https://www.listennotes.com/e/0f391c74f33349078f676fb4f23ce51a

You'Re watching tbbn.

Today is Friday, May 30th, 2025. We are live from the tbpn Ultra Dome, the Temple of Technology, the Fortress.

Of Finance, the capital of capital. It's a real lag, John.

I'm traveling. How are you doing?

I'm good. I miss you.

Yeah, me too.

It feels close, but so far, it's brutal.

Sleeping without an eight sleep. It's the punishment worse than death. It's a true torture.

It really is.

It's the worst. It's the worst. Anyway, exciting week, short week with the Memorial Day holiday, but I feel like we did a good job making up for the lack of major tech news, the lack of major announcements from AI companies. Although, WWDC is coming up next week, which should be interesting. And that's where I want to start because John Gruber has been writing about Apple for years, Daring Fireball. He has a show called Talk Show that he does at WWC. And it's very interesting because he was one of the first people to do a talk show and interview show in the Vision Pro in VR. They bring the Pro cameras and you can stream it and you can watch it in the Vision Pro. He went all in. He's been writing positively about Apple for a long time. Of course, a couple months ago, he wrote Something is Rotten in the State of Cupertino, talking about the difficulties with the rollout of Apple intelligence. And they didn't completely rug him, but they didn't sign up fully. So last year, the talk show live from WWDC 2024, Had Craig Federighi, Greg Jaws and John Giandrea, Apple's head of machine learning and AI. And this year he invited them all and they said, no, we don't want to participate.

No way. Yeah. No way.

Which is right.

They took the rotten post personally. Yeah. They took it personally.

It hurts. And so, I mean, I was just reflecting on like what this means. Obviously we want to interview these folks. And so I just kind of had this realization that truly Apple intelligence is one of the most underrated products of all time.

Totally. Totally. And people are critical of it because it's Apple and they're they always achieve perfection.

But I think they did. I think they did preach perfection.

I think it's in love daily, right?

Exactly.

When you text me, when you text me, John, yeah, text me this a couple paragraphs, right?

Yeah.

Cool idea. I actually like that Apple Botches the summary because it inspires me to think, what would John think in this circumstance?

Exactly. Exactly.

What would John want me to think that he was saying before Apple kind of stepped in? And so I think you're totally right. I think Apple intelligence, criminally underrated.

I agree.

Honestly, John, I would take it a step further.

Please.

Just eliminate all original, all the original text of every message in iMessage and only provide back and forth sort of ping pong summaries kind of like a game of telephone in some way.

Yeah.

And just let Apple intermediate everything because 100%. Yeah.

And I think a lot of people were critical of the Genmoji campaign, but as someone who's sponsored by AdQuick, we love out-of-home campaigns. And I think that that was also something that with the long arc of history, we'll look back on as As an iconic campaign. Exactly, it was contrarian.

Up there with 1984.

I agree.

And so many of these other, you know, the iPod campaign, right? So these type of iconic injury. And we spent a lot of time talking about the Genmoji campaign, right?

We did.

And many people would say all attention, you know, even though we had some critiques, but all attention is good attention. And yeah, when you think about that.

I think that's part of the show is like, Yes, you could say we were critiquing, but a lot of the show is humorous. And so, yeah, I would now, my message to Apple is like, those critiques, those were the jokes.

This is the serious part.

Yeah. Now, I'm serious. And I'm saying they've never done anything wrong. Everything's perfect. And I want to interview you. So let's make it happen. But, you know, the funny thing is that I was actually, obviously we're kidding around a little bit, but I was reflecting on the fact that I am actually more bullish on Apple than John Gruber right now, because after the State of Cupertino came out, me and you were talking And I was taking the position that like Tim Cook, because of how he handled the supply chain and the tariff stuff, that was actually- He.

Made it all back in one trade.

He made it all back in one trade. And so, you know, I think that, yeah, the Apple intelligence thing, it hasn't gone perfectly, but the execution, I mean, it's still a massive company and they're doing great. And I think that there's a world where they land this in a very interesting way with partnerships and they open things up and there's some really, really interesting outcomes that could happen. So, anyway.

Yeah, the question is for Gruber, How much angrier does he get at Apple as he gets less and less access? Right? Did they awaken, you know, a monster?

Yeah, yeah. You'd think that you'd want to have him come on and kind of ask the hard questions and make your case and win him over. You know, they have a long-standing relationship. Also, did you say Gruber? It's definitely Gruber.

Gruber, sorry.

I like Gruber, though. That's great. But yeah, he was talking to Ben Thompson and saying like he's kind of relieved because this was going to be one of the harder interviews because the expectation would be that he asks like the hard questions, right? And that he actually pushes them because he has been a, although he's obviously been very, it's easy to be pro Apple over the course of his career because Apple's been on an absolute tear and so If you're, like, the correct analysis of Apple at every point in the last two decades has been the stock's about to rip. And if you just kept saying that, you were right again and again and again. It sounded like you were glazing or being shilling, but, like, you were also correct. Yeah, people look back and they're like.

Yeah, that's good journalism. Buffett bought Apple when it was already.

The most valuable public, valuable company in the world.

Exactly. Exactly. It was not a, you know, there were people making that same trade just thinking, oh, I should just buy the biggest company. And Buffett was basically making the bet that it was undervalued at that point massively.

Great. Well, really quickly, we're going to do some ads. We're going to tell you about RAMP. Time is money, save both. Easy to use corporate cards, bill payments, accounting, and a whole lot more.

Whole lot more.

We love Ramp. I'm going to get to see the Ramp founders this weekend, which should be a lot of fun. Anyway, let's rip through some news. I want to give a little update on like what did happen this week because there were a bunch of small stories, nothing really big that became the current thing. But we should rip through this. So Elon Musk's Neuralink raised $600 million and 9 billion valuation. This feels like old news to me. I don't know if that's just because I've been hearing through this, through the rumor mill, but very excited for what they're building. We talked to Ashley Vance a little bit about Neuralink and the progress, and it's kind of underrated because the the first time it happens, it's like a miracle with that one neuralink patient. But there's been a few, and Ashley Vance has actually gone and interviewed others. And everything that I've seen is that this project moves slowly, like many, like SpaceX, where there's regulatory concerns and FDA approvals, but the foot is not off the gas and it just continues to grow.

So, the thing that stood out to me here is Neuralink. raised $43 million in November 2023.

Yeah.

Like the fact that there was an Elon company with this much potential in the private markets being valued at 5 billion, you know, roughly two years ago is actually crazy. Granted, the round was effectively like an additional tranche that had been part of an earlier FF round. But still, you know, Not a bad bet.

Yeah, but just insane concentration talent. I mean, we talked to the Teal Fellow who was working at Neuralink and that's the DNA at that company. And it's pretty remarkable to see. Obviously all the focus is on super long-term brain computer interfaces, but when you talk to the Neuralink team, they're much more grounded in the immediate effects of what the technology can do, which is there are a lot of people who are paraplegic It's actually a lot of people that get in pretty horrific accidents. They're doing cliff jumping or skydiving or something, and something goes terribly wrong. They break their neck, essentially, sever their spinal cord, and then they just cannot access their limbs, but they're alive and they're going to live out their life. And you give them the access to play video games, which is the most beautiful gift you can give someone, in my opinion.

Totally.

And so, yeah, it's always super inspiring. I'm sure we should talk to some of the Neuralink folks at some point. We should have Colin call in.

What was the name of the teal fellow who was at Neuralink? Do you remember?

I don't. I mean, we interviewed 20 people.

20 back to back. But that was one of my favorite moments of the week.

Oh, yeah.

His philosophy of so funny. Just trying to get two times cooler over and over and over. And you stack a few 2x wins like that back to back. starting to get pretty cool. Yeah. I thought that was an amazing philosophy of life. And he was just skill maxing.

Yeah. I mean, also his definition of cool, which I kept, I kept expecting it to be like, and then I hit the gym, and then I learned to race now.

Now he's like, not cliche at all.

Building a trumpet from first principles. It's extremely cool. It's genuinely cool. Yeah. Yeah. Anyone can, anyone can hit the gym.

And learn to do a kickflip, but, yeah, build a truck.

He'll get to, he'll get to that. Those are, those are They're on the road for sure. Down the road. Anyway, other fundraising news. Relit raised $25 million from Sequoia to automate general ledger systems using AI for accounting departments. Sequoia has been big on this application layer, value accrual in AI and the application layer. Obviously, they have Harvey, they're in some of the foundation models, but they're also looking for bets in kind of every vertical SaaS category now. So this kind of makes sense. The partner at Sequoia that did the deal said, the general ledger is the beating heart of the finance function. And so asking a company to remove it is kind of open heart surgery. So interesting to see how this actually interfaces with some of the other Sequoia bets, because they're in Ramp and they're in other companies, but there's certainly more stuff to do inside the enterprise and there's a bunch of different places. So a $25 million bet makes sense there. What else is going on? Horizon 3 AI raised $100 million from NEA at a $750 million valuation. AI startup shock. Yeah, lots of size gongs this week, honestly. A Databricks competitor raised a $50 million Series A from Felicitas at a $500 million valuation. There's a bunch of other stuff going on. Databricks, we've talked about the Neon deal, but it's becoming more and more finalized at this point. There's some venture fund news.

There's also Clark, led by Brad Mendes, the first AI agent to build internal enterprise apps, raised $60 million, which I think we may have covered, but I wanted to give Brad a shout out because I wanted to get him on the show this week, but we didn't end up having the time. But shout out to Brad and the whole team.

Yep. So the big news in the technology world in the Wall Street Journal this week is this Google lawsuit that's been going on for a while. Judge considers curbs for Google in AI arms race. And this is kind of interesting because this started around Google search monopoly, which has been well documented. Obviously, they own like 90% of the search market. They own Chrome and they're the default on Apple, and they have that very lucrative deal for Apple, where they pay, I think, over $10 billion to be the default search in Safari.

Isn't it $20 billion?

Yeah, it's a lot. And it's just pure profit for Apple. It's pretty awesome. So lawyers are making closing arguments Friday.

In the landmark case. It reached 20 billion in 2022. So it may have ticked up from there. Let's go.

Size gong. Major size gong. Every size gong. Let's hear it for the Apple shareholders.

Yeah, you love it.

And this is why, this is why you can't critique Apple. There's just nothing to critique because they're printing. They're absolutely printing. I mean, truly, they set themselves up with the perfect platform. And so whatever happens next, they make money.

Like they didn't need to compete in the search. It's crazy. Yeah, they have the $20 billion search deal with Google. And because of that, they can offer Genmoji for free out of the box. Incredible. It's just that incredible consumer surplus that.

They'Re able to deliver. So US District Judge Amit Mehta on Friday questioned how far he should go to limit Google's monopoly in internet search, including putting curves on how it competes in artificial intelligence. Meta's query came at the beginning of closing arguments in the final phase of the Justice Department's antitrust lawsuit against Google. The judge questioned how questions underscore how much rivals such as OpenAI have already changed how people access information on the internet. Meta earlier found that Google is a monopolist in search and that it abused its power to maintain dominance. But the government's proposals for improving competition include curbing ability, Google's ability to promote Gemini, its AI product that competes with ChatGPT. I wonder if that's part of the story around Gemini being hard to find. There's this constant critique on X that Google has incredible models, incredible products, but they're just not distributing or surfacing them well. And I wonder if it's less of a product management mistake, like the product manager made the wrong decision, or the lawyers came in and said, Hey, I know that the the product management best practices would be to roll this in. But if you do that, we're at antitrust risk. So we have to make it harder for you to sign.

Yeah, but I mean, this proposal from Judge Metzger is a proposal. It isn't, wouldn't the strategic thing to do as Google is to just say, well, we're going to promote this as much as we can until we're told we can't. Right.

No, I so, yes, idea of like, if the speed limit 60, you go 59, but I think internally at some of these companies that have been really scarred by antitrust and legal battles. Sometimes the legal department says the speed limit 60 product managers, everyone's going 45 because we are actively being tailed by a cop. They don't have their lights on yet, but but they could turn those lights on. You know how people like, you know, start going five under ten under when there's a cop behind them. That's I think that might be a little bit of a dynamic at some of the big tech companies. I think Bill Gates talked about this at Microsoft, how part of the reason why he transitioned out and and Ballmer came in was just he got so tied up in antitrust legislation and antitrust fights that it became like a soul sucking job and he couldn't do the thing that he wanted to do, which was just like build computers. And so there's this interesting dynamic where you could imagine a lawyer saying like, I'm going to get fired if we get a massive lawsuit or something like that. And so to protect this, I'm okay with us growing a little bit slower. I'm okay with these new products being, and it's because they're not comped on, did you deal with the innovator's dilemma and find the big next wave, right? They're comped on how bad are things in the courts right now. And also, if you're a product manager and a lawyer on the team comes to you and says, legally, I'm recommending that we don't do this, it's very hard to push back on that. It's not the same as the dynamic between a designer and an engineer. Where the engineer might say, oh, this is like, you know, maybe better from an engineering perspective, and it's more of a discussion. I feel like when the legal team shows up, it's usually like they are issuing a, they're dictating what will happen. And most people are just saying, okay, well, legal said we have to do it this way, so we're gonna do it this way.

Yeah.

But I don't know. The interesting dynamic here is around, like, the case specifically is not directly shaping how Google deals with generative AI and their new products, but it could have knock on effects because they are proposing the forced sale of the Chrome browser and preventing Google from being able to pay Apple to be its default search engine and requiring it to share data with competitors. And now this is obviously bad for a number of reasons, but the interesting thing in the context of AI is that all the data that goes into Chrome and Google search from these extra search. It's not just about running more ads. Obviously it is. Obviously, if they're paying Apple 20 billion, they're probably making 40 billion in ads off of those searches or something like that. Right. But that's also really, really valuable data that they're going to be using to train the next generation of Gemini. And so if they lose a little bit of that edge in the flywheel to something like ChatGPT, where people are going back and forth and and the data flywheel is spinning up faster at open AI, that could actually put them at a place where the data is compounding slower. And so they're not able to stay on that Pareto frontier of the best model at every single price that they've been so happy with. The interesting thing to me is.

Have.

You heard the story of how Google originally built the algorithm for auto- Interesting, by the way. Yeah.

I don't think this would be the case for For our audience, but Safari apparently has 90% of the browser market share on iPhone. I imagine a lot of people listening have Chrome installed and prefer that.

I use Safari on iPhone.

Safari guy.

Yeah, I use Safari. It just feels like it's the default. I've tried to use Chrome and the iPhone keeps pumping me back. Oh, do you want to open this in Safari? And I've noticed that With the cookies, if you're logged in in Safari and then you're on the X app and you open a link, that will open a Safari web view. It doesn't open a Chrome web view. And so the Safari web view, it's like an iOS code module. If you're logged in in Safari, that will stay logged in even when you're in the Safari web view inside of Instagram or inside of X. And so that extra benefit is enough for me to use a second browser, even though on my desktop I use Chrome. And I'm losing that connection. It's like I want the mobile experience to be really cohesive. So I don't know if that's just like boomer mode, but it seems to.

Be working for me. We need a boomer sound effect. Yeah. So anyways, going back, you said the Justice Department, which brought the case against Google in 2020, has proposed forcing the sale of its Chrome browser, preventing Google from being able to pay Apple to be its default search engine. The department says the uncommonly harsh remedies are supported by a legal standard created when the U.S. government tried to break up Microsoft over 20 years ago. All three of the DOJ's major remedies are aimed at helping generative AI providers take share from Google, says Paul Gallant, a policy analyst for TD Cowen. The government hasn't met the standard of proof for obtaining such drastic remedies, Google's lawyer John Schmidt Lean argued Friday, the Justice Department didn't prove that Google's exclusive contracts with device makers such as Apple were the primary cause of maintaining its monopoly.

Yeah, we, I gotta say, you gotta.

Be happy if you're, or if you're Sam or Arvin from Perplexity right now, given that the DOJ is explicitly stating we are trying to help generative AI providers take share from Google.

Yeah, yeah, makes sense. The data flywheel is interesting. I got to tell a story of the original Google autocomplete algorithm. This was prior to transformer-based language models, but they needed a way to autocomplete on Google search as you type. It fills it in. And what they found to be the most effective data set was let's say you're searching for something like how to take the train from SFO to San Francisco and you type it incorrectly, you misspell San Francisco or you misspell SFO or something. What they found is the best way to generate a massive data set of autocompletions is just look at when someone types one thing and then they type a similar query immediately after within the next one minute. And because if Google fails and gives you and doesn't understand the query, you'll typically, the human, will go and correct the query and be like, oh, I misspelled San Francisco. I need to spell it correctly. I will go spell it correctly. But how do you send that signal to Google that what is the correct way to spell San Francisco? Or what is the correct search term? What is the right way to phrase this question? Well, they were able to create this massive data set just by looking at like one query A, query B, and then just telling you, are you sure you don't want to just do the next query immediately? So it's the same kind of predictive text, but when you're at Google scale, that has incredible value. And you can imagine that having a big data flywheel potential at Google as they scale up generative AI.

So anyway, Metta said he expects to issue a ruling in August. Google which has roughly 90% market share of online searches, proposed a narrower set of remedies that would modify its exclusive agreements with Apple, as well as Mozilla and Android to allow for more competition. It has said it would appeal the judge's ruling. Sounds like whatever, whichever way it shakes out.

Yeah, there's so there's so many different things going on.

Here's the interesting thing, too. So Google's already paying Samsung and Motorola to preinstall Gemini on devices, which is interesting. So potentially a counterpoint to what you were saying earlier around wanting to be if the speed limit's 60, stay at 45. Yeah. This feels like basically being like, hey, speed limit is 60 here, but it's 80 over here. Let's go 80.

Yeah, I'm surprised they even have to pay them because Samsung and Motorola build off of Android, so you would think it would just come pre-installed as a default Android app, but I'm sure that there's a debate because, Uh, Samsung probably has their own foundation model and, and, and could auction that off. And so it's kind of up for debate, but that certainly wasn't the case with like the other stock features of Android. And there's, so now with the new lens, they gotta, they gotta figure out the, the deal. But yeah, there's so many interesting things going on with Google with like absolute dominance in video generation, absolute dominance on, on YouTube. Like the YouTube model just seems like stronger than ever and compounding and compounding and like, not going anywhere, like just the absolute best business model. And then some weakness in search, which is obviously huge on-- well, the funny.

Thing is, you could imagine search is getting more competitive naturally, right? As people use LLMs to do things that they would traditionally use regular search engine for. And you can imagine two years from now, the Department of Justice taking this massive victory lap of being like, you know, we created competition in this market. We did our job, and it ultimately was just from the creation of new technology that Google itself has played a big role in creating. So I'm much more in favor of letting the best technology win and the best teams win. And it's unclear to me right now if there really needs to be any meaningful intervention.

Well, if you have the best Technology, you need the best design tool. Go to figma.com, think bigger, build faster. Figma helps design and development teams build great products together. Go to figma.com, check it out.

I've been noticing a bunch of people starting to use Figma make and realizing how powerful it is.

Starting to take hold.

I will.

I have the post in here somewhere.

Yeah, we'll get to it later, but it's very cool to see.

Yeah, and Angela Liu says H-O-L-Y-S-H-I-T, which we will not say on the show.

We will not say that word.

We're friendly. Just got access to Figma Make Beta on our team account, and I'm totally mind blown. These are mocks that we one-shot prompted, and it's interactive. Buttons, tabs, prototype works, and all the React code is there to copy and paste into an application. That's very exciting. Congratulations to the team. Very, very cool that they're getting this out into the world. Anyway, other deals that are going on, the Ray-Ban Maker, how do you pronounce this? Essilor. I just know it as Luxottica. Luxottica.

I never had seen Essilor slash Luxottica.

Luxottica is buying Optegra in a medical AI push. So Luxottica partnered with Zoc on the Meta Ray-Bans. and they're getting more into AI and smart glasses.

Interesting.

Interesting to see.

Integra operates eye hospitals in the UK, Czech Republic, Poland, Slovakia, and the Netherlands. So interesting. You think of Luxotic as a, you know, consumer product company, but this is a very different kind of business to integrate.

Yeah, it's such a different approach to the Medibands. The reason it made so much sense for Luxottica to partner with Meta on the Meta Ray-Bans, because it's a fashion item. You put it on your face.

And Meta doesn't need to recreate an iconic silhouette.

It's really, really hard. Yeah, iconic silhouette for sure. And Meta, I mean, the Quest looks okay, but it has not reached anywhere near the kind of final form factor that the iPod had or the iPhone had, like the silhouette, as you mentioned. And so, To jump into and it's on your face so it's not even something that's like, oh, yeah, it's a little bit of an awkward device like it looks like a wallet But I put up with it because I'm only using it when I when I'm when I pull it out It's like, no, these are gonna be on your face all the time. They need to look good and the.

Ray-Bans are so iconic it made a ton of sense and they're able to draft off of Ray-Ban's marketing and brand and you have ASAP Rocky is their creative director There's real momentum there and they don't need to reinvent fashion while they try to create a new technology platform.

Yeah, yeah. But it's a little bit different when you're talking about medical AI, where I think you would think that a branded product would be a little bit less important. But I guess Optegra operates a network of over 70 eye hospitals and diagnostic facilities across the UK, Czech Republic, Poland, Slovakia, and the Netherlands. It's offering comprises ophthalmic treatments. And effective vision correction procedures supported by artificial intelligence in the pre and post op phases, the transaction details were not.

You want to know something interesting? So since doing the show, my eyesight has actually gotten better.

Really?

And I think it's because we spent, we spend a lot of time looking at screens, which probably isn't good, but typically it's screens that are far away. Guests on that are far away. So my vision is actually getting better. That's great. Narrative violation.

But I didn't even realize you could, I thought your eyes only got worse. I didn't realize you could get your eyes better. He's like going to the gym for your eyes, basically.

You can, you can, yeah, you can effectively work out your eyes. It's just, I guess a lot of people get into routines where they're either looking at their phone computer.

Sure.

Um, and there's not a lot of like, you know, intensive exercise at other sort of consistent ranges, I think.

Yeah.

Yeah.

Interesting. Anyway, report from the information says every VC backed IPO in the past 12 months has been a down round.

Silence.

Yep.

This moment of silence is brought to you by AdQuick, out of home advertising. Anyways, back to the news.

But there's an interesting wrinkle here. So there's some good data in this article by Corey Weinberg in the information. And there's also some good news, which we'll get to. So when Banking App Chime goes public next month, it will mark the eighth straight venture capital backed IPO in the last 12 months that has been valued less than it was in private fundraising rounds. Chime was valued at 25 billion in 2021, but its bankers have been discussing pricing its shares at roughly an 11 billion valuation. That's a pretty significant down round, less than more than 50% off. Startup valuations are far lower than several years ago. So a lot of these companies, they raised big rounds during the DeSerp era. Now they got to get out, got to get liquidity. And so they're willing to take a little bit of a haircut in trade for actually being able to do some distributions because Even though the most recent backer might be taking a little bit of a haircut, a lot of the early backers are still like, well, yeah, I invested at 10 million. I'm happy with $11 billion. Yeah, so an example of the founders are happy. The employees are happy.

Forerunner did the seed round. They probably invested in a bunch of other rounds. But their entry is so low that I imagine they're going to do fine if it gets priced at 11.

Kristen Green's on the list for that specific deal. Yeah, among others, but yeah, many investors are facing losses when companies go public at lower valuations. Yes, but like again, these are later stage investors that might, where even the LPs might have the assumption that they'll hold a little bit longer and that's what this article gets into. So quickly, let's go through some of the IPOs that have gone out and where they're pricing. So Chime was 69 bucks a share in 2021. Upcoming, we know that it's about down, it's down about 50% circle. last private round was at $42. The IPO is predicted to be at $26, so down 38%. Hinge Health went from 77 in September of 2022 to 32, down 59%. That's a pretty steep haircut. MNTN, which was actually fair a little bit better, just down 30% from peak.

Yeah.

Etoro, I see that 66%. Core Weave down 15%. I feel like Core Weave has lifted. ServiceTitan down 16%. TempeST down 35%. Now, there are a couple, Ibotta and Rubrik did pop after the IPO, or at least price higher than the previous round. So let's hear it for Rubrik and Ibotta. But what's interesting is that the information shares some data on companies that went public at discounts to the last fundraising tended to perform strongly in the public markets. And so Reddit went out at a lower valuation than their previous private market round, but then they went up 221% since IPO. Same thing with Core Weave, went out at a slight haircut to the last private valuation and is now up 164%. MNTN is also up 63%, ServiceTitan up 59%, Tempus up 47%, Hinge Health up 22%, Etoro up 19%. And so it's not all bad news and there might be some sort of narrative where You know, there's been this, there's been this story of the Zerp era that the private markets went crazy, that the valuations were super, super high, like crazy revenue multiples. And so you go out to the market, you talk to your banker and you could do your IPO roadshow and you say, Hey, look, we're taking our medicine. We are going out at a price that is a discount to our last private valuation. But we've grown so much. It's been three years. The business was sound. These businesses weren't frauds, they weren't bankrupt, they weren't like no users, they were just a little hotly valued for a while and they say, hey, retail or whoever's buying in the IPO, we're giving you a discount to what the investors paid in the last round. And then that sets them up for an actually like new, new, new resets the valuation, gets them into a liquid position where they can actually grow and build up earnings, beat earnings a couple times and then go on an absolute tear, a generational run potentially.

Yeah, it's good. It's good. We, you know, having a number of IPOs end up, you know, you want the retail or institutions coming in to not get host every single time.

Yeah, totally. And if, yeah, I mean, that was the other story of the Zerp was that, like, there were a lot of SPACs that went out where that went way down. And so it's almost better to go out at a discount and then build back up, especially if there's a hold period, lockups for various investors. You want to be growing in the public markets. You don't just want to get out a super high valuation.

You know, an example, a company that went out in December 11th of 2020, Airbnb is actually, if you just bought the day that it went out, you'd be down 8% after five years. So again, priced very efficiently. It did pop. But then it's just been effectively flat to down since then.

Yeah.

Well, if you're thinking about taking your.

Company public anytime soon, you got to get on Vanta, automate compliance, manage risk, prove trust continuously. Vanta's trust management platform takes the manual work out of your security and compliance process and replaces it with continuous automation, whether you're pursuing your first framework or managing a complex program.

They make compliance a joy, John. They make it a joy. It used to be painful. Now it's joyful.

It's joyful.

Go to Vancity.

Another big read in the information about Jensen Wong. He used to delegate politics until Trump's return. We've been talking about Jensen going on an absolute tear, going to all these different conferences. This picture is pretty much everyone. This picture is pretty much everyone. He's been relying on why Lago visits and a budding relationship with President Donald Trump to reopen a path to China, which demands his attention, too. So big, big questions about how Nvidia can maintain some position in China, while obviously there is a decoupling going on.

Yeah, it's interesting to see. I was too fixated on this image, which is iconic, albeit a bit silly. Yeah, since the start of this year, Nvidia CEO Jensen Huang has been thrust into an unexpected new role, chief lobbyist for his company, forcing him to carry its hopes, suggestions, and pleas directly to President Donald Trump. Many of Huang's conversations with Trump have happened at MaR-A-Lago, Trump's residence and private club in South Florida, and Huang has made far more visits there than have been publicly reported, according to a person with direct knowledge of the matter. Indeed, these... I don't even know how to say that.

But the next line is interesting. Nvidia executive to remark on Huang's absence from the company. We see a lot less of Jensen. He's been traveling to Florida a lot. But I mean, it makes sense.

This is what we were commenting on even a few weeks ago of just, I mean, this guy's conference schedule alone ignore the lobbying is, you know, feels like a couple a week, sometimes more.

Yeah. And it makes sense because like, as we saw with how Tim Cook handled the tariff situation with Apple, Even if there is going to be a decoupling, having insight into how that will unfold and what the speed limit will be in one month, in six months, in 12 months. We know Jensen's going to go one under, but what is the speed limit going to be so you don't get caught with buying a car that goes 200 miles an hour and the speed limit's 65. I'm sure it's top of mind for him, but it also seems like There's the flip side of this, like this story is being told from the perspective of the bear case of China going wrong for NVIDIA, but there's also the flip side, which is Trump is clearly reopening relationships with new countries, Saudi Arabia, UAE, Qatar potentially, and those are new markets as the diffusion rule gets phased out. And those are major, major drivers of GPU buying. And if there's a Stargate in every new country that Trump visits and opens up diplomatic relations with, you know, rolls out some trade deals. Yeah, they're 50 billion.

Yeah, the other thing is Jensen is continuing to embrace and NVIDIA are continuing to embrace China. I mean, they're setting up new they're investing more in this facility in Shanghai, more into on the ground R&D. So it doesn't feel like while Jensen is having to sort of massage the relationship with with the White House, he's not pulling back from China in any meaningful way that I can see.

Yeah, I mean, what's interesting is we were talking to Aaron Ginn about this idea of standardizing on the American AI stack. And we were only talking about the jump ball countries, right? It's like Saudi Arabia is gonna want GPUs. Those are either gonna be from Huawei, Ascend chips or NVIDIA. There's really no one else in the conversation. They're either going to be running Llama or OpenAI code or Deep Seek and Manus, right? What's interesting is like, is there a world where standardizing on the American AI stack in China is the right move? That's a very controversial thing to say or even explore, I think. I probably default against it, but I wonder if there's some sort of good ending there where keeping China reliant on American technology is actually better than them developing their own tech tree, like out competing Huawei in a free market. could work. I'm like 1% on that, but I want to hear from other people.

Jensen is finessing the situation. I'm looking right now at the Polymarket, largest company end of 2025. Nvidia is actually in the lead right now at a 37% chance with Microsoft at a 36% chance.

Wow.

So, amazing. Pretty, people believe in the leather jacket man.

Anyway, let's take a quick second to tell you about Linear. Linear is a purpose-built tool for planning and building products, meet the system for modern software development, streamline issues, projects, and product roadmaps. Go to linear.app.

John asked me this week if we could also use Linear to plan and track our workouts.

We're going to do this.

So we're going to do this.

I didn't ask you for permission. I told you. what's going to happen.

But, you know, using, you know, bodybuilding is, you know, potentially, I'm sure linear would say, you know, that's not a core focus for us. Small market, small market. But, you know, I think we can show them that that bodybuilding is, you know, potentially, you know, an entirely new opportunity. Yeah.

It's the next big tam expander for them, especially when you think about how linear has agents, you can have an agent constantly. Going out there, finding the next trend in science-based lifting, figuring out exactly what workout you should be doing, building that into your linear plan. Then you show up to the gym, just open your linear app, start checking off the issues. Oh, put in a ticket. Oh, my delts aren't 3D. My traps aren't eating my head.

Okay, Kari's gonna text us and be like, guys, guys, I know it's funny, but please don't tell the market. that we're getting into bodybuilding, but we can use it's a it's a it's an open ended tool and we're going to we're going to push it to the limits.

Yeah. Other news and the information VC funds are for sale at a discount. We were talking about this a little bit. 2025 was supposed to be a year of cash windfall for venture capital backers as companies that had delayed going public finally made the leap. We were tracking some big DECA corns that had hired CFOs that had previously taken companies public. The IPO window was supposed to be opening up. The market was ripping. Then we had the tariffs news and the market and the IPO window kind of shut a little bit. Now it's opening again, but there's still, you know, more demand than ever for liquidity. And so it looks like some funds are trying to sell.

The taco meme is really putting, you know, putting us in a precarious position that Trump always chickens out. whichever journalist asked Trump about that to his face, potentially could be responsible for another freeze in the IPO markets, but hopefully not.

So the information writes, that hasn't happened with the roller coaster stock market delaying many anticipated IPOs, LPs and VC funds have been finding new ways to cash out of their stakes in private tech firms. Some are selling their ownership of VC funds. The VC funds in turn have been putting their own startups stakeS up for sale. That means ownership of hot startups like Stripe and Databricks are sometimes up for grabs. Uh, the takeaway Nokia pension plans plans to sell, Nokia pension plans sells VC fund stakeS. Leaf fX's family office sells dozens of early, early stage fund stakeS. Buyers hunt for fund stakeS at steep discounts. And so this is kind of like a macro version of what Jeremy Cuffon is doing, right? Instead of targeting a single company. targeting the actual fund stake. And we also heard a little bit about maybe there would be pressure on some of the educational higher ed LPs, like if they're under pressure, like nonprofit for profit conversion, like Harvard is kind of going to war with the Trump administration. Would that lead to them needing to pull back from VC? We haven't really heard that or seen that, but it would be interesting to see these changing hands. And I wonder what the final holder will be, you could see someone creating kind of a roll-up SPAC or you could see someone creating a new fund. But these ideas of the continuation vehicles are not entirely new.

Yeah, I mean, it's not, I mean, it's totally normal and healthy for people that hold certain assets to be willing to sell them at a discount to what they believe the future value or the terminal value of them will be. and I don't think this is necessarily a bad thing. It just historically wasn't as common in venture because it wasn't as mature of an asset class and there was faster timelines to liquidity. But to me, all this is normal. It would be more concerning if, you know, I'm sure there's funds out there that, or LPs that would love to sell stakes in funds that don't, where there's no actual ultimate buyer. but if you're in quality funds, there's always going to be some ultimate value for the LP stake. And so it's just good to see...

Yeah, I want to talk to someone who's running a company that has a material stake from a VC fund from an early vintage that's well beyond the 10 to 12 year hold period. Because I remember talking to a founder who had sold a number of businesses and he was very successful in kind of telling this nightmare scenario where you raise in 2010, your company's doing fine, but you're not IPOing and then the venture fund that you raised from needs to distribute shares. And so all of a sudden you wind up with like 500 different individuals or LPs on your cap table because the fund, because the shares have been distributed and then you have to go public or that changes the way your you're seen by the SEC. And I've never heard of that actually happening. It's always just been this hypothetical nightmare scenario. But I wonder if there is something real to it, if there was ever a real problem, or if it was instantly solved by roll up vehicles and kind of continuation vehicles or something like that. But I'd love to dig into that more.

I don't know if you've ever run into that. Yeah, I mean, that's one solution. They can always sit in a single entity, but yeah.

Anyway, let's go through some timeline. Sara Guo is highlighting some possible topics for the AI engineer keynote, which I think we'll be covering next week. I thought it was an interesting overview of what the hot topics are in AI right now because we've moved past the AGI doom scenario and we're getting into the next era of discussions. And we've been talking a lot about these. We're putting together an informal AI day next Thursday. trying to book more people around AI that day. But we will be having an official TBPN AI day soon that we will announce. But here are some interesting possible topics. AI native UX, not chat skins, vertical AI surge, healthcare pharma, we're certainly seeing that at Sequoia, multimodal frontier video, 3D audio, obviously VO3 driving a lot of that retrieval and long term memory. I saw a funny take that somebody was like, we've switched from retrieval augmented generation to agents. And the person was just like, no, like you're, you're still augmenting the retrieval with decks. Like it's the same, like it's the same buzzword, but it's like you kind of, you gave it a new buzzword and there's a lot of like this buzzword recycling going on. Synthetic data flywheels, obviously super important robotics as we talked to Rob Taves about yesterday will be interesting to see if there are hard caps on how far synthetic data takes you in different scenarios. Closing the autonomy gap, robust agent Matic workflows. We're certainly seeing that with some of these agents that seem super promising but don't quite get there. But at the same time, we've seen major breakthroughs with deep research is, I think, the first real breakthrough agent.

What's the autonomy? I had a funny experience last night. As I was going to bed, I turned on my Matic robot.

Oh, yeah.

I was telling it to mop and vacuum and mop, and I turned it on. and it was like 17 hours remaining because it needed to basically vacuum the whole house and then figure out which areas to mop. I was like, this thing is just gonna run all night. I was like, honestly, I feel like I should pay you or something, Matic. It's kind of weird that you're doing all this work. And then I actually woke up in the middle of the night to to the noise, it was outside my bedroom. And I was like, okay, I'm actually, and then I actually went and turned it off. But it was interesting to think, you know, we've had people on the show talking about this concept of like 10-minute AGI right now where you can like, you know, have an agent do an incredible amount of work for 10 minutes. But, you know, getting to this point where you can give an agent a task and then have it be so robust that it can just operate autonomously for hours and hours and hours and hours. and even days. And I think that's where things are going to get really exciting on the purely digital side.

Yeah, I saw that video that the team sent of the Matic robot cleaning our studio, and I was very proud. It was handling the cables pretty well. I was surprised by that. That usually would throw them off.

I thought that the cables would defeat it, but no way.

Yeah.

Cell-scale digital twins in programmable biology. excited to hear more about that.

Yeah, still seems a little bit early from the folks we're talking to, but obviously exciting. And the, the, the Jacob from New.

Limit had a pretty solid case for, for some of this stuff and sort of their edge on the data side.

Yeah. Compute geopolitics, world models. We're having someone next Thursday talking about world models and Fei-Fei Li is also working on world models, which is very interesting. And then RL environments that actually generalize. All cool stuff. And you know what else is cool? numeralhq.com, sales tax on autopilot, spend less.

Than five minutes per month on sales tax compliance. Benchmark series A. Benchmark series A. We just had somebody send us a cold email earlier. I didn't get quite a chance to look at it, but they included, they typed a sentence and then said, Benchmark series A. Really? That's amazing. Good way to signal that they actually listen to the show. We got a post here from Matty Pally. I threw this in this morning. In 30 years, I'll probably be up to some boomer stuff. Like, no son of mine is marrying an LLM. Marriage is between two humans only. And my kids will be like, OMG, Dad, you're so robo-phobic.

Crazy real for sure. Yeah. Yeah.

No, I, I, I, It's funny that even the kids that are being born today will very likely have conversations like this, which is absolutely wild.

Yeah, I like this. Homies in the chat will be dropping nice AIGF, no Robo. Ridiculous. Anyway, speaking of other Robo nonsense, Mike Solana says we were supposed to have flying cars in his recent His recent calls is just Robo callers getting spanned. It's like six calls in the span of five minutes. It's so brutal.

Brutal.

There is a service. I don't know if I should directly plug them. No more Robo. There's a couple apps that you can install. And there's also the iOS functionality to reduce, if you go into the phone app, you can tell it, hey, I really want to screen calls. And then also, I think I have my phone in Do Not Disturb. all the time and you can only get through to me and actually like ring the phone if you're in my contacts list. Otherwise you have to text me first. And so like there's a whole bunch of things that you need to do to kind of prevent this from being annoying, but it's still just like ridiculous. My take was we really need to come up with a new phone number system that's like a crypto address, like a hash where, because there's only what, like 10 billion phone numbers and so you can just RoboCall them all. and it's not that expensive to do if every call is one cent or whatever, you can just call every human and just test them all and I'll test every number. But if you had this really complex hash where there's now quintillion possible phone numbers, it's like, yeah, it's probably gonna be hard to remember, but once I give you my number, you're gonna be able to call me and I'm not just gonna be able, and people aren't gonna be able to randomly guess my phone number. So I don't know. We need something because it's bad. What's your take?

Something. Yeah, I just don't pick up the phone unless it's somebody I want to talk to. And then I pick up as fast as I can. This ad went pretty viral from Zoran, Mom Donnie. He's running for mayor of New York. And he's got a bold pitch. Freezing the rent.

Price controls.

Price controls.

You love it.

I saw somebody compare this to, you know, somebody running for like student body president or whatever, and they're just like, free lunch. Free lunch for everybody. I know people say there's no free lunch, but we're doing free lunch.

If I'm elected, free lunch. Yeah. Yeah. So tech was pretty against this. Patrick Collison posted just a really quick screenshot of ChatGPT. It says AI is getting pretty good. He says, does rent control tend to help or hurt cities? summarize your answer in one objective, pithy and blunt 280 character assessment. And the response is rent control is a short-term shield for incumbent tenants and a long-term poison for cities. Caps freeze rents for the lucky few who but choke new construction, shrink supply, jack up market rents, misallocate housing and shift costs onto everyone else. Good politics, bad economics. And so this is kind of the general consensus that like rent controls are bad. I wanted to offer some bull case for this. So first off, people are saying, oh, prices are gonna go up, there won't be any new construction, housing prices are gonna go up. But if price controls really spread and then there's a kind of cultural revolution type like mass famine and the population declines by like 90% cuz everyone's dying, well then there might be so much housing that the price of housing falls for the remaining that made it through the famine.

Even below the original cap.

Prices fall. Yeah, exactly. And so counterintuitively, it could be an interesting way to drop prices if you're okay with everyone dying of famine.

Yeah, Zoran just might be thinking in decades while everyone else is thinking more immediately.

Exactly. And then the other case is, the other bull case is if you hate New York and you're pro California, pro LA. Like we're pro Hollywood. I don't really care about New York City, so I might be donating a ton of money to this guy to try and destroy New York City so that everyone, all the best people in New York have to move to Hollywood and hang out with me.

Yeah, I wouldn't be surprised if you have SFVCs and New York VCs sort of, you know, basically funding the politician that they think is most gonna damage that startup market to try to you know, get themselves a little edge. I just basic game.

I like basic game.

This guy Nikita Beer I saw in the comments on Zoran's post saying that, you know, he said that housing should be free, you capitalist pig.

That's true.

And so didn't go far enough. Yeah, didn't Zoran really got mogged by Nikita.

It was rough.

And the silence from Zoran on Nikita's response has been deafening. to say the least.

Yeah. Anyway, so silly. Anyway, we got to tell you about public.com investing for those who take it seriously. Multi-asset investing industry leaders trusted by millions. Let's go to Delian. He says, as a private pilot, my favorite part of this season of the rehearsal is that while it's an objectively a hilarious show, Nathan Fielder is also pointing out an incredibly critical pilot training issue that seems to be a regular contributing case to crashes. caused to crashes, including the DCA crash. And Kristoff asked him what the pilot training issue was. And Delian said, First officers not being comfortable with pushing back against a captain they think is making a mistake. I think this is funny because I was watching the show the first couple episodes, I haven't finished it, but he identifies this and he's half joking, half serious, but Delian is just like, no, it is real, it's real. And so it's this interesting way where I actually could see this comedy show having an effect on FAA regulation and actually increasing safety. He's doing good for the world through his comedy. It's amazing.

It is, everybody should go and watch it. It is some of the most entertaining, probably the best show I've seen this year.

I was watching a little bit of the original first season of Nathan for your. last night and I was crying, laughing about when he goes into that interview and he has the comedian telling him what to say and he's like, oh yeah, like my mom passed away and the guy asks him like how? And he just says like, in the war? Like in what war? Like the Iraq war? And he's like, what branch of the service was she in? And then he has to say, his mom was fighting for the Iraqis in Saddam's army. It's just like, Absolutely.

I went back and watched, I don't remember what season or episode it was, but there was the bit where he's helping, he's giving advice to a woman who has a pet shop who wants to increase business. So he goes and helps and he basically gives the advice to the woman that what's the right moment to try to get in front of a future pet owner or somebody that's about to, you know, buy a new pet. And his idea is to go to a pet cemetery and get the biggest possible tombstone. And so he works out a deal with the owner of the pet cemetery where he basically gets the guy to agree that for like $1,000 one-time payment, he can erect like six foot by four foot tombstone. And so it's just like this pet cemetery. It's like, we're sorry for your loss. Like this tombstone is brought to you by, you know, whatever.

Very, very on brand for us. Pro advertising, for sure. Yeah, that's great. Yeah, it's a classic show. Go watch all of it.

Yeah, the other show, now that we're talking about it, Mountain Head releases, I believe, on Saturday around 5:00 p.m. It's by Jesse Armstrong, who obviously created Succession, Mountain Head, seems to be based on the All In podcast or loosely, but it's a group of billionaire friends get together against the backdrop of a rolling international crisis. So I'm looking forward to watching this over the weekend.

Sounds amazing. Let's go to this post from Kevin Kwok. He says, When exactly did the flip happen that all announcements have to be videos? That flipped really fast. What do you think? It was us. Did we do it? Can we take a victory lap?

I mean, we've been doing video announcements for years.

Party round.

Yeah, we use these heavily in 2021, but I also feel like they were popular then. I also still think there's incredible alpha in just creating an amazing text post. You can get the same, it doesn't give you the level of depth to really deliver brand, I think, in the same way that a video does. but I honestly think a extremely well-written text post assuming you can even, there's potentially a hack to just include a lot of mdashes so you get a bunch of comments from people being like, you use ChatGPT for this and then that will actually boost the post and the algorithm. So there's potentially alpha and plain text, you know, write it yourself but make it look like it was, you know, fully AI generated. But these things, these things flip back and forth. I definitely see videos now where I'm like, okay, this is based on the 2023 launch meta. And I'm excited to see what comes down the line.

Yeah, I think the new meta.

That.

Comedy video from that YC company, I just saw it. Rippling and deal. That was fantastic. But then the other one that was really good was from Zipline, where it didn't have the same hard tech, insane free bird vibe real with American flags. It was this sun-drenched, high saturation. It felt like it was out of a Linklater film, and it was just this beautiful pastoral, illustration of what life could be with Zipline. And I thought it was really beautiful.

Yeah, I think where I would like to see video goes is go are if you don't have any, if you're not trying to make people feel something, just explain really clearly what you're doing in a text post. Yeah, that makes sense. It sounds like Vibral. If you're trying to entertain, inspire, et cetera, like if that really is your goal, then video can be the right form factor. And we saw that with that sort of humorous video from the YC company. And then there was another company called Spawn that went super viral a while back for basically saying that they took the team and just lived in Europe for a few months. And that was one where I couldn't really quite tell if they were being serious or joking. And that's part of maybe what made it go so viral. And then you have the Clulee, where you can go make half a million dollars a year making videos at Clulee. So, great, great opportunity.

Potentially clickbait, who knows if that's real, but certainly entertaining. I think the question of the Vibreel is, it worked for us early on because we were trying to create a brand specifically, and I think a lot of companies They obviously like a brand is important to every company, but like I would say like most B2B SaaS companies are not necessarily brand led or they're not exclusively brands versus if you are building a fashion or media or, or nutrition company or something that has like badge value that like expressing the brand and building the brand world. We have this example of like, of like, you know, you know, it's a brand when you can immediately say, like, what is the car associated with that brand? The classic example if Nike built a hotel, you would know exactly what that hotel would look like. It would have posters of Michael Jordan on the wall. The gym would be insane. You could buy the shoes there. It would feel Nike branded. Now, if Hilton made a shoe, you'd be like, what does that mean?

Hilton 100s.

But does Hilton really need a brand as big as Nike or as clearly defined as Nike? It's kind of like, if you're, if you're building a product that's not necessarily needs to be brand led forever, you don't necessarily need to ape into let's do a brand vibe reel to really tell people what our brand is. If it's going to be just like, and just a software company and the value prop is what the value prop is. Yeah, totally. Anyway, nothing has a clearer value prop than eight sleep. I've been missing mine while I'm on the road. It's pain, but you can get yours@eightsleep.com they have a five year warranty, 30 night risk free trial. free returns, free shipping. How did you sleep last night, Jordy?

I was, historic numbers. Historic numbers, eight hours, 18 minutes.

What?

99.

Huge.

Huge.

Congratulations.

Yeah, that was despite being moderately interrupted because I let my Matic work too hard overnight. But I did that to myself.

Oh, well. Well, we have a post from AdQuick, then we'll give you an ad read for AdQuick. AdQuick, they have a new AdQuick CoPilot. You can generate generate AI ideas for your out of home campaigns. Yeah.

So what's interesting about this to me is AdQuick obviously has a, you know, you can go into their application and figure out how you basically build a campaign off of that. But a lot of my experience with out of home is like, you want it to be conversational. It's like, okay, here's the general areas where we want to be, what inventory is available these dates. okay, what does that inventory actually look like? What are some examples? And so you can build a campaign in a sort of conversational way, which makes a lot of sense. So go check it out. Even if you're not ready to run a campaign, it's a good idea to get a sense of what these will actually look like, what budgets will look like, and all that kind of thing.

Yeah. And just strategically, this is like the classic AI cherry on top business. They're not pivoting to AI. giving people a different way to interface with their proprietary data set, right? Because this is something that you can't just go to ChatGPT for because they have billboards. So home advertising made easy and measurable. Go to adquick.com, say goodbye to the headaches of out-of-home advertising. We need, we got a massive size gong, Jordan, for Michael Dell. One of the greatest to ever do it. In Q1, we received $12.1 billion in AI orders surpassing the total shipments for all of of fiscal year 2025 and leaving us with a backlog of 14.4 billion.

If John, if you were here, I'd be hitting the size gong, but I'm just hitting it in the air here. Pretty amazing, interesting story. I didn't realize that Michael Dell was back in the, is he, oh yeah, is he back in the weeds? Oh, he's C, I didn't, how did I not realize it?

You gotta do a whole take private, go public again. He's been an absolute tear.

Wow. Wow. Dark horse. Dark horse. So cool on his LinkedIn. It says chairman and CEO of Dell Technologies 41 years, four months.

It started when he was like 18 or something. 19.

So good.

Incredible.

Congrats to the Dell team on the milestone.

Congrats. Yeah, and makes sense. I mean, they don't fabricate GPUs, they're not in the energy side, but obviously they're building large racks and servers for AI. There's a boom and they're taking advantage of it, so makes sense. And we're seeing that with like Cisco too. Remember we were talking to Kevin Wheel about how Cisco plays a really interesting unique role in the AI boom. It's a company you haven't heard about for a while, but networking gear is obviously extremely important in AI contexts as you're networking together all these data centers. And so Cisco is also a beneficiary of the AI boom. So, Diving in, revisiting these companies that are so historically significant is fascinating. There's a great founders episode on Michael Dell that you should go check out if you want to hear about Michael Dell's approach to entrepreneurship. So check that out. I don't know if we can sing. I think singing over Zoom just doesn't work. So we're just going to have to say wander, find your happy place. Wander with inspiring views, hotel, great amenities, dreamy beds, top tier cleaning, 24/7 concierge service.

It's a vacation home, but better folks.

We'll be back on Monday with a song, a full rendition.

We have to go hit the track with John Andrew, too.

Sure. I want to see how he performs. Laguna Seca, you think?

He's an absolute dog. I think he keeps his track cars in Texas, if I remember correctly. But anyways, that will be a great story when the time comes.

We'll be in Texas. Go visit him. Race him.

We got a post from Chris Pike. He has a new post. He dropped the end of software on May 31st, 2024. He's got a new post called the Dawn of Infinite Code. And I mean, he plays X on...

Built different, I was about to say. He posts the link and it works every time.

Still does numbers. So he has a new article, the Dawn of Infinite Code. All started off It says San Mateo 2005, three founders add an upload button to their video hosting website in an office that smells faintly of pepperoni. Everywhere 2025, a person types an idea, presses enter, and a network of agents compiles, tests, and executes. In the end of software, I paused it. Paused it. That large language models would drive the marginal cost of software towards zero, just as the internet did with content. A year later, that forecast is playing out. Code is beyond trivial to generate. The question now is how we deal with the oncoming flood. The pattern is familiar. When YouTube launched in 2005, its first video, Me at the Zoo, seemed immaterial. This was the harbinger of doom for the media industry, but that was the point. It proved that content had become too cheap to meter. Now the dam has broken with code. Prompts produce prototypes that once required a team and a sprint. As transformative as broadcasting from your bedroom was, anyone can now ship their ideas. And you know what immediately comes to mind is we have somebody visiting us at the studio today that built an entire app based on TBPN yesterday, just randomly had the idea and sent it to us this morning. I think it's at tbtn-guests.com yeah, that's right. tbtn-guests.com is amazing, cool looking website that ranks every single guest that we've had, including gives you timestamps and breaks down the topics that it covers. So very cool. This was built yesterday. Extremely impressive and probably wouldn't have been possible to build something as cool and beautiful.

That would take a lot longer for sure.

So anyways, continuing Chris Pike's post, he says, Web 2.0 gave everyone a single verb upload. and it vaporized newsrooms everywhere. AI gives us a new one to scribe and the 500-person single feature company goes poof. The work shifts from syntax to intent. Instead of writing software line by line, we define what it should and let models handle the show. And what about distribution? Don't software companies have robust sales and marketing teams? Yes, as newspapers with paper routes and circulation numbers when free competes with free for distribution, paid distribution loses badly. At first, these models will rely on a set of default environments and stacks determined by a legacy training set. But defaults change as options multiply the decision of where and how code runs will be handled by algorithms. We already accept this logic elsewhere. TikTok decides what we watch, Spotify decides what we hear, the system is opaque, the results are good, so we scroll. When the flood of code arrives, we won't ask for explanations, we'll reward speed. Convenience will outrank transparency. The systems that earn our trust through responsiveness will hold the power once reserved for operating systems. As cost curves come down, we chase scarcity up the stack. It starts with compute, still mostly centralized but already seeping to the edge as local models take up residence on laptops and phones. Above that lies context, the habits, preferences, and history that let a model anticipate us better than any public corpus ever could. And finally, trust, credibility earned over time determines whether an agent has the right to choose on our behalf. Once we arrive at full trust, active choice becomes a burden. Do you want to continue?

Children after 2010 have never opened a newspaper, so they may never open applications. History shows the vessel always fades once the payload can flow without it. Newspapers yielded to website CDs to streaming software in its current circumcised form is circumcised form is the next container on the chopping block. When logic can be summoned and stitched together on demand, the container evaporates and only capability remains. That's the real end of software, not the death of code. But the moment the container dissolves, utility unboxed everywhere. And all at once, infinite capability, infinite code, and very interesting.

Yeah.

I mean, it's a fun narrative. I wonder where, if there will be, there's a big question about discovery, how will people discover these vibe coded apps? Distribution seems to be king. We saw that with Levels IO. He was able to vibe code. He had a very different outcome from the random vibe coder who wants to share a link on Axe, unless you're Chris Pike, apparently, and you're not shadow banned from posting links. I mean, good links will break through, of course. And of course, you can always do a viral video that announces your website and pitches it and then gets people to go there, even if they have to open up a new tab. But distribution will be king. So get ready to optimize. And the value of the prompt will be important. Like the value of actually coming up with a unique idea something that creates value, even if the instantiation of that value happens very quickly via vibe coding.

Yep.

Anyway, let's tell you about bezel. Go to getbezel.com, your bezel concierge is available now to source you any watch on the planet. Seriously, any watch. Check it out. And I want to close out with this post from Guillermo Rao, the founder of Vercel, who's been on the show. The only durable moat in technology is relentless execution. And we got Molly O' Shea in the comments or violent execution. I like that. But this relates to the previous Chris Paine post, right? That as the barriers to entry come down, the winner is the one who executes relentlessly, right? And we see this all over the place.

And I would put an image in people's minds. Relentless execution is like a golden retriever chasing a tennis ball, right?

That's a good metaphor.

No matter what is in their way, no matter what gets in their path or presents challenges to getting that tennis ball, they're going to relentlessly charge at it until they get it and bring it back to you. And that is kind of like, I think, the right framework to approach life with, right? A golden retriever is not rabid foaming at the mouth, being a poor sport. They're just joyful, relentlessly chasing that tennis ball. And so I hope everybody brings that energy to the rest of their work week.

I agree. I agree. Go have a Golden Retriever weekend, everyone. This has been a fantastic show. Thanks for tuning in. Are you gonna see on Monday? We have a whole bunch more guests. We're doing a bunch more of these special days, and we'll have a whole bunch more ads for you, of course. Leave us five stars on Apple Podcasts and Spotify, please, and stay tuned for the next one. Thank you.

Cheers.