TBPN Live - Samsung Invests $70B in AI Chips, The Cubanator Joins, Apple: Behind in AI, Ahead in Revenue | Mark Cuban, John Kim, Eugen Alpeza, Ari Herbert-Voss, Alex Konrad, Carl Eschenbach & Pat Grady, Jim Cantrell, Tom Hulme

Starting point is 00:00:00 You're watching TVPN. Today is Thursday, March 19, 2026. We are live from the TBPN Ultradel, the Temple of Technology, the Fortress of Finance, the Capital of Capital. Let me tell you about ramp.com. Time is money. Save both. Easy use corporate cards, bill of pay, accounting, and a whole lot more. Don't test me with the soundboard.

Starting point is 00:00:18 Don't go soundboard for a soundboard with me. You know I got to. That's a narrative violation. No, no. We're having some fun. We're out of control. We got a great show for you today, folks. Carl Eschenbach, is Eschen back at Sequoia? We'd love to see it. We had the pleasure of chatting with Carl a couple months ago. And I've always been a big fan of his, but we'll let him introduce himself. Let's pull up the linear lineup. A linear, of course, is the system for modern software development. 70% of enterprise workspaces on linear are using agents, and you should be too. We also have Mark Cuban coming on the show. And what a fantastic return to form for us, because the first time we had him on the show, we discussed.

Starting point is 00:01:00 And we can talk about Cuban in a second, but of course we have our Lambda Lightning round. And Alex Conrad from Upstarts Media is joining as well. Anyway, last time we had Mark Cuban on the show, we were debating ads in LLMs. And since then, we've gotten a bunch of data points about ads in LLMs. And I think that some of his takes have probably aged well. Some of our takes have probably aged well. It'll be an interesting time to reevaluate what's actually happening. There's been a lot more points.

Starting point is 00:01:29 I don't know, John. We just have more evidence. I said that ads would be fine. Well. And now the world is ending. Yes. Now we add, now we have ads in the L. It's not because of the ads, though.

Starting point is 00:01:40 It's not because of the ads. It is much more complicated than that. But here's a white pill. Samsung is investing $70 billion to advance their fab capacity. They're getting back in the AI chips game. They've always been in the AI chips game. So brief history of Samsung, you probably know them from the phones, from the TVs. They, of course, are a major player in HBM, high bandwidth memory.

Starting point is 00:02:09 They are a massive company, over a quarter million employees. They're close to touching a trillion dollars in USD market cap. They pull in around 200 billion USD a year in revenue, maybe 250 billion this year in revenue. Really good. All that's USD. When you look up Samsung, you get South Korean won. but I like to think in USD because I'm an American. And it's actually kind of complicated thinking in foreign currency.

Starting point is 00:02:37 They're the global leaders in memory and OLED displays as well. So a lot of the displays that you see in other electronics, even it has a different brand name. It's still Samsung actually making that OLED display. But they're second in smartphones to the iPhone and Apple. And they're second in the semiconductor foundry business to TSM. C, semiconductors still make up 30 to 40% of their business, and they supply HBM to NVIDIA for the H100 and Blackwell systems.

Starting point is 00:03:06 So it's not like they're sitting out the AI bull market. They are doing great. They are definitely participating. They're incredibly important in the AI buildout. But if TSM is bottlenecked and TSM is sort of risk off and they're not going to be, you know, guiding to like insane KAPX numbers while every American hyperscaler is, well, that creates an opportunity for Samsung. And so Samsung is stepping up and they're announcing that, hey, we're going to put another $70 billion to work on this particular business. So Tesla has

Starting point is 00:03:36 been working with Samsung on the founder's side in AI for a while. So Samsung's never really been on the frontier with a direct competitor to the H-100 or the Blackwell chip. That's been more of like AMD's game and AMD also fads at TSMC. So there hasn't really been this like neck a neck battle between TSM and Samsung. But it's like you can do AI inference on a Samsung chip, and we know that because Tesla went to Samsung years ago

Starting point is 00:04:03 and said, we need a chip that can take in pictures from the road, decide where the lines are, and decide they want their chips with the dip. They want their chips with the dip, and now Samsung does too. That's all you know. And so the FSD system, if you

Starting point is 00:04:19 have a Tesla, you might be familiar with like HW3, hardware 3, that has been deployed into millions of cars, and it was fabbed on Samsung's 14 nanometer process, which is a lagging node. We're not in the three nanometer, the crazy frontier stuff, but it's working and it's on the road. And according to a U.S. regulatory probe, there were 3.2 million vehicles, Tesla's, on the road in America with FSD systems that were basically all running Samsung chips inside. And so now, to be clear, Tesla, just like any foundation model lab company, they have training,

Starting point is 00:04:57 and then they also have inference. They're a little bit different than many of the labs that you know and love because they do training in a data center using what's called the dojo chip. And that is fab at TSM. But so they train the system. They take all the data in from every Tesla camera, every road, all the information that they have. Every time that there's a disengagement, that's feedback to the reinforcement learning system. It says, hey, we were in FSD mode, but then someone grabbed the wheel or someone stepped on the brakes. You made a mistake.

Starting point is 00:05:30 Understand what happened to get you to that point where you made that mistake. And so all that data gets collected in the Tesla data center, runs on these Dojo chips. They do the training, and then they deploy the model onto the Samsung chips in the actual cars. So if you're driving a Tesla, you have a Samsung chip in there that was trained, and the model was trained on TSM chips. And so the Dojo D1 is one example of their training chip. That was Fab to TSM on 7 nanometer. And it's completely separate from the in-car fSD chips. So with the backdrop of NVIDIA's massive GTC news cycle,

Starting point is 00:06:02 they've done so much press around GTC and so many different launches. You know that NVIDIA is just going to suck a lot of the air out of the semiconductor discussion this week. Out of the clean room. Out of the clean room, yes, which is recycling all of the air every three seconds or something like that. So Samsung dropped this update. It was pretty quiet. We were actually struggling to find it.

Starting point is 00:06:19 There was one Wall Street Journal article about it. But it has not SEOed well. Maybe they need to do some more podcasts or something. But they did, in fact, I mean, Jensen's doing a whole fleet of shows and interviews and all sorts of stuff. Podcaster says companies should podcast harder. Yes, yes, yes. Yes, the solution to everything is more podcasting. We're talking to my book here.

Starting point is 00:06:44 Yeah, but I think this is like particularly important, especially this morning. The, I guess the CCP put something out in the last 24 hours basically saying, hey, Taiwan is going to have an energy crisis due to the broader global energy crisis. So we need to reunify. Peacefully. There's an opportunity for peaceful reunification. But peaceful reunification, even if it was completely peaceful. And all the Taiwanese people just say, hey, we want to be part of China.

Starting point is 00:07:10 They all vote for it democratically. That's going to be rough for the American. American chip buying industry, if you're a buyer of chips that are fab there. And so having another chip on the board, metaphorically, to make physical chips is probably a good thing. You know, I was writing yesterday. I'm very excited about Samsung, very excited about Intel, very excited about all of the new fabs, the gigafab, the TerraFab is the one that Elon's talking about.

Starting point is 00:07:38 Launches in five days? Oh, wait, five. Did he actually say that? He said seven days. Seven days ago. Okay, so wait, five days like the plan launches? He just said TerraFab launches in seven days. Okay.

Starting point is 00:07:51 So I don't know what launches mean. And Tyler, what's the lead time for an ASML lithography machine? Yeah, I mean, it's at least like, you know, three, five years. Yeah, it depends on which like tools. So the TerraFab will be ready in five and the ASML machine. It's possible. It's alien technology. It's possible.

Starting point is 00:08:10 Elon's going up and back to space all the time. Maybe he got some of his own. An extreme ultraviolet lithography machine on the moon, on the back side, on the dark side of the moon. They just had them stacked up there in crates potentially. So, yeah, Samsung's been doing well over the last five days. Stocks up 11% during time when the NASDAQ is down 2.2% and geopolitical tensions continue to rise. The compute bottleneck, we know it's important. We've been discussing this constantly.

Starting point is 00:08:37 And it's going to be very constraining over the next few years. So every increase in CAP-X in the supply chain is a step in the right direction. And so Samsung gets the first gong hit of the... Great congratulations to everyone over at Samsung making a big bet. Who else is making big bets? Cursor is making big bets. Before we talk about cursor, let me tell you about Phantom Cash, fund your wallet without exchanges or middlemen, and spend with the Phantom card.

Starting point is 00:09:08 And let me also tell you about label box, RL environments, voice, robotics, e-vals, and expert human data. Labelbox is the data factory behind the world's leading AI teams. So, cursor is out with Composer 2. Composer 2. It is frontier level at coding, priced at 50 cents per million input tokens, and $2.5 per million output tokens. It's also, they have a fast version.

Starting point is 00:09:32 They say we're able to significantly improve the model quality and cost to serve. These quality improvements come from our first continued pre-training run, providing a far stronger base to scale our RL. So it performed quite well on, what is it, cursor bench. Yes. Which is a funny bench, but. Which is, well, yeah, TBPN. TBPN performs quite well on TBPN bench too.

Starting point is 00:09:58 Yes, yes. It's a little silly to design the bench and then publish the bench that your score on your own bench. But, I mean, to be fair, like, they're putting GPT 5.4 high and medium. above them. So it's like they didn't, it's not one of these graphs that's just like, oh, look, we made some arbitrary X and Y axes and like we're in the top right corner, of course, because the axes are like good and cool. On TVPN, TBPN Bench is like technology podcast, publish at least three hours of content every week. Yes, yes. Naturally, we are right at the top. And it's actually there's no one else on there.

Starting point is 00:10:38 Yes. But yeah, I mean, this seems fair. It is a little bit odd to read this because the cost, the cost is on the X axis and it's inverted. So the further you are to the right, the cheaper you are, which makes sense because people associate an X and Y graph with you want to be in the top right quadrant. And they certainly are. And it does seem like in terms of this Pareto Frontier, you want to be on the frontier. You want to be pushing out across every single curve. Maybe if you are interested in sparing no expense,

Starting point is 00:11:13 you'll go with the GPT 5.4 high or medium model, and you can align cursor to GPT. I'm sort of surprised that Opus is not doing as well on cursor bench. That feels surprising based on like the general vibes around Opus 4.6 generally, but cursor has specific needs for specific customers. And I don't know, what else do you think is going on here? Yeah, I mean, the cost is really big. Like, this is basically like 10x cheaper than an opus.

Starting point is 00:11:47 Yeah. So I think also, you know, cursor has kind of been, like not really a dark horse, like everyone knows about it. But in the coding race, it's like, everyone's like, oh, okay, there's code versus cloud code. But like, you know, if you imagine that, you know, cloud code and codex are kind of like these environments for getting a ton of like really good data.

Starting point is 00:12:05 for training coding models. Kurtzars had that for way, way longer than Open Air Anthropic. Yeah. So you should imagine that, like, at least, you know, in the near term, they actually have, like, really, really good data that they can, you know, train these good models on. And obviously, like, this is a very specific model. They've said it, like, you're not going to write poems with this model.

Starting point is 00:12:23 It's this very, like, specific kind of almost like point solution model where it's just coding. Don't listen to them, Tyler. Write a poem with the model. Poem bench. Poem bench. Yeah, I would be interested in my own. would be interested to know, like, how many sacrifices were made, because it's at a certain

Starting point is 00:12:40 point, like, I remember talking to an AI researcher, actually a semiconductor, uh, uh, uh, entrepreneur, who was saying that, like, he actually thinks, he actually does believe that importing, like, the Odyssey and like Homeric epics is key to humanoid robots learning to walk. Yeah, well, I think, If you look back at just like the general history of like machine learning AI, like the lesson is that like big general models always beat these small specific models. Yes. But if you kind of zoom in on the time scale, like you can still train, you know, GLM, some open source model on a very specific task like accounting or something. And you can like he'll climb and you can actually make it better than the frontier models right now. At that specific thing.

Starting point is 00:13:25 And especially at cost. Yeah. Yeah. Yeah. Yeah. Very much so. But like on the long term, if you zoom out what actually once here, it seems like it's basically all. always going to be these big, you know, general models.

Starting point is 00:13:36 And I wonder if that's true. I mean, we talk about this a lot where the big general model outperforms the smaller model. But at the limit, like, if you were to think about like a Python if statement, just like flow control that is truly deterministic, like, yes, if you piped the same question of like the if statement, like, is this number bigger than this number? You pipe that into 5.4. It's going to get it right all the time.

Starting point is 00:14:02 It's going to be very expensive compared to a. if statement, which takes like no, no compute whatsoever, right? But the if statement is 100% accurate, like, unless there's some bit flipped from cosmic radiation. Like, it is deterministic. And so if you're in a world where the small model that you've built, the classifier that you've built, whatever machine learning pipeline or small model you've built is actually functionally at 100%, well, then there's this upper bound that even like the bigger, smarter model doesn't get you any benefit at all, so you're just purely in cost control mode, I would imagine. Yeah, that's reasonable, but I think...

Starting point is 00:14:39 This is not a great example, because coding is more... Coding, we're not at 100% saturation on just coding. Yeah, and just literally one-time. I mean, look at the chart. Like, the best-performing models are sitting at 65%, 663, 64%. So there's clearly more room to saturate this particular bench. I'm actually super interested to know about what goes into Cursor Bench at this point, because I feel like when I see every benchmark, it's like 100% now.

Starting point is 00:15:07 But that's just from, you know, the old models. Legendary poster, S.Colp, says all S-H-I-T-S and giggles on that headline till Anthropic or Open A-I decide to cut off their access to cursor, referencing the Bloomberg article. Cursors are taking on Anthropic and Open A-I with a new AI coding model. Would that matter? Like at this point, if they have Composer 2, and it's a small model, but it's good at writing code,

Starting point is 00:15:34 and it performs well on cursor bench, and the cursor users are satisfied with the Composer 2 model, and they do, Curser does get their access cut off, and when you install cursor, you roll it out to your organization, you just get Composer 2, and you know what? It's, you know, maybe there's taste that wouldn't pull you towards the office. Yeah, we just say at this point right now, I don't think we have any visibility until, like,

Starting point is 00:15:58 how much of cursor's revenue right now is tied to, using opening I are anthropic models? Well, I think, like, in some ways all their cost is, but is all their revenue? It depends on the perception of the user base, because the revenue might be, well, I pay for cursor. I don't really care what they use under the hood. There's a lot of people, this was Ben Thompson's argument

Starting point is 00:16:21 for a long time, was that for a lot of people, they just show up to chat GPT, and they wouldn't care if the model was powered by Gemini, because they're just like, I just ask it a question, and I get an answer. And so if you're a cursor user, it's possible that you're in the same boat where you don't really mind what happens under the fold. What else is going here? George says I'm hearing tons of complaints from cursor customers at enterprise companies, a silent change, but almost all models cursor uses behind Max mode.

Starting point is 00:16:50 Devs, he used to manage to spread out monthly credits over a month, see all of it used up in one to two days. Oh, interesting. Are furious and switching. It does feel like there's a little bit of like an economic war here. Yeah, and this is what came up, you know, earlier this month around the lab sort of subsidizing. Yeah. So, you know who's... They're not in an easy position, but they're such a talented team.

Starting point is 00:17:14 Yeah. Well, you know who's great at Pareto Frontier Pushing Models? Gemini 3.1 Pro. It's here. And it has a more capable of baseline. It's great for super complex tasks like visualizing difficult concepts, synthesizing data into a single view, or bringing creative projects to life. And let me also tell you about graphite,

Starting point is 00:17:33 which is owned by Cursor Correct? Code review for the age of AI. Graphite helps teams on GitHub ship higher quality software faster. Nikita says, we're rolling out summaries for articles now. Just tap the summarized button if you want to know if it's worth your time to read it. Yes.

Starting point is 00:17:48 And yeah, it's basically grok. Turn this into a regular tweet. I am excited about the listen button. I've had this, you know, on my commute. There's so many moments where I'm like, I wish I could just have somebody read this article. I actually wound up doing this with a number of Will Menitis long form essays. I would copy them, put them into 11 reader from 11 labs and have it read it to me in sort of a...

Starting point is 00:18:13 A silly voice. A silly voice. It was a silly voice. Well, I was trying, I was actually trying to use GROC. I was trying to use GROC in the X app to just take an article pasted into GROC and say, hey, can you read this to me? It said, cannot find the post. It couldn't kind of actually get the context. It is sort of crazy reflecting on the fact that there is so much software out there that, you know, we talked about DoorDash. Like how can you completely reimagine the experience with agents on the platform?

Starting point is 00:18:44 And there's so many different things that you can do to sort of like re-architect what your product is if you've built a successful software company. But then there's all these like little things where like, yes, every product probably does need to, an LLM that can summarize text and expand it, and that's what GROC does. And also, everyone wants text to speech, and everyone wants speech to text. And everyone wants, okay, if there's an image, I want you to have a text version, so it's searchable. Like, I've complained about if I see a funny meme on X and then I want to go search for it later, impossible.

Starting point is 00:19:18 Like, there's just no way you could ever do it because that is stored in Twitter's database or X's database as like image number, 7, 6, 2, 5, 4, 2, and someone's comment was just like, this changes everything. And I'm never going to remember what they said. But now you can run every image through a image model to understand what's actually going on, and then potentially service that in search. Now, that's going to be a long project to actually implement that,

Starting point is 00:19:46 make it fast, make it cheap, make it affordable, make it like fit within the business model of X or whatever social platform is out there. And, I mean, you can see Instagram, like, struggling through these things right now as search becomes more important and as ranking becomes more important. Meadow's already seeing the lift from machine learning applied to add ranking and whatnot. But this is a response to, you know, every article people would post, people would always say, GROC, summarize this. And now there's just a button. I wonder if this button will be gated by X premium, because I recently learned that, you know,

Starting point is 00:20:23 you can only ask GROC, like at GROC, is this true? You can only do that if you're paying for X. And sort of underrated how well X has seemingly, I don't know how big the subscriber base is, but that was a crazy idea to have a paid social network. Dalton Caldwell from formerly YC partner actually launched a competitor to Twitter back maybe a decade ago that was paywalled only.

Starting point is 00:20:53 And the whole pitch was like better content, more substack model, no ads. And he never really got it, you know, to perfect product market fit. But it was an interesting idea. Yeah. And I think it's because people are deeply addicted to X. Yeah. It is very valuable to them to be on there to participate. Yeah.

Starting point is 00:21:17 And the paid functionality, the way that it was marketed and the way that it generally worked, was like you were going to have a bad time on X. Like if X was valuable to you and you didn't pay the $10 a month, it was going to be like significantly less valuable to you. Probably, you know, you might, depending on what kind of business you're running or what you use X for, it might be the equivalent of like losing thousands of dollars a month of value or you could just pay the $10.

Starting point is 00:21:42 So it was a good trade. Yeah. But it was also just, it was weird how the targeting never seemingly got dialed to the point where you could actually target the CEOs of companies. who are on X? Like, I mean, you see Travis Kalanick on X, like, replying to things.

Starting point is 00:21:57 It's like, he's raising money, he's growing a business. Like, there's a lot of value in advertising to him because he's going to be picking a corporate card soon, or he probably already has, or it might be in that market. He might be picking a payroll suite. Like, there's all these things where if you could deliver that to that audience, it would be incredibly valuable, and the CPM should be, like, through the roof.

Starting point is 00:22:18 But I think for privacy reasons and for variety of other reasons and sort of like really monetizing that long tail has been very difficult across every platform so they've just gone with scale and the products that have sold the most on social networks have been very broadly marketed. And the criticism that we saw from the Oscars

Starting point is 00:22:37 is always like YouTube ads are generic. It's just like for a pillow or like injury or like something that applies to every single person. But there's always this like hyper-targeted opportunity there. Yeah, the other thing is the paid program with X has seemingly worked in that we know a lot of people that happily pay and have no plans to churn, but it would be a failure in the context of like meta scale, right? I think the last reported number that I saw was something like they had like one to one and a half million paid subs at $10 a month.

Starting point is 00:23:14 Oh, on X? Yeah. So you're talking about somewhere in the range of 100 to 200 million of like ARR. And if Mark, if Zuck had launched a product like that, he would just wind it down, right? Yeah. Reels went from zero to 50 billion of run rate in like a handful of years, right? That's what a home run looks like. And so I think it makes sense for X, but it certainly is not a home run from a, you know, consumer application standpoint and they still need the, you know, the overall business. Yeah. Olivia Moore had some extra context there around monetization of, you know, via ads versus subscriptions. So Neil Patel,

Starting point is 00:23:56 who is the founder of NP Digital, a New York Times bestselling offer, shared. This is how ChatchipT ads are performing. He's like the SEO guy, I think. Okay. He said the data is only from five businesses, but these businesses also run Google and meta ads.

Starting point is 00:24:12 Compared to Meta, ChatGPT's rough lead quality is 256% higher. on the flip side, lead quality is 49% lower than Google. I mean, that seems like a miracle to be in between two hyperscalers on, on like day one, basically. But on the bright side, due to ad costs, it's substantially cheaper from a CPA perspective than meta. And this was sort of what we were talking to the good folks over at Ridge about, was that

Starting point is 00:24:39 at least in the early days, like being early to a new ad platform that you can potentially scale on, can drive a bunch of new conversions. But Olivia Moore said, a big story that most people are missing in the AI race for the consumer, Chet, GPT versus Claude, is ads. Right now, most consumer AI revenue is coming from power users who are willing to pay high subscription costs. This currently skews positive for products like Claude, but this will not be the end state. Google makes $460 per user per year in the United States mostly on ads. I didn't know that their ARPOO was so high.

Starting point is 00:25:13 Meta makes around 250. I mean, I guess those Google ads are really, really valuable. And it's so intent-driven that it makes sense. I would argue, or she would argue, that Chattip-T's ad-based Arpoo's will be even higher as they will ultimately have deeper, more frequent user engagement. Even at the $460 level, monetizing everyone in the U.S. via ads is $152 billion in annual revenue. By contrast, if you're able to monetize even 5% of the population at $200 a month subscription, which is a stretch, that's only $40 billion.

Starting point is 00:25:44 That's actually a crazy difference because $200 a month subscription is like. super high. Like, you know, you're talking 20 times like Netflix or something else that's, you know, premium and like really important. Yeah, the $200 subscription at the time was crazy. Yeah. But even at that point, some of the people that were more kind of just like AI pill generally were like, oh, it's actually possible that someday you could spend $20,000 a month. I was like, give me the $20,000 month. And it sort of came via API, but it was heavily subsidized. So she says, I suspect this will be even more drastic outside of the United States where user are even less willing to pay or directly pay for subscriptions.

Starting point is 00:26:22 And the earliest data from a very small rollout shows chat GP ads are already outperforming meta in effectiveness. This just gets better over time. So interesting. The question about Wilmanitis and the article summaries, should he move to substack? He's threatening Nikita. He says, I'm defecting to SB. STCK. He won't even type it out. He says they pay more. And Nikita didn't reply. I think people would

Starting point is 00:26:52 follow Will over there. I think people would read his articles anywhere potentially. But if you're looking to advertise, why don't you head over to app love it? Profitable advertising made easy with axon.a.a.c get access to over one billion daily active users and grow your business today. And let me also tell you about Gusto, the unified platform for payroll benefits and HR, built to evolve with modern small and medium-sized businesses. So, Carl says it's time to return to the place where I know I can have the most impact. I'm beyond excited to rejoining Sequoia as a partner. Here is what I shared with the team on how I am approaching my next chapter.

Starting point is 00:27:31 He wants to serve the ecosystem, fire to win. Let's see what this means. Being a servant leader does not mean I have lost my edge. In fact, the fire in my belly burns brighter than ever. The difference now is that I'm not using that fire to light my own path. I'm using it to light the spark and others. So their fire burns brighter. Leading from behind, I have no interest in the view from the front of the room.

Starting point is 00:27:52 I will leave that to our great leaders, Pat and Alfred. I want to lead from behind empowering each of you. Egoless impact, contagious energy, mentor and build great leaders, always ready to serve. Carl's the man. We will have him on in just 30 minutes. we can wait to cover more of that story. But first, we gotta go to his Allen & Co. Photo Shoot, dual wielding coffee, jump rope, and a faded Sequoia T-shirt.

Starting point is 00:28:23 Andrew Reed says, immediate overwhelming response to the VC pushup debacle. Welcome back to Sequoia, an all-time great partner. That's a great, that's a great photo. He's looking great there. Let me tell you about Plaid. Plaid powers the apps you use to Spence Bayor and Invest, securely connecting bank accounts to move money, fight fraud, and improve lending now with AI. This is an interesting story.

Starting point is 00:28:46 This is an interesting story. Apple is way behind an AI and still making a fortune from it. Let's see. Thanks the question. Are they actually behind? It's AI revenue is set to top $1 billion this year, reassuring investors wary of rivals, sky high spending. And keep in mind. They have a chart here showing gross revenue from Gen.A.I. Apps as well as Apple's commission. So look at this.

Starting point is 00:29:10 The beginning of 2025 was really the boom of Gen AI app growth. 400 million. Is this monthly app store revenue? Wow. They're really cooking. And then sort of a flatline. Yeah, it's so interesting that it actually dropped. Yeah.

Starting point is 00:29:31 Well, we did read that article a few days ago about how Apple has been pushing back against some of the vibe coding apps. And there's this question about, you know, where are the bounds? Obviously, Apple's had pretty strict App Store rules around adult content and, you know, what else you can do, even just the app reconstituting itself, pushing changes because they want to review every line of code that goes into the app store. If someone's pushing 10, 20, 30,000 lines of code a day, that's a lot of code for Apple to review. It's going to slow things down. So that could be a little bit of what we're seeing. Maybe they've capped out on their ability to review all the vibe-coded apps that are flooding the app store.

Starting point is 00:30:14 But let's go to the Wall Street Journal and dig into this. Apple's on pace to surpass one billion in AI revenue this year, a tidy sum that demonstrates the company's AI advantage, even as it struggles to deliver an AI strategy of its own. Its Siri chatbot is still weak by modern AI standards. What Apple does have that the other AI players don't is a dominant position making devices. However fancy open AI, Google, Anthropic, and XAI make their chatbots, iPhones are still a primary way to deliver them to customers. That means they typically pay the App Store tax roughly 30% of subscription fees in the first

Starting point is 00:30:45 year and 15% a year thereafter. The rates vary. Gen.A.I. Apps paid Apple nearly $900 million in App Store fees in 2025 with almost a billion of revenue and very, very, very little CAPEX. Three-fourths of the revenue Apple rakes in from Gen AAPs and its App Store come from chat chbt. Oh. Next at about 5%

Starting point is 00:31:08 is X-A-I's GROC. There we go, GROC. Cooking. I mean, there's so many different like funnels. They did the essay competition. They did the video competition. And I mean, I've talked to people

Starting point is 00:31:22 that are just still, they're like, you know, like people that are in the Apple ecosystem, they're like in the Tesla ecosystem. And so they're like, yeah, I talk to GROC on my way to work. I'm not kidding. Yeah, GROC in the iPhone app store

Starting point is 00:31:35 is that did 12 million last month? Yeah. And I know, I know like the true like AI heads will be like, GROC's behind on this benchmark or model or whatever. Tyler, is that a correct characterization? Yeah, GROC did more revenue. GROC did more revenue last month than it than Claude in the iPhone app store. But, but like I've started having conversations with, I mean, I'm using CHGPT,

Starting point is 00:31:58 but I wanted to just, I wanted to get up to speed on, on Taiwan and, the, the, just the, like, what was the reason for the original Civil War and stuff. And so I was just having a conversation back and forth. And at no point was I like, oh, it really needs to be like, you know, GPT 5.4 pro. It's like, these are things that exist just like with one search to Wikipedia or one search to any. It's probably baked into the weights of 3.5. But so, so, like, if I'm just going to be, like, chatting with someone who's, like, reasonably smart, I would say GROC is there.

Starting point is 00:32:37 And so what do you think? Yeah, but you could be talking to someone who's really, really good. No, not if you're asking basic, basic knowledge retrieval questions that any model is going to have one shot and just be 100%. Yeah, but you're just describing stuff

Starting point is 00:32:50 that you could just like actually Google. Yes, but I can't Google via voice in my car on the drive. And for someone who's driving a Tesla and has a GROC integration right there, they're just like, sure. Like, this is great. Okay, yeah, that's fair.

Starting point is 00:33:04 It's like not. And like the frontier use case is important. Like that's where the action is happening. That's what's driving the next order magnitude of growth. But like there are plenty of people who are like Google's search overviews are amazing. You know? And they're like, that's like, that's my level. And like that's good.

Starting point is 00:33:20 And they're like. Yeah, but like I don't think those people have actually tried like GPT 5.4 pro. But I don't. So good. It is. It is good, but it's slow. And truthfully like, like you can fire off the exact same query to. to 5.4 Pro and 5.4 and 5.4, 5.4 fast. And if the query is simple enough, the answer will be

Starting point is 00:33:46 exactly the same. Because if I ask, if I ask 5.4, 5.4 extended thinking, like, what is the capital of California? And it thinks for 10 minutes. And it just tells me, say, like, Sacramento. See? There you go. That's why you need to think. A lot of people. I told you. I run my life on I'm saying, I hallucinate a lot. But people have said I have the mind of GPT2. It's true, it's true. But, but so, so I think, I think for, I think for some use cases, you know, a smaller model, something's a little faster, something that's not, you know, absolutely frontier is fine. So, I don't know.

Starting point is 00:34:25 What, what do you think? Okay, so imagine that, um, do you think there's something else going on? Imagine that you could get 5.14 Pro spark. So it's on, you know, Subaru's chip. Yes, yes, yes. Would you hit that every single time? When would you not use it? I would not use it if I was doing a deep research report necessarily. Because I just want extended reasoning for certain things.

Starting point is 00:34:43 No, but I'm saying like you could have that. Oh, oh, oh. So 5.3 codex sparrow, which is like super fast, extended reasoning, but it's still super fast. But it's still super fast because it's on service chips, right? If you had that, yes. If money is no object, absolutely. Like I, you know, I'm happy to pour out the glass of water for it to get the best and the best intelligence possible.

Starting point is 00:35:04 I just think that even if you just care about speed, there's still better. I think, in my opinion, like right now there are better models than... Okay, so walk me through it. Like, 5% of App Store revenue seems really high. What's driving that? Yeah, not everyone is, like, extremely tapped into, like, the current model that came out two days ago. You've got to use it. So you agree with me.

Starting point is 00:35:26 I think you agree with me. About... About the fact that, like, good enough intelligence is still, like, a good business to the tune of 5% of app store revenue from... Gen. AI apps. Yeah, but I'm saying that, like, you, because you know about the stuff, you should, like, there's better things that you can use. I told you.

Starting point is 00:35:42 I told you I'm talking to Chad GPT. Look, don't, don't shame me. I'm not the one. But I'm just saying, like, I don't even think you should be shaming someone who's talking to an old model. I'm not shaming them. I think, I'm saying, you're model shaming. It could be better.

Starting point is 00:35:54 Like, you could have a much better experience. Okay, okay. So you want to evangelize the frontier. You want to evangelize the frontier. Yes. But, I mean, I'm just, I'm just wondering, like, if we did, we need a new touring test. So we need to have random people come in and they get to talk to 5.4 or, you know, 4-0 or something. And can they tell the difference?

Starting point is 00:36:16 And which one do they prefer? They might prefer 4-0. They might prefer 4-0. This is the new New York Times writing test. We should put one of these out and be, can you actually tell the difference? This is interesting. Because I feel like a lot of people say they can, but they probably, unless they're really, really grinding and they're trying to do something that requires

Starting point is 00:36:36 like a really long reasoning chain, it's totally possible that they're just like, yeah, like it's gave me the right answer. It looks good. That's a narrative violation. Anyway, let's continue. Apple's revenue from generative AI apps rose from about 35 million in January

Starting point is 00:36:53 to a high of 100 million in August. Do nothing win. Do nothing win. Create an app store over a decade ago and just keep reaping the rewards. They sowed, and now they're reaping. Sales have fallen from their peak, partly because Chatsypte downloads have declined, according to the data. As a proportion of Apple's total sales, $1 billion is small, yet GenA apps are, GenAI apps are a growth driver for Apple's services

Starting point is 00:37:23 business, which investors have focused on in recent years because it has grown faster than device sales and boasts higher profit margins. Apple's dominant share at the top of the smartphone market affords it another luxury. Time to get its own AI strategy rights. So they're making money while they figure everything else out. Apple's AI plan runs counter to strategies of competitors that are spending hundreds of billions of dollars on chips and data centers to build frontier language models. Apple is spending a fraction of that, aiming instead to use all of the personal information. People store on their iPhones together with the chips that it designs itself to power an on-device AI strategy. That strategy could prove

Starting point is 00:38:01 a winner if, as some AI researchers have suggested, access to user data and strong privacy makes on-device AI the dominant way consumers access the technology. Apple investors want to see progress from Apple's own AI strategies, such said Charles Reinhart, chief investment officer of Johnson Asset Management, an Apple shareholder. Quote, if they can act as a toll road for providers of AI, then they'll probably end up looking good long term for not having the big CAPX overhang. Now, I have to imagine that Apple is not capturing any revenue from enterprises, developers, ClaudeCodeC, any of those developers, they're probably not,

Starting point is 00:38:45 even if they are winding up using like a chat GPT subscription in Codex, they're probably setting that new subscription up on desktop. It's a toll road on the actual side. But it's a toll road on consumer, which is consumer sales. All the more risen to get into ads, honestly, because Apple does not tax those. Yeah, and AI is exciting for Apple because they need a new product that they can just randomly bill you like $2.99. Yeah. Anytime they need a cat, like some additional cash.

Starting point is 00:39:17 299? Like don't, don't you get just random bills from Apple, like you're in there. $2.99? Like $2.99? Yeah. Like I feel like every time I check my email, it's like Apple is charged you. random amount for some for some subscription. No, I do get emails from Apple,

Starting point is 00:39:34 but it's always like two days after I bought or rented a movie on Apple TV. And it just says, like, you rented this movie. And I'm like, yeah, I know, I click the button. It's fine. You don't need to email me. In other news, Rolls-Royce has scrap plans to go all-electric by 2030 as, quote, drivers prefer V-12 engines. Would you look at that?

Starting point is 00:39:56 I mean, and this is just a total shop. Total shock. Yeah? Total shock. Yeah. Drivers totally had to experience, you know, being forced, EVs forced upon them for the last few years to know that they preferred combustion engines after all. Of course, I'm kidding. I think a lot of people were just, you know, sort of saying this over and over, manufacturers were not listening.

Starting point is 00:40:21 Everyone said, Elon has been saying the roadster reveal will blow your mind. if it has a V12 and a we've been talking about this. If he drops a V12 that would be completely break the internet.

Starting point is 00:40:37 Yeah. It would be incredible. Anyway, let me tell you about public.com, investing for those to take it seriously. Stocks, options,

Starting point is 00:40:43 bonds, crypto, treasuries, and more with great customer service. Let me also tell you about MongoDB. What's the only thing

Starting point is 00:40:49 faster than the AI market? Your business on MongoDB. Don't just build AI. I own the data platform that powers. Let's talk about this Tesla, that you were following yesterday. Oh, yes.

Starting point is 00:40:58 Did you drop this in the chat already? I sent it to you. I know, we shouldn't, we shouldn't, we shouldn't share the actual picture. But I saw a Tesla that was a very funny mix of, it had the anti-Elon club on it, but also an 1199 license plate. And it was a plaid and it just like mixed

Starting point is 00:41:16 every possible political ideology together. And it had a vanity plate. That was very sci-fi. So it was mixing like, I do wanna go to Mars, but not with Elon. Yep. And I want the cops to do it. The license plate basically said, beam me up. Beam me up.

Starting point is 00:41:30 So they want to go to Mars, but not with Elon. They support. They have an incredible amount of disposable income based on the enforcement. They enjoy high trim levels, but they do not agree with Elon's actions. Well, maybe they work for a rival AI lab or something. And so they're extremely sci-fi pill, but they just don't like, they just feel like they're competing with X-A.

Starting point is 00:41:52 California has now spent. over $100 million on a new bridge to nowhere. It is a wildlife bridge, which I've driven by hundreds of times. I've been seeing it. I've been experiencing the traffic that it causes. I'm not against the concept of a wildlife bridge. In fact, I think it's fantastic. It does feel like in a concrete jungle, this is beautiful.

Starting point is 00:42:21 Totally. This has a lot of opportunity to actually improve the visual aesthetics of this particular part of the state. Caleb Hammer says, bro, this state cannot be real. Isn't Caleb Hammer? It's very real. Isn't Caleb Hammer? He's like a finance.

Starting point is 00:42:36 Yeah, he's got like the number one. He's like the one person you'd come to to be like, should I spend $100 million in a bridge? And it's actually quite a bit more than 100 at this point. And the funny thing is like it's just kind of a bridge, but it doesn't, it's lacking the entrances for the bridge. I feel like if they had built just like even just a little bit of wood to like, like smooth it out so that it looks like there's at least going to be start of a of a of a of a of a

Starting point is 00:43:01 ramp to get on the bridge like the bridge looks solid the actual center part looks solid it doesn't feel that hard to finish this bridge i'm i'm optimistic that this gets done in the next hundred years like top apparently colorado built a built a wildlife bridge or a low cost of 15 million dollars no that's not bad like functionally uh something very very similar the the interesting thing is Apparently the bridge is in some part for cougars. Cool. And the wild thing is like on one side of the bridge, you have a bunch of residential homes.

Starting point is 00:43:35 And on the other side, you have a bunch of cougars. And so they're now going to, the cougars are going to be able to go hang, basically hang in all the backyards. So we'll see how this goes. But I'm excited for this to be finished up. It's been, as long as I have lived in Southern California, They've been working on this bridge, and it's about time. Well, former partner of TVPN, Fall, a generative AI model hosting service that you know and love,

Starting point is 00:44:05 is in talks to raise $300 million to $350 million at an $8 billion valuation. Annualized revenue has hit $400 million up from $200 million in October. That's the TV effect. No, they've executed incredibly, really, really insane. So congratulations to them. Growth. And look forward to having them on after they move past the advanced talks. Yes, yes, yes.

Starting point is 00:44:33 What is Miles Brundage saying? He says, I'm a bit worried that Anthropic is an org-wide case of AI psychosis that makes them think Claude is good enough that they can ship random product features without breaking things, but they, in fact, do keep breaking things, and they're not online enough to notice people complaining. Yeah, I don't know about the last part. They seem very online.

Starting point is 00:44:53 Yeah, and I don't know too much about the issues, but there is, like if they're truly competing in consumer, like there are, like low-hanging fruit, like text to speech on deep research reports is not a feature that exists yet. It feels very obvious, but it's so interesting being this dynamic where, you know, you can ship things so fast, and yet there's some obvious product improvements that are just sort of stuck in the queue because there's a lot. a lot to do, and it's an exciting time. What else is Anthropic doing there? Hiring for a policy manager who will be in charge of chemical weapons and high-yield explosives. This reads like you're going to be building high-yield explosives, which sounds like an Anderol job posting, but it is in fact for a policy manager who will be hopefully stopping people from... No, no, no. I think I read this as somebody whose job it is to decide how Claude is used to create chemical weapons and high-yield explosive? I think it's, I don't know.

Starting point is 00:45:56 I think it's probably like this person decides, like, where's the edge? So if you're asking, like, okay, I have a firework and I want to make sure it doesn't go off, like, should I, you know, throw in the trash or put in the recycling or take it to a special place? Like, Claude should answer that. But if you go to it and you ask it, like, how do I build a C4 or something like that, like, there's all these like policy edges where if you're talking about counter strike and you say like let's plant the bomb it shouldn't flag that as okay you're actually trying to plant a bomb it's like you know

Starting point is 00:46:30 you're asking about a video game we we know how to interpret that appropriately but there needs to be like a human in the loop to decide like where that frontier is and where that particular trade is anyway or if says what terrifies me is if AI were to cure cancer and save 50 million Americans imagine the backlash from hardworking scientists who wanted to cure cancer themselves they will be involved for sure well it's interesting that that was yeah that was like at least one person's response to the australia dog story where they were like yeah this we've been able to do this for a while but like don't do this which was you know an interesting response to somebody who you know went on a multi-year journey to try to save their dog seemingly is having some success outcomes very very odd

Starting point is 00:47:17 Let me tell you about Railway. Railway is the all-in-one intelligent cloud provider. Use your favorite agent to deploy web apps, servers, databases, and more while Railway automatically takes care of scaling, monitoring, and security. And let me also tell you about Vanta, automate compliance and security. Vanta is the leading AI trust management platform. So, PG says anything made before 2028 is going to be valuable. And he's quoting an opening eye employee who he says implicitly discloses their timetable. Anything made before 2028 is going to be valuable. That is such a vague. Someone was hanging out with Paul Graham and was like, let me vague post-IRL. I require content. I require content. This is going to be very valuable. I'm not, I wouldn't get this up.

Starting point is 00:48:05 We are working on something that will be incredibly valuable. One of our team members had a major breakthrough, and it could be the blockbuster product of the year. I'm very excited for this. This actually, we were just messing around. around this morning with an existing product that we had. Had a breakthrough. We had a major breakthrough. We.

Starting point is 00:48:23 I had nothing to do with it. We, and when I say that, I mean, Ben, had a breakthrough that I think will change one of America's past times forever. I think so. I'm actually so confident in this that I'm willing to vague post about it. We need to get him a patent for sure. Get this young man a patent. I'm on a patent.

Starting point is 00:48:44 I have a patent. It's great. When you get a patent, you can. also like frame it, get a little tombstone. It's very nice. Regardless of what happens with the business, it's like a good moment in your business career to like have a patent. Do you have a tombstone for your patent? I don't. I need to order one. It has been issued like my name's on it. I'm like the seventh name on the list or whatever, but I'm technically on it. And so I should, I should get my, you know, plaque or whatever. I'm sure you can just buy them. There's probably something

Starting point is 00:49:08 that. What did PG give? He gave some more context. He said this was after I mentioned the idea of buying rare old things as a hedge, since the one thing AI won't be able to do is go back in time that we know of. I don't... I mean, that's the whole plot of Terminator. Just say you're not AGII pill. Just say you're not AGI-I pill.

Starting point is 00:49:27 Everyone's saying it's going to be like Terminator, it's going to be like Terminator. Well, what happens in Terminator? They invent time travel. And so you're going to be able to go back. You're going to be able to go back to time. Yeah, I feel like after we get AGI and we can build incredible things,

Starting point is 00:49:38 like those will be really valuable too, right? Yes, yes. but yes I mean He is correct That like like You know images in chatGBT And and V-O3 and mid-jurney Like don't decrease the value of the Mona Lisa

Starting point is 00:49:52 Like that's just like obvious And everyone agrees on that Except you You're like I would actually like to go to the mid-journey Louvre Maybe that's why Banks Maybe Banks he sort of intentionally Kind of kind of revealed himself I don't I didn't follow that story

Starting point is 00:50:07 Like how did that happen because I feel like most people would just be like that. Hopefully Scotto is watching and can fill in. But I think he was just kind of like caught in the act. Really? But maybe he's confident enough, hey, AGI's coming. Okay. AGI's here.

Starting point is 00:50:26 My stuff's still going to be worth a lot. Yeah. Even I don't want to be in the shadows anymore. Yeah. That's very interesting. So anyway, Paul Graham clearly does not believe in the Terminator thesis of the AI future. where time travel is off, is possible. Imagine being in the Terminator future

Starting point is 00:50:43 and creating the time machine and just being, okay, I'm going to go back in time and paint new paintings that then I can acquire over time and have new Mona Lisa's. It's just like, no, we sent you back to save the human race. It's like, but I got to hang out with Leonardo da Vinci. I got to build my art collection.

Starting point is 00:51:04 He said, I don't put too much weight on the specific year, but the shape of the idea is interesting. And I agree. It is an interesting thing to noodle on. Similarly, CrowdStrike, interesting idea. Your business is AI. Their business is securing it. CrowdStrike secures AI and Stop Breaches. Let me tell you about TurboPuffer as well. Serverless is back during full-tech search, built from first principles and object storage, fast 10x cheaper and extremely scalable. Martin Screlli. What's he say? He was coming on Monday for the great debate. The great peptide debate says good music is the last mile of AI.

Starting point is 00:51:39 And Lil Wayne has some thoughts on AI music. Let's play this clip. Let's play this two-minute clip from Lil Wayne on a podcast. Let's see. Here we go. How you handle AI in this business now? Challenge. The challenge.

Starting point is 00:51:55 Bro, it's wild. I love it. AI is a better thing. I love that AI is what it is. Yeah. Because, man, I love to be able to stand right next to whoever AI is, he, she, whatever, or whatever, AI is staying right next to. I'm still a bell.

Starting point is 00:52:09 I know. Ain't that's something. I'm going to keep telling me what you do again. Yeah. I'm going to run your list of it. I do this, I do that. Yeah. I love it.

Starting point is 00:52:20 I love the challenge of it. The first time I was seeing somebody, it was, my friends was a little worried. They was like, man, bro, they got this AI stuff where you can just ask it to give you a verse like little Wayne. And so I did it. I said, let me have verse like little one. And he gave me her best shot. Yeah. I done it on a couple of devices.

Starting point is 00:52:39 That's you know not only a phone a computer or calling for a commercial I was shooting for the Alexa thing I want to hear and I have a thing called proto and home they got his own little robot thing As Turner give me one and they are you suck You suck I'm gonna be okay I fuck with that Yeah, who was it beanie Siegel I think he had to start using it because he like Losing his voice a little bit yeah another rapper to maug basically that's his take That's so funny that's great

Starting point is 00:53:08 Well, how are designers feeling about AI these days? Samir says, bro, it's so over for designers. Google Stitch is insane. Google launched a new generative AI design tool where you can sketch something out on a piece of paper, turns it right into an image. And there was a lot of back and forth over, you know, how this debate's playing out between Google and Figma.

Starting point is 00:53:36 Hadley Harris says 12 years later the VC who passed on Figmas seed because Google could kill them is finally feeling seen Lots of amazing work done of course in the interim Very very silly will be interesting to see how Google pipes this into the other tools. I was on Apparently this is in Engagement bait. Yes, other people are testing similar prompts and getting much much much worse results. Yeah, I don't know I I was on I was on AI. Google.com, or just AI.org, not even AI.orgul.com, looking at all the different Gemini features. And it feels like the next challenge is just integrating all these different things.

Starting point is 00:54:21 They have so many great models, Nanobanana, V-O-3, Notebook L.M., Gemini, Flow, AI mode. There's so many. And actually, piping a workflow from one to another is certain. going to be like the next question and does this all live in the Google search box? Is it in Google apps or something? Either way, they're certainly investing in AI across the organization and Ryan Peterson says whoever named Gemini at Google really named it. In the mythology, the immortal twin gives up his immortality to save the life of his mortal twin. It's just like Google giving up 100% of its free cash flow to make sure deep mine survives. That is not actually the reason for the name Gemini, but do you know the name for why they

Starting point is 00:55:16 pick Gemini? Because they had two internal teams that they brought together. Yeah, the twins. It's a good name. And it's really, they were suffering from a bit of a naming crisis for a while with Bard and Palm and they were definitely shipping a little bit of like, not the not necessarily shipping the org chart, but shipping some of the internal naming schemes that were like abbreviations. Even like Nanobanana Pro, like the image models used to just be like Gemini 3.1 image.

Starting point is 00:55:45 Yeah. Whatever. They didn't get 3.1, but. There we go. There we go. Like Nanobanana was famously just like the internal name. Name. Yeah.

Starting point is 00:55:54 And sometimes these internal names can really fly in consumer context. Mostly, I don't know what it is, but something about like nanobanamo. banana like really sticks out, it's so funny. It doesn't sound like an AI product. Like when you have Siri and Alexa and then you have Rufus and Argus or something, Sparky from Walmart, like doing another human name can actually be a disadvantage because you just get lost in the clutter. But if no one's really using the nanobanana name, I know strawberry was used by Open AI before, but no one had really like used that as a public name. It certainly like helped them, help them break out. But Gemini has been an interesting name.

Starting point is 00:56:32 It's good. It's good. It balances both. It sounds like a product, but it sounds, but it's just one word. It sounds like a, it doesn't anthropomorphize too much, but a little bit. It is reference. There's a lot of deep knowledge there. Quickly, let me tell you about 11 labs. Build intelligent real-time conversational agents.

Starting point is 00:56:51 Reimagined human technology with 11 labs. And let me also tell you about vibe.com. Where DDC startups. Oh, my God. for DDC brands, B2B startups, and AI companies advertise on streaming TV, PEC channels, target audiences, and measure sales just like on meta. So without further ado, we have our first guests of the show. Pat Grady and Carl Eschenbach from Sequoia Capital in the TV.

Starting point is 00:57:14 Let's go on, guys. Carl, Pat, how are you guys doing? We're doing great, man. We're doing great. It's great to be with you guys. Yes. Thank you so much for hopping on the show on this day. Walk us through the decision.

Starting point is 00:57:28 How long have you guys been talking? about potentially reunifying? What's the thinking? What's the role? Sort of break it all down for us. From the moment, Carl left Sequoia. Yeah.

Starting point is 00:57:40 Come on back. Come on back. Yeah. No, well, maybe I'll just start. And then my partner, Pat here, can jump in. And I say my partner,

Starting point is 00:57:48 because I've been his partners for 10 years. Like I joined almost 10 years ago, you know, at Sequoia, which was an incredible journey at that time to join. and then I decided to step away back into an operating role, you know, after an eight-year total journey at workday, five years on the board and then back into the operating role. And I remember, quite honestly, when I left Sequoia three years, three months ago, I said to Pat and the team, you know, I'll be back.

Starting point is 00:58:18 I'll be back. Because I always had wished, I dreamed, and envisioned of coming back here. and I thought it would be a great place to end my career and provide the impact to the partnership, my partners and companies that we work with. And, you know, when I stepped down at workday now two months ago, we started talking without exaggeration that day. Yeah. Typical fashion, the news was out and it was within, no exaggeration, five minutes. I started getting text from Hat Grady every single partner at Sequoia.

Starting point is 00:58:57 Not only did I get them, but my wife started to get it. And they were like, when is Carl coming back? We want you back. And I will tell you, it's a blessing to be back. And here we are today. Okay. Walk me through the bull case for becoming or, you know, going back into venture capital in 2026.

Starting point is 00:59:19 There, you know, it's maybe a little bit of crazy narrative, but there's a narrative around like AI is eating everything, AGI is here, the big labs or dominates. are dominating they're going to eat everything. You know, Google's going to beat every company now. Open AI is going to beat every company. Sequoia has investments in the labs. What's the opportunity that you see for new startup formation, new growth rounds?

Starting point is 00:59:42 Where do you want to see opportunity, either in AI or in software or outside of AI? Like, what is your thesis for why it's going to be a rewarding experience to be a venture capitalist now? Yeah. Well, it's a great question. but let me step back and go back two months ago first. When I stepped down at workday, I did open the aperture up and look at everything. I looked at everything from going back into an operational role at very large companies, big public companies. I looked at operational roles in private companies.

Starting point is 01:00:17 And some of those private companies are the AI companies that our friend here at Sequoia at the time had invested in and being part of it. And then I also start to think about, well, what would it be like to go back into Venture, go back to Sequoia at a time of what I would describe as the most massive technology disruption we've ever seen in the history of mankind? And to be part of it across many companies as opposed to a company is something that really excited me. joining someplace like Sequoia, and I'm biased when I say this guy, the greatest venture capital firm on planet Earth ever, mission living, the greatest partners in the world, people that are smart as hell, that even I at this age get to learn from every day, and to be part of something so iconic in the midst of this massive tectonic shift and get to explore across so, many different companies as opposed to a company is something that really excited me. Just in my first two weeks, while I haven't been fully on board yet until today, I've met with so many companies for Sequoia, right, and got to experience the vibrancy of what's happening in the AI world. And I just felt like Sequoia is a place that gives me the opportunity

Starting point is 01:01:45 to stay engaged, partner with incredible people like my young man here next to me, Pat breaking and his team, and just be part of a generational shift, both as Sequoia inside the building and outside the building when it comes to building tectonic technology companies that will stand the test of time. And I get to be part of that across all different dimensions of this crazy environment we're living in. I love it. It's amazing. So you talk about this massive technology shift, which we can all agree on, is part of your thesis of coming back and judging from the letter that you sent to the team. It felt like to me, like some element of your approach is like, hey, a lot of stuff is changing in terms of technology and markets, but a lot of what you're trying to bring is kind of wisdom around maybe what isn't changing, which is human nature and leadership and how to work with people and how to get the most out of your team. teammates. Is that like part of your mindset at all in just coming in and helping both portfolio

Starting point is 01:02:51 companies and the, you know, the entire team at Sequoia just level up? Yeah, no, clearly I think that's a big part of my personal mission going forward. If you read the letter that I sat down and wrote one evening to Pat and Alfred in the partnership, and by the way, I wrote that letter after looking at a lot of operating roles and talking to a lot of other venture capitalists and firms, and I kept coming back to myself and saying, why not Sequoia? Why not Sequoia?

Starting point is 01:03:23 That was always my grounding point when I was thinking about what to do next. Fast forward, I wrote that letter because I have a mission in life, a mission in life to give more than I get. A mission of life to not focus on success, but focus on significance and impact to others, including my partners, the partnership and founders. And now being in the industry for 38 years, and having been here for six and spent 32 years in

Starting point is 01:03:54 operating roles, I felt like I was uniquely positioned to bring wisdom, to bring advice, bring coaching, to bring mentorship to younger founders, right, to help them achieve their personal and professional goals. and the Sequoia platform in my incredible partners like Pat and Alfred and Andrew and the rest of the team here, give me that opportunity to bring what I'm most interested in that is serving others to the table at a time that while, yes, technology is moving at a pace and rate that we've never experienced, there's also a need to help people understand how to build, scale, grow, lead. inspire, motivate others. And that's something I'm super passionate about. And my incredible partners here are security, he said, hey, this is a place for you to do all of that, both inside and outside the building. And that's why I'm here. And I couldn't be more excited about it, guys. Nowadays, it feels like, and also in the data, companies are growing faster than ever. And you get,

Starting point is 01:05:01 you know, what would have been maybe 10 years ago, like five years of company building compressed down into two years when you're kind of mentoring CEOs or leadership teams, you know, in the portfolio, like, what advice are you giving them around that new dynamic, which is that, hey, you might raise three rounds in a year, you might be adding headcount at a much higher rate than you had to. And meanwhile, like, again, the technology is shifting at this insane exponential rate as well. Yeah, so it's a great question, and I'd have Pat chime in because they're seeing this, you know, obviously every single day, companies are raising a round, and one, two, three weeks later, they're raising another round, and then they're trying to figure out what the hell to do with that capital. But let me go back and just share with you one of the principles that I've always focused on when it comes to business strategies.

Starting point is 01:05:59 and I always have talked about speed as being one of the best business strategies ever. This is long before the current environment we're in. And I say that because I use, if you will, a sports analogy. If Pat is faster than me and is operating or executing faster than me or running plays faster me, it's hard to defend. Speed is a business strategy. And right now, we're in the midst of everything happening very, very quickly. So speed becomes more important, but you can't be sloppy. You can't do things in a wasteful way.

Starting point is 01:06:34 You can't just spend unlimited capital because at some point that capital will have to be replenished and now have to come to people like us and others. So I think there's a dimension of how to leverage speed as a winning business strategy, but also be smart about how to do it and where you're spending those dollars. And I'll let Pat chime in and what they're seeing. I saw it today on some of the calls that Pat was leading across the partnership. Yeah, I think that's well put. Speed is, you know, one dimension of a vector.

Starting point is 01:07:03 The other is direction. I think the other thing that's interesting right now, it's not just how quickly things are happening. It's how dynamic the entire market is. You know, our partner Constantine has this framework that there are some revolutions in computation and some revolutions in communication. A revolution in computation is about the way information is processed.

Starting point is 01:07:20 A revolution in communication is about the way information is distributed. Cloud, mobile, Internet. that all of those are revolutions in communication. It was about the distribution of technology or of information. This is a revolution in computation. It's the processing of information. The result of that is that the raw ingredients that you have available to build your product change every day, or at least every week, possibly every month.

Starting point is 01:07:45 But they're changing fast, right? And so you can be running like heck in a particular direction. If it turns out that that technology floor shifts underfoot, all of a sudden you've got to change direction. And so I think one of the big sort of things we have learned from the founders who are doing it best, the Daniel Nadler's and Winston Weinbergs of the world, is that talent density matters so, so, so much. It's not about being bigger.

Starting point is 01:08:11 It's about being better. It's about having the densest possible talent pool so that when the world shifts, you're not caught flat-footed. You go in the right direction. Makes sense. It feels like an amazing time to join Sequoia. the technology industry is bigger than ever, venture capital is bigger than ever, we're in an AI boom.

Starting point is 01:08:31 But I'm wondering if you could turn back the clock for me and tell me what the vibe was like when you joined VMware, because it feels like a very different time. And what was that like? Yeah. 2002, right? Yeah, wow, you're taking me way back. Yeah, I joined VMware in 2002. Yeah.

Starting point is 01:08:54 And when I joined, I think there was a couple hundred people with probably 95% of them being engineers. I think it was a deeply technical company by great founders out of Stanford and Diane, you know, the founding CEO, great people. And I remember joining, you know, this young little company, basically no revenue and saying, hey, we're going to go change the world. Like every startup says, we're going to change the world. And I remember Diane Green saying to me, we're going to virtualize. these little x86 computers, and the entire world's going to run on top of them. And I remember talking to Diane at the time.

Starting point is 01:09:31 You mean like mainframe partitioning, L parts? That's where we do. But there's, oh, we're going to do it on these little boxes. And I'm like, who's going to hell? Who the hell is going to do that, right? And then I went home and I thought about it. And then I had to go quite honestly to my wife and say, all right, I just left working for a Bay Area company living in Pennsylvania

Starting point is 01:09:53 computing. I committed. I wouldn't do that again. And I thought about this more. And I said to myself, wow, if Diane in this incredible technical team can do what they say they're going to do and turn that slideware into software that virtualizes these servers and allows you run multiple applications operating systems simultaneously, I said to myself, if they can do that, the next thing I said is I can figure out a distribution strategy and I can sell that shit, silly. Yeah. And the journey began there. The first year or two, no one was buying it. It's like all start up to go through that phase.

Starting point is 01:10:31 Like, wow, you want me to take all of these 10, 20 servers, physical servers, separated by physical computers and put them on one. And if that server goes down, then everything goes down. And I'm like, yeah, that's what you should do. Yeah. You should go real well at the beginning. But then we hit an inflection. And then once we hit an inflection with this technology. at the time that they brought to market was called B-motion, which would actually allow you to take a live-running virtual machine and move it across physical servers without the application ever going down.

Starting point is 01:11:08 So now you had some redundancy and backup. And when we started to show people that, first they're like, wait, this isn't really working. So we had to unplug cables to show things were actually working. And then it inflected. and quite frankly, one of the greatest, you know, professional journeys in my life being there 14 years, a couple hundred people, to 20,000. Yeah. And I was so blessed to be there because I got to do so many different roles in the company. I was never CEO.

Starting point is 01:11:36 I was always, I guess, number two guy at the time across three great CEOs between Diane, Paul Merritt, and Pat Gell Singer. And got to be a CFO for a while of a public company, got to help run. product when our CTOs left. It was just an incredible journey. It gave me a view into a startup and it gave me a view of how to scale companies. And it really helped me get a deep understanding of how to build global operations to build a mass scale at the enterprise level. Why did you, what made you push through that one to two years where you felt kind of silly even selling the product just given that there wasn't it didn't feel like you weren't feeling that pull from the market eventually you got the breakthrough with V-motion but what was

Starting point is 01:12:29 the signal that made you keep just running down opportunities? It's a great question. Very simply the engineering talent at the time and if you look at now the industry is proliferated with incredible talent from VMware. It's everywhere. from CEOs to head of sales to head of engineering, it's everywhere. What gave me the conviction in belief to keep pushing through is whatever that engineering team said, the product would do, it would do. And I kept saying, I remember at the time telling people, right, we moved from forecasting numbers or revenue or bookings to forecasting,

Starting point is 01:13:12 literally our forecast calls was how many proof of concepts have we started? Because if someone tested it, it worked. It literally worked. And we were having to... So the market was like, we don't believe you. And you just had to actually convince them to let you show them. Yeah. Well said. I had passion. I believe I had conviction. And I witnessed the technology working as advertised. A lot of times it doesn't in startups. And it takes a while to get there. And I knew it was not if, I knew it was when the market would tip. and come our way. And then when it tipped and inflected, it took off very quickly. And then we started to see all this competitive pressures come in with open source technologies like KVM, you know, OpenStack, HyperB by Microsoft. But we just stayed focused. We stayed convicted about the technology

Starting point is 01:14:07 and we proved it time and time again to anyone who tested. It was an incredible journey. and I'm so grateful for that 14, 15 years. Amazing. Yes, it's such a wild, right? I mean, you've at various points, been working at organizations with 100 people, up to 20,000. Do you feel like you have a good calibration on when you're talking to a founder or an executive, just asking yourself the question,

Starting point is 01:14:35 can you see this person thriving in an organization that's three orders of magnitude bigger? I don't even know if that's a relevant question anymore for a venture capitalist to be asking, but I'm curious if you feel like there is a set of patterns that you've discovered or a set of skills that are on display from younger, more up-and-coming executives and founders that sets them up for success at massive scale. Yeah, I think I can identify patterns and I have enough experience for the six years that I was here with Pat and a team learning from them about what to look for a founder and whether or not they can scale. One of the things I absolutely look for is self-awareness and founders.

Starting point is 01:15:19 And are they honest about what they're good at? Because a lot of these founders, quite frankly, are just incredibly intelligent human beings. And sometimes they think they can do everything, when in fact they can do everything, but not great. and they have something they're really good at. And do they have self-awareness to say, but I need to go get to other people as part of the company to help me scale so I can focus on what I'm really good at?

Starting point is 01:15:43 Do they have the ability to know when to turn things over to others is critically important? And I've seen that happen time and time again. And then the other things, quite frankly, that I personally look for, I'm sure Pat, I heard Pat talk about this this morning, there are attributes and characteristics of people that I look for that are way beyond the intelligence side of the equation. You can't teach grit, you can't treat strive, you can't teach a great attitude, you can't teach determination, right?

Starting point is 01:16:16 All of those are things that are innate and part of people, and you want to see that in these founders. And when you find someone who has that passion, drive, desire, relentless ability to fight through challenges, issues, and opportunities, and then they have the intellectual horsepower on the other side. When that comes together, that's a beautiful thing. I actually reminded the first time I don't know, Carl remember this, first time I met Carl of 2010. And it was Doug Leone, Fred Luddy, who's the founder of Service Now, and I went to visit Carl, get some advice on stealing service now.

Starting point is 01:16:50 And I was probably in my, I was in my late 20s, Carl was in his early 40s. Still in his late 20. Look how young this guy. It bothers the hell out of me, guys. how young, how smart, helping run one of the most iconic partnerships in the world. And here I sit turning 60 this year. He pisses me off every time. So I have pages of copious handwritten notes from this meeting, which I'm sure exists somewhere around here.

Starting point is 01:17:16 But the one liner that's stuck in my brain, which is very consistent with what Carl was just saying, was attitude, was attitude and will determine skill. And I always think about that. and we kind of morphed that into the expression, attitude is the ultimate input. And I think when you see these founders, I'd also kind of use the Ray Dalio line, you know, pain plus reflection equals progress.

Starting point is 01:17:39 When you see these founders who are willing to take a risk and experience the associated pain and then reflect very honestly on what happened and what they can do better next time, progress is inevitable. And so I'd almost say there's this, and like you have to have the right attitude to put yourself in that position,

Starting point is 01:17:57 but if you're willing to take risk and experience a little bit of pain, and then if you're willing to be intellectually honest with yourself and self-aware and sort of clinically diagnose what you can get better next time, those founders are just going to keep frankin. Yeah, really well said. While introspection. Yeah, taking a side on the introspection gate. I was going to try and tease something up, but you really delivered.

Starting point is 01:18:21 Talk to us about both of your agreements or disagreements. or disagreements maybe, I'm sure you've been debating the future of enterprise software, the future of what's going on, the SaaSpocalypse, public companies, like just the nature of business changing. What remains true now that hasn't changed and won't change for decades versus what maybe has changed in the last few years and needs an update in terms of how people think about how businesses grow, how businesses flourish in when they're working in the technology industry? Yeah, I'll have past start and then I'll give you my perspective because I've answered this question about 17 million times in the last three and a half years.

Starting point is 01:19:07 And I have a different perspective than probably most. Yes. So, you know, it's funny. The first thing that comes to mind is a line that I learned from a man named Carl Eschenbach, who was a partner in Sequoia from 2016 to 2022. And the line is people do business with people. And I think there's a there's a foundation model maximalist point of view that the labs themselves are going to do everything and every nook and cranny of the economy. And I just have a hard time imagining that version of the future coming in fruition because people do business with people. And I think that between a job to be done and the raw capabilities of a model, there's a lot that needs to happen.

Starting point is 01:19:51 but like shape it into the path of least resistance for you to travel down as a user to get to the right answer with the least amount of paint. And there's probably a person in between who's going to do that work. And as a customer, you want to do business with that person. And so I think people do business with people is going to remain true. The shape that that takes in terms of what the businesses are is probably going to change. I think in the world of software, you know, the first wave of the on-prem to cloud transition was this transitioning of systems of record. the work days and sales forces and service nows in the world. The second layer on top of that was the systems of engagement.

Starting point is 01:20:28 Those systems of record might own the core database, but then there are a bunch of different workflow applications that reside on top. I think what we're going to see with the wave of AI software is this third layer on top of those, which some people call a system of intelligence. I don't want to call it that. It's the layer that does the work. It's the agents getting deployed that may or may not need those workflows beneath them, but certainly need access to everything that's sitting in that system of record.

Starting point is 01:20:56 I think that's what we're going to see. And as a result, I think those system of record companies are relatively safe. They may not catch a lot of net new workloads because a lot of the net new workloads might go to the AI native companies. But I think they're overall pretty safe. I think some of those workflow-based companies in the middle are in trouble because they're neither the system of record, nor the agenic capability that's getting deployed. And so they'll have to figure out how to become, like that agent harness, so to speak, for whatever job it's to be done. And then I think those AI-native companies on top, the basic thing they need to achieve is figure out the context of this organization, figure out the guardrails, come up with some

Starting point is 01:21:31 sort of an eval framework, some come up with some sort of a value function, basically wrap all the context around the capabilities of the foundation model to achieve the outcome that the business person wants. And so I think there's a very important job to be done for those, that new layer of companies. And again, people want to do business with people. Like there's a lot of value in, you know, having somebody you trust, take your hand and lead you into the AI future. Yeah.

Starting point is 01:21:56 Yeah. So first of all, it's the first time I heard someone else ask this question to someone else other than me. And then you see you for the answer. Actually, it's so funny. I was sitting here thinking, I am going to respond, but you're going to hear something very similar from me. Sure.

Starting point is 01:22:14 I'll start in kind of reverse order. I do think there is power to use Pat's exact analogy of a system of record, a system of action, a system of engagement, and I will use system of intelligence. Because I think if you have the bottom three and you can layer on an agentic or agent strategy and you back that up with the data, the context of the data, and you own the business process work through, you're in a unique, unique, unique. position to have a great enterprise AI software company that stands a pest of time. I don't think we're in a world where does AI win or do the incumbent SaaS companies win? I think we're in a world of and. And I think some people are the beneficiaries of AI like a workday, like a Salesforce. And then others maybe have more headwind because they can be disintermediated.

Starting point is 01:23:16 because there's AI and you don't need access to all that data. I personally believe all of the challenges with software companies as scale and what's happening in the stock market are completely overblown. I've been saying that repeatedly. They're not going anywhere. Incumbency is incredibly powerful in the enterprise. Incumbency is even more powerful for in a company like Workday, who I was blessed to be there for over three years, has a 98% grocery tension.

Starting point is 01:23:46 rate of 11,000 customers, 65 plus percent of them being Fortune 500 companies. Yeah. Not going anywhere. And the other thing we can't forget is why I say and is because what matters in the enterprise is scale security and compliance. And some of these big SaaS companies have that. That all being said, the pace and rate of change that can happen outside of the big incumbent and big SaaS companies who are innovating like crazy.

Starting point is 01:24:18 We're innovating at workday like I've never seen. Have an opportunity to start completely fresh. You can start from scratch. Get to leverage all the technologies and all the models that are out there and build agents and agentic solutions faster than anyone else. And they're going to be able to go in the enterprise and provide value day one either on their own or on top of and through some of these SaaS companies.

Starting point is 01:24:43 So I think it's an and opportunity that's going to happen in the enterprise in the market as a whole. There's going to be some winners. There's going to be some losers. But I think the current narrative out there of the SaaS companies being overblown or being in trouble is completely overblown. And obviously on bias, I think workday stands in a very unique position to completely continue to crush the ERP market, both on the HR and finance side. and I couldn't be more bullish on the opportunity ahead. But at the same time, I'm super excited watching these young, talented people now that we get to invest in and how quick they can iterate, iterate, leveraging AI and just completely disrupt markets,

Starting point is 01:25:27 legist markets that don't have all of the data, don't have the context and don't have the workflows. What advice would you give to a Fortune 500 CEO that maybe is an incumbent that's thinking about buying versus building? these agenic products? I tell them to do both. I think if you have all the data and you have the context of the data and you can build a great engineering team that comes with an AI background,

Starting point is 01:25:56 build as much as you can, at the same time, go do acquieshires, go by technology companies that are completely AI native from the beginning and bring them into your organization. I wouldn't say do one or the other, it's both. You know, at workday, in the last six months I was there, we fought four AI companies, and they become part of the core fabric of the company that Anil and Garrett and the leadership team there get to take advantage of.

Starting point is 01:26:23 At the same time, they bring in their talent on their own as they build out their organization. So I don't think you say it's one or the other. You have to do both. I think it's also one of the classic questions of what do you want your best people focused on? Do you want your best people building the same sort of stuff? get out of the box from somebody else who spends all day long thinking about that thing? Or do you want your best people creating competitive advantage for your company? I think if you're a Fortune 500 company right now, kind of a no-brainer to go with Hardy

Starting point is 01:26:52 for everything related to legal, kind of a no-brainer to go to Sierra for everything related to customer support. You know, you should probably try something like Expo to work on pen testing. Like these are excellent companies with excellent people who spend all day obsessing over a particular problem. Why take your best engineers and ask them to go do that? go do something that's going to be unique to you. Okay, last question.

Starting point is 01:27:13 By the way, Pat makes a great point. Having spent a lot of time with CEOs of Fortune 500s around the world, there's this whole narrative, we're going to do it ourselves. We're going to build our own agents, and they will do some of that. Why, if you can go to a Harvey, right, get what they've already built and leverage it and very quickly get a return on that investment. So there's a time cost of value equation here for the enterprise. Go with that solution.

Starting point is 01:27:36 Let them go focus on financials or insurance. or retail or whatever is CPG. Let them focus on what their poor businesses. I build that technology if someone could do it much faster outside the company. Last question for Carl. A few years ago, you were spotted at the Allen and Company conference sporting.

Starting point is 01:27:55 Two coffee cups. You have an incredible amount of energy. Are you a two coffee in the morning guy? Or were you bringing an extra coffee for a friend? Well, the answer, I hate to use the term again, but both. Okay. I was bringing a coffee.

Starting point is 01:28:12 I was bringing a coffee back to the room for my amazing life of 35 years, Anna. And, and, and, and I used to drink probably eight to 10 to 12 cups of coffee a day. There we go. Wow. There's all I have all day. That's amazing. And that's all I have all day until I get to dinner when I eat dinner. It's my one meal a day. That's fantastic.

Starting point is 01:28:37 Wow. Incredible amount of meat. Well, it's clearly working. Thank you so much for taking the time to come chat with us. It's so good to see you guys both back together. Yes. No, thank you. I just want to say, listen, I just want to thank Pat.

Starting point is 01:28:50 I want to thank Alfred. I want to thank the entire partnership here at Sequoia for allowing me to join them and serve alongside them. Our partners, our customers, our companies we're investing in our founders. It's a true honor to be back. I'm super excited about the journey ahead. And I think, you know, Sequoia is uniquely positioned to continue to be one the most iconic, mission-oriented venture firms of all time, and I'm proud to be back part of it.

Starting point is 01:29:19 Yeah, we're excited for you. Thank you so much. Incredible stuff. We will talk to you soon. Great to see you guys. Have a good rest of your day. Cheers. Goodbye.

Starting point is 01:29:26 Let me tell you about 10 to 12 coffees a day. That's incredible. They're the makers of Devin, the AI Software Engineer. Crush your backlog with your personal AI engineering team. And let me also tell you about Finn, the number one AI agent for companies. If you want AI to handle your customer support, go to fin.aI. And without further ado, we've been running along, but we will bring in Jim Cranthrell from the Phantom Space Corporation. Sorry to keep you waiting, Jim. Thank you so much for taking the time to join the show.

Starting point is 01:29:54 How are you going on? Please, you have a fascinating background. Give us a little bit of your background leading up to this company, and then I want to talk about how you're thinking about the business and also just the orbital economy more broadly. Yeah, so I've been in the automotive and aerospace industries for north or 35 years, and they had positions everywhere from the front space agency to early at SpaceX. I was the guy that took Elon to Russia to buy Russian missiles, and when that didn't work, we started SpaceX. And after that, I'm on my 12th space or automotive startup, and fan of space is the last one of them.

Starting point is 01:30:35 But several of them have gone public since. in and, you know, Phantom Space was sort of the ultimate of all these startups that looking to solve the problem that really needs to be done. Yeah, that story of going to Russia to try and buy the ICBM has been told and written about in books. But what does the current narrative get wrong? What's your side of the story? What were expectations like going into that meeting? Did you, was it seen as a long shot at the time? Or did you think that it was likely to work? Yeah, no, it was a complete long shock. When Elon called me, he had just left PayPal and had this idea of trying to inspire

Starting point is 01:31:15 humanity to become multi-planetaries he still talks about. And he wanted to do what amounted to a stunt to show that we could send creatures to Mars. That turned into something that we put together, which was a growth chamber to land on Mars, on a lander. And we needed Russian rockets to buy it. So by the time we got done dealing with the Russians, they didn't want to some. sell to us. They were just being Russians. And Elon announced all of us in a sort of shocked way.

Starting point is 01:31:44 We heard that he wanted to start the company SpaceX and build the rocket ourselves. So I will tell you that very few people gave us a snowball's chance in hell to make that happen. Now we can see 25 years later where that ended up. But, you know, everybody bet it against Elon and the rest of us. And, you know, what the story's gotten wrong and what it gets right, what it gets right was, you know, this determination of this guy who knew nothing about rockets who decided to learn everything he could. And he got a bunch of space cowboys around him. And most of us were sort of revolutionaries and thought at least and wanted to stick it to the system and do something that everybody said we couldn't do. What the story got wrong, sometimes people write and say I was never

Starting point is 01:32:28 part of it. I don't know why that got written in, but I was definitely employee and had founder's stock and the whole nine yards. So here I am. There you go. What's the modern version of the long shot in space? We've heard about colonies on the moon colonies on Mars, space data centers. What do you think is the most practical problem right now in space that people are maybe undercounting? So I think there's three tracks that it's going down. One is, you know, the military use of space, which we see ongoing today, right? Wake up and read the news every morning.

Starting point is 01:33:01 The second one is planetary settlement, which is what SpaceX is trying to accomplish. And everything that Elon does, I believe, is aimed at that planetary settlement goal. And then the third is kind of more recent, even though I've been thinking about it for almost a decade, is putting compute in space. So now AI is the killer app that enables this,

Starting point is 01:33:21 much like the internet came along was a killer app that enabled Starlink. So I think what we'll see is the next generation of AI in space, these so-called space data centers, but not in the way that the common narrative is going to portray it.

Starting point is 01:33:34 Okay. How will it be different? I mean, Jensen Wong at GTC earlier this week was standing on stage saying that, you know, he will be providing chips. Elon has given some outlines around what that might look like. We've talked to StarCloud,

Starting point is 01:33:48 a startup that's planning to put data centers in space. There's a whole bunch of other people that are approaching this problem, but how do you think people are getting it wrong and how do you think you fit in? Yeah, so, you know, Navidia, to just address that, is nailed the silicon, right?

Starting point is 01:34:04 So they're going to be the winners on that, I believe. I think so. Among others, right? But they're going to be one of the primary winners. And thank you for doing that in the video. Number two is there's going to be a camp that I think is mostly hype that says we're going to put, you know, these large language model hyperscalers into orbit.

Starting point is 01:34:22 If I'm generous, that's maybe 20 years out, right? And it's like flying a big factory on a huge rocket that doesn't exist there. So everything that's going to happen in the future is going to be distributed data centers on a much smaller scale. Everything's more expensive, right? So we're at least 10 times, maybe 100 times more expensive to do something in space today. So the really killer app I see is to put AI inference in orbit close to where the data tsunami is being generated. And today it's a tsunami and tomorrow it's going to be a mega tsunami. And what the problem is is being able to get that data back.

Starting point is 01:35:01 So there's this funnel that restricts how much of that can get back. So AI is a natural way to reduce that data load and get it back to the earth. Maybe later we'll solve the issues with, you know, with the earth power and heating. But, you know, Phantom, you know, we've been at it for 10 years, one form or another, writing patents, writing about it, speaking about it. It's nothing new for us. And we're putting together micro data centers to address exactly this along with data backhaul from the satellites to really exploit what we think is the next killer app there

Starting point is 01:35:37 and create a space app store environment and an ecosystem for others to implement their creativity on. So where do you think in the supply chain or the rest of the orbital economy, there is enough maturity that you will never really need to build. I imagine you're not going to build a new rocket, but are you planning to do connectivity through Starlink? Like where will the partnerships happen? And then where will your, you know, your core value prop live within the supply chain? Yeah, it's a great question because this is exactly where I think all the differences between the approaches become evident. So it's my belief that in order to be successful in this

Starting point is 01:36:19 you really have to vertically integrate much the way we did in the early days of SpaceX. We saw that building the rocket and building your satellites and then implementing in their case, Starlink and now their version of X-A-I in orbit, you really have to have that. And, you know, SpaceX is going to dominate a lot in that phantom space. We have exactly the same playbook. So those who don't have that vertical integration are always going to be at the risk of what amounts to a very scarce launch supply, Even today, you know, there is a perception that there's more launch than we need, and it's not true at all. It's very scarce.

Starting point is 01:36:55 You know, Phantom, we're building something we call the Daytona, which is quite a bit smaller than anything SpaceX builds. So we're more the taxi. They're more the, you know, the freightliner. And so, you know, we find a real market for that. People are buying our things. And there aren't that many people that can actually build launch vehicles at work. It's a really tough business. And it takes five to ten years.

Starting point is 01:37:15 So that's going to always be scarce. The rest of it is really a matter of supply chain control. Yeah. So the, you know, the silicon, Navidia, you know, all the rest of the satellite parts, suppliers, but you have to have somebody to put it all together, operate it, and manage it. So we think of ourselves as building the railroad to space, and then ultimately in the space app store on top of it so that people can build their own code. Yeah.

Starting point is 01:37:39 Talk about some of the tradeoffs of the launch vehicle decisions. Are you thinking about reusability? Is that less of a factor at this scale? How frequently do you want to be launching? Like, I imagine taxis, they go everywhere. They launch from everywhere. Like, how else play out the taxi analogy for me a little bit more? Well, so launch is like the critical cost of getting any data system into orbit, period.

Starting point is 01:38:06 And so, you know, it's incumbent on companies to control that cost. And if you can build it internally, if you can gather the capital and the talent to do it, you're in better shape. So there's there's a trade on getting that cost down between building them super large on Starship and reusability and then mass production. So mass production like the car you drive probably cost $100 million and you maybe paid $100,000 for it. And there's a huge cost reduction through this your number. So at Phantom, we're going to apply both the reusability and the mass manufacturing because we're smaller we can do that whereas Starship as an example won't

Starting point is 01:38:45 necessarily be mass manufacturing, but probably mass used. So reusability is a core thing more for the logistics of these things. And then the other side of it, here's the other choke point, the business is launch sites. We are really out of range capacity, launch range capacity in the United States. And we will be, I think, at that limit within five years. And so companies that control that range capacity are going to be in a position to control, you know, the railroads or the shipping lanes, as it were. Interesting. So are you looking at land?

Starting point is 01:39:18 What's the process to create more capacity? So it's very complicated. It's very bureaucratic and it's very political. So we have five different launch ranges in the United States. Yeah, exactly. Music to my ears. Yeah, right. So most of them are all federal ranges that we built during the Cold War.

Starting point is 01:39:39 And there's one in California, one in Florida. We know at least about the one in Florida very, very nicely. of King Canaveral. And so there are so many pads you can put on there. Most of what we're using today is legacy from what was built in the Cold War. We're grandfathered into the, you know, the, the, the bureaucratic process that approved those pads. So any new ranges, you're going to run into huge opposition. There have been people who tried to build new ranges on the coast of this country. And everybody not in my backyard comes out of their home to oppose it, right? And even in California at Vandenberg, you know, SpaceX recently got in a lawsuit. We had about half their

Starting point is 01:40:17 capacity, which was governed ultimately by the Coastal Commission in California. And, you know, SpaceX had to sue and they got a little bit more capacity. But that's what we're heading into. So that's where you see these launch ranges around the world coming into play. And the problem for U.S. companies is we're restricted from taking our launch vehicles to these foreign countries without government approval. Yeah. Yeah. It's so tricky because I can imagine. Imagine living next to, you know, a launch pad and the first time the rocket goes up, you're like, wow, that was amazing. And then if it's like, we're going to be launching those every 20 minutes. My windows are shaking a lot of time.

Starting point is 01:40:54 I actually would like. Exactly. The novelty wears off pretty quickly. Cool. And so, yes, obviously, as a country, we need to figure out where these go that are not disruptive and are scalable and are tied to the supply chain, maybe actually near a railroad. Who knows? A physical railroad. Give us any predictions around the moon economy over the next decades? Yeah, I have to say I'm surprised by the SpaceX pivot to the moon.

Starting point is 01:41:23 It's been something, you know, since I was first in this business that a lot of us saw as a logical pivot. And it was almost like a religion between, do we go to the moon first? Do we go to Mars? So you thought it was logical like 20 years ago. And so you're surprised. It's a great training ground. How late the pivots happen. Yeah.

Starting point is 01:41:43 It doesn't really matter, honestly, in terms of their technical capability. Sure. It's more incremental in terms of the development of the technology, which is probably why they did it. I really don't know why they did it. It could be a business decision. There's certainly, I think, a more near-term economy there. I think of Mars, you know, as being so far away, eventually that economy will have to form on its own, kind of like this new world where we all sit, formed as an independent economy from Europe 500 years ago began forming.

Starting point is 01:42:13 And Mars someday will have its own manufacturing bases. And you might have products labeled made on Mars, right? But that's a longer-term thing. And that's really where Elon's mind always was, you know, from the early days that I was with him, that it was all about Mars. And to me, the moon is just a stepping stone on that way. And I think there's a lot of people who see that as, you know, sort of an effectively a theological choice.

Starting point is 01:42:37 where it's really not. It's just a technical issue. Economic opportunities on the moon that you think are interesting. We've heard about regolith and maybe like a mass driver, but there's so many opportunities, obviously, besides tourism. The most obvious one is Healing 3. So this is pushed out by the sun. There's something like 18 kilograms of it in the United States.

Starting point is 01:43:02 And this is a strategic resource. It's a byproduct of nuclear weapons manufacturing. Now, what are you going to do? you use it for? Well, you can use it for a couple of things. Clean fusion energy. It's one of the few fuels that doesn't create radio activities of byproduct. So that's obviously desirable, right? The second part is for quantum computing, which a lot of these have to be near zero. So this is one of the few substances that you can cool to near zero temperature. And the other is an absolute zero in temperature. So there's a huge demand on that. Now, once you start mining it, does that, you know,

Starting point is 01:43:36 that price collapse, probably to some degree. The second one that I see is this mineral, this rare earth mineral kinds of deposits. And we honestly don't know enough about what's up there. We have a pretty good idea from the Apollo missions, but there's probably a lot of rare earth deposits, my guess, and I don't know geologists, but I would guess there's probably some rare earth minerals that, you know, mining from the moon, which would be more palatable than tearing up our beautiful earth, would be as long as we can solve the transportation problem. And all comes back to the rocket, by the way.

Starting point is 01:44:06 Yeah. Thanks a lot of sense. Well, thank you so much for taking the time. Yeah, great to me. Pleasure. Congratulations. And we'd love to talk to you soon. Yeah, come back.

Starting point is 01:44:14 All right, soon. Have a good rest of your day, Jim. Let me tell you about Cisco. Critical infrastructure for the AI era. Unlock seamless real-time experiences and new value with Cisco. And let me also tell you about console. Console builds AI agents that automate 70% of ITHR and finance support, giving employees instant resolution for access requests and password reset.

Starting point is 01:44:34 So without further ado, let's bring in Tom home from GV. Tom, how are you doing? What's going on? Hey, guys, can you? We could hear you for a second. Can you hear us? You're back. You're back. Where are you calling in from? Can you hear us?

Starting point is 01:44:49 No? Yes, no? Check. It's okay. Can you hear this? I'm going to tell you about Sentry. Century shows developers what's broken and it helps them fix it fast. That's why 150,000 organizations use it to keep their apps working.

Starting point is 01:45:04 Could you hear that? Because I could also, I can also potentially tell you about the New York Stock Exchange. Because if you want to raise the, if you want to raise capital, Tom, you got to do it at the New York Stock Exchange. I'm sure a lot of your companies are aiming for IPO. Let's get them live on the New York Stock Exchange. We'll get Tom back as soon as he's ready. There is some breaking news that we do got to talk about, which is that Jeff Bezos. Nothing gets me more fired up than a personalized ad. Okay, breaking news. Jeff Bezos in talks. to raise $100 billion for AI manufacturing fund. The Amazon founders traveled to the Middle East, Singapore, and fundraising effort linked to Project Prometheus.

Starting point is 01:45:42 That is incredible. Very, very, very exciting. Breaking news. Advanced talks, I don't care if it's just advanced talks. I'm hitting the... Congratulations to Jeff Bezos. Meeting with some of the world's largest asset managers to raise funds for the project. A few months ago, he traveled to the Middle East to discuss the new fund with sovereign wealth representatives.

Starting point is 01:46:02 more recently he went to Singapore to raise funding for the effort as well. It's being described as a manufacturing transformation vehicle. It's going up against TK., right? I mean, TK.K. is not as directly focused on manufacturing. Like this is something I asked. Yeah, but transportation vehicle, right? No, no, he's just saying manufacturing transport. Like it's a vehicle, a fund for transforming manufacturing.

Starting point is 01:46:28 Oh, yeah, yeah, yeah. It's like an investing vehicle. It's aiming to buy companies in major industrial. sectors such as chip making defense, aerospace. Let's try again with Tom. Get out of here, Tyler. It's Tom, time, Tom. There we go.

Starting point is 01:46:41 Hey, Tom. Can you hear us? Oh, no. No, we don't have audio. We don't have audio. Okay, guys, try it out. Nothing? Okay.

Starting point is 01:46:50 We, you can hear us. We can't hear you, but we can tell you about this post that we enjoyed from Llariken. Got my horse to water. Now for the easy part. I'm going to continue. The fund is aiming to buy major industrial sectors,

Starting point is 01:47:08 such as chip making defense and aerospace, who would dwarf the size of some of the world's largest buyout funds and rival soft banks, $100 billion fund. I got to wonder, how much do you think, how much do you think Jeff is pitching in himself? I could see him, you know, anchoring. He's like, I'm good for 30, you know, something in that range. Yeah.

Starting point is 01:47:28 Yeah, yeah, he's got some fun. But this is such a white pill. Yeah. Why? I mean, the whole, this is like, you know, we need to manufacture, you know, basically we need to reindustrialize America. Yeah. We're not going to do it by just copying everything from the past.

Starting point is 01:47:44 There's some element of transformation that needs to happen as well as new efforts. Yeah. And this is tremendous news. Yeah. And I mean, there has been like a venture capital boom in reindustrialization, but most of the funds that we talk to that are in. that category are 50 million, a couple hundred million, certainly nothing at this scale. And this has got to be incredible news for the founders that we talk to that are part of

Starting point is 01:48:15 the reindustrialization effort because they have a new potential investor. Did you see that Open AI has acquired Astral who will be joining the Codex team? So finally, Open AI has Astral Codex, which of course is a great play on Astral Codex. 10, the blog, which has some fantastic articles. Let me tell you about ACTA. Octa helps you assign every AI agent a trust identity, so you get the power of AI without the risk. Secure every agent, secure any agent. Will Open AI increase the cost of chat GPT? This is a Kalshi market out there. It's a 46% chance. If either chat GPT pro or plus have a price increase after January 2nd, 2026, and before January 1st, 2026, and before January 1st,

Starting point is 01:49:02 So in the year, based on the current pricing of $20 a month and $200 a month. What do you think? I feel like, this is a financial advice, but I feel like they're not going to move the price points because they're focused on so many other things. Yeah. The only thing is I could imagine an additional plan on top of it. Yeah, totally. An incremental plan versus changing. Max is right there.

Starting point is 01:49:26 We got pro. We got plus. We got pro. Sign me up for the max plan. And I think they also have a light plan. And I would imagine, there's a go. There's go, yeah. So I would imagine that there's more tiers within there.

Starting point is 01:49:38 At the same time, give Tyler the goat. Ripping the Band-Aid off of price adjustments is extremely painful. And probably the earlier you do it, the better. Like Netflix, yeah, yeah. Like Netflix has been increasing prices. And when they're at $9.99 a month or something, it's very normal. And then when you go up, it's all, you raise the price.

Starting point is 01:49:57 But if you just raise it every couple months. Anyway, we're good. Third time. Third time's the charm. How are you doing? I'm very good. Can you hit me, guys? Yay!

Starting point is 01:50:06 Fantastic. Introduce yourself. I'm Tom Hume. I'm one of GV or Google Ventures' is managing partners. I'm based in London. I'm going to hope I'm going to bring as much energy as Carl. I love it.

Starting point is 01:50:18 I love it. It's a tough back to follow, but it's a good start. AI, good or bad. What's your strategy? What are you investing in? What are you seeing? As you can imagine, 80 or 90% of what we're doing is AI. In truth, it's hard to imagine a credible founder

Starting point is 01:50:32 that isn't leading with it at the moment. And so we're viewing everything. I mean, in your email that you sent out today about Samsung, about some of the effects of the what's going on in the Middle East, we're even looking at that through the ends of AI and how it should affect our investing strategy. Sure. And then in terms of the portfolio founders that you're talking to,

Starting point is 01:50:52 how interesting is the sovereign AI efforts? How interesting is just finding amazing entrepreneurs that are going to run through walls all over the globe? and they just happen to be there, so you're the first point of contact. You meet them early versus maybe going to an American entrepreneur that has some traction, you're going to help them, if you join the board, you're going to help them go global. What are you thinking? Yeah, absolutely.

Starting point is 01:51:14 So one of the things we're really proud of is that we actually are sort of a global firm, primarily the US and Europe. And so we offer founders soft landings in Europe, if they're American companies, and vice versa. So that's something that works well. But in truth, we're finding we're meeting more and more technical founders. and Europe has an edge on that. To give you an idea, I think 35% of the world's AI researchers or master's programs are actually in Europe.

Starting point is 01:51:40 We've got four of the top technical universities globally in Oxford, Cambridge, Imperial, ETH, Zurich. And so there is this incredible sort of talent building up. So historically, I think Europe's bottleneck was probably human capital and financial capital. The financial capital is now global. Your guy shows global. now the human capital is really growing.

Starting point is 01:52:02 And we're seeing a real multiplier effect in two ways. So the first is the very best founders are starting companies over and over again. So we have investments in, for example, Sneaks founder, Guy Bejani, who has now done Tesla. Most of our European founders are actually repeat founders. And the second thing is we have an equivalent of the sort of PayPal Mafia happening. So we were early investors in GoCardless. We've now seen other companies coming through as a result of that. like Monzo.

Starting point is 01:52:31 And I think you're just keeping that non-fly effect. Got it. Yeah. Does Europe broadly give England enough credit for DeepMind? Because it got rolled into Google so quickly. But I feel like the UK punches way above its weight in terms of AI research with DeepMind. And maybe that's under-discussed? Oh, I think it's under-discust.

Starting point is 01:52:55 And I think it's easy. People will often post-rationalize it and say how, great it would have been if it had stayed independent. But we've got to remember, when Deep Mind was really going in 2012, it was really hard for people to see. The 2017 Transformer paper had not been written for five years. It's early days. And so it's easy to look back and say it was obvious. It wasn't at the time. But I think the person that single-handedly has done as much for tech in the UK as anyone else is Demis or Sabis, the founder, he insisted on keeping a base here because he knew the technical talent was here. And now we're seeing that kind of multiplier effect. So there's great

Starting point is 01:53:33 Neo Labs being funded right now. You've got reinforcement learning companies like recursive superintelligence. Like Tim Rock Tashow is based here. Richard Sosha and Timo from the West Coast. You've got ineffable, Dave Silver's new company. You've got world model companies. And these are all coming out of GDM. So you're getting this multiplier effect. They deserve huge credit for that. For every one safety. Yeah, yeah. Yeah, when you look at a NeoLab, it feels like there's a thesis where it's just, okay, you have a brilliant technical mind. They're going to go explore. Yeah, the price might be high, but you're underwriting it as, you know, a venture style bet. There's a chance that something great comes. But do you have, at least from conversations, do you have an idea of how the Neo Labs might plug into the broader AI ecosystem, either through partnerships with big labs, or are you talking to Neo Labs that are saying we can leapfrog on certain vectors or maybe we can launch our own consumer product that's happened before. How do you think about where the Neo Labs fit in post the research phase?

Starting point is 01:54:43 Yeah, one of the questions we've been asking ourselves is the current S curve of technology that we're on, perhaps the third phase of large language models and diffusion models, like the existing companies are doing very well and they're often chasing benchmarks. you become what you measure. And so they're really driving fantastically into that. The question is, what are the other S curves of technology that could be explored? And the two that I think are really interesting, are world model companies at the moment. So AMI Labs just got funded.

Starting point is 01:55:11 Our portfolio company, Odyssey, is going in that direction from diffusion. And then reinforcement learning, we think there's huge work can be done. So if novel breakthroughs can be made on either of those, we think that they actually can be complementing to the, existing companies and they could work in addition. So very rarely they may go full stack and create their own product. Often they will actually service their unique intelligence through API. Yeah, that makes sense. Predictions around the next breakout prosumer products from Europe. We've seen the lovables, the granolas, anything that is like very on the radar in London,

Starting point is 01:55:51 everybody's talking about but hasn't necessarily broken containment and gone super global yet. I mean, one that everyone's really excited about, but it's partly we just exited the business to Apple is. People keep asking me what Q is. So Q is a Israeli company, second biggest acquisition by Apple. It was in stealth until the acquisition, right? Correct. So I was on the board for the whole of that period. We led the seed and did the same. Series A with our friends at Kleiner, Aleph and Spark, and that company is going to do something very special. A few years ago, we had a thesis that actually voice is really interesting because you can communicate in voice at about 150 words per minute. It's high input information because

Starting point is 01:56:40 of intonation, etc. versus typing, which might be 90 words a minute. We invested in Neurilink, which is like the very invasive version of high throughput. And then we started exploring what of private ways to communicate by voice. And Q is that company. So I'm asked about that a lot at the moment. The other one I'm excited about from a consumer perspective is an investment in nothing. Any ideas or guesses? It's probably not your information to share.

Starting point is 01:57:07 But, like, you know, Apple is making, you know, huge push and effort right now. They need to show the world that they still got it on the software side is Q something that, you know, Apple, you know, you know, iPhone users will get to experience in 2027? Is it the longer term? Will they ever kind of like, well, do you think there'll be like a moment where they're like, wow, this is, you know, a huge? Yeah. I think it's going to be a wow moment and I think it'll be in 2028. And, you know, our belief is actually that if you believe the software, the models actually kind of overshoot requirements, then a lot of the value will actually accrue to the physical devices because it'll be the distribution point. So we, for example, invested in nothing.

Starting point is 01:57:53 They just sent me their new products, which I'm going to unbox after this. But they, incredible smartphones, we believe that actually the value accrues to the distribution and so whoever owns those smartphones. Yeah, Carl being in the position of, you know, they've proven they can make beautiful products, but then also being able to be really flexible and quick around implementing AI all the way down to the hardware level. It's a very cool position to be in. Yeah, it's fun time.

Starting point is 01:58:22 Well, thank you so much for taking me. You brought the energy. You brought the energy. My apologies, but I also brought the technical issues. So apologies, next time. It happens. Well, thank you so much. That was great to meet you, Tom.

Starting point is 01:58:33 Come back on soon. Good see you both. Take care. Have a great day. Thanks. Let me tell you about Figma. No matter where your idea starts, Figma May, Claude Code,

Starting point is 01:58:40 code, or a sketch, the Figma canvas is where ideas connect and products take shape, build in the right direction with Figma. And without further ado, we do actually have Let's play. But what do you want to play? Let's play if we have time before the Cubinator joins.

Starting point is 01:58:58 We're working on some. We're working on it. I don't think we have time. I'll try to. Let's rotate that screen and I think we're almost ready. The one small news article here that we can talk about is that META has signed a 10-year lease for a 15,000 square foot townhouse on 697 5th Avenue to open Meta-L Lab, New York. It's first Manhattan flagship retail location. They painted it completely blue, apparently. The store will focus on hands-on demos and meta-AI glasses and VR headsets.

Starting point is 01:59:30 So they're getting into the retail space. Well, let's figure out... Still working on it. I can tell you about WayShirt Capital. Okay. Tell me how this. They invested in SpaceX at apparently 200... Okay. It's fake news. brutal. That was quick. Brutal. I was like this seems too good to be true.

Starting point is 01:59:56 Pitch book is, but we can tell you about Rivian robotaxies. Oh, yes. Uber and Rivian have announced a deal. Uber is going to invest $1.1.1.1.5 billion in Rivian and deploy up to 50,000 R2 robo taxis. I'm so interested to see what Rivian can actually do on the robo taxi side. I have friends that have owned Rivian.

Starting point is 02:00:20 have said the autonomous driving is fantastic. It just feels like this. I will be very interested to see if more than a couple companies can really crack it quickly. But it's a young company, very agile, and lots of opportunity. Well, I believe we have Mark Cuban in the Restream waiting room. Let's bring him in to the TVPN Ultram.

Starting point is 02:00:46 Mark, can you hear us? Yes, but you're too loud. Hold on one second. Oh, too loud. Too loud. Okay. Whisper. No, because I'm my ear.

Starting point is 02:00:54 It's great to have you. It's great. Thank you so much. Hey, guys. Hey, guys. Okay. Oh, well, I'll deal with it. Good to see you.

Starting point is 02:01:02 How's your 2026 going? We haven't talked this year yet. What's life like? I'm loving life. I got no complaints whatsoever. Yeah. That's amazing. Yeah, I'm loving life.

Starting point is 02:01:11 That's great. Are you, so you're not disappointed about the rollout of ads in LLMs thus far that? Is that safe to set? your year? Because I saw the first ad. It was for the Wall Street Journal, and it was just a little little bubble at the top. Hey, you might want to take out the journal. Seemed innocent enough to me. But how are you feeling? I haven't seen it at all. So it hasn't ruined my life at all. Okay. What's your information diet? Because it's hard not to log on the internet and not start black pilling these days. Yeah, no shit. So my first stop, my first stop is a site called

Starting point is 02:01:44 MimiO Random, which kind of gives me an update on all the what's happening in the world. My second is Drudge Report because that gives me the hyperbole on everything that's happening in the world. And then after that, all the different newsletters and emails that I get that try to keep me up. How are you processing the flood of cold emails that appears to be thoughtful, but is AI generated? Because you are notorious for your response rates and getting back to so many people that have reached out. But it feels like maybe an impossible task now. No, I do what everybody else does.

Starting point is 02:02:24 I bought a Mac Mini. You did. Yeah, for sure. And I'm still learning. So you're just like, you hit me with AI. I'll hit you with AI back right away. Right back, right? Because it's not even like the cold emails,

Starting point is 02:02:38 because that's pretty obvious. Yeah. You know, it's pretty easy to see. It's people subscribing me to shit. And, you know, the good news is email has an unsubscribe button. So you just got to train it to hit the unsubscribe button. Then I just review it and all that shit. So it's still a work in progress, but at least I have a path.

Starting point is 02:02:56 Well, the issue with us is that historically, if you had a podcast and somebody wrote you an email and said, hey, I really appreciated this moment where you're talking about this one thing. Totally. You're like, oh, they actually listen to the show. And then you could actually tell like, hey, at least they press play and at least they found a moment. But now AI just does it instantly. So there's no way to clock whether somebody actually. Yeah. And that's okay, right?

Starting point is 02:03:19 Because they're going to, the response rates most likely will be so low. We're in that trial and error phase where people are like, we're going to try it, see what happens. You know, maybe we'll get lucky and then they'll get bored and then it'll drop off. Yeah. Is owning a Mac Mini a green flag for entrepreneurs these days? Talk to me about what you're seeing in early stage startups in this AI era. Like where are the interesting builders? What patterns are you seeing that are like, oh, I didn't think that this person would be going down

Starting point is 02:03:47 the founder road, but they are now. Yeah, agents for everything. Yeah. You know, it's just because once you figure out how to do agents, then you can do them a little better than most other people. And then you can turn that into what would have been a SaaS business in the past is now, you know, we'll create your own marketing team and we'll, you know, do all these different things for you that you no longer have to do and we'll charge you X number of dollars a month. That's it. And I'm seeing dozens of those, you know, typically one for every industry you can ever possibly imagine. And are they growing revenue faster? Are they growing profit faster? Neither. They're just still trying to get some, just trying to get some traction at all.

Starting point is 02:04:26 Yeah. You know, because if you're growing revenue quickly, you're probably not coming to me yet. Okay. You know, you're, because, you know, the marginal cost to start is so low and it's so fast, and you're using the agent. So if they can get anybody to give them a credit card, you know, or sign up or, you know, now there's a little bit of a battle to use, USC for payment, the payment rails, you know, to make it so that, you know, just give me your wallet. And that's, you know, it's going to end up being a scam. And a lot of people are going to get ripped off there. But I think the real thing right now is agents for verticals and trying to turn that into

Starting point is 02:05:04 replace all your employees. So you can start up or you can cut costs. Do you think any of those agent-focused sort of like niche, at least niche to start businesses, would be a fit for Shark Tank? Yes and no. Yes, they should be. No, people won't understand them. And the other sharks wouldn't understand them.

Starting point is 02:05:23 Sure. But yeah, I mean, effectively, anything you can do. But you only need one shark to understand. Yeah. Yeah, right. And I'm not on the show anymore. It's time to go back. It's time to go back.

Starting point is 02:05:33 No chance. No chance. Yeah, what was the anatomy of, I think we talk about what makes for a great company all day long. We'll talk about that throughout the show today. But what makes for a company that will put on a particularly captivating Shark Tank appearance? You got to remember, it's for TV. You have to be entertaining.

Starting point is 02:05:56 And if it's not entertaining in some shape or form, it's just not going to work regardless of the quality of the business. If you don't have a charisma, you don't have a compelling pitch that's entertaining, it doesn't matter. You could be selling dollar bills for 50 cents and it would fail. Interesting. How important is like the visual component? There's a lot of physical products, but at a certain point, it gets too big for the studio. How important is it like the physical product presentation? Well, the good news is that producers will work with you on that.

Starting point is 02:06:26 And they make them practice over and over and over. And of the hundreds, if not thousands of pitches I saw in 15 years, we only had one really just choke, right, where they couldn't spit it out, maybe two, which is amazing. And it's a testament to Mindy and the producers there, how hard they prepare them. Yeah. How do you think Shark Tank and shows like it will change in the era of AI, video generation, you know, endless content? Is it stronger than ever because it's a known brand?

Starting point is 02:07:05 Or is there some weakness there? How do you see that playing out? It all depends on platform. It's like you guys, right? You know, it really just depends on reach of the platform and quality of the product. Sure. I think for Shark Tank, it's not going to have to change because of AI or technology, simply because you're really communicating to a family audience.

Starting point is 02:07:25 And the message you're communicating isn't, hey, here's a bunch of businesses that are great. The message you're communicating is that could be you on the carpet. The American Dream is alive and well. Sure. And so that's really what makes Shark Tank successful, not the quality of the business. Yeah, what about sports? I've seen some robots playing tennis. They're going to be playing basketball soon. I don't think I'll be watching robot basketball. But how do you think live events, sports, basketball will change over the next decade? I mean, maybe for the referees like you've seen in tennis, but that's it. I think in reality, more people will want to go to in real life events than before. Because if, you know, if you're just managing agents, looking at output, looking for exceptions, You're going to want some human touch, right? You're going to want to be able to engage.

Starting point is 02:08:15 And I think that's really, really important. And I think that's where sports will grow. I mean, you're starting to see that now with, you know, what happened with the Olympics, the World Baseball Classic, you know, became much more popular. Because people want that disengagement from all the stress that's happening right now. Yeah. At the same time, it feels like there's almost an opportunity for not to bring it back to targeted advertising.

Starting point is 02:08:40 but AI can tell me, okay, my favorite team's in town. I should go to this particular game. I should remind me at the right time instead of just signing up for the newsletter. Yeah, that's not really AI. Yeah, that's just targeting. Yeah, that's just targeting, right? And I'm going to tell you what, in terms of, I'm going to take it down a different path because I'm contrarian on this.

Starting point is 02:09:00 And that's with robotics. I think everybody's making this push for humanoid robots. I think they might have a five-year lifespan and then they'll fail miserably, maybe 10. Yeah. You mean the companies or the device or the individual hardware? Or the both. Both.

Starting point is 02:09:17 Both. Right? Because I think everybody defaults to, well, we live in a human world and humanoid will take the place of humans for various functions, particularly in the home. And I think there's just no chance. I think if you look at warehouses and what Amazon does, they're not humanoid robots carrying boxes, the robots that are designed to fit the environment. And I think, you know, I've heard people say, well, a house is a house.

Starting point is 02:09:45 You need a humanoid. I think houses are going to be redesigned completely so that whatever the optimal robot is that allows it to simplify the house, that's where houses will go. So I'll give you an example. If we had robots that looked more like spiders that could, you know, but had hand, you know, the ability to. carry and lift things, more like ants, I guess, maybe.

Starting point is 02:10:12 Right? But you could create a house where the pantry and the refrigerator and the washing machines were hidden behind the garage, if you even have a garage. And that way you could redesign it. So all the living space was for people

Starting point is 02:10:29 because you know that the robots aren't going to be full-form humanoids. They're going to be whatever the optimal shape is and they're kind of co-designed. You design the house to fit the robot, you design the robot to fit the house. And I think that way you could go a lot less expensively on both. You know, the humanoid founders will tell you,

Starting point is 02:10:51 you know, how do you solve stairs, right? Like, it doesn't work for wheels. But if the robots are really great, people, you could put like a mini robot elevator, right? Like if it's on wheels, like you can just put it. Yeah. Like, you see like the old. the house, dumbwaiters, right?

Starting point is 02:11:08 Where there's just the thing where you pull it, you put it in there, and you pull it up, and it goes up to the next floor and somebody opens it up, you're going to see, you know, a mechanized equivalent to that, right? Where it recognizes the little, you know, ant robot that's coming up, and it opens up a little door that leads to the size of whatever it is, it needs to carry or whatever. And then it goes up the dumbwaiter, does this thing on the next floor, the next floor of the next floor and does what it needs to do. I don't think, you know, stairs are an issue at all. Yeah. How do you think about just these types of ideas, AI products, generally getting

Starting point is 02:11:50 rolled out and then hitting it, bumping up against like human guard rails? Like when I hear that, I think like that sounds incredibly sci-fi and potentially possible from an engineering perspective, but then, you know, you try and, you know, remodel your house and then you're stuck in permitting for two years and so that tends to slow the progress down a little bit. Is that a real technology? Okay. Yeah, of course it is, but that's all technology during the interim period, right? Sure.

Starting point is 02:12:16 There's always a transitional period where you go from the old to the new. Like back in the day, you know, there wasn't enough electrical outlets for your PC. There wasn't enough, you know, you had to go into the walls to run all the Ethernet cables and all that shit, right? You know, and so the houses and offices weren't to for that because it wasn't considered when they were built, but they adapted. You found ways to adapt, and it'll be the same thing with homes. It'll be the same thing with offices. We'll find ways to adapt. I think, you know, the biggest challenge going forward is going to be, as we go from an

Starting point is 02:12:51 LLM world to a worldview world, a worldview world of AI, where, you know, we're taking in video and learning from the video and extracting rules from the video, a lot of the things that we're going to do are going to be outside and are going to be going to have to consume in the interim, at least, either satellite bandwidth or 5G bandwidth. And I don't think there's going to be enough bandwidth when you're working with video-based AI models. Interesting. You mentioned maybe a garage not existing in the future. Is that a way to say that you're excited about self-driving cars? What are you, what are you thinking is going to happen there? You know, I played around. I have a Tesla and I upgraded. for a couple months, and it terrified

Starting point is 02:13:37 the shit out of me. Not that I didn't trust it. Oh my God, because when you're going 25 miles an hour, it's no big deal. You go in the highway and you're going 70 miles an hour and there's a median right there in the middle of the highway. I was like shaking.

Starting point is 02:13:53 I was like, you know, Elon's cool or whatever, but you know, I ain't trusting them that much, right? You know, and I'm not waiting for the... You know about Mad Max mode? No, no, no, no, no, no, no, no. But you set a delimiter where it's like how much above the speed limit are you willing to go? And if the speed limit is 65 or 70, you know, you've got to go the speed limit.

Starting point is 02:14:15 And it was scary as fuck trying to go 70 miles an hour. And I don't want to be there when somebody paints some adversarial, you know how there's graffiti in the weirdest places? Wait until there's adversarial graffiti. Oh, yeah. Yeah, like somebody, if somebody paint, if they put it up on the brick wall, It's road running. And when they paint it to look like the road, road running. Wiley Coyote is going to.

Starting point is 02:14:37 No, that's like ridiculous shit, right? Wiley Coyle Coy is going to paint the tunnel and then you're slam into it. There's somebody somewhere trying to figure out how to fuck up self-driving mode, right? For sure. And because it's just taking video input. Yeah. And, you know, what, it could be some pattern. And all of a sudden you're seeing this pattern on medians or, you know, overlaid,

Starting point is 02:15:02 on stop signs or whatever. Yeah. You know, because, you know, somebody's got to do that because it's just too easy not to. Yeah. Hyper-realistic camo wrap gets confused, you know, eventually. Whatever, right? You wrap your car in camo, and if the camouflage is effective, you're going to confuse the AIs. And you just might, and there could be a predator, right?

Starting point is 02:15:25 Alien versus Predator. Yeah. Predator could show up, right? It's possible. And if Arnold isn't there to save us. It's possible. I mean, speaking of Arnold, are you, it sounds like you're in a good mood, you're optimistic about things. Does the question of AI Doom come into your mind, these runaway robotics, are you worried about that at all?

Starting point is 02:15:45 No, not even the least, but for the reason I just mentioned, right? In order, like right now, LLMs are basically bimodal with some video, right? where it's almost all text and pictures with some video. You can't model the world with that. You just can't. AIs right now doesn't understand the consequences of its recommendations. It has no idea what happens next. A two-year-old kid with a high chair and a sippy cup knows if it pushes the sippy cup over the off the high chair.

Starting point is 02:16:16 Mom's coming running and the kid's going to start laughing at mom, right? Large language models don't understand. Every time. Every time. Every time, right? And there's no large language model ever, right? Ever. And it's hysterical, you know, unless your mom, right?

Starting point is 02:16:34 But you get the point, right? The large language models we have today can't do that. And so we have to evolve to models that can capture the world and physics and deal with the latency of capturing or not having access to video that you can't see. and so you have to try to model that. And not only does that take up a lot of processing power, but it takes up a lot of bandwidth, as I mentioned before. And so the Terminator is taking over.

Starting point is 02:17:04 I just don't see how it's going to happen. Now, you can have localized brains for military applications and power get better and manual dexterity will get better, all that. But that's not going to allow you to take over the world, right? That's going to be application specific. So I'm not afraid of that at all. I also think that we're talking about agentic applications. I think particularly for small, medium-sized businesses and some large businesses,

Starting point is 02:17:33 they're not going to have that skill set. It's not going to be natural for them to do that. And I think kids coming out of school today that have taken some Python, don't have to be comp side majors, but have done agentic AI. Like when I go talk to schools, that's what I tell them. You know, get into Claude, you know, teach yourself all the agentic stuff. And then go to small businesses because they're not going to understand how to do any of at all. Yeah. What advice are you giving to friends, portfolio companies, etc., around navigating as a

Starting point is 02:18:05 business leader during a time where we have major global conflict? I don't know what exactly you were working on in 2002, 2003, 2004, but there's so much, I mean, right now, everyone's hoping for a quick end to the conflict, but it's hard to lean. on that? You know, it's funny. In 2002, when we attacked Iraq, I created something called

Starting point is 02:18:32 the Fallen Patriot Fund, which, you know, was just money available that I funded myself, money available for the families of soldiers who didn't return or soldiers that were, you know, horrifically,

Starting point is 02:18:46 um, injured or disfigured or whatever it may be. And we paid out millions of dollars, but the bigger point was, the way the media world was, and we kind of just trusted what was presented to us by the gatekeepers, right? You could have an opinion, whether it was right or wrong,

Starting point is 02:19:02 but hey, there are WMDs, right? Weapons of mass destruction, and we kind of trusted. Now there's so many information sources and social media, and we really only consume what the algorithms show us, and each one of us has a different algorithm. Like the three of us are out. algorithms are like fingerprints. No two are alike.

Starting point is 02:19:26 And because of that, everybody's got a different perspective on what's going on in Iran and what's happening around the world. And to me, that's scary, right? It's hard to know what's real and who to trust and now with AI, video, you know, what's been created.

Starting point is 02:19:43 And we really are in across our fingers and hope things work out for the best because I don't know that there's a way for anybody to really participate in decision or make a good decision. Yeah, we're all trying to predict the future together, but based on

Starting point is 02:20:02 different, well, based on wildly different kind of influences. Exactly. You know, we don't have access to the information we truly would need in order to make a cogent decision or even have a decent opinion. I mean, we just don't. And we spend more time trying to filter

Starting point is 02:20:21 and determine what's real and what's not so that it's almost impossible to really do anything but just hope and pray we have a couple questions from the chat the first one is about cost plus drugs can you give us an update there how is it crushing it crushing it's this yeah rush it does fantastic like give us a sense of scale give us a reminder of the strategy reintroduce the company sure so what kind you go to cost plus drugs.com and you put in the name of the medication. If it's one of the thousands we carry, then we actually show you our actual cost. Then we show you that our markup is only 15%. And we charge you $5 for shipping and handling, and then the credit card fee. And in doing so with only a 15% margin, unless it's like a $4 Walmart drug, we're almost always the cheapest option

Starting point is 02:21:19 for anybody. So if you're underinsured, if you don't have insurance, you know, even to compare it against your co-pay or co-insurance, we may be cheaper than your co-pay. And because of that, our business is just skyrocketing. So that's part one of our business. Part two to our business, for my company, for my employees and their families, I went around and talked to the CEOs and CFOs of a lot of hospitals and found out where they were getting ripped off by the insurance companies. If you think about this, and I don't think many people do, whatever your deductible is, if something happens to you and you can't afford it, even if you have great insurance, you might have a $1,500 or $2,500 deductible, which is big company good insurance,

Starting point is 02:22:06 but if you don't have that money and 40% of people don't have $400 for an emergency, when you go to the hospital, as an example, they literally end up loaning you the money. And that, as a result, we've turned hospitals and providers into subprime lenders. Think about that. Yeah. Right. And then you have the denials and then they underpay, lay play callback. So anyways, so I went to local hospital, Beller Scott and White, who's a really forward-thinking hospital system.

Starting point is 02:22:35 And I said, look, I understand where you're getting ripped off by the insurance carriers. I'll pay you on time. I'll pay you what we committed. I won't claw back. I won't lay pay. In exchange, I want two things. I want a better price. I want it as a reference price of Medicare, 100 to 100% of Medicare.

Starting point is 02:22:52 unless it's really complicated. And more importantly, we created a site called Costplus Wellness.com. And we are going to post this contract on cost plus wellness.com so that any business, you guys, TBPN, anybody any size

Starting point is 02:23:10 that wants to direct contract can reach out to Baylor Scott White and get the same pricing that we get. And it's just blown up. I mean, it's just incredible. We have more than 9,000 providers. and what we're trying to do is teach companies who self-insure in particular that they can take control of their expenses. You don't need to be dependent on the insurance company to come up with the right deal because they won't.

Starting point is 02:23:36 They'll steal from you. Did you ever, I mean, it's such an interesting company for you because I feel like when you have as big of a presence as you do, it could be a book or a course. or a protein powder. Did you look at anything else? Yeah. Or have you burned out on that stuff? Protein. Protein powder.

Starting point is 02:23:59 The Cubanator protein stack. I'm in the, but no, this is obviously much more, much more important. Yeah. I just thought, you know, nobody looks at health care and says, you know, the economic side is great. Yeah. We're doing it the right way in this country. It's the exact opposite.

Starting point is 02:24:15 And so if you're going to try to disrupt, go big or go home, right? Another question from the chat. What's the most? underrated business you've seen in your career. I think they're talking about something that like you, it is the moment you saw it clicked and you were like, okay, this is like a wildly mispriced asset or something that could really fly.

Starting point is 02:24:35 Streaming. Yeah, we called Internet Broadcasting. I sat down with a buddy of mine, 1995 at a California pizza kitchen, and he was like, how can we listen to Indiana basketball in Dallas, Texas? And this is right when the internet just started, right? It was brand new to everybody.

Starting point is 02:24:52 And I'm like, let me figure it out. And so we started a company called AudioNet and got the rights to, you know, hundreds of schools, hundreds of radio stations, TV stations. You know, back then the copyright laws were different, created our own internet jukebox and unlived a number of internet radio stations and started streaming until we sold it to Yahoo. That was the most obvious thing I'd ever seen in my career. What was the domain name negotiation like? Jordy's a big fan of great domain names. Great question. Great, great, great question.

Starting point is 02:25:25 So when we started, it was AudioNet, and I just registered it. Nobody had it. Yeah. But then-Odiomnet.com or audio.net. No, audionet.com. Okay. I like it, yeah. Because we were just doing audio in 1995.

Starting point is 02:25:38 Yeah. And then by 97, we started to do video, and AudioNet wasn't going to cut it. No. And so I found Broadcast.com because we wanted to broadcast everything and anything. and found the guy and paid him $8,000, and he was thrilled to get the $8,000. Wow. Yeah, this is 1997.

Starting point is 02:25:55 Did he ping you after that? Did he ping you after? Yes, he did. Yes, he did. But wait, it gets better. Well, wait, there's more, right? And so I'm like, oh, shit, this is nothing, right? It's an automatic traffic generating.

Starting point is 02:26:08 And so I started going out there and just glombing up and just grabbing all kinds of URLs so that we could put content. on them and then drive it back to broadcast.com. So literally, I own Final4.com. I own baseball.com. I own sandwich.com. You name it. I bought it.

Starting point is 02:26:30 I bought it for, I would buy like just packages of URLs, right? And this is because people were just going to their browser and being like, sandwich.com. And they would type it in. Yes, exactly. Exactly. Exactly.

Starting point is 02:26:45 Exactly. Everything was a portal, right? everything was a front door. And so I was like anything that generated traffic. And I've done it since. Like I own Mr. President.com. I own democracy.com. I mean, all kinds of...

Starting point is 02:27:00 He privatized democracy. He checked democracy private. I was worried. That's the most American thing I've ever heard. I love that. That's incredible. Okay. The last question for the chat and we'll let you get back to your busy day.

Starting point is 02:27:13 I want to flip it around. What's a business that you've seen your career? that you wanted it to work so well, but for whatever reason, the business just didn't achieve what you had in mind. And why? Yeah. And you don't need to be specific about this particular company. I mean like a technology or maybe an anonymized company, something like that. Yeah.

Starting point is 02:27:37 God, I'd have to think about it. You know, what was not the motorized skateboards, they weren't called it with the two wheels. Hoverboards. Hoverboards. Yeah. Hoverboards. So I agree. I want the hoverboard future.

Starting point is 02:27:48 Everyone traveling on hoverboards ever. Yes. So a buddy of mine, his son was an engineer, and I was like, okay, this kid could try to come up with some new ways to do hoverboards, make them safer, etc. And so we started a company that did hoverboards. And there were so many more patents already in place than I ever imagined. We couldn't get past them. And it failed miserably. Yeah, yeah, yeah.

Starting point is 02:28:11 That was a very interesting boom, the hoverboard boom. It sort of came out of Schengen fully formed. because there was a massive supply chain and they were all over, but there was no one brand. They were like a ton of different brands because really what was going on was there was one amazing supplier in China that had like 20 different companies

Starting point is 02:28:28 that were reselling it all. And they were making a killing. Oh yeah, they were. Oh yeah. Yeah, what, is it still possible to create a widget and make like $100 million from it or does the cot or do the clones come? Because I know the guy who made

Starting point is 02:28:44 like the, the fidget spinner, like his claim to fame. Right, right, right. That's cool. Yeah. But, but he didn't. It got knocked off like that. Yeah, I mean, it was, it was a kind of thing that, like, was a hit product. That's all on Amazon. Okay.

Starting point is 02:28:56 All on Amazon. So I started talking some Amazon resellers like mid-24 because I was just curious about some things. I'd see some things on X. And as it turns out, if you're an American seller, it may have changed. So correct me if I'm wrong. If you're an American seller, you can have one company. that sells on Amazon, right? But if you're Chinese, there's no limit.

Starting point is 02:29:21 And you don't even have to have a nexus. So if you're at that American company and you're making sales and making money, then you have to pay taxes and define your nexus and, you know. Oh, so you're just screwed because you're- You're screwed. Yeah. Because so these Chinese companies, to this day, as far as I know, these Chinese companies don't have to have an nexus, don't pay the taxes,

Starting point is 02:29:43 even though they're supposed to, right? you can literally have a Chinese bank account and Amazon will send the money right to that Chinese bank account. And I was proposing to these guys and talking to some legislators at the time that Chinese companies should have to post a bond before they can sell the product and post it on a website that whatever, whatever.gov so that, you know, the fidget spinner guy could come in and say, you know, we're have an agent now that continually continuously checks to see if there's a knockoff with their product and then can challenge it. And then at least there's that $10,000 or $25,000 bond that offsets the risk for that fidget spinner. I know one widget company that bought the next five most popular widget companies in the category that we're knocking them off. And they just continue to operate them. But they have enough ranking on Amazon and they have the scale. But it's just wrong.

Starting point is 02:30:39 It's just wrong that, yeah, because any, whether it's China or any country, If you're outside the United States, you immediately have a cost advantage, not the manufacturing, but just from an IP and from an Amazon cost perspective. Why in the world is it cheaper for a Chinese or Vietnamese company to sell on Amazon and to easily knock off than it is for an American company to sell the original product? That makes no sense. And yeah, and legislatively, you could fix it in a heartbeat. You got to post a bond.

Starting point is 02:31:12 $25,000 bond, depending on the business. size of the market maybe more and then give everybody 90 days to check it and all of a sudden the whole industry changes and American manufacturing skyrockets and the because that that cost of knockoff isn't just about the cost of losing sales it's the administrative the legal cost that there's just so many nuanced things that you have to spend money on yeah we have a we have a knock we have knock off issues and like we we spend thousands of dollars to have our lawyer like chase them down and send takedown requests. From our merch, like just T-shirts and stuff.

Starting point is 02:31:50 Yes. Oh, yeah. Merch is crazy. And then IP too, right? All the DMCA take down notices because they're just scraping and, you know, reposting, all that shit, right? It's easy to fix if, you know, someone has the guts to do it. What is the anatomy of using your likeness once you've made an investment?

Starting point is 02:32:10 What does the best relationship look like? I imagine it's very open and transparent, but I imagine that anyone who's been associated with you at all is trying to, like, slap your face next to their product and, like, pump it all over. And maybe you haven't invested yet. And you just said, oh, it looks nice. And then they're like, clip it. He said it looks nice, you know. Yeah, I mean, it depends on the company. You know, usually I don't even care.

Starting point is 02:32:34 But two things. One, you know what Synthesia is? Synthesia.com. Yeah. Yeah. I think so. Yeah, they've been on the show. Oh, yeah.

Starting point is 02:32:41 Yeah. Yeah. They have the avatars. They have Victor and all those guys. Well, I was their first investor, so I send them there. Okay, yeah, yeah. Your dog, that's a unicorn. Come on now.

Starting point is 02:32:55 And I gave them a lot of money. I gave them a lot of money. And this was 10, 12 years ago, way ahead of the curve. Yeah. There we go. Okay, so Synthesia. Yes, and so it's easier. So I'll push them to them.

Starting point is 02:33:14 Or like I'll just fuck around. Like you saw it was Sora. They had, so I just, I put one picture of me out there. But I was playing with it because I want to learn all this stuff. And they, Sora had this thing where you can put conditions on how, when it can be used. And so I made the, yeah, so I made a condition. I made a condition so that at the end of every video that used my likeness, it showed the logo for cost plus drives.

Starting point is 02:33:42 So smart. So smart. And it's been used like hundreds of thousands of times and I know we've seen a bump as a result. So smart. John did the less commercial thing. He said, portray me as a bodybuilder. It's funny.

Starting point is 02:33:54 It's a lot less commercial. But of course, the source is kind of falling behind now. So they kind of, I don't know if the opportunity is gone. No, it's good. It's going to, it'll just get added into chat chit.

Starting point is 02:34:05 And then you've got a billion people just, just pump and cost plus drugs. Yeah. It's always, it's always crazy to me to see it. Like I tell it, you know, because it has terms of service. You can't show drug use and done it. And so there's pictures of me like doing lines of coke and shit.

Starting point is 02:34:20 And I got, you know, so it's kind of crazy. Ridiculous. What, when is the right time for a company to apply to Shark Tank? Anytime. You just don't know. I mean, they have open auditions all the time. So if you go to Shark Tank's website, it'll give you all the information there. Yeah.

Starting point is 02:34:38 And you just got to go out there and have fun. Go out there and have fun. you processing the peptide boom both FDA approved non participant non participant I'm not a believer in that shit at all like every single LLM that I put it into and ask for you know show me the trials and show me the research you mean the non-fda approved just just the the stuff coming off the boat right or are you short Eli Lilly no no no no no the insulin like the real because when people talk peptides you're talking supplement stuff right yeah yeah yeah yeah Right.

Starting point is 02:35:12 But yeah, the ElyLYLY stuff. Yeah. Not, not OZMPIC that's running Super Bowl ads and like very heavily regulated. Yeah, that makes sense. No, because that stuff's going to come down in price.

Starting point is 02:35:22 And now, you know, Lilly and Novo are smart now with their GLP ones. Yeah. They're working around the PBMs and doing direct to patient, direct to company. And that was brilliant. That was really smart.

Starting point is 02:35:32 Yeah. Yeah, yeah, that's very cool. Well, thank you so much for taking the time. George, Yeah.

Starting point is 02:35:35 It's always fun. It's always fun. It's always fun. It's always fun. It's fun, guys. Enjoy the rest of your time. soon. Goodbye. Let me tell you about Shopify.

Starting point is 02:35:45 Shopify is the commerce platform that grows with your business and lets you sell in seconds, online, in store, on mobile, on social, on marketplaces, and now with AI agents. I think we got a future shark tank. Oh, yeah. I came up with the name. Oh, you did? I'm not going to say it. I'm not going to say it, but the domains available.

Starting point is 02:36:02 Okay, we're getting the domain. We're getting the domain. You're going to shark tank. You're going in the tank. I'm so excited for this. I'm so excited for this. Not you, Tyler. Not you, Tyler. Keep working on our other launch. Once, yeah.

Starting point is 02:36:15 All right, we got to hit the size gong for Roe Kana. Ro Kana. He's put up $609 million of trading volume while in Congress. He's fighting back against the elites by trading against them. I was talking to Jordy about this this morning and I was like, volume is hard because you can be a very high volume trader. But I was reflecting on like, there are years when I, just don't really trade that much.

Starting point is 02:36:43 Or the only volume is the buy. The buy and then you just hold or something. But he's putting up big numbers. But maybe he's running like, maybe he has like 10K in like a high frequency trading. Yeah, long short strategy. It's just, that's just eking out milliseconds of a tick. Yeah, maybe he's running, maybe he's got a stack of 40 open, open cloth. Yeah, maybe he's been poaching from Jane Street.

Starting point is 02:37:06 And so he's putting a lot of volume. This is just insane volume from the anti- elite champion. It is crazy. He said, you guys want to dance? I'm going to trade against you. How do you even have time to consider 37,000 trades? Like, he's, I mean, how many minutes are in a year? In a year. That is a lot of volume and a lot of trades. 500,000, 525,000 minutes in a year. So you can spend, you know, 20 minutes on each trade if you're working 24-7. That is... He's been in...

Starting point is 02:37:44 I'm sure there's something else going on. He's been in office for nine years. It's probably just like a investment manager. Yeah, he says it's his wife's money prior to the marriage. Well, he's averaging 11 trades a day since he went into office nine years ago, roughly. Yeah, and the stock ban legislation... And that's no days off. That's weekends, holidays.

Starting point is 02:38:06 I don't know. Let everybody trade. Why not? Yeah, sounds like he's locked in. all times. What else is going on? We're removing sanctions on Russian and Iranian oil, which is... Craziness.

Starting point is 02:38:25 Craziness. Didn't think that would happen. Let's watch the public latest ad. Is that what we should pull up next? Pull it up. Actually, let's pull up this video first of our president. Of Gary Tan. But this is Gary Tam when he sees a YC applicant using G-Stack.

Starting point is 02:38:41 to ship 500,000 lines of code daily. This is one of the best videos of all time. Just pulling out a lot of cash. I'm ready to invest. Brickly counting it. It is so funny how lines of code became like not important at all. They're the new eyeballs. And now it's new eyeballs.

Starting point is 02:38:56 They're the new eyeballs. Oh, the lines of code are the new eyeballs. I hope not. I was thinking that we should do a challenge here in the studio where every member of the team has to race to generate 10,000 lines of code. whoever can do it fastest. Okay, but how do you define a line of code? Like, I can just write a for loop that says...

Starting point is 02:39:17 I will run a four loop after the lines of code are generated, and I will pass every line of code individually through GPT 5.4 Pro, and I will ask it one question, does this count as a line of code? Okay, but can I have the same line of code 10,000 times? You've invented the for loop, potentially. Print this.

Starting point is 02:39:39 That would be the most efficient way to do it, potentially. Let's pull up this ad. It might even be faster to just clone a repo, right? Clone a 10,000 line repo. Do you have to write it from scratch? We need to write the ground. Download timing. Pull up the ad, sir.

Starting point is 02:39:53 Yeah, what are we watching here? Well, looking at your portfolio, you've got diverse equity exposure, broad market ETF, some fixed income. However, I am seeing a gap here. College basketball, baby. I'd recommend a three-leg party, maybe sprinkling a few player props, to even things out. I've got a strong read on an early upset.

Starting point is 02:40:16 If you're looking for a broker that's not also your bookie, we invite you to try public. It's a good ad. Really solidifying public's position as, you know, which has been very, very, very consistent. Yeah, very, very smart these days. Taking shots. I wonder if they're going to actually run that on TV for March Madness.

Starting point is 02:40:35 Would that be the right place to run it? I feel like March Maddennings viewers kind of want a bookie. Like they want to better on. Madness is just like kind of everyone. Yes. Yeah. Makes sense. Anyway, we have our start to the Lambda

Starting point is 02:40:48 lightning round. Let's take a look at the beautiful cloud. And let's tell you about Lambda. Lambda is the super intelligence cloud, building AI supercomputers for training in the print that scale from one GPU to hundreds of thousands. And we'll bring in our next guest. John Kim from Paraform.

Starting point is 02:41:07 How are we doing, John? Thank you so much for taking your time to come chat with us. Please introduce yourself in the company. Hey guys. One of the founders and CEO Pariform. We are a gentic hiring platform that makes hiring exceptional talent as easy as pressing a button. I love it.

Starting point is 02:41:23 So our first product. We got your button in the mail. We did. We got your button. We got a package from you guys. It was very well. It was a great, great execution. The only issue is the chocolate completely exploded everywhere.

Starting point is 02:41:35 It made it way more memorable, honestly. Yeah, yeah, yeah. I will never forget. Maybe there's alpha. There's deep alpha there. Anyway. But no, it was awesome. Tell me, thank you for the company.

Starting point is 02:41:45 So who is the customer? Who's paying for this? Is it large corporation, small companies, startups? Or do you have the applicant pay sometimes? How does the business model work? Yeah, yeah. So it's the company's paying. We started out helping startups, you know, build their founding teams.

Starting point is 02:42:02 Now we have sort of S&B, mid-market enterprise. So, you know, everyone from like a fast-growing startup all the way to, like, public companies like Palantir are our customers. Yeah. Wow. And you raised some money. How did it come together? Yeah. Yeah, we raised a $40 million round led by scale venture partners. I need to wait for the long. Congratulations. And yeah. Yeah. And help me understand how you're going to deploy capital to accelerate at this particular moment in time. Because I imagine that actually building the platform has gotten easy. but engineers are still expensive. You still have to do top of funnel work. There's SDRs to hire.

Starting point is 02:42:46 How are you thinking the shape of the business evolves over the next 12 to 18 months? Yeah, no, definitely. I mean, it might be just like a classic answer, but we obviously want to continue to do best in class growth. We grew a ton 10x star revenue last year. We want to continue to grow at a fast pace. So in order to do that, you know, we need to grow our team

Starting point is 02:43:08 and all that stuff. Yeah. And also accelerate our product roadmap, right? I think in particular, I'm personally spending a lot of time this year, you know, really just narrowing down, focusing on our product roadmap. So, yeah. Oh, one more thing I will mention is we sort of started in tech and helping tech companies hire, you know, EPD talent, sales, design, you name it.

Starting point is 02:43:31 Actually, we launched a new vertical. So we're helping law firms hire as well. You know, our goal is to be a universal hiring platform, not just. for tech companies. So obviously deploying capital going across industries as well. Yeah. Talk to me about the legal hiring market. Are you sourcing those people in a different place? Like what's different about that that requires investment and changing the and generalizing the platform? Yeah. Yeah. Walk me through that. Yeah. I mean, since we're a recruiter marketplace, you know, in order to scale to another industry, we basically need a sort of a new set of supply,

Starting point is 02:44:08 like legal recruiters. You know, actually a lot of a lateral attorney and partner hiring at law firms are driven by recruiting agencies. So the like boutique, you know, sort of heavy hitters like who do all the recruiting. So yeah, I guess to expand, we need a new set of supply. So we're doing that. How are you thinking about just areas of growth in the U.S. economy broadly? If you stack, rank, like there's a lot of energy around reindustrialization right now.

Starting point is 02:44:35 we were hearing that electricians might be the LeBron James of the next era. But that feels like, I don't know that electricians are on the internet the same way that software engineers are where they might have public GitHub profiles, blogs, LinkedIn profiles. How are you thinking about solving those next verticals? Yeah, no, I think definitely, like, defense in government seems to be a huge area of growth. Yeah, like you said, manufacturing, like anything that's sort of. sort of like atoms, not bits, I think it's like sort of all. So going to grow a ton. I actually think like travel entertainment, like those industries, media, you know, is going to do well as well. So yeah, we're looking at what sort of verticals to go after, like to your point based on what,

Starting point is 02:45:21 you know, what we think are exciting. I also read the anthropic report where they published sort of like what jobs are going to be like sort of replaced by AI, all that stuff. Oh, yeah. Yeah, that's spider chart and then under Carpathie posted something similar. Yeah, I mean, I think actually, like, you know, my point of view is that the economy is not like a zero sum game. Like it's actually like an abundance game. So I actually, you know, I'm not too worried about like AI replacing people. I mean, like sure, there's some like adjustment on like how we need to upskill ourselves and differentiate. But I think every technological revolution like humans cleared out of way.

Starting point is 02:45:57 Well, I mean, that's the interesting thing about that chart is that what's missing from that chart of things, jobs that will go away is like the new job. that will be created. Exactly. Like, live streamer, business technology news was not a thing when I was a kid. There's zero chance that I ever could have put that down as like, my future career will be live streamer. That wasn't the thing. What, in conversations with investors for this round, what were, like, what kind of exits

Starting point is 02:46:24 like were kind of reference in the recruiting space? I know there's like public recruiting firms, like, you know, these companies scale to like massive, massive revenue. You know, they're not, no, I'm talking about like non-tech companies that are just like we do staffing and recording. Oh, yeah, yeah, yeah. In different kind of verticals, but there's massive, massive companies in the space. Are you comping to those and saying like, hey, we, there's, we can be a billion dollar business, uh, just based on serving a similar kind of, uh, sector. Yeah. I think, um, not, I'm not too sure if that was like the focal point of all the conversations, but I think it's a great question nonetheless. Like, I think, well, yeah, and I just, I just, I just,

Starting point is 02:47:03 I just say that because there hasn't been like, like every company has to hire a bunch of people, but there's not like the face of recruiting. There's nothing that like there's not like a perfect comp where Figma's like, well, look, you have Adobe. Yeah. We're just going to like eat some percentage of Adobe. Yeah. I think the way I look at it is if you look at the total amount of like dollars spent on recruiting, broadly speaking, like where does it go to? And actually, the biggest spend category is external recruiting. So outsourcing recruiting, working with recruiting agencies, staffing agencies.

Starting point is 02:47:39 So like that's sort of the biggest spend. But actually, if you look at traditional like VC dollars over the last 10, 20 years, it all went to like HR software or recruiting software. And that category is actually not that big. It's like maybe 10 billion. I'm not sure exactly. But it's not the biggest category. Yet 90% VC dollars went in there.

Starting point is 02:47:59 I think there's a little bit of a, you know, I think mismatch there. But obviously the market we're going after is, you know, like, you know, labor market itself, right? I think I've heard somewhere also like we're shifting from paying for like software to paying for work. And instead of sort of building paying for tools, we're paying for outcomes. And I think Paraform is like very much aligned with that trend. So yeah, we're going after the biggest market in recruiting. Yeah. Amazing.

Starting point is 02:48:26 Well, thank you so much for taking the time to come chat with us. Congratulations on the new round. Great to get the other. Good luck. Thank you. Talk to you soon. Great to see you, John. Have a good one.

Starting point is 02:48:34 Cheers. And we will continue our Lambda Lightning Round with Eugene from Edra. He's the co-founder and CEO. You know some exciting news for us today. Eugene, how are you doing? Good.

Starting point is 02:48:47 How are you guys? Thank you for having. Exiting stealth. I love when companies exit stealth. Welcome to the show. Welcome to the public eye. Please introduce yourself in the company. My name is Eugene Alpes.

Starting point is 02:48:57 I'm the CEO. and co-founder of EDRA. With EDRA, really what we saw is that models are smart enough to do basically any work inside an enterprise, but the only problem is you have to tell them exactly what to do, and nobody has that, right? Like, you can't go to any company and tell them, just tell me exactly. I don't know what to do. How am I supposed to tell you what to do?

Starting point is 02:49:18 Yeah. So what we do is we built an agentic learning system that just hooks up to their existing systems of record, figures out what. their people are already doing, writes it out for them so they can see it, and then we use that to actually automate and do the work. Makes sense. So talk about what it means to, like, just hook up to their internal systems, because I feel like there's 20 tools in every category, and every company has 30 tools, and you multiply that

Starting point is 02:49:47 together. That's a lot of integrations. Writing new integrations is easy, but is there a platform that you can sit on top of? For sure. So we have a couple of core systems that are really good for us, so we do service. We do Service Now, Jira, Outlook, of course. Can't forget Outlook. Salesforce, Zendesk are some of the main ones.

Starting point is 02:50:04 Sure. That are kind of the first ones that we are working on top of. Yeah. And then I'm sure if there's a client that's big enough, that could maybe prioritize an integration or something. But how big are the companies that you're working with at this point? Yeah. So, I mean, our specialty is like the larger the company, the better it is. The messier the process, the more of a challenge they have.

Starting point is 02:50:23 So if you look at some of our customers that we went public with yesterday, it includes ASOS, Hushman Wakefield, HubSpot, right? So pretty big. Yeah, what was the process like for closing those deals? Do you mean at conferences? I imagine that it's not some sort of direct response ad. You were in stealth. So how did you get those deals done? Yeah, well, we've been around the block for about 10 years doing a one of these things. So we had enough of a network. And I think it's just a very compelling pitch, right? Like, I'm not telling you, let me come in and it's going to take three to six months and we're going to figure out what we're doing. I'm literally saying, hey, just give me one static cut of your data.

Starting point is 02:51:02 And in a week, I will show you new things about your own company's operations you didn't know about. And if I can come back with that one week with something new, don't hire me. But so far, it's been going pretty well. Are you throwing frontier models just at every problem because you're in a high growth phase? You want to have the best possible product. Are you already offloading certain jobs to lagging models, open source models, cheaper models? How much of the Pareto Frontier are you using these days? Yeah, so we need the smartest possible model to help us figure out what people are doing

Starting point is 02:51:37 and what the process actually is. But then if you do that well enough, you don't need a super sophisticated model for it. So we just need something that's good at instruction following, and it tends to be fine. That makes sense. And how much did you raise? I want to hit the gong. So we raised $30 million. Woo.

Starting point is 02:51:58 She was Sequoia? Did you let anyone else get a, get a slice? So Sequoia let our A, we had ATVC and ASTAR who they let our seed already. So we have continued support from them too. It's a murderous row. Fantastic. Well, congratulations on the progress. Congratulations on exiting stealth.

Starting point is 02:52:19 And thanks for taking the time to come chat with us today. Yeah, great to meet. We'll talk to you soon. Thank you. Have a good one. Let me tell you about Restream, one live stream, 30 plus destinations. If you want to multistream, go to Restream.com. And we will continue our Lambda Lightning Round with Ari from Run Civil, who's in the Restream waiting room.

Starting point is 02:52:38 Let's bring Ari in. How are you doing? Howdy? Good to meet you. Hey, good to meet you. What's happening? Please introduce yourself and the company. Yeah, happy to.

Starting point is 02:52:48 All right. So I'm a man on a mission where we're trying to automate hacker intuition. I guess I can start with a briefing to myself. So I was titled by the first security hire at OpenEI back in 2019. I was a grad student Harvard doing my machine learning PhD, and I saw GP2 come out, and I was like, wow, this would have been really useful back when I was a miscreant teenager doing insane operations on the internet. So I ended up bundling up a couple of demos of things that I would have made as a miscreant, and I sent them to Sam Altman, and I sent them to Jack Clark, who was the head of comms at the time.

Starting point is 02:53:21 And then kind of the rest is history. They liked it enough that they invited me to come join, and so I was there for three years. I was a core researcher on GP3 and on the Codex model. I also built our first monitoring system for when we started offering the API as a thing that customers were then using to make sure the customers were following our terms of service. And I left the company in part because we just didn't have a good answer for when the bad guys have access to everything. Black pill moment for me was when we were doing this anonymized review of model outputs, I saw somebody trying to lock a file system.

Starting point is 02:53:51 And that could be totally benign, you know, educational, Like, how does encryption work? Or it could be somebody, you know, fuzzling around with malware and ransomware specifically. And there's really no way of telling that particular intentionality. And at the time, like, our thought was, well, what if we focus on the monitoring, we just block people that are doing bad stuff. But realistically, when you have something that says explosive as language models have become, you're not going to be able to, like, play whack-a-mole. You kind of have to get offensive with it. So I started this company.

Starting point is 02:54:18 I have very blessed to work with my co-founder of Lott, ANSQ, who built the red team at Meta. And then I have a team of really strong engineers that we pulled from some of the top security engineering teams in the industry. And we're focused on building something that will make the internet just broadly more safe. And it's just really rewarding to see a payoff. Are you seeing more danger and risk from like large scale state actors or like the script kitties who are just trying to like wreak havoc or is it both? Because I feel like some of the some of the, like some of the, like some of the, the, like some of the, the, the, the, the, the, like, some of the, the, the, the, the, the, the, like, some of the, the. problems with like the new AI security threats, like there's new capabilities, but there's also like a lot more cost than just like running some PHP script that like guesses WordPress

Starting point is 02:55:02 passwords like back in the old days. Yeah, I'd say it's actually kind of a combination of the two. Like you have two types of threats and oftentimes it also depends on the type of organization that you are too. So for our customers that are smaller startups, like they're not really seeing any of this kind of stuff. For our larger enterprises, we've been asked by a lot of them if they They basically want to replace us with their bug bounty. They want to replace their bug bounty with us. They want you to win all of the bug bounties. That's great.

Starting point is 02:55:29 Just don't make mistakes. That's your immediate tam, and I'm sure much further. Talk to me about model distillation. There's been a lot of news about it's not as serious of a threat. It feels more like a business threat, but I've always been shocked by, you know, the stories about different open source companies where it feels like they trained on an American lab. and that seems like something that the lab should be able to detect. Is that hard?

Starting point is 02:55:56 Is that something you can help with? Yeah, so that's not what we do, but it is something that I've dealt with previously. And it is something that everybody does and everybody kind of knows about as sort of a bit of a dirty secret. But I'll also say that when you do distillation, the model that you get out of it is going to be like a net less good than the model that you're distilling off of. That's just information theory one-on-one. So it's something that is somewhat of a business threat, but it's not as big of a business threat as, say, like, stealing the actual model itself. Yeah, just actually breaking into the system. Can you give us your pitch to like a startup or a scale up on the customer side and then all the way up to a lab,

Starting point is 02:56:32 like how you how you sell the product right now? Because I understand like the opportunity at a high level, you know, basically every, all these companies are distributing intelligence. It's hard to understand how it's going to be used. A lot of people are going to use it for things that they shouldn't. You want to stop that. But like what is the specific pitch in this moment in time. Yeah, I'd say, like, for one thing, focusing on, like, security means a bunch of different things to different people. And right now what we're saying is that smaller teams need different things from larger

Starting point is 02:57:01 teams. And the benefit of the way that we've built our product is that if you have more attack surface, it's just much more interesting for the type of things that we can find for you. So we've been moving up into enterprises, and we have a lot of strong response from enterprise teams that are large. They have a lot of the old code bases that go back 40 years. There's a lot of cursed things in their environments, trying to get, like, additional coding tools in is kind of tricky.

Starting point is 02:57:23 And I actually have kind of an interesting hot take for you, if you'll take it. Please. Love it. Okay. So obviously there's a lot of movement in security in terms of, like, the markets, especially when Anthropic dropped some of their news about some of their vulnerability discovery stuff. So I think a lot of people are concerned about whether or not, like, are the language

Starting point is 02:57:41 model lab just going to solve security? And what I think is interesting is if code gets so much better in terms of security, the main question is, like, does that mean that hacking gets harder. And my answer is no, because I think that speed is what's going to kill us. So the space of these large possible attack vectors requires a lot more data than simply the code. So if you look at the code, I like to kind of think about it as you're looking at the bones of a dinosaur. You're going to find a lot of interesting things about structure, but you're going to miss a ton about things like muscles and whether or not they have feathers and also like bad behavior and like broader things

Starting point is 02:58:13 like that, which are also very important for understanding the ecosystem. And that's true of code as well, and it's true of computers. It's easier to find bugs. with the code. You know, having the bones is very helpful for us, even knowing that these dinosaurs exist, but you miss so much other stuff, and that is where the real delta lies here. Authentication, for example, is famously difficult to suss out. Like, there are not very many – I don't think there are any good authentication scanners out there. But the way in which we've built our product, it's very good at finding off bugs.

Starting point is 02:58:39 And in fact, one of our strengths that we've heard continuously is that, like, we're very good at finding these weird, esoteric things that have existed in bug bounties for, like, the last 10 years. and we get like pretty nice payouts from that, which is always fun. That's really cool. Talk to me about your take on the forward deployed engineer pattern, model, trend, boom, whatever you want to call it. Are you in favor of that? Are you employing that at this time?

Starting point is 02:59:06 That's a good question. I think that forward deploy is important if you're working with enterprise because a lot of them, there's a lot of a human factor. Like at least in startups, what we've learned is that people just want to solve the problem, Like the CTO comes to us and it's like, hey, I have this deal blocked by SOC2. I really need to get a pen test and we're like on it, got you fam. And we get them. They're on their very way.

Starting point is 02:59:25 But with these enterprise companies, it's a lot more of a political process. Security in general is kind of, it's partly the people. And it's also, of course, the software too. And what you can do with a forward deploy engineer is you can provide more of that layer of trust. You can communicate a lot more with the proper stakeholders so they don't feel like their job is going to be taken away. There's a lot more of the human factor that you're able to introduce when you have that. think that's why it's so popular. We're something somewhat similar there, and we find it to be particularly helpful with bringing people on board and being able to serve them faster.

Starting point is 02:59:54 I'm looking at some of the customers here, cursor, turbo puffer, notion, base, tan, thinking machines. Congratulations. The business sounds great and makes a ton of sense and obviously you're raising money. But I'm curious if there's a almost like direct-to-consumer play at some point because everyone is going to be vibe-coding. Like we are a 10-person team. We have like three systems and, you know, we're sort of, you know, security second, maybe. Maybe we're working on that. Yeah, I have a black pill for you then. So tell me. Yeah, I don't think so. Because people don't like paying for security. Okay. But isn't there another way that you can make it so cheap or bake it in or partner with a lab where, you know, I'm vibe coding something and I, you know, get runcible installed or

Starting point is 03:00:40 it's modeled into the system? Yeah, the SOC2 example is relevant because that's somebody that's like, I have to do this. Because I'm selling the software. It's hurting my revenue. But we've heard so many things about somebody's using OpenCla, they're vibe coding something, they're running their business on it, and increasingly it's turning into a system,

Starting point is 03:00:55 and at some point they need to think about security. But think about it. We bought hundreds of thousands of dollars of like camera equipment before we bought cameras to secure, you know. Oh, yeah. That's true. And even when we were doing that, we're like,

Starting point is 03:01:10 uh, like, yeah. I think there's some truth to that. But let's also think about the economics of who buys these tools to. So if you're paying like 20 bucks, 200 bucks for like a month-long subscription, how much you're going to pay for additional security on top of that? That's something that you're going to have to sell direct to that company. And there's not a lot of companies that really offer code security-related stuff. So I think for companies that are making the bet in that space,

Starting point is 03:01:33 they're focused a bit more on like the ecosystem, which is we're focusing a bit more on like the overall ecosystem within an enterprise that has a ton of ancient code that is going to require a lot more in order to fix. and they also have just these enormous attack services that needed some help. Yeah, yeah, that makes sense. Well, how much did you actually raise? Tell me about the funding round.

Starting point is 03:01:51 We want to ring the gun. Oh, yeah. So we raised $40 million, which we love. Okay, man. Congratulations. So Coastal had the round. We had participation from S32, conviction, a lot of Gil.

Starting point is 03:02:06 We also had a bunch of angels as well. So Nakesha Rora, who I know is on the show. Yeah. Friday. That makes sense. Jeff Dean. Wow. Ian Goodfellow.

Starting point is 03:02:14 Wow. Ian Goodfellow, too. It was pretty good. That's incredible. Congratulations. That's the lineup. Thank you so much for taking the time to come break it down for us. I can visualize the Coachella post.

Starting point is 03:02:23 And thank you for securing the American software ecosystem. We appreciate that as well. Get every bug bounty that is out there. You deserve it. We'll talk to you. Great to meet you. Have a good life. Cheers.

Starting point is 03:02:33 Thank you. Goodbye. And we have our last guest of the show. We will leave the Lambda Lightning Round and bring Alex Conrad in to the TV. in Ultradome from Upstarts Media. How are you doing, Alex? Good to see you again. He's back.

Starting point is 03:02:49 Hey, I'm back. It's great to be virtually in the Ultradome. Yes, I love that poster behind you. That's a TV. Oh, is that a TV? We're high tech here, John. Okay, that makes sense. Anyway, what's new since we last talked? Tell me about the shape of upstarts, how it's going,

Starting point is 03:03:08 what type of beat you, I don't know, how have you defined your beat? I think everyone knows your beat from before with the minus list of course, but what's changed? What's remained the same? You know, it's been almost exactly a year since we launched and you guys had me on the show, which is awesome. And as you know, year one startup, everything is crazy, but a lot of fun. You know, we've launched a podcast. We have started doing some feature stories.

Starting point is 03:03:34 We had one on William Hockey from Collum that was a lot of fun a couple weeks ago. And just and having a lot of fun experimenting. You know, we haven't been to the Ultradelman person, but that's a year or two. stretch goal. Yes. What about lists? I remember we talked about this and I was like, I know you can't do the Midas list, that's left behind, but I feel like there's a big gap in the tech media landscape around lists. Market maps do well. People are split on them. We had a lot of fun with a medist list list list list lists of AI researchers. Have you, are listicles just cringe or are they just actually not that interesting to you? Or are they bad business? Because I feel like there's a, there's something

Starting point is 03:04:12 there and you're the guy. You know, I'm sorry to say we don't have that list for you yet, but maybe that'll be a thing this year. We are trying to do really service-oriented coverage for founders. I think, you know, the reality is founders are super busy, right? And so are builders at startups. And so our podcast is one commute length. It's a, you know, if you're not listening to TBPN yet, you're driving to the office, you can, you can tune in for upstarts each week, you know, 35 minutes. And then similarly, we tried with our article this week, a illustration where this woman Natalie Frato actually drew how data centers connect to this new GPU startup so that people could visualize it. And the hope is that it just helps people

Starting point is 03:04:54 understand the info super fast. That's very cool. That's very cool. I read about this company Giga today that I don't even know if I should call it a startup. They haven't raised any money, but they're in the AI boom. It's just a business. Are you seeing like these knock-on effects of the AI boom show up. And are you getting pitches from those folks? Or do they see themselves as like outsiders and they're happy to remain outsiders? Or do they want to cross over into the tech ecosystem?

Starting point is 03:05:23 Like how do you think about the broader ecosystem and knock on effects of the AI boom? Well, startup can mean anything these days, right? Like I remember when we all started our career, a startup was venture backed. It was maybe less than seven years old. It wasn't hiring people for a billion dollars. It wasn't worth a trillion dollars, free IPO.

Starting point is 03:05:42 That's right. And now startup, I think, is more of an aspirational goal. You know, some days upstarts feels like a startup. Some days it feels like a small business. I think similarly with these companies, in my mind, if they're trying to be really high growth, they're trying to move fast. And if they're serving a tech savvy audience, that's good enough to be a startup. Yeah, I love it. What do you think about the scoop economy, the big labs? There's so much drama, so many personalities. There are some journalists who've gone out and carved out, like, you know, they're just the scoop masters. Is that something?

Starting point is 03:06:14 Scoop athletes. Trying to go to the scoop policy. As someone who's, I don't think ever had a scoop, I don't know if I just haven't felt a rush. Is it addictive? Like, what are the pros and cons of getting into that side of the business? There is a huge endorphin rush, right? If we're chasing endorphins and avoiding cortisol, I think, like, you know, when you do

Starting point is 03:06:34 publish that scoop, it can feel really good. You know, for a while, our biggest story at Upstarts was last summer we wrote about a startup that had left Open AI, and they had raised a ton of money to do an RL reinforcement learning company. And when you looked at the spike in subscribers we got, that felt really good in a way. But you don't want to play that game all the time. I think it does end up being like chasing a rush that is not sustainable. And so I think, you know, I will let Katie Roof and those self-described scoop athletes chase it for the love of the game. For me, it's only really relevant if there's something concrete, like a lesson or an insight for that wider ecosystem.

Starting point is 03:07:11 versus just the horse racing of, hey, these guys from Open AI raised even more money than those last guys. In terms of the horse race, obviously you ran the Midas List for many, many years. Is there, are there any venture capitalists who are underrated right now? Or do you think that there's any misconceptions in the venture capital community? Because I feel like the strategies have shifted so much and we're seeing bifurcation between the small funds and the megafonds. There's folks who are venture capitalists, but they're trading in public markets all day long or running private equity shops now or buying hospital networks. The strategies have evolved so much. Like what other stories are interesting in the venture capital landscape broadly?

Starting point is 03:07:57 Well, first, I think this year we're going to have to start a upstart spotted on the street, VC thing, because I spotted Keith Rabe at a restaurant earlier this week in New York, and I gave him the eye. And I said, come over and we shook hands. And, you know, the poor founders who are with Keith were like, who is this dude? So I do think we should do a segment of just where I spot VCs around New York City and San Francisco and awkwardly wave to them. TMZ. But in terms of, yeah, that's right. Yeah, that's right. That's the coverage we need, right?

Starting point is 03:08:25 We need the gossip coverage of VC again. But on a serious note, I mean, I love the domain experts, like the really nerdy guys who are not posting a lot on X who are just really well regarded. When I ask around, like, hey, who's really smart on GPUs? And so my advice to VCs usually is like have a thing you're known for. If all you're known for is posting on Twitter slash X, that's probably not a defensible strategy in the long run. Yeah. What is, what advice do you give to VCs that might be due for their first Midas list appearance now that you have a bit of space and you're not involved in the process? Interesting.

Starting point is 03:09:04 I mean, at the end of the day, venture is a results business. I mean, you guys have had guests on recently who talked about what is real and what is not. And I think the numbers generally do speak for themselves. You get a big exit that is kind of the mic drop that I think Midas is a lagging indicator to notice. I think for VCs who feel like they have that portfolio that's not recognized yet, the first thing I would say is be top of mind for your founders. You know, often journalists like me or at the big shops, like we'll talk to a founder and we'll be like, who are the couple of VCs who backed you who we should call to get to know your business better?

Starting point is 03:09:37 If you're not one of those first two or three VCs that the founder mentions as a reference, that's not great. So I'd start there with like, are you top of mind for your biggest winners? Yeah. So it's a mix of results plus the mic drop moment in the last 12 months plus kind of founder brand. Is that a good way to think about it? Well, yeah, I mean, the Midas list is data only. So the founder brand doesn't matter as much. But I think if you're feeling like, hey, I want those flowers.

Starting point is 03:10:06 It's data, yeah, well, it's data only, but it's not just like blended IRA across every investment. And there's still like crazy internal politics at every VC firm of like, oh yeah, that associate who was here for two years and it was the one who actually got that deal, but then left, like, that's my deal now. Hey, I mean, you guys know, venture is a tough game like that. You know, my wife just left VC to go into operating back at a startup clay. And, you know, she will have her deals that she sourced and she was involved in. and will she be in the history in years? We don't know. And I think similarly, if you source the deals

Starting point is 03:10:41 and you moved on to another firm or you went back into operating, like history will, I think, give you the credit in the long run. But yeah, for Midas, that can be tough because like five people at Sequoia claim each, you know, big deal. Yeah, yeah.

Starting point is 03:10:52 I heard that there's one firm, I don't know if it's benchmark, but they have like a ledger that when the deal closes, they all agree on the allocation, okay, you brought it in and you're going to be on the board. You get 80%,

Starting point is 03:11:03 but I worked the deal with you, so I get 20% and we all sign and then we know who got the system of record. A system of record, ERP, basically. But I don't know if that's employed at every VC firm. Has there ever been in your memory a situation where sort of like a VC stake was discovered in sort of an IPO prospectus or like an S1? Because I imagine usually the VCs are taking plenty of victory laps throughout the process

Starting point is 03:11:32 once things get close. but I'm wondering if there's ever been like the quiet VC, not on Twitter, not posting, and then all of a sudden the IPO comes down and they're like, wait, they own 20% of this company? This is crazy. Yeah, I mean, a firm that was historically under the radar was Sutter Hill. Oh, yeah. So Michael Spizer, when Snowflake went public, he got tons of credit, deservedly so. But they had been totally under the radar.

Starting point is 03:11:58 And so those more incubation type ones, those are really interesting. Yeah. And I think otherwise the thing to know with the S-1 is it's usually like a big dog at the firm whose name is attached. But that doesn't mean they necessarily were the person who did the deal. What happened earlier in my career is people would be like, did you know I sourced that deal? And then I left that firm and I hadn't been in the game long enough to like know any of this trivia. And so that was terrifying for me. Over time, I started to know all the trivia.

Starting point is 03:12:25 Like Airbnb was sourced by this person and then this person was on the board and then this person was on the board. And I don't wish that date on anybody's head. That's hilarious. Do you view venture capital and the startup ecosystem as like a buyer's market or a seller's market? Like, is it a good time to be a startup versus it's a good time to be a VC? And where are we in the cycle? What an easy question, right? You're saving all the easy ones for less.

Starting point is 03:12:56 I think, you know, we continue to be in that have and have not market, where, I think you see crazy valuations for companies that have traction. And then I hear from so many startups that are still like, how do we meet these guys? Like, how do we get anyone to pay attention to us? And it's humbling for me that, you know, even though I'm saying, hey, we want to cover startups that aren't getting that coverage, even then there are most that I just can't help or get to. And so I think like I would encourage people to get away from the buzzwords, you know, especially get away from Silicon Valley. And there's still plenty of companies that are not getting funding. Yeah. Will you ever write a book?

Starting point is 03:13:33 About what? About you guys? No, about your experience. I mean, this is a common path, I feel like. Sometimes it's like, you get a scoop, and the scoop grows into a book. Sometimes there's, you know, a composite profile of an industry or career. I don't know. It sounds like, no. It sounds like you have no appetite for a book at all. But is that? Well, I think I just don't have the bandwidth. I mean, like you guys, I think I'm in the arena every day. now putting points on the board. And I think, you know, books seem like a beautiful stretch goal if upstarts really scales. But, you know, I wanted to write books in the past maybe, but I think you need to really have the idea. You can't just reverse engineer it. It's like saying I want to be a founder and not knowing what company you want to actually create, right?

Starting point is 03:14:18 And then it just has to sort of work. And you have to love the topic, right? Like, how boring would it be to write the world's 10th book about Nvidia? Like, do we need that? I don't think so. I don't know. I don't know. I don't know. I, I, Jordanie's like you're going to do it. I would, I mean, I would, I would, I would, I would read an entire book just about the the leather jackets or the leather jacket or the leather jackets or just an entire book just around the history of DLSS just, just deep learning super sampling. Well, you know what I think you guys are speaking to that I do think about a lot?

Starting point is 03:14:50 Yeah, yeah. Like information is so crazy right now. You know, you guys have these amazing guests on every day. Yeah. You're grinding. Yeah. And, you know, so many people are out there putting out. good information. Is there room for that person who just kind of disappears for a long time

Starting point is 03:15:05 on a crazy project and comes back with like a big fish? Yeah, yeah. No, we talked about this a while back with, there's this YouTube channel that I love called the Corridor Crew. And they talk about visual effects, specifically, very niche. How would that ever be on, it just would never be on TV, but it's turned into a TV show, multiple episodes every week, fantastically successful. They build a whole business. They have a studio and team. And their dream was always, okay, we're good at visual effects. We want to do a movie or be the VFX crew on a movie. We're capable. But they kept building up their YouTube business. And every time Hollywood would come to them, they would say, yeah, we'll give you, we'll give you 300K. And they're like, but our business is

Starting point is 03:15:49 making a million dollars now. And they say, okay, it doesn't work. And they come back and be like, you know what, we're ready. We got a million and a half dollars for you. And they're like, but our business is doing three million. Like, we can't step away. And so this tug of war is always happening. And I feel like the book is the same thing where, you know, can you really turn off the podcast or can you turn off the reporting or can you turn off the, all the other right-in reporting?

Starting point is 03:16:09 Yeah, I mean, I've just come back to this as like legacy media brands are the best place to go, you know, fishing. Sure. Because they will say, yes, you can, we're going to pay you a salary and you're going to go work on the story. Yeah. And you might not be able to show any real results for. a longer period of time.

Starting point is 03:16:27 It doesn't work that well for a substack model where you got to show value. Yeah, yeah, yeah, yeah. Value every week, I feel like I got to win the week. You know, I feel like I'm proving myself to my audience every single week and day. I will say we are doing these quarterly profiles now. Hockey was the first one with column.

Starting point is 03:16:46 That's great. I think where you have a really cool one cooking for the second quarter. And the dream is that maybe like we put four of those in a little booklet. That'd be cool. That, you know, can be on a coffee table. and maybe we start to get into print and have fun that way, but it's baby steps.

Starting point is 03:16:59 Yeah, yeah, that makes a lot of sense. Do you have sort of a media critique take on the future of investigative journalism, where that might exist, what the funding model for that is? Because it feels harder than ever to have a journalist go and spend a year on something that may or may not work out.

Starting point is 03:17:18 Yeah, I think the short answer is that I agree with Jordy in a lot of ways that you do need the sort of big shops that can still weather the storm for someone. You know, at Forbes, I would be hunting that big cover story and do it and then recharge my battery while other people were kind of putting up the singles and the doubles. And now it's like, you need the singles and doubles and then you look for the home run. So I do think a team is important there. But one thing I would challenge people to think about is, is there a way to fund a project for a year?

Starting point is 03:17:49 Like, you know, like whether it's Patreon or substack or something like that, where you do a year or even a multi-year subscription where you say, we're going to give you money up front, do the best, craziest thing you can do in that time. And, you know, it can be as few as one thing, but then we'll be happy. Because I think the challenge, as you said, with substack is substack sends you the numbers.

Starting point is 03:18:08 You can see them going up and down. You never want to see them going down. And so could we create that headspace via some sort of crowd-funded model? I mean, I would love to see the innovation. Yeah, the same stuff. Well, thank you so much for taking... Let's hit the gong for a year.

Starting point is 03:18:22 Is it actually the year, the first year? Did you hit the anniversary yet? We're a week away. A week away. We're going to hit me. Thank you so much. Great to catch up, Alex. Great to catch up.

Starting point is 03:18:36 And we will talk to you soon. Goodbye. Cheers. All right. See you guys soon. Here's some advice. Don't buy AirPods. You need the Sony WH-1000 XM-5, W-H-C-H-720N, W-W-S-2020.

Starting point is 03:18:51 WF-1,000 XM-5-C-H-520s. Get those. Just pick those up. That's what Dylan the tell wears. I can't believe how mainstream this posted. Whoa, I didn't realize 150,000 likes. What is Sony doing? They need to rename their products.

Starting point is 03:19:07 Just call it like the Sony, I don't know, head pods or something, headphones. They should, I mean, people say XM-4s, which I think is the, like, the last three digits. That's what they refer to them, or the last three characters. That's what they refer to the headphones as. And so we will just call them XM-FORs.

Starting point is 03:19:24 But rough with the naming schemes. Anyway, thank you for watching. Leave us five stars on Apple Podcasts in Spotify. It's been an honor. We will see you tomorrow. We'll see you tomorrow. Goodbye. Boom!

TBPN Live - Samsung Invests $70B in AI Chips, The Cubanator Joins, Apple: Behind in AI, Ahead in Revenue | Mark Cuban, John Kim, Eugen Alpeza, Ari Herbert-Voss, Alex Konrad, Carl Eschenbach & Pat Grady, Jim Cantrell, Tom Hulme

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