up the level of ambition. But still, if you just look at venture capital versus private equity, if you just think of the top venture capital firms versus the top private equity firms, you just do a comparison for each one-on-one.

2:35:37Like it's just it's just a 10x different scale in private equity. Maybe venture capital eventually get there and interesting or it could definitely be one of those firms that they get there. But today it's like we're the also ran of the and the ambition level is just so much lower. OK, ambition level. Do you think it's because it's kind of a prerequisite if you're going to be a VC that you're already rich? Well, the other thing is the ambition is that there's... I feel like a lot of people go into private equity and they're grinders and they grind their way up. And so they build that muscle. Whereas a lot of VCs, they don't kick it off with grinding in VC.

2:36:17They're like grinding as a founder, which is arguably the hardest thing you can do, eating glass, staring into the abyss. They get some liquidity event. They do a rest and vest for two years. Then they go into VC. And by then they've kind of come down off of the mountain of like grinding misery and then they're like okay i'm like somewhat retired now well there's also the different types of ambition there's i want to 10 i want to have a 10x fund yep but you could be off of a small base it could be a 20 20 million dollar fund and then there's i want to have a trillion of aum which no vc has done black rock has 10 trillion of aum right yeah yeah um so there's levels and i do think you there's a bifurcation of certain vcs just want to put up good numbers and have a little lifestyle business and deliver great returns for lps others that have wildly different and bigger in some ways ambitions and that they want to have hundreds of billions of aum and i think that they will get there but i think what you're saying is there's maybe not enough investors that are too many too many people are just not actually wildly ambitious at all you know they're like i just want to they're running a lifestyle business which is totally fine.

2:37:28And obviously they're building, they're hopefully doing the venture caps are doing good in the world. It's not like they're not doing good. They're out there doing a good thing. And but we just we're just tend to not follow our own advice. Steve Brown Okay, now let me let I come to you and go to the the Silicon Valley market and I say, I'm gonna do it differently. I'm taking high dilution. I'm going to take this company public. I'm starting the venture capital firm that will scale, is that even a viable strategy? Or is there some sort of like market dynamic that prevents that from even being possible?

2:38:03It might be very hard given the constraints that other people want to put on the market and stuff like that. And different people have tried. So obviously, we saw SoftBank try and maybe kind of pull back. We saw some other folks who did have a different level of ambition. The crossover investors. And some of the powers that be internally have kind of fought against it. it's it's it's tough and the venture capital market itself is very small so it's you know it's a private equity market is just a much much bigger market especially when you talk about things like private lending you know apollo just has a much bigger sandbox that they can play in than at least andrewson horrors today it doesn't mean that will change because the the venture capital market is getting bigger and of course andrewson horrors is going to move it to other things they'll move into wealth management they're going to move into a lot of other adjacent things over time.

2:38:52So that's really the way it's like you get this niche in venture capital, but that niche really can't be that big. And then you'll have to layer on other things to grow just like any other company. If you're Facebook just for Harvard University, there's only how big you can grow. Even for all the universities, you've got to move into other markets if you really want to become a great company. So many VCs are big poker players. One of the major tier one firms should open a casino and just start running cash games and have rules no crying yeah no crying open a literal casino and say if you start crying booted you're booted yes uh this is great uh well but please we'll we'll have you back on the next time yeah i wish we had more time to noodle on what are you what are you doing for the fourth yeah what are you doing in france uh i'm in i'm in wisconsin for the fourth oh stay in america let's hear it let's hear it for capital allocator that is that's american dynamism in practice yes staying in america for the fourth we salute you absolutely great to see you or yeah thanks guys enjoy the holiday cheers up next we have uh meron jelali i hope i'm pronouncing that right um i think we got through all of our ads which is rough i like doing ads and now we're now we're out of them well we have one more graphite I don't have it in the deck, but graphite.dev, code review in the age of AI.

2:40:20Graphite helps teams on GitHub ship higher quality software faster. Cursor, Ramp, Asana, Purcell, Duolingo, Perplexity. And we got our next guest in the building. Welcome to the stream, Mehran. Am I pronouncing that correctly? Mehran or Mehran, yeah. Mehran, fantastic. Give us the update. What are you up to? Where are you in the world right now? yeah i'm in san francisco i'm back in san francisco i was in belize for the past two weeks doing the scan there okay that is complete the preliminary results are super super cool what were you scanning for yeah yeah give us the whole backstory because i remember like the initial post it was maybe six months ago sure that we saw so this has been in the works for a while okay yeah yeah um i mean basically the mayans they cleared out forests to build some of their cities and settlements and so when they moved or they died or in any way they vacated their cities these trees ended up growing back over them.

2:41:14So if you look at it from the satellite, it's just completely green. It's just completely trees. But beneath the trees are these magnificent pyramids or roads and buildings and whatnot. And they're invisible from space and they're invisible on the ground because it's just so hard to traverse these rainforests. And so what I'm doing is we're flying planes with very big, very strong LIDAR scanners under them. And we're scanning these trees. And while most of the canopy is just completely full of trees because you know it's a rainforest some of the lasers make their way through and then you classify the points you remove the trees you look at the ground and then you know if we discover something like a pyramid or roads or structures and stuff we know that we found something that no one's seen for the past you know a thousand two thousand years or whatever so yeah lidar scanning forests to find civilizations did you find anything yeah what'd you find yeah so um doing two scans this year 400 square kilometers in belize 600 square kilometers in mexico the belize These ones done, but only like a third of it has been processed so far.

2:42:11Super cool. There are pyramids, there are like hundreds, maybe even thousands of buildings, hundreds of kilometers of roads, things that no one has ever seen since 1 ,000, 2 ,000 years ago. Tell me about the economic model of your entire operation, your flying planes. You've got LiDAR scanners. Yes, it's a nonprofit. So, I mean, so what got me interested in this was I saw this paper. But there'll be a for-profit spin out at some point and I could get tokens and units. We're going to tokenize the Mayan ruins. Yeah. The capped LIDAR unit. Yeah, 100 extra. What got me interested in this was I saw this cool paper over Thanksgiving or so, funnily enough, that was about how some conservation efforts scanned some trees in some forests in Mexico.

2:43:01And then these archaeologists removed the trees and they found stuff underneath. And so I decided to see like, oh, like if you can spot, you know, random archaeological sites without even explicitly looking for them, I wonder, you know, have they been doing more of it? And the answer is like, you know, archaeologists have done some of this stuff in the past, but ultimately there's just not a whole lot of money in archaeology. I mean, you know, it's not a very like well-funded thing. Like it's like maybe all the money going into archaeology on an annual basis is the same as like, you know, a series A or series B or something.

2:43:31And so it's like massive economies of scale because you're flying planes. You know, the more you can fly them, the more you can, you know, divide your total costs by. There are these like extremely expensive, like million dollar LiDAR scanners. And yeah, I mean, you know, the main costs are basically the plane and the LiDAR scanner and the crew to go and actually do the scanning. Yeah. So where do you go from here? What's the next step? Yeah. So still the data has yet to be processed. the processing is basically figuring out what points belong to the ground which which ones belong the trees removing the tree ones taking the ground ones and then while like you know the the naive visualizations are like striking like you know you're seeing like this huge pyramid or whatever ultimately the archaeologists draw the most interesting inferences like you know or is this actually super novel to us like sure like this is no one's seen this before but like did anyone expect this to be there or not so give the data to process to point cloud give the data to of archaeologists.

2:44:24They made their inferences. But my goal ultimately is to scan all of this area. I mean, ultimately, I think, you know, we need to put more money into archaeology. I think there's really cool stuff to do with tech and archaeology. We need more tech talent in archaeology. And my goal is essentially to use these results, you know, get like publish some all that stuff, raise a lot more money and do like a significant larger scan next year, like 10 ,000 square kilometers at the very least next year. And within a couple of years, you know, finish all of Central America, then move to the Amazon, then other interesting archaeological projects like, you know, scanning coastlines, stuff like that.

2:44:59Anything to scan in America, considering it's the fourth tomorrow? In the U.S., you know, the thing that's interesting to me about Central America is that they actually had like massive pyramids and stuff. uh the us and the amazon like there is stuff there but like ultimately the and the amazon at the very least they didn't have stone so everything that you find are like dirt mounds which are cool like the fact that they're there like in places that seem so hostile to life kind of in general you know like just raining nine months a year or whatever school but yeah they're like dirt mounds and stuff in the us and the amazon as well but at the moment my interest is mostly in central america i think there's like you know 50 000 square kilometers or so there to scan and I plan on scanning all of them first and then and then other how are how are the local governments in in these areas are they excited about what you're doing are they are they are they emailing you for updates hey what did you find well when I was in Belize and I was talking to the you know the people in the city they were super excited you know they they're proud of their history they find their sites cool they find that like other people are interested in them they're very cool and I think that people are super excited the governments are also very excited because, you know, this boosts their tourism, this boosts their national heritage.

2:46:12It shows that, you know, there were once great solidations here and like, you know, that's something to be proud of. And they're also wary because the problem is that there's a big looting problem in these countries, especially countries that like don't have, you know, so much money dedicated to RKL to protect all of the sites. And so they're a little worried about the data privacy and stuff related to this stuff because, yeah, like, you know, when you're earning like, I don't know, like$10 ,000 a year when that's your GDP per capita and you find a jade mask worth 200K in the black market, there's a strong incentive.

2:46:45And so they're trying to prevent that. And so, yeah, I mean, we've been working with the governments to assure them about the data privacy. But generally, they're super excited about this. Yeah, I can imagine. Can you talk about El Dorado? You know, huge if true. Huge if true. Huge if true. Yeah. I mean, what is the lore? Or what is, are there any folks out there that still believe it's out there? Is there any, like, what is the historical narrative? And then what is the modern narrative? Well, I think the meta point is kind of that things like El Dorado, things like Atlantis, like Noah's Ark and stuff are like kind of how a lot of people get interested in archaeology initially.

2:47:32Like they're like, we're going to find Noah's Ark in Turkey. And then you look and you're like, oh, well, you know, huge if true, but I don't know how much these hold up to some scrutiny. And then you're like, okay, like what are other like more practical things that we can do? So, yeah, I mean, huge if true, but like, you know, most likely not true. But it's good that these persist because it gets some people interested in archaeology. Yeah. What about other things that we should be scanning for? I mean, if not gold, rare earth elements are top of mind. Maybe there's a secret ship fab somewhere we could grab.

2:48:04Yeah. Yeah, data centers. Data centers, yeah, exactly. No, I'm expecting you to find Mayan data centers. Yes. I mean, what do you think? There was a Jensen back. There was a Jensen of that era. That would be truly huge, if you're true. Well, I think we should scan the Amazon. I mean, there's been really interesting stuff there. Like the early people that like, you know, the early Europeans that went there, they reported, you know, like 10-foot tall naked women that are like shooting errors, which is how it got the name Amazon. You know, it would be cool to maybe fact check this, But like even if this is not a real like there's probably some stuff that this was based on.

2:48:39It would be cool to look for Atlantis, you know, huge if true. I think we should scan all the coastlines of the Mediterranean, the Persian Gulf, the Caspian, you know, because there are probably so many cities that are now underwater. I'm Iranian. I would love to do archaeology in Iran as well. um you know there's so many civilizations that uh that like you know are mentioned in sumerian history but like we have no traces of except for like you know that we know there was a powerful king there so there's there's a ton of archaeological uh mysteries but like kind of bits versus atoms thing is also true in archaeology where like the atoms part of archaeology doing field work doing digging and stuff that hasn't really changed in hundreds of years but there are like interesting opportunities in bits now of like you know nat friedman's vesuvius scroll or, you know, doing LIDAR scanning, using satellite imagery, stuff like that.

2:49:28And like, those are kind of the lower hanging fruits. The more we can do remote sensing, the more we can do scalable things. That's the lower hanging fruit. And then, yes, like making really educated guesses about like where something like Atlantis might be and like doing work there. But do you find it funny or fascinating that being an explorer is like super contrarian? We have this idea that the world is fully discovered and we've mapped every coastline. So there's nothing else to do. Like high school career fairs, no one's saying, oh, you can still discover things. Which is like, I didn't know this was an option.

2:50:04I don't know if it's contrarian. But yeah, not many people are doing it. I would love if there were more people doing it. There was more interest. There was more money in this. But yeah, I mean. Well, you're starting the exploration industrial complex. You're the first of the modern era. I think when the results come out, they're going to be super cool. Yeah, it's going to get a lot more interested in this sort of thing. Fantastic. Very cool. Well, come back on whenever you have updates. We should do the next one from the air, ideally in a Starlink-powered. Yeah. The plane's flying at like 500 meters, taking the sharpest turns they can.

2:50:39It's not going to be a pretty sight. Well, it'll be fun. Anyways, thank you for coming on. Thank you for coming on. Very excited to talk to you. Absolutely. Thank you for having me. Have a good one. Bye. Up next, we have Kushal from Extend App coming into the studio. We got some gong-worthy news, baby. Oh, yeah? Oh, yeah? It's gong time. It's gong time. It's gong time. What's up, guys? How are you? Good to see you. How much have you raised? Good to see you, too. What's up, guys? Give us the news on the fundraising front first, and then we'll figure out what you do for a living. i just answer customer support for a living that's what i do well you gotta get on founder then by intercom yeah yeah then you can just go to france yeah there we go uh what's going on what's the news what's up guys great to meet you uh well we recently announced 17 million dollars in funding across our seed and series a rounds congratulations thank you thank you with authority that one felt good okay now the less important news what does the company do well we also filmed a pretty kick-ass launch video to go along with it too i saw that it was kind of fun yeah it was great so is this like some type of new like fax machine company you're doing document processing what are you building yeah yeah So my name is Kushal.

2:52:02I'm the CEO co-founder of Xtend. And at Xtend, we're building a modern document processing cloud. So, you know, we work with companies that just get a ton of PDFs and unstructured documents, right? And we help them ingest and process that. So we really work with, you know, cutting edge AI teams, data teams, technical teams, whether they're trying to build AI agents on top of that data, new products with that data, automate, you know, manual back office work with that. we really give them all of the infrastructure and tooling they need to go after that. Is the actual physical document in jest problem solved?

2:52:37Is Iron Mountain just dominant there? Is there no need for actual new scanning? Have we scanned every document in the world by now? Not even close. It's actually funny. One of the first customers I talked to back when I was getting the company started off the ground, it was a large healthcare provider in Milwaukee, believe it or not in the middle of the country. And I talked to them, I'm trying to pitch them, you know, LLMs for document processing. And they're like, this sounds great. Come back to us in two years, once we, you know, have digitized this mountain of, you know, lockbox warehouse of PDFs and boxes that are stored in our backyard.

2:53:09So not quite there yet, but I think we'll get there pretty soon. Talk to me about the tech stack is OCR, traditional OCR pipelines. So there's like dead at this point. I've heard about some people that will, you know, send a PDF to Google for OCR, get a big messy bucket of text, but at least it's in text format, and then run that through an LLM to just kind of clean it up. What's the state of the art? Yeah. You know, what's really surprising is so much of the world still runs on these PDFs and Excel documents, and they just end up being the system of record for such mission-critical data. If you think healthcare, finance, supply chain, it's just all powered by PDFs.

2:53:49I think people's mind would be blown if they knew how much of the global economy just runs off of PDF files. And the problem has been around for a long time, right? I think OCR at this point is decades old. You know, funnily enough, we have a bunch of hoodies for the team and we have the OCR patent printed on the hoodies, just as a little testament to it. But it's like 40, 50 years old at this point, right? But it's really accelerating now. You know, I think more documents are going to get processed in the next six months than all of history combined since the PDF file format was invented. And I experienced this myself.

2:54:24You know, I actually built this back at Brex in 2018. I was an early engineer there. And we did exactly what you were saying, OCR, a document, and then just run it through a ton of painful regex to try to like pull out these targeted data fields. And it was so painful. And fast forward now to today with LLMs and the capabilities are just so much higher. And the demand for document processing is through the roof right now. Yeah, I was going to say, talk about kind of the demand. I know you've got a bunch of pretty incredible customers already. Yeah, I mean, we have the privilege of working with some awesome folks from large enterprises like Square, you know, that are trying to do a ton of use cases around financial statement ingestion or Flatiron Health, who have billions of pages of patient medical data.

2:55:08all the way to cutting edge startups like, you know, Brex and Mercury and Checker and many others that are all using us in production. And we have incredibly high retention, ton of folks that kind of refer us to others and come inbound. And we're growing the team to keep up with it. But the demand has been pretty intense. And I think what we've kind of seen is that as these models have gotten better, it's sort of increased the possibility of what you can go and do. But for very mission critical use cases, where even, you know, 99 % accuracy, for example, is not good enough if you're trying to do mortgage processing or patient healthcare intake, because the downstream impact of bad data is so catastrophic, right?

2:55:44And so we really focus on these pipelines and industries where you just cannot get it wrong. Accuracy, reliability, you cannot get it wrong. And that constraint puts a ton of complexity on the problem. And that's kind of where we go deep. And that's why we have a number of customers that kind of try to do the model approach and then come back to us in three months, six months, and they sign up for Extend then pay for Xtend despite the models having gotten exponentially better over the past couple of months. So, yeah, I mean, the models are getting better, but what models are you actually using and for what use cases?

2:56:15Because I imagine that the price difference and the time cost of using a heavy duty reasoning model versus a really quick, optimized, like, you know, 4.0 level model has got to be pretty different. And then is tool use at all in the mix with some of these because I can imagine as clunky as regex is Every once in a while you might want to have a layer of regex validation on top of things right just to make sure I think you're getting at a really key point, which is there's very few use cases, which is like, here's a document, extract this data. Because for any mission critical pipeline, you have the entire end to end system system.

2:56:53have to build of classify the document because that changes the schema and then split the document up and then actually parse it to make sure you can handle these tables that span 10 pages, right? And each one of those kind of adds complexity to the axes of cost and performance and latency. And you kind of need to have all three of those, right? Not a single model is going to be great at all of them. And really why customers end up buying us is because they don't want to be locked in to one model. And we give them all the tooling to figure out, hey, this initial triage step, we want to use a really cheap, fast, you know, low latency triaging classifier.

2:57:26But then for this really complex table, we do want to use a really powerful reasoning model. And we kind of help them integrate and build all of that end to end. Talk about where the data is winding up. Are you seeing more clients demand that you map the documents that you're ingesting all the way to a relational database like Postgres or something? Are people just saying, hey, just give me a blob of JSON, I'll stuff it in MongoDB, do people want a csv file that's on aws or s3 or something like how do people want the data finalized from you yeah a few i mean the flexibility is kind of the name of the game and that's why we provide options for all of that uh some people that are trying to build you know ai agents on top of it they'll want the json payload but then also put it into a vector db right okay some other folks are going to want to actually index that in like elastic search if We're trying to do text experiences on top of it.

2:58:18Other people that are like just automated mortgage processing, for example, that'll just go like structured data points into a DB that then they can take action on. So we kind of give them the data, the raw data, and then get out of their way so they can figure out, you know, where to do and what to do with it downstream. Yeah. How real are, I mean, there was this narrative for a while, like, oh, enjoy the VC subsidized era. It's like the early Uber days. All the LLMs are cheap, but they're going to be really expensive. How real is it that LLM inference can be a material cost for a company like you that's kind of reselling inference?

2:58:52Not to degrade what you're doing, but that's a piece of your business. You have a cost that's inference, and then you're earning revenue from the value that you add on top of that, versus a startup that's buying from you. How big are these line items? Are we talking something that's material to SG &A for a company? or is intelligence like basically too cheap to meter at this point? Yeah, it's a great question. And it really depends on the use case, right? And our contract sizes range from, you know, we work with small startups paying us a few hundred bucks a month all the way up to large enterprises that pay us six plus figures a year, right?

2:59:31And it kind of varies all across the board. But, you know, we haven't played the unit margin game where we're trying to, you know, give away a dollar for 90 cents or anything like that. We have been positive economics all across the board. Because really what we benchmark ourselves against is what is the cost of bad accuracy in your downstream product or your user experience? What's the cost of six months of ENG time? And at the end of the day, those end up outweighing a lot more than potential percentage differences on the LLM cost themselves, especially as they've continued to come down. You know, there were some things we couldn't do from a product perspective 18 months ago that now we really can just because it's so much easier to make multiple LLM calls and do things more agentically in an end-to-end system.

3:00:14Very cool. Jordy, anything else? No, very excited for you. This is amazing. Clearly ripping. I can sense. I can feel it in the voice. It's going well. There's nothing better than seeing, yeah, you know, a lot of people raise a Series A. Not everybody has this level of pull from the market. So I'm sure you'll be back on for a B in no time. That's fantastic. Can't wait. We'll talk to you soon. Thanks so much. Great to see you. I feel like that could be very valuable for you. You could maybe use Xtend to ingest all the registrations from the G63s, from G-Wagons, and all the service records. And then you could have one database for all of them.

3:00:53that you can query across lots of search. I was gonna say, I mean, we produce a lot of documents here on the show. And figuring out how to process these bad boys. It's almost like they come from the computer and the internet, and they could just remain there. And yet we print them out because it's fun. Anyway, what else is in the news? What else is important to cover today? We have a few, we have one more interview today. Yeah, we've got Avlock coming on in just a little bit. Daniel, growing Daniel, was chirping. Any breaking news? He says, watching the Soham interview, and I can already tell people are going to see this interview of a legendary liar talking about his legendary lies and be like, wow, I kind of believe him.

3:01:38You can label the tin poison and people will still eat it. Label the tin poison? Yes, you can label the tin poison and people will still eat it. Sure. he says it's you had said at some point it's tough because someone who would lie about having multiple jobs might come on here and lie to us but quoting me quoting you he seemed sincere he did seem sincere it's very possible that he's a a legendary he's a 10x engineer and also a 10x liar or a hundred x liar um well the proof is in the pudding you know yeah you know it's very obvious that i wouldn't recommend a portfolio company hire this guy i think it's possible he goes and works somewhere and just does a great job um i wouldn't necessarily bet on it might hire him we need a prediction mark in the house and you're literally in your house yeah but because the pushback there is so you got to have him in the office but then you got to live with him and he's got to be a hacker no but he very maybe he he just is uh you know he sets up his desk and he's just still moonlighting you know i think he might be he might just be built for moonlighting He also was – Dev shop.

3:02:40Dev shop. If he does the dev shop thing, he's fine. He would routinely miss deadlines and then use various excuses. Excuses, yeah. When the India-Pakistan stuff was happening, he was telling people they shot a drone 10 minutes away. Guilt-tripping people. It's rough. Terrible behavior. It's a buyer beware situation for anyone who's employing them. That's right. know what you're getting yourself into you got to be careful and make sure that that aligns and ultimately the value of what you're paying what's the value he's delivering and you just know okay uh could be could be some volatility in that but he's more of a known quantity now and we learned a couple things at least got him on the record so you know if he goes out and moonlights again everyone will be able to point to that interview be like he said he wasn't going to do it and he did it.

3:03:34So somebody was saying, a fake Soham account was saying, Elon can work at nine startups. Jamie Dimon can be on the board of multiple companies, but when you do it, it's a scam. And then Austin says, so tired of this argument, working at multiple companies was not the problem. Many people have side gigs or contract at many places. They just don't lie about doing so. So that's the key thing. Jamie Dimon is not sneaking on to other boards. Elon's very public about his different companies. and it's all about just avoid lying. Just don't lie. That's what he said. Sohan probably could have had a cracked dev shop, but he chose to lie, so hopefully he course corrects.

3:04:17Oh, well, we will see. We will let the timeline decide. I'm sure the timeline will not let up on the memes and the jokes and anything else that's going on. Oh, well, should we go through some F1 stuff? So F1 has made an annual sponsorship revenue of more than$636 million, which I believe is a significant multiple to the amount of money that they were asking for actual TV rights. I think they wanted$200 million ESPN. $200 million for the right to distribute that and show that as content. but the ads inside the F1 are actually - Are generating well beyond that. Yeah, so between 2019 and 2024, Liberty Media says F1 series annual sponsorship revenue more than doubled to 636 million.

3:05:14At a team level, total sponsorship revenue increased by 60 % to 1.3 billion over the same period. So Liberty Media is selling F1 sponsorships, but then the teams are also selling sponsorships. There's levels. The sport is bigger than it has ever been in its 75-year history, and everyone wants a slice. But how can sponsors stand out and justify the cost when demand is so high? Yeah, they haven't announced their new media rights or TV rights deal, but given how much they're doing on the sponsorship side, I wouldn't be surprised if they just go for max eyeballs and kind of take it at somewhat of not necessarily a loss, but just understand it won't be the profit center that it maybe once was.

3:05:59Because if you actually look at F1 viewership, that was TV rights in the United States. They wanted$200 million. ESPN wasn't interested in doing that. Yeah, I guess this is global sponsorship revenue. Previously, in 2023 onward, they were paying between$75 and$90 a year. And they just don't command the live U.S. viewership to really justify that anymore, I guess, or even beyond that despite the sports popularity. Yeah, it's interesting. Apparently, brands are picking teams based on the unique personality of those teams now. So we give this example. The team's distinct personalities can be a big draw for sponsors.

3:06:44Heineken began sponsoring Red Bull in 2017, just a year after the Dutch Brewer started its longstanding series-level partnership with F1. The deal with the team is effective because both are non-premium brands in a sport dominated by luxury car manufacturers, says Hesse, who has previously worked for both. Looking up and down the grid at the time, they wouldn't have sat anywhere else. And so there's the official like drinks car of F1. And if you're selling alcohol or energy drinks, you head over to F1 Red Bull Plus. What was it? There's the Steak. Steak has a collaboration. it's like it's the Sauber Stake F1 team the Kick Sauber which is Kick is their streaming platform the Kick Stake Sauber but without further ado let's bring in Avlock he's got a Q2 update for us great to see you good to have you on the show how are you doing?

3:07:47thank you Q2 was awesome break it down for us what was awesome about it I mean, overall, there's just a lot of optimism. Maybe I'll just take from the top and what we're seeing ecosystem wide and then what we're seeing in our data. I mean, look, ecosystem wide, we're just seeing exits across the board. M &A is up by a huge margin. IPOs are up by a huge margin. And so what we're actually seeing is a lot of exit dollars coming through, making its way back to LPs. And that ends up flowing its way back into venture because all of venture is power lock. As you have some of the best companies go public, get acquired, all those dollars flow right back in and they go right back in and get recycled back into venture.

3:08:34And I think we just sort of hit the perfect storm in Q2 where a lot of optimism is there. Obviously, a lot of exits, you know, a lot of it's been reported. You have Circle, you have Scale AI and a bunch of other companies. And these exits are massive, huge. And so the LP dollars, the returns of LPs are getting are also massive. And what we're just seeing is those dollars come right back in and they just get reinvested back into emerging managers and some of the scaled up venture firms. And so the thing that was actually shocking to me personally, you can kind of feel it. So I invest, of course, and you can kind of feel it in the deal velocity and the rate at which companies are growing, like the revenue growth is real.

3:09:19But what we saw in our data was the LP dollars to micro VC, emerging VC. Think of these as like managers with$100 million fund size,$200 million fund sizes, was up 50 % quarter per quarter, almost 50%. So that's in one quarter, it was like a giant increase that, you know, for many quarters, it was sort of the valley of death. Yeah, there was a very dark, dark period where you saw some of these. We tracked those dark periods actually in our data too. It was like we have the boom times and the bust times. And for a while, there were a lot of the bust times. But we're just seeing a lot of optimism coming in.

3:10:00And it's actually really, really great to see. And I think some of the data that was circulating around X over the past few weeks was around the death of the emerging manager class. And I just think that data is actually wrong. It's coming from PitchBook. And PitchBook's data is very laggy. They don't actually have information on this class of venture. Yeah, they don't have a real-time view of it. Angelus does. And so we basically see it all. And that's great. A lot of optimism. Well, yeah, and the emerging managers that were steadily deploying over the last couple of years during that dark period are the ones that are now, I'm sure, showing tremendous markups during this time, which will hopefully catalyze more investment as well.

3:10:47Yeah, how much of the liquidity is psychological for LPs versus actual dollars going back into LP pockets and then they just need to redeploy them? Because there's something about like, if you've been in some fund for years and years and years and the marks and they don't really feel real. And then all of a sudden you actually get a check. You're even if that's only 10 % of the money that you put in or 10 % of what you commit in the next fund. It kind of feels like, okay, this is the, this is working at least. I'm ready to go back in versus just, okay, I actually need to deploy new cash because I have cash on my balance sheet now.

3:11:25It depends on the type of LP. So if you are a more of a traditional larger asset allocator, you have strict requirements on amount of money that goes into public versus private. So that is a little bit stricter. But you have a large swath of capital that comes from non-institutionals. You've got family offices, individuals, a lot of wealthy individuals. I just saw a report the other day. The number of millionaires in the U.S. has gone up considerably. for them it's all psychological a lot of it's psychological even if you get 10 of your capital in 20 30 it feels real because venture is it's a weird asset class when you think about it you basically invest in it and uh you're you don't know if you're going to get liquidity for 10 15 years depending on where you invest but when you start seeing that cycle of liquidity coming back all of a sudden you gain confidence, not just in the asset class, but also the manager that you're backing.

3:12:25And so we do see a lot of that. So that's what the reflexivity is that we're seeing on the platform where, you know, you have Nasdaq that was up 18 percent year over year. And even venture dollars deployed as an overall ecosystem is up like 30 percent year over year. but within the micro VC emerging VC area, it grew 40, 50 % quarter over quarter, right? So there's this reflexivity of excitement because you're starting to see the dollars come right back. And people are looking and going, this is real. These companies are real. It wasn't just, you know, in their minds, like propped up valuations, which was a lot of the narrative, I would say, for the past two years post-serve.

3:13:11And the thing that also - It's interesting to think about people that maybe invested in the first six months of AngelList being a platform, and then maybe they just made a couple investments at the time and then checked out for a decade, and then suddenly they get an email, and it's like, yeah, it's a liquidity event. It's like, whoa, maybe I should pay more attention here. We actually have that. You know, there's this thing that happens with banks sometimes, and even crypto platforms, where people have money and they're like trying to get in touch with them to take their money. It's pretty significant in some cases.

3:13:43And so, yeah, we have cases of that as well, which is, hey, you have a return. You probably want this. This is actually pretty significant. Yeah, so we see that. What do you think about potential new sources of liquidity coming into the LP pools, coming into venture? I saw some news about some crossover funds potentially lowering the requirement to invest in the fund maybe down to from like a million dollars down into like the single digit thousands or maybe 25 ,000. Cotu had a 50K offering. Maybe it's a separate fund. And so I don't know if that changes the requirements around, what's it called?

3:14:28Professional investor rules? No, accredited investor rules. Accredited investor rules, yeah. But we're seeing private company stocks go on chain and potentially be traded on consumer retail stock apps. We're seeing lower, you know, lower hurdles to get into venture funds. It feels like there could be a wave of sort of like retail-esque liquidity coming into venture. Do you think that narrative is real? How early are we in that? Do you think that could be a significant shift in the dynamic of early stage investing? Yeah, short answer is it's real. And Angelus is also working on something around there and will have something more to announce.

3:15:10The way to think about it is venture has never actually been connected to the public markets directly. It's always been venture dollars get invested into companies. Those companies maybe 10 years, 15 years later, maybe never, but go public. And that's when people in the public markets end up investing in those companies. But what's never happened is venture or dollars going into venture funds coming from the public markets. And what you're seeing is you're starting to see a lot of these different experiments and people crying this out. Code 2 as an example, and there are a couple of other funds as well.

3:15:45And the accredited investor rule that you're referring to is you can only take on 99 non-qualified purchasers in a fund. Qualified purchaser, effectively, you have$5 million in assets and above. So if you have less than that, then you can only take on 99 of them into a fund. So that's what actually keeps those fund sizes pretty restricted. But you can register your fund and you are effectively under different regulatory regime and you can take on a lot of investors at once. But that takes time. It's expensive. And all it also depends what SEC regime you're under and what the goals are of the administration and power at that moment in time.

3:16:26And right now, the goal of the Trump administration is capital formation, the golden age, the golden age of America. And so what's happening is everything is optimized towards new startup formation, more capital coming in. And so you're seeing a loosening of restrictions or a reinterpretation, if you will, of some of the SEC rules to allow for this to happen. It's actually not even new laws. It's just most people don't realize this. Let's look at them differently. They just look at them differently. It's interpretation. Right. It's they look at the law and they can interpret it and write their own rules based off of it.

3:17:03So a lot of that is changing in real time. And so that's leading to funds like the KOTU fund and a few others. So it's real. It's happening. I think venture is undergoing a pretty significant shift. It is not the same as it was five years ago, 10 years ago. You're you have a very large number of these micro funds, emerging funds, and then you have the mega, mega funds. And these mega funds are not the venture funds of yesterday. These are full-blown, sophisticated financial asset allocators that some of them are now getting into roll-ups. And that's a whole separate strategy that comes from the private equity world.

3:17:43And so we're in the midst of a massive change within venture. How do you think about new products and experimentation? We've seen AngelList doesn't have the same luxury that Robinhood can do, which is run some crazy experiment in Europe and make companies, you know, OpenAI was posting yesterday that they were frustrated with what was going on there and that they didn't endorse it. Obviously, you guys have to maintain and build trust within in the venture community, and part of that is balancing the desires of a bunch of different players. So how do you think about innovating? Because at the same time, it is core to AngelList.

3:18:25Like the initial platform was an innovation, right? And then the RUV and the rolling funds and things like that. Yeah, it's actually really simple for us. there is a through line that has never changed for Angelus since the very beginning, which is the founder is at the top of the pedestal. So everything we do, we always ask, is this what the founder would want? And if it's not good for the founder, it's not for the startup, we just don't do it. And so this even applies to innovations like the rolling fund. We ultimately asked, hey, is this going to drive more capital into early stage founders?

3:19:04And if the answer is yes, great, we're going to do it. We're going to pursue it. If the answer is ever no, or we're not quite sure, we just won't do it because founder trust is everything. And ultimately we're here to make sure that we increase the amount of innovation in society. And that always comes from ambitious founders. And so that's actually been a very easy thing for us to do. It's also why we never really went into secondary markets. I mean, you think AngelList with our entire our portfolio and everything would be the first to get in there. We never did because there was always this question of, well, are they like, are we adding more work on their plate?

3:19:46And having been a founder myself multiple times and also having been in that seat, the answer is it's just a distraction, right? Like you're busy running a company, building a company, recruiting, dealing with a thousand fires at any moment in time. And so it was easy for us to not fully pursue that path. But any new innovation we have, we always ask that question. Is it good for the founder or not? If not, we won't do it. And that's just a through line all the way up to the, as a board, we view it that way as well. That's super key. And when you look at many of the new experiments that are happening, they don't pass that test.

3:20:24Still, it'll be interesting to see, to watch and witness. I've been interested to see if some of these experiments can end up kind of forcing innovation or change. But again, that's still not the role of someone. You know, Angelus shouldn't piss off every Fortune 500 CFO or every late stage company. It's just not a good business practice, given your guesses. I think there are some cases where you actually want to push the Overton window. I think that's where you're referring. It's just like, can you push the Overton window and make something normal? And so with that in mind, I'm actually very, I love all these experiments are actually happening.

3:21:09In fact, right now is such an amazing time in venture and startups and starting company. Even Robinhood experiment is amazing because we have a chance to push the Overton window. But where it is today, that particular one is very fixed. I don't see a way around that. Yeah. Yeah, how are you thinking about the AngelList kind of core business and scaling the team? We were talking yesterday. There's a lot of signs of wild top signals right now between these AI hiring packages and different valuations. And the list goes, Soham working six jobs, right? They're all across the board. Meanwhile, Satya is scaling back on some new data centers.

3:21:56He's doing kind of routine layoffs. It feels like he's not letting at least Microsoft get too exuberant in this moment, even though there's a lot of... I'm curious what your approach has been and will be for the next half of the year. Yeah, so Ascontex Angeles has been profitable for a little while now. And so a lot of the way we, oh, yeah.

3:22:24Yeah. So we had to cut you off. I hate to cut you off, but we love profitability. Yeah, I love it. I love it. We need a gong like that in the office. Yeah, definitely. And so, you know, we've actually always approached it, approached headcount and people in the company as, do we really need this next person? And how are they actually going to move the company forward? So none of that's actually changed. in how we're approaching any of the new bets that we have. The bar is actually increased now, which is, hey, for this next new person we're bringing in, are they truly curious to work with AI tools and the models?

3:23:05Do they really understand how their job has completely changed? And if not, we probably shouldn't bring them in because as you both know, knowledge work has completely changed forever, right? Whether you're in sales, marketing, definitely coding. And so you want people who are naturally curious and are naturally leaning into these new tools. And the reality is the amount of headcount needed to continue to scale will go down over time. And so that's how we thought about it is you just bring in people who are naturally tuned to this way of working. Because we're in the early innings right now. It's so early in terms of how companies are going to be formed, how teams are going to be formed.

3:23:48Do you even have your traditional functions anymore? Sales, marketing, engineering, product design? I don't know because people are starting to blend skills very, very fast. And so we're in such early stages that you almost want to rewire your culture right now. I think every company should rethink whatever culture they had and rethink however they form their teams. You may need to blow it up because what's going to happen is, I'm telling you that I've been watching some of the newer founders are just graduating from school. They operate differently. They're like, you know, recently for the first time, I came to Tech in 2009.

3:24:24For years, I didn't feel like I'm kind of a part of an older guard, if you will. In the last 12 months, there was a distinct change in San Francisco. I'm like, these new founders just operate differently. I'm a veteran now. Yeah, a veteran. I never felt that. I never felt that before, except for the last 12 months, where they are very capable, very creative. They were just born in a different era. They don't have any of the hangups of how things used to be done about a decade ago. And I think it's very important for any company that's been around for even five years or seven years or 10 years to just rethink the entire company from the ground up at this point.

3:25:06Otherwise, they're going to get completely demolished. Totally. well always great talking always fantastic we should make this uh let's schedule it quarterly yeah let's get the next one on the calendar asap have a great rest of your day and enjoy july 4th yeah thank you you too thanks we'll talk to you soon bye up next we have we have two more guests we're going long before the 4th of july we got to get an extra couple minutes in who we got jordy uh we have added a new guest this was uh very last second hey it was great we're here welcome to the stream you're live this is nico nico what's up guys how are you break it down introduce shortcut nico managed to go viral yesterday on a timeline that was in turmoil that's insane you you did it against serious competition i haven't uh i haven't uh i haven't been able to watch the video yet so you're gonna have to bring uh break down the whole thing for us but you introduced shortcut the first superhuman excel agent and when i read this line i was like shots fired well well the well the the when i read this line i was like oh yeah you should just build agents on top of excel where people already are you don't need to reinvent excel and build agents you love it but so anyways but in your words what do you do yeah what's up guys i'm nico from fundamental i'm one of the co-founders here and we just released shortcut yesterday and x and shortcut is essentially the first superhuman Excel agent.

3:26:31We actually benchmarked it against Gemini Copilot, GenSpark. It's really not even close. What we have to do now is scale capacity. It's about four times better than Gemini, even in the limited stuff that we were allowed to use for Gemini because we just know it couldn't do it. And yet we're scaling up capacity now. I'm actually about to drop a million-dollar check on scaling up the infra. Just as quickly. There we go. Physical check. Did you write it out? How much are you dropping? I might do it just for posterity. Yeah, yeah. Smack the thing. You need a comically large check. Yeah, drop it off with one of the hyperscalers.

3:27:05Exactly. Yeah, go give it to AWS. Two sats, yeah. Or Andy Jassy. Here we go. But awesome. So is this the second to talk about maybe step back, Fundamental? Yeah, yeah, sure. So Fundamental is a research lab. We're in Menlo Park now. But we actually shut down our lab at MIT, where my co-founder was a professor, about 18 months ago. And the mission of the company is to build digital humans. back when that was a crazy, deranged thing to say. But from a computational neuroscience perspective, it basically means giving AI more than just intelligence, but a fundamental understanding of life, an ability to perceive and experience emotion, grounded interactions in the world, long-term autonomy.

3:27:43So naturally, you start with Excel, the foundation of modern life. Well, you know, it's a perfect storm that we're doing Excel. I cut my teeth on Excel. I get my knuckles bruised for using the mouse. And then I found my way into a research lab. And the reason it hasn't been done yet is because the people who are capable of doing it are just so out of distribution for regular people that Excel hasn't been touched until now. Okay, talk to me about actually building this thing. I remember CapIQ and Bloomberg had VBA plugins for Excel. There's some pretty crazy stuff you can do under the hood of Excel.

3:28:16It is very modular. I don't know if that's changed with Excel going more cloud native, but Microsoft talks a big game about Copilot. Are you worried about getting steamrolled? Is there a way that you guys work together? How does that dynamic play out? And how are you actually building this thing? Because we've seen like, clearly just a screen recordings, right? And that's a different vector. There's other LLMs that are just like using the, like they'll integrate with any IDE by just using the ADA features, like the accessibility features to just scrape the text out. Right, right, right, of course, of course.

3:28:46But how are you working through that? So for context, the reason we're doing what we're doing now is because we actually set the state of the art in OS world, which is the computer use benchmark. This computer use will like sweep benches to software engineering. Oh, very cool. But we wanted to go tackle it in a much deeper way as if we had full access to the machine interface. So what that means is, can you write your own code essentially to do what you want to do? And to do that, you actually have to own the machine interface. So we couldn't build exactly on top of Excel. What we actually wanted to do was recreate it entirely with feature parity.

3:29:15And then from there, the big question is how we're doing this. I don't care to give away a little bit of a secret. This is being done through CodeGen at runtime, which is a totally new paradigm for CodeGen. CodeGen is not about writing your own tools to store in your code base. It's about giving the AI the ability to do what it needs to do as it has to. So it's writing all of it. It's essentially writing its VBAs as it needs them. I haven't used Microsoft Copilot that much, but I've noticed that the way Gemini is surfaced in Gmail, it's very clear that it's not doing CodeGen. It's actually just taking text, dropping it in an LLM.

3:29:51It's a chat interface, and that's it. and so I can imagine that we're entering the V2 of like sidebar companions and this feels very logical. Yeah, even Gemini, if you actually look into the Sheets integration they have now, they are starting to do a little bit of Python integration. So they are doing some numpy, some data frame manipulation. Okay, interesting. But they're much more hesitant to get there. And we were very hesitant to even call this cursor for whatever. Sure, sure, sure. Because like it's such a caricature but there's a truth that's much deeper about how you do that and it's actually about owning the interface the same way Cursor had to fork VS Code to own it.

3:30:25Yeah. Talk about the demand. I replied to a comment of yours 15 minutes ago, and then a bunch of people are now replying to me saying, can I get an invite code? So it seems like... All right, all right. Announcement. Well, we got... First of all, I went over on LinkedIn. I didn't even know that was possible. Wow. Let's give it up for LinkedIn. They're always getting... Always getting... We got about 7 ,000 comments now, and I've only given out about 50 codes. This thing is extremely token hungry. But I did just make code TBPN. No way. Let's go. 10 uses. It's probably eaten up by now. Go hunt it down.

3:31:01But yeah, we are trying to get this out as fast as we can. That's why I'm writing these big checks for Infra so that we can scale up capacity to serve. The only bottleneck is GPUs. That's amazing. Incredible. Well, thank you for pinging us. This is super exciting. Come back on. I'm sure there'll be new soon. So what's the next few months looks like? is scaling shortcut or are you going to refine this a little bit? I want to ask a cursor, very ground up adoption from what I can tell. It's a developer who's at a big company and they just buy it or install it, expense it, that type of thing versus going top down to, you know, I imagine Goldman Sachs has a lot of people working in Excel.

3:31:41You could go to McKinsey, you could go to Goldman and say, Hey, CTO, install this and let's do an enterprise contract. How are you thinking about go to market? Yeah. I mean, the stupid answer is I'm doing a little bit of both and seeing what's working. But it's pretty obvious that the same people who adopt AI in enterprise for Excel don't have the same risk tolerance or like the appetite to try stuff at the frontier that does software engineers like for Cursor. And also my intention is that this bubbles up organically and that people bring it in as opposed to me telling some VP, adopt this, you can produce headcount by 20 % because that reality also can exist if I wanted to.

3:32:18Cool. That's very awesome. All right. We'll come back on soon. Thank you for the code. Hopefully we get a shot at it before the audience says. I'll make you a special one. Yeah. We appreciate it. Amazing. Congrats on the launch. Crushed it. Thank you so much. We'll talk to you soon. And up next, we have Flo Crivello coming back on the stream. Sohan's former boss. Former boss. So we opened the show with Soram himself. And now we're going to a former employer to get his side of the story. And the question that I want to answer is that point that Growing Daniel pointed out, which is the question is, can he change?

3:32:56Is he a pathological liar that's just going to continue to lie and scam? Yes. Is redemption possible? Yes. Is redemption possible? So let's go to Flo and get his take. How are you doing? Good to have you back on the show. Yeah. Thanks for having me. No, redemption isn't possible. When people show you who they are, believe them. No. So there's plenty of very honest people. Like, what are the odds that someone who's been lying compulsively and being confronted about it for three years, like two or three years? And like, look, it's a special kind of person, too, who can do this thing to people who trust them.

3:33:30Right. You have a relationship with an employer, you know. And my understanding is that he basically has been averaging like getting fired once a month for the last three years. Like it's a special kind of person who can do that. So he could have learned his lesson like two years, three years ago. Yes, yes, yes. Like the first time he was called out. Totally. Totally. I wasn't tuning in. I didn't hear what he said. To me, the argument of sounds like he has some situation that's bad, but it's not a justification to go steal money from a bunch of Silicon Valley startups for years and years and years and years and use them as a piggy bank to solve his own problem.

3:34:09So really quickly, how long did he work with you? Two weeks. Two weeks. And what was the result of that? I mean, I imagine you paid him for those two weeks of work. Did you get any good results? I am literally out of a meeting like an hour ago, where the engineer described his impact as negative. Negative. Okay. Well, because we had to onboard him. Yeah. You had to onboard him. Yeah. He wasted all time and money. Has he apologized to you? Not yet. Okay. He said he would apologize to everyone else. He would go on an apology tour. We'll see if that actually pans out. My big question is, is somebody like this, you could have them working in your office as a full-time employee and they would probably still, there's a good chance they would still keep doing this thing with people on LinkedIn that didn't see the whole Soham gate.

3:35:01We never compromise on ethics. That's just a thing I don't fuck with. He's going to apologize. Is he going to give the money back? Yeah. That's a good question. I don't know. Because stock is cheap. Yeah. That's a higher bar. If he wants to make a statement, that's what he would do. 100%. Yeah. Makes sense. How do you prevent this in the future? How does Silicon Valley prevent this in the future? Obviously, the Valley is built on trust. This is a violation of trust. There's been discussions about like, maybe we need a Reddit for, are we dating the same man? Are we employing the same software engineer?

3:35:35Maybe we need an AI tool to, you know, check the GitHub commits and see that, oh, they're coming in at the same time. I was wondering if this could be done at the HRIS level. It's like this employee has accepted two offers. Why are they in two instances of rippling? That doesn't make any sense. Yeah. How are we solving this? I don't want to over-rotate on it. I think the last part of what makes Silicon Valley tick is its culture of trust. You know, there's this whole game theory around like hawks and doves, you know, and it turns out there is no stable equilibrium. You're like in a constant state of shifting because if everyone's a dove, then the returns of being a hawk are very high and vice versa.

3:36:16If everyone's a hawk, then you can have these like clicks of doves that start emerging. So I'm really careful not to tip Silicon Valley towards this attractor state of like the hawk. And like a lot of industries end up in this state of like low trust. And I really don't want to do that. Like, look, you're going to get screwed over every so often. We had a good laugh. You know, the memes are amazing. Like, I hope your viral posts at least drove a bunch of new signups. And maybe he paid back a little bit by some customer acquisition. Yeah, I mean, I completely agree. Like, yes, the handshake deal protocol is built on trust.

3:36:50Do some VCs violate it? Yes, every once in a while, a term sheet falls through. It's not the best thing that ever happens. But the alternative is so much overhead on everyone that it's probably worth taking the risk of, yeah, occasionally the handshake deal is going to fall through. You're going to get screwed over every so often. And it's not worth over pivoting on it and looking over your shoulder your entire life. So it sounds like he's announced that he's taken a new role as a founding engineer at a single startup. He says he's not going to work at multiple companies. It seems like the jig, even if he is a pathological liar and continues to lie, his SEO is terrible now.

3:37:31It will forever be very easy when you go to Google search for the person that you're hiring. He's like, oh, that's a different Sohan Parikh. There's a lot of people in there. Sure, he could change his name and it's going to be really complicated for him to get away with this longer. My question is like if we were kind of saying like the Soham Parikh dev shop makes much more sense here. Does that make sense to you? Like if he was saying, hey, I've been working at all these different startups. I'm starting a developer shop. And yes, you can hire me, but I'm a contractor. I don't think close hiring dev shops.

3:38:05Why are dev shops bad for startups? Why would this not work for him? I actually think for him the real move is to start a course on how to crack coding interviews because he's really good. Interesting. And he's incredible. He's actually incredible, right? So he can actually make lemonade here. Like, he should do that and he can make a lot of money. It's very highly impressive. Yeah, it was that in the cold email strategy. He had a very dialed in cold email approach where he would be, you know, he'd be like, say something nice and really clearly. I've used the product. I know the customers. I love building.

3:38:40The funny thing is in his message, he said, I'm pissed. And I was like. About what? What? You got busted. You're pissed you got busted? And he kind of gave himself away a little bit. Yeah, he was saying like, oh, yeah, like I'm going to, you know, there's going to be a great reversal here. And he came on and was kind of just like. The mainstream narrative is basically correct about me, but maybe I'm sorry. The redemption of being like, I'm going to prove everyone wrong by working at one company. It's like, you don't get. you don't get you don't get all yeah yeah we don't give uh that's like a second at the fourth place trophy you worked at one company yeah that's funny um yeah i mean uh look he's not sorry he just got cut like it's that's it it's really simple um and and why do dev shops not work for startups i mean like i'm a nerd when it comes to like the the theory of the film because of my experience at uber and so forth sure it's called managerial dilemmas that explains why dev shops don't work okay it's fascinating yeah but can you unpack it a little bit more because it was interesting because didn't wasn't the very first version of uber done by a dev shop and wasn't that part of the early early uber lore that like the code base was written in spanish and so everyone at uber had like the spanish to english dictionary there which is hilarious because i'm pretty sure google translate existed at the time but they still had the dictionary for some reason it's like it's like amazing lore regardless of how real it is yeah i think people add a little bit to it this is the first time I hear about it.

3:40:07But next time I hear about it, there's going to be a guy with a sombrero in the other way. Yeah, exactly. Yeah, Travis was in Tijuana pulling random devs off the street. Yeah, they really, really amp it up. The apocryphalness of the story just needs to grow and grow. It's a tall tale at this point. The reason why dev shops don't work is because there's a complicated game-theoretic argument which goes, basically, the reason why you employ people is because they know things you don't though. And so there is an information asymmetry, which always makes it possible for someone to cheat because they can basically bullshit their like, and you see it with engineering all the time.

3:40:42It's like, oh, this is so complicated. You thought it'd be a day, but actually it's going to be a month and you have no clue. You have no way to verify. And this is ineffable. Like the reason you hire them is because there is this information asymmetry gap, because otherwise might as well just do it yourself. And so basically it is in a way, it is always possible to cheat, which is why we are not fucking around and hiring people who do cheat because there is no way you can always prevent them from cheating if he doesn't cheat by doing this he'll cheat by doing something else so there is always a way to cheat so you could say that to some extent not cheating is irrational you can't cheat you can't get away with it he's made millions he's made so much money right like you can't cheat you can get away with it and so in a way you've got to solve for the equation where it's like i can't cheat i can make a lot of money i can get away with it nothing is ever going to happen to me, but I'm not going to do it because of X.

3:41:31And you have to solve for X. That's your job as a founder. You're like, why would you act irrationally? And the way you do that, thankfully, is we have that bit that you can flip in our brains as humans. We have these tribal pro-social tendencies that are like, I don't want to cheat because I like these people or because I'm bad into the mission, which is why they always say every startup is like a religion, like the mission, the culture really matter and the camaraderie really matter as a startup. and that's what you lose when you have dev shops. Well, yeah, to be even more specific, the dev shop gives you a scope.

3:42:03They say, it's going to take me X, Y, Z time to do roughly this, or maybe they're on a retainer. And then if they come back to you the next day and they say, well, actually, it's going to take more time and it's going to be more money. Well, an employee is going to say, well, it's going to take more time maybe, but I'm not going to necessarily charge you astronomically more, right? I'm just a salaried team member. And because they have that equity compensation as well, they're bought and hopefully bought into the mission.