Palmer Luckey, Brian Armstrong, & Brian Chesky LIVE | OpenAI Joins the Browser Race, AWS Outage Aftermath

You're watching TVPN.

Today's Tuesday, October 21, 2025. We are live from the TVPN Ultradome, the Temple of technology, the fortress of.

Finance, the capital of capital.

There we go.

It is good to be back.

Massive news out of OpenAI. They launched their browser. It had been rumored for weeks now, months now, the browser wars have been heating up. Perplexity has been in the game.

Thought they were over.

DIA browser browser company has since been acquired. Since all of this was going down, OpenAI dropped a trailer announcing the browser, which is called Atlassian.

By the way, the Atlassian browser company acquisition just closed today. It was announced.

So Atlassian owns a browser and OpenAI's browser is called Atlas.

They keep doing that.

Wow.

Remember they did that with Google IO?

Oh, yeah. IO I Yo Iyo interesting. Taking shots. Well, I'd love to play the video from OpenAI announcing their browser Atlas. Do we have that video? Let's play it. Here we go. I like the different sound cues. This one that sounds kind of like typing on an iPhone.

I mean, it sounds exactly like typing.

Is that what it sounds like?

Okay.

It's funny. I like the sound cue in videos, but I turn it. I certainly turn the sound off on my actual phone so I don't actually hear it very regularly. So this is. You're on all trails. Is that the website? So you're in Coursera and you're saying, cheat on my homework. Wall Street Journal summarizes for me. Interestingly, the Wall Street Journal actually has AI summaries now in every article. Just a couple bullet points there. I don't go to them a lot though. Interesting. That looks like a Gmail. That is going to be kind of a wild interaction. I feel like the war, the browser war is less significant than the war for like my actual email because every time I'm in Gmail I'm getting all these different pop ups from like Apple Intelligence is saying like, do you want to rewrite this with Apple Intelligence? And then Gemini was like, do you want to. Do you want to rewrite this with Gemini? And now OpenAI is going to be like, actually we. You want to be the ones to rewrite your bullet points into paragraphs and vice versa.

Comments like, we will pay you $100 a month to use this.

We will pay you to rewrite your bullet points so it does some online shopping. Tyler, were you successful in your prompt? Give me your review. You actually downloaded this thing.

You're absolutely right.

Yeah.

How would you describe chatgpt Atlas?

Yeah, I Mean, people are saying, you know, it's a new browser. It's not just a browser. I think it's kind of a whole new way to really use the Internet.

Browse the web.

To browse the web, to browse the web. But so the first thing I tried, I wanted to try to use the new agent mode.

So you were successful. It's not a beta, you just were able to get it immediately.

Yeah, you can download it.

So it's not like Sora where you needed like an invite code or anything?

Yeah, no, I think it's. Right now it's only available to pro and plus users, but you can, you.

Have to log in with an OpenAI account.

Yes.

Okay. So it's just the same thing under the hood and maybe no different inference thing. We can get into that. But yeah. What was your experience?

Yeah, so the first thing I tried was I just want to use the agent mode stuff. That's like really what's new here? So I tried to buy a unitary robot and then it went.

And so you open it up. New tab just gives you a prompt. You can type a URL or a prompt. You type a prompt?

Yeah.

And your prompt is by.

I think I said 100.

Gabe in the chat says it's Mac only.

Oh, interesting.

How many of ChatGPT's 800 million weekly.

Actives do you think are the weekly actives, though? This is just for the paid users. And I bet a lot of the paid users are on Mac disproportionately, for sure.

And definitely pro users.

Yeah. Name pro user.

I don't know, but they're trying to go out. I mean they're trying to go after Microsoft.

You think they're going after the PC master race, the desktop? Desktop.

I mean, just a lot of people use PCs in the workplace.

I don't know. Not that Linux. Yeah, well, you're trying to deny, you're.

Trying to deny that people use PCs.

I mean, I wouldn't be surprised if the penetration. I mean it's the same thing for like when you launch. And when you launch an app, you know, you always like Sora is iPhone only.

Yeah, but iPhone Android is different than Mac PC for trying to build large scale Internet products.

I don't know, I think it's pretty close. I think it's closer than you'd think. Anyway, Tyler, what else did you. How successful were you?

Yeah, so I think I said go buy Unitree robot and then first thing it goes to is Amazon.

Okay.

It failed, I think.

Yeah. How did it fail?

Yeah, so I think it was when it was trying to do customize, like it loaded the page correctly. I was a little bit surprised by. Because people have said that it's been blocking a lot of the search.

Yeah.

But you would imagine that this is. Since this is happening in a browser, the user agent is just like Chrome browser or Chromium or something like that. And it looks and it feels like a normal user interaction. Yeah. It's puppeteered by AI. But in general you're like. If you're the web server, you're just seeing. Oh like, like Tyler who sends in, you know, one Amazon get request to some web page every once in a while, just send in another one. This doesn't look suspicious. It's not like, you know, some deep research query where it's firing off like 50 queries really quickly. So.

Yeah, that's true.

Seems reasonable.

Okay, so some reason, for some reason didn't work. And then.

What do you mean it didn't work? Like it's not available on Amazon or.

It said so I think it was like during you can press customize and it just said error, error. And then it went to a different site.

Interesting.

But then eventually, I mean it went through the whole flow and got to the button where you do like confirm. But one cool thing they added was when you do the agent, you can choose to either be logged in or logged out all your accounts.

Interesting. Yeah, it's a new browser, so you have to log into everything. You'd have to log into your Amazon, log into your Netflix or whatever.

I think you can very easily port them over.

Oh, from your previous Chrome. Oh yeah, you could save all the cookies or something. Interesting. But Chad agrees with you, Jordy, by the way, everyone says Enterprise runs on Windows and you're right about the PCs. I still disagree, but we'll see.

You're right.

I didn't say you're right. I said chat thinks you're right. And yeah, I disagree with you. And chat now fight me. Well.

I think we can like the facts are the facts and then we can disagree on whether or not it matters.

Yeah. Okay, so does it matter? Will this be successful?

So here's what I believe. Yes, I believe that ChatGPT, especially once it had some close to live search, was five to 10 times better than Google search for a lot of searches. And my prediction is that Atlas in its current form might be 1.1 times better than Chrome. Yeah, and that will not be enough to get large scale consumer adoption. Granted, I haven't used the product yet I need to get access to it after the show today and play around with it. But it's very different creating a product that's 10 times better and competing in that market, competing in search, which was like the browser, basically a monopoly. And so again, I'm not. I'm not super bullish, but I think it's great that they're taking a big shot on goal and it makes sense strategically.

Well, if you're looking for a product that is 10 times better, go to ramp.com Time is money save. Both easy to use, corporate cards, bill payments, accounting, and a whole lot more all in one place. I agree with you. I think that if I'm comparing what the deep research experience of the GPT5Pro experience is to a Google search, it's like you're searching for the story, the history of the browser wars, let's say, and you might wind up on one web page that kind of put it together, and then you might be on a Wikipedia page and you have a million tabs open and you're tabbing through them all. And GPT5Pro just summarizes that all into a nice little dossier. And it's a great experience. I agree with you on the 5-10x better than the typical open a million tabs now.

Yeah, and historically you could get what would have taken you navigating through a bunch of Reddit comments and kind of summarizing them in your head. You could get that in a quick block of text.

That was remarkably the question is on the browser, like, how sticky is the actual browser? Because if browsers just aren't sticky at all, people will upgrade to something that's 10% better. But I have a feeling that they're pretty sticky. I feel like people feel like their bookmarks are locked up, all their logins are in one thing. But at the same time, I've been running this weird combo for years of Chrome on desktop, Safari on mobile, and I feel like that's very. I guess it's somewhat common. Some people do it, but it doesn't really make sense because you'd think you'd want one unified browser across desktop and phone. And it's not like I'm not using Mac products. I'm using a MacBook and an iPhone. And so you'd think I'd either use Safari on both or Chrome on both. But I've never been a single browser, so I feel like getting me to install this and use it is not that difficult. I don't know. I feel like as long as it has most of the same features and it's feature complete, which, like, I need PIN tabs, I need to be able to hit Control T to open a new tab. There's a few other features that are probably valuable, but as long as it has those, I'd probably be down to try it. And then if I tried it, I'd probably stick around for a while. But the question is like, would I, would I like, what's the activation energy for people? Because you need to convince them, like, hey, let's.

Kari had a good post earlier at 9am saying me, my daily driver in a peaceful corner of the Internet. Safari power user.

Safari power user. I wonder if I should switch to Safari on desktop. What do you use?

Chrome Desktop. I've gone back and forth from using Safari on mobile and Chrome on mobile.

Yes.

I think they're both fine. I just don't do a lot of web browsing on mobile. You don't like here or there.

But that's so funny. I feel like I'm always opening web pages on my phone.

Yeah. Opening them briefly. Oftentimes it's within another app.

Yeah. And so that's why Safari is fine. Because if you open, especially if you're in the Safari web view, if you open a link in X and it's the Wall Street Journal and you're logged in in Safari, that login transfers, whereas if you're logged in in Chrome, it doesn't transfer. That I'm aware of or. That was a big problem for a while. I do wonder. The new Liquid Glass update to Safari is terrible and I really don't like it.

And I. I mean, Liquid Glass, we can all agree, is just terrible.

It's rough.

Like, I've had my new phone for a while now and the only nice thing I have to say is that the camera is a little bit better.

Yeah.

And everything else about the phone in totality, in my view, is worse. It feels cheap. It's getting dinged up constantly already. The software overall, every single time I'm a new place in iOS, I'm like, oh, thank you. You made it worse. Yeah, you made it different and it's worse.

Yeah. I mean, the fact that they went from. To open a new tab in Safari before Liquid Glass was just one click, I think, and now it went to two clicks, something like that. So every time I wanted to view the tabs, I have to click two buttons and I'm like, this is just so frustrating. And I was using someone who had Google Chrome in Liquid Glass on a new iPhone and the Google Chrome app on Liquid Glass still has the ability to just launch a new tab with one button with one click. And I was like, this is so refreshing. Maybe I should go back to this. But anyway, whole bunch of news around OpenAI. There was a massive long read.

And first of all, let's say thank you to Sam Altman for constantly creating content.

It really is like the bull market in tech news. There's always something to talk about.

It's always something, always something to talk about. Well, today Jim Cramer read the piece we're about to talk about. He said Wall Street Journal's OpenAI piece makes you realize Sam Altman must succeed or he could be a real problem for an otherwise sterling industry.

Yes, well, OpenAI has been doing a stream. Let's hope they're using Restream, iO1 livestream, 30 plus destinations, multi stream. Reach your audience wherever they are. Let's, let's look through the Wall Street Journal piece. There was one big scoop in here that I wanted to get to. I had a whole take on this. My, my thesis, and we were sort of debating this, is that Sam Altman is kind of becoming the preeminent deals guy of the modern tech era. He's done so many deals at such a huge scale for such a young company. He's a young founder, young CEO, he's only 40. And now the deals are so big and they're coming so quickly that a lot of people are asking, the timeline's wondering.

People are nervous.

Is Sam just so good at deals that all the counterparties are actually making mistakes by tying their fortunes to OpenAI and OpenAI's aggressive forecast? Like, if OpenAI doesn't deliver and Oracle does all the build out, but then OpenAI doesn't have the revenue, is Oracle in trouble? That's the narrative right now. And so we love to use the fox and the hen house analogy around here. It's too funny to pass up.

I could go out on a limb and say it might be our favorite analogy.

It's such a good one. We're big into animals here with the horse, the fox, the fox and the hen house.

Yeah, that's our next sculpture. We have the horse.

We need a fox at a henhouse for sure with a fox going inside, smart looking. And I want that. But that is a pretty good analogy here. So, you know, if OpenAI ends up getting the fox's share of value from deals with Nvidia, AMD, Oracle and SoftBank, like, I don't exactly feel Bad for Jensen, Lisa, Su, Larry and Masa. Because the primary job of a CEO is to defend the henhouse from foxes. Like that's your job if you're a CEO. So just saying, oh, Sam's a good deals guy and he's like getting these people to do these crazy deals, like that's not enough. Like the other, the counterparty needs to make good decision making here and not risk themselves. So the question is like, why are these deals happening? I don't think Sam's gotten any foxier now that he's 40. He's going to be a silver fox. But he's always been a great dealmaker, right? He's always been a great deal maker back to the YC, back to solving logjams with founders and VCs. There's so many examples of this throughout his career, even going back to his very first company where he got looped, acquired when the company wasn't doing that well, sort of one of the first acqui hires that turned out really well. He made well financially out of that. But now everything has 3, 4, 5 extra zeros after it. And I think that the reason isn't just that he's become a better dealmaker, it's that he actually has more leverage now. He has a lot more leverage.

And so when you have the breakout consumer product of a generational tech trend. Yes, you have a lot of leverage.

Yes.

And you've been saying, and yeah, so I was saying to you off, off air, if you're running AMD or Nvidia or Oracle or any hyperscaler, you know, computing company, and you have to go in front of shareholders and they're asking you, are we doing anything with OpenAI, the company that scaled to over a billion users, the thing that will be.

The next Google, the next Amazon, the next Microsoft.

And if you have to go in front of them and say no, yeah, they're, they're, they're like, okay, what exactly do you do?

Why are. Yeah, yeah, why are you missing this? And so I think that that's a good framework. But what do all the hyperscalers have in common? They all have monopolies over whatever market they're in. Whether that's search, whether that's online shopping, whether that's, you know, the enterprise software stack that Microsoft has been able to monopolize in many ways. And this is the teal monopoly thesis. And Sam Altman, you know, he's basically created a new aggregator. So ChatGPT has captured enough of the consumer market that they now control Demand at a very high level. This is what Google's done, this is what Meta has done. And that affords them an incredible amount of leverage in every negotiation. And that's why Meta has taken so much of the lion's share of the ad market and the D2C market is because they aggregate the demand, they aggregate all the users, all the eyeballs, and then they can point the value where they need to, and so they can take a large cut. And Ben Thompson was actually road testing this theory that OpenAI could swap out the underlying model and still accrue the vast majority of value. So hypothetically, you could have chatgpt, chat.com powered not by GPT5, but instead powered by Gemini or Claude, and still see positive user growth monetization. Because at the end of the day, most of the users aren't there for the model there for the app, they're there for the ecosystem, they're there for the platform now. And so if OpenAI is potentially going to be successful in building monopoly at the consumer AI layer, what's the strategy for the supply chain? So Joel Spolsky in 2002 coined this phrase, commoditize your complements. And so anything that complements your product, you want to be a commodity so that it can bring more content onto the platform. This was the thesis behind Meta Open sourcing Llama. If it's free to generate content, there will be more content. The platform, just like the Sony camera, you know, commoditized the video production, made more content for YouTube. And so that means that OpenAI needs to partner with multiple cloud partners, Microsoft, Oracle, multiple data center builders. You got Oracle, Core, Weave, N Scale, Crusoe, multiple GPU designers, Nvidia and AMD multiple foundries. Maybe tsmc. Sam was saying TSMC does not produce enough chips. Maybe Intel's coming into the picture, maybe Samsung, There's a whole bunch of other options there. And so not every link in the supply chain can be fully commoditized though, because some of these, like there are only so many GPU designers. And so if there's only two major players, AMD and Nvidia, you need to balance them out so that they have similar margins and are in direct competition.

Yeah, and the memory market historically was a commodity market, but the players like SK Hynix are now telling shareholders, like, we actually think that this new computing wave will make that no longer the case.

Yeah, yeah. And so there's a little bit of this desire to create an anti Nvidia alliance. Right now, Nvidia Ramped full year revenue in 2023 was 27 billion. In 2024, 60 billion. And 2025 is 130 billion. So just a massive revenue ramp. But during that time, net profit margin went from 16% to 56%. Like insane profit growth. And so, so you have this company that's just printing cash. All of their partners are looking at them, whether they're Google with the CPU program, Amazon with Trainium, OpenAI say AMD pretty good.

Sam's looking at that margin profile, just being like, Jensen, I'm so happy.

Does it have to be 56%? Why can't it be 30%? That's what AWS makes. That's what Azure makes. Roughly. That's what GCP makes, roughly. Everyone's doing 30% around here. You're doing 56% in your hen house.

Why don't we get a. The whole industry is losing money. Money. And they're like, we're trying to find the one guy that's making money.

Exactly.

And they see Jensen absolutely printing.

If only there were another plate that we could eat off of. And Jensen's just there with like the massive Thanksgiving dinner on his plate. And you're like, I'd like to eat off of that plate, actually. And so everyone's kind of incentivized to like work against Nvidia. And that's the, and that's kind of the evidence of the leverage that's happening right now. That was in this Wall Street Journal article by Berber Gin. And so the quote says, as part of the deal, Nvidia is also discussing, discussing guaranteeing some of the loans that OpenAI plans to build to take out to build its own data centers. People familiar with the matter said a move that could saddle the chip giant with billions of dollars in debt obligations if the startup can't pay for them.

Papa Jensen. Co signing.

Yeah, yeah, yeah, exactly.

Co signing the lease.

It's like co signing a mortgage with your parents. Basically. It's like, you know, I have 13 billion in revenue or whatever OpenAI has and they have 130 billion in revenue. It's high margin. We're losing money. They're. So let's put them on the, on the loan docs and they'll back us up. And now this can work out great. If there's huge demand, it can be fine. But at the same time it is a risk that otherwise you'd probably just say, nah, I'm good, I don't need to do that. Why would I sign for your loan? Well, the reason is because you have to stay in the game if you're Nvidia. And so dual sourcing is nothing new. It's best practice. But we just haven't seen a push like this happen so fast before. Like exerting pressure deep into the supply chain is usually a hassle and something tech CEOs only do when they're backed against the wal wall. And single points of failures creep up really unexpectedly. I think about Apple in China. Like if you were in 2005 and you're Apple, are you really thinking like how advanced, how far in the future would you have to be thinking to be like, wow, I'm really indexed to China. I should start spinning up India and Vietnam. Like that was not something that was top of mind for Steve Jobs and Tim Cook two decades ago. Now it seems obvious. Now it seems like, why aren't you doing this? Why didn't you do this earlier? But these things creep up on you. OpenAI is just unique in that they're actually looking years out and saying, well, we want leverage at every single piece of the supply chain all the way down the stack right now. And we want to set up these deals so that we're there when it matters, in 10 years, in five years, whenever.

We should get into the article a little bit because there's a couple things.

Yeah.

What else do you want to read from here?

One, allegedly when Jensen had heard that Sam was considering buying TPUs.

Yeah.

That is when he.

That's another point of leverage.

They had been basically Sam and Jensen, high level, had been working on some type of deal wasn't really going anywhere. Jensen hears that Sam was considering TPUs, he called Sam and that's when this $100 billion investment, obviously it's structured and in tranches, but that's when it got announced. So it was in reaction to this interest in TPUs. Obviously Jensen wants everybody to stay and.

I don't think that's a headed. I don't think that's like oh, hypothetically we could run our system on TPUs. I think that they could do. Doesn't seem like the CUDA lock in. It just does not seem that crazy because we saw this with inference max from semi analysis. GPT OSS runs on AMD chips right now very efficiently on a token per dollar basis. And so you're like, what does it take to get GPT OSS or the frontier model running on TPUs? Like I don't know, a year, A billion dollars. And so is it worth it? Absolutely.

So Sam is straying A little bit. He's interested in other options. Jensen says, whoa, whoa, whoa. Okay, we can figure something out here. Figures out this hundred billion dollar investment.

Yep.

Two weeks later, Sam announces a deal with amd.

Yep.

And this is obviously Jensen would have been incredibly frustrated by this for a number of reasons. His reaction, very presidential, was I saw the deal. It's imaginative, it's unique and surprising considering they were so excited about their next generation product. I'm surprised that they would give away 10% of the company before they even built it. Anyhow. It's clever, I guess. So that is a very presidential way of saying he's incredibly frustrated in my view. So imaginative, unique, surprising. I'm surprised. It's clever, I guess. And I mean, so anyway, I mean, look like Sam is doing what he needs to do, Jensen is doing what he needs to do, but they're playing, you know, a game of high stakes chess with the global economy. As far as I.

High stakes hen house control in the farm on the barnyard. It really is a barnyard because the fox goes in the henhouse and then the piggies are at the trough and the trough is on the barn. It's all barns and we're, you know, we're just riding by on our stallions.

The world is a farm.

The world is a farm. This is a great analogy. We need to flesh out farm theory fully. Yes. On the whiteboard for sure. We need to have the full farming map mapped out before we move on. Let me tell you about Privy wallet infrastructure for every bank. Privy makes it easy to build on crypto rails, securely spin up white label wallets, sign transactions and integrate on chain infrastructure all through one simple API. I have one last thing on the Wall Street Journal article if we can pull it back up and scroll down. Berber Ginn is a journalist who writes very fact based, heavily sourced articles. But and it's not an op ed, it's not his hot takes, but if you look at the images that they chose, you can really understand what's going on here because the picture of Jensen that they picked for this article is insane in my opinion. It's like him looking through this blurred glass and he's like, and he's kind of touching his temple, like stressed. It's just like a crazy, crazy image. But I don't know.

Yeah, he looks kind of like he's like wiping a tear away from his eyes.

Exactly, exactly. And I mean all the images in here, like Even the SoftBank OpenAI one, they picked like the darkest image possible to show Sam and Masa just like completely in silhouette. It's very ominous. There's a lot of editorial stuff.

This line from Lisa Su asking Sam, can I call you an AI icon?

Yeah, that's a flat point.

OpenAI has truly been at the center of the universe. She went on to tell the crowd. Everyone listens to what Sam Altman has to say, which is true. Yeah, they picked a much cooler picture for Lisa than Jensen.

Interesting. Yeah, yeah, yeah. Lisa looks very heroic. Shot from a low angle, very highly lit, very, very brightly lit, smiling, you know, looking authoritative with her Rolex, of course, but also her microphone there. She looks great. And Jensen is looking a little bit stressed, but who knows? These things can flip back and forth on the turn of a dime. But we are clearly in the deals guy era. Let's go to this Jeremy Giffon post. But first let me tell you about cognition. They're the makers of Devon, the AI software engineer. Crush your backlog with your personal AI engineering team. So, Jeremy Giffon, we've highlighted this before, but it's more relevant than ever. Jeremy says we're firmly in the deals guy era. The Deal Guy era. You can raise venture to make deals. You can do foreign policy via deals, build frontier tech via deals via deals. We're trading off stability for dynamism. It's unclear if it's long term optimal, but it's certainly more fun. I couldn't agree more. And Will Manitis quotes it and says, the modern deal guy has no permanent capital, no permanent loyalty and no permanent location. The modern deal guy is playing a repeated game. He doesn't care who he pisses off, there's going to be another hand. The modern deal guy is much more a broker than a principal. He has an ecosystem around him that he identifies as his guys.

Yeah, you should just be able to be good at fundraising. And that made you a good deals guy in tech.

So, yeah, I mean, it's interesting because how much can you really learn from Sam? Because obviously he is very good at deals. But he also wound up building an aggregator, building an incredibly dominant consumer product that gave him a lot of chips on the table. And then he's very effective at playing the game on the table. But it would just be a very different story if he was third in the market and he was like, Yeah, I got 8% market share and my apps are at number 25 instead of number one and number two.

Anyway, so I posted right before we got on the show. I posted Atlas isn't just a web browser. It's an entirely new way to browse the web.

How you doing?

And a number of people, people don't get the joke.

Oh no.

And are like quoting it and saying like, this was written by ChatGPT.

Yeah, yeah, yeah. Which like, that's the whole joke.

That's exactly what a browser does.

I do have an interesting update on this. So ChatGPT is famous for using em dashes. You can actually go into your custom instructions and tell it to never use EM dashes, remove all EM dashes, just use commas or periods. If you don't like EM dashes, it doesn't really work. But I think it, I think. I think it might steer it a little bit in the right direction. The other thing is that we've been identifying that whole. It's not this, it's that. Not a blank, but a blank syntax. I was wondering, like, what is that syntax? Apparently it's called antithetical parallelism or contrastive construction or correlative conjunction pairs. And so that's what you want to avoid because ChatGPT is truly beaten that dead horse completely. And if you use antithetical parallel parallelism, even just accidentally, everyone's going to be like, you use ChatGPT. So you just have to remove antithetical parallelism from your vocabulary, from your writing style, which is unfortunate, but sometimes things get killed off. And so the meta of writing has moved on. And so just just don't use those because no matter how authentic your writing is, people will just assume that you use ChatGPT.

Will Depew says, enjoy the vague posting while it lasts, my friends. Soon all the AGI companies go public and Rune and AI's shit posts are also called security violations. Martin Shkreli says jail is not so bad.

That's incredible. I hadn't thought about that. Yeah, apparently.

I mean, it's funny because OpenAI's somebody at OpenAI Legal is going to probably send this to Will his post to him and be like, hey, look, a lot of people are going to read into this that going to go public soon.

I mean, I don't know, something about the culture there is pretty awesome because they still haven't made will delete the Infinite Chest post. Which is just incredible because posting that while Sora is on the roadmap is one thing. You might just be like, hey, well you might want to delete that. We're going to do the infinite jest thing. But leaving it up is extremely bold. That from a corporate comms perspective, I think that's actually a bull case before we move on, let me tell you about figma.com, think bigger, build faster. Figma helps design and development teams build great products together. Our next post comes from Jacob Rintamaki. Genuine question For VC firms invested in both Frontier Model Labs and coding application layer companies, how do you handle the discrepancy between the two? Since coding is number one on the roadmap of the labs versus other products which are more adjacent, what do you think? He's asking John Ludig, who did not reply, for how Founders Fund thinks about Cognition plus OpenAI and then Thrive is also in OpenAI and Cursor.

Right.

And so there are a number of firms that are in the foundation model which actively wants to be in coding and then also the application layer. I mean I think that, I think.

It'S somewhere, I think it's. If you're a company. I don't, I don't think there's huge downside here. Like VCs tend to give immense support for their winners.

Yep.

And I haven't, I haven't historically seen a lot of, you know, ultimately I can see some potential upside. Right. In helping sort of like manage some of these relationships. Right. And back channel things like that, but not a whole, not a whole lot of downside.

I think, I think something similar might have happened with, with Palantir and Databricks which had some same investors actually. But there was this thesis of like well if databricks is the database layer and they allow you to do AI on top of that, like what is Palantir's role? And they kind of wound up partnering and Databricks does do some forward deployed implementations. They do build some software but Palantir has been much more focused on actual go into an organization, build a piece of enterprise software effectively that uses Palantir systems to speed things up, but then leave behind basically an enterprise software system that lives within the organization, whereas most of the database companies are not really thinking about it that way. So I think that there might be more of a bifurcation between what the coding application layer companies are doing actually solving business problems, building software that then stays there and they act almost like more like new McKinsey than like new coding application. And so like they're kind of squeezing the market from both sides. And I would see OpenAI as disruptive to like the, I don't know, I don't even know if it's like the aws's of the world or something, but it's more like at the tool layer swap it out whenever you want. And at this side you're more like partnering with an organization that's, that's helping you.

I think as venture investors want to become, want to invest in companies that become so successful that they launch products that ultimately compete, compete in a bunch of different adjacent categories and ultimately compete with a range of their portfolio companies. Like that is ultimately what success looks like. And I do think, yeah, I mean something I've been thinking about a lot is this dynamic between anthropic where all of their revenue comes from having the best coding model, which is very different than having a lot of revenue because you have a hit consumer product because you have to keep training the next model. Whereas otherwise your application layer companies that are using the API, let's say they're a coding based application company, they'll just switch to the next best model. Right. It's like very easy to switch out. And so I don't think the kind of competitive overlap between investing in, for example, like anthropic and investing in. I just think it's like totally fair game for VCs to back both out of the same fund.

Well, before we move on, let me tell you about Vanta Automate compliance. Manage risk, improve trust continuously. Vanta's trust management platform takes the manual work out of your security and compliance process and replaces it with continuous automation. We were talking about VCs. We have a post from a VC here, Mark VC. He says pets.com is the go to slur of the dot com era. Let's see, pets.com burned $182 million of total investor capital before going bust. Today just US online pet goods sales is $38 billion a year, 70 billion total market cap including portions of Amazon and Walmart.

Lesson should have invested more in pets.com I mean that's the takeaway here. Like this was, in my view, that's my takeaway is like makes sense to take a really big swing at what became a $38 billion market. Yeah, like I have no idea if.

Like they were set up for success. I mean when did Chewy start? We had Ryan Cohen on from Gamestop yesterday. If you haven't gone to listen to that, please give it a listen.

2011.

2011. So and pets.com is probably over 10 years earlier. Yeah, a lot of these ideas they just took, they just took a decade. They took the Internet actually getting built. They took stripe, they took online payments Getting built. The friction was just so crazy that you know, what are you underwriting? If it's 100. I mean, how much do you think they burned in their last year at pets.com? probably 150 million of that was burned in the final year. Maybe 100 million of that. And so you're not just saying 182 million and then you could have stuck around for an extra decade. It's like probably be spending 100 million a year for 10 years to get to a place where you're ready to actually ramp the business. Yeah, there were a bunch of examples of that. MP3.com was basically Spotify.

I didn't know this, but Amazon actually bought a 54% stake in pets.com no way. Early 1999.

54%. Wow, that's. That's a lot. I wonder how much they paid. It was probably worth it. I mean, I wonder if they got like some assets during bankruptcy, some customer lists. I wonder, I wonder what the final accounting on the pets.com is.

Well, pets. Pets.com is now owned by Petsmart, which also owns Chewy.

Oh wait, they bought Chewy? I thought Chewy was public.

I think they.

Oh, they did like a take private or something. I thought Chewy was one of the companies that got out and was doing well in the public markets.

Yeah, but it was acquired. It was acquired, but maybe taken public later.

Interesting. Oh, weird, huh? Well, it's a $15.29 billion company on public.com, investing. For those that take it seriously, you got multi asset investing, industry leading yields and they're trusted by millions. AWS was down yesterday. We were fighting for our lives on this stream. We thank you for sticking it out with us. We had to do some in person interviews, some. Some phone interviews, but we had a fun time and we kept everything up. There's a little bit of a postmortem. US east is down from Syriac and it's. What exactly is this image from? Is this like Rome falling? Is this the fall of Rome? This is very blackpilling. But services have been largely restored. The outage disrupted services from retailers to airline social media apps and financial services companies.

Do we know what caused it yet?

Wasn't it database migration or something? Some data?

Oh no.

DNS configuration I think was the. Was the root. Something like that. The outage began around 3:00am Eastern Time when Amazon made a technical update to a widely used aws database service, DynamoDB. The update, which included an incorrect domain name service or DNS information for DynamoDB, kicked the database offline in Amazon's critically important Northern Virginia data centers. With Dynamodb out in the region. Other AWS services began failing too. In total, 113 AWS services were affected by the outage, 32 of which had been restored by 10:30am ET on Monday. Not a bunch of stuff we use. Not a bunch of vibe coding Tyler had done.

Yeah. All of the internal software that Tyler's built for the show fell down. Was down.

Yeah.

Really exposed. Insane dependency. Yeah. So we're taking it on prem.

Yeah, yeah, we gotta go on prem for sure. This was my take yesterday. We're going back to sticks and stones. We're going back to pen and paper, baby. The computer revolution is over. AI is dead. We're going back to typewriters.

Did you know the story of James Hamilton?

No. Who's that?

This is this account tuxedosam.

Okay, break this down for me while you pull that. I need to tell you about graphite code review for the age of AI. Graphite helps teams on GitHub ship higher quality software faster. If you're creating fault tolerance in your organization, you probably need graphite to review your pull requests.

So Tuxedo Sam says, meet James Hamilton. He's Amazon's literal, very top engineer, the brain behind all the data centers. He lives in a custom yacht and does not give an F about any rto.

Wow.

And then apparently he was behind, I don't know, like, wait, what is this? Is this real?

There's a community note on this. So it claims that his yacht made landfall. And this is like AWS update, fake news. But it was digitally altered. It was fake news. Moving on. We got some massive news from Brian Armstrong who's coming on the stream in just maybe half an hour. An hour.

This was making waves yesterday after we got off the show.

This was a hot topic.

There is a podcast called Up Only that retired at one point.

Yes. So I listened to this during the last crypto boom in like 2021, 2022, whenever he was going. And I found it hilarious. Kobe was very, very funny. And it was just a very interesting, nuanced take. They were very much like, like self aware of the hype, but then also like giving, you know, some advice, but not direct financial advice. I thought it was just a great show. I think a lot of people loved it. And they put up an nft and it said, if you. If anyone buys this NFT for $25 million, we will produce another season of up only. And so the holder of this. This is the. The image here. The holder of this admission ticket can compel Kobe and Ledger status into performing like monkeys. Eight episodes of Up Only tv. It does not convey any sponsorship rights. So this is not a sponsorship deal and we are allowed to call you idiots for buying it or ignore you completely with zero mentions of your existence during our eight episode season. We get to pick the guests, but if we like you, then we might ask you for some suggestions. And Brian Armstrong chimes in saying, the rumors are true. We bought the NFT up only to your coming back.

Million dollars.

Yes.

And of course.

And so Kobe said, sir, you fire whoever approved this. And Brian says, don't spend it all at once.

And then of course, today, Coinbase and Coby announced that Coinbase is acquiring Echo.

Yes.

Echo is an angellist like platform for crypto. So they help groups of people invest in different projects. They paid 375 million and so you can think of this as a effectively a $400 million acquisition. But yeah, I thought it was wild. The announcement that they paid $25 million for eight podcast episodes was circulating group chats yesterday and didn't fully make sense. It felt like a really big price to pay for goodwill. Right. Obviously people wanted it to come back. It makes a lot. And then the pushback was like, I saw a post, somebody saying, imagine you're a Coinbase employee and you just got passed up for a promotion or your bonus wasn't as good as you were hoping. And then you see that the management team spent 25 million on NFT. Obviously it makes a lot more sense. We have Brian Armstrong coming on to talk about the deal in about an hour.

We do. Quickly, let me check the poly market on what price will bitcoin hit in, in 2025? The chance that it hits $1 million is at 1%. Can you imagine if that happened? That'd be wild. Consensus seems to be that it might hit 130, but it might trade down to 90.

We will see.

And there's a lot of activity there.

Atlas, Atlas would be a beautiful name for a baby browser.

Yes.

Atlas said, please, for the love of God, ask cesky what his 1 rep max bench press is.

This got 1.3 thousand likes.

That will be our opening question.

That'll be our opening question.

We have what is your one rep max?

But the, the, the, the, the, the Coinbase thing was a real, was a real roller coaster because, yeah, it seemed like paying $3 million for a single episode, 25 for eight. That seems crazy. But I came around and I wind up loving it. Because if Coinbase, Coinbase is a huge company, $88 billion market cap, and I Feel like in crypto, there's a lot of companies that the general world is just not that aware of. And a $400 million acquisition, that's incredible. But at the same time, like, there's a lot of. There's companies that are raising $100 million seed rounds, all Sam Altman's out there doing $100 billion deals. And so how do you actually break through with the news of a $400 million acquisition? If you just want to tell the broad community like, we exist and we did this, well, one way is to structure it such that a beloved founder podcast host is coming back and you're doing this funny thing in this nft. And even if it's all just structured in a normal acquisition and it's all.

Part of the headline number, best crypto podcast by far.

Yeah, I agree, I agree. And so, yeah, I just, I think that it's such a creative way to get a little bit more attention on an acquisition. That would have been cool, but it wouldn't have been like a, like, it wouldn't have had a narrative to it where people were like, this is crazy. Oh, it actually all makes sense now. So it was a really good Mission Impossible rip off the mask. Coinbase isn't actually crazy. They're not actually paying 25 million just for a podcast there. And when I think about a founder coming into Coinbase, you know, the typical thing is like, resting, investing, like bringing back a podcast, letting them do eight episodes. That seems like a really, a really fun way to kind of welcome themselves to the Coinbase ecosystem. They can do a lot of cool stuff there.

So I, TJ Parker, was highlighting one of the most iconic quotes from a tech CEO this century from Chesky said, I was basically going room to room, just pouring out this stream of consciousness manifesto, like Jack Kerouac writing on the Road. I basically said, we're not just a vacation app. We're going to M dash. We're going to be a platform, a community.

What's funny is, like, I think of Airbnb as a community from like day one. I think it's actually like, started as a community and then like maybe became more of a platform.

What does community mean now?

Community means, like, it's a place to meet people. And like in the early days, like in 2012, 2013, when I was using Airbnb, like I met my co founders on Airbnb I met in YC because there was one listing for an Airbnb that was listed and it was like, we're going through YC And I was like, I want to live with people that are going through YC because they'll be on the same cadence of, like, staying up all night, working, basically, and, like, not. Not going out. Whereas if you. If you live with somebody who's, like, making six figures at Google, they're going to be like, let's go to the park on the weekend.

Yeah. My pushback would be that maybe the last 10 times I've used Airbnb, it was to book a vacation rental.

Yeah. Yeah.

And I don't necessarily want to meet the host.

That's what I mean about it.

Some people are exc.

And then just.

But that being said, my lovely aunt, she likes to book a studio on a house when she's traveling in Europe on a bigger property and actually meet people.

Yeah.

I mean, also once I was traveling and I was staying at an Airbnb in New York, and it was very clear that the owner of the house wanted to meet me and wanted to make friends in that context, I was like. Like, I don't know. I'm here on a business trip now. I kind of don't want to make a new friend. You seem cool, but I'm good. And so the community thing is a little bit tricky to massage. When it works, it can be amazing. I'm literally still friends with people that I met on Airbnb. It very much was a community. But, yes, over time, these things evolve. But this was the start of the Airbnb 2.0 era. The question of.

Of.

Can you book more than just vacation homes on Airbnb? Is it a thing where you can go and get surf lessons or a dog walker? And then there's all these questions. Practice intermediation. Yeah, haircut. There's a whole bunch of different things about Airbnb being like the IRL app. Obviously, Yelp has been kind of, like, sequestered and relegated in the Google flow. Airbnb still has a big audience, but Ben Thompson's critique was always, well, people just don't open the Airbnb app all that much. So most people, they don't really have DAUs. They have yearly active users, people that open it once a year and think, oh, I want to go to the mountains. Let me book a vacation. Oh, I want to go to the beach. Let me book a beach house. It's not like an app where people are opening it every day and saying, oh, I have some free time. Let me find an option on Airbnb. So this was like, it's a big second act and so I'm excited to talk to Brian Chesky about it. I'm also excited to tell you about Julius AI. What analysis do you want to run? Chat with your data and get expert level insights in seconds. Sorry Jordyn.

Nathan Lambert shared a quote from Chesky said we're relying a lot on Alibaba's QUIN model. It's very good. It's also fast and cheap. We use OpenAI's latest models but we typically don't use them that much in production because they're faster and cheaper models.

Interesting.

We can ask them more about open source.

Yeah. Well there's also news in Anthropic world. Apparently anthropic spent 2.66 billion on Amazon Web Services in first three quarters of 2025, around 100% of their estimated revenue. I thought their estimated revenue was way higher than that for the first three quarters. That seems off to me. I thought they were at a.

They're scaling just so rapidly.

I thought they were at like a 10 billion run rate or something. Do you know roughly, I think they're.

Going to get there by the end of the year.

I thought they were going to be like around 12, but it was going to be next year.

And is that ARR or like projected full year next year?

That's like the projected next year.

Oh, weird. I felt like Anthropic was doing well north of 2.66 in the first nine months.

I had just heard that they had got at some point in the last six months they'd got to a $4 billion run rate and they were pacing to get to north of 10.

Well, I mean obviously this is a referendum on inference costs. If inference costs fall and, and usage model usage is still growing, the business looks great. If inference gets more expensive for some reason it's a dangerous place to be, but interesting to see new data points hit the timeline. So thank you.

Yeah. And something do you know this might be a dumb question, but is Anthropic how much are they leaning on AWS for training?

So they use Trainium but then they're also using Nvidia and basically every foundation model lab is using a ton of different hyperscalers.

They have to go wherever the this article is implying that Anthropic has basically zero percent margins. But it's possible that some of like they're paying to serve and run the models for customers, API clients, but they're also training. Training.

Sure, sure, sure. Yeah. So if 2 billion of this AWS bill was training Claude 6 and then that's going to make 20 billion in inference. That's a wildly different thing. But I mean truly, like no one is saying like oh, bombshell. I thought these foundation model labs were crazy profitable.

It's like, no, we know that Jensen, we know that they all have terrible. Jensen's the only one making money right now.

That's not the question. The question is will they turn it around? On what timeline will they turn it around? Will they use fall? Will they use the Generative media platform for developers? The world's best generative image and video and audio models all in one place. Develop and fine tune models with serverless GPUs and on demand clusters.

Patrick Collison and Jeff over at Stripe as well as Haley on their team announced that you can now fundraise in Stripe Atlas. It's very cool. Stripe Atlas generates a significant amount. I forget the exact percentage, but a meaningful double digit percentage of all C Corps created with Stripe Atlas. Now you can fundraise in the product. You can generate, send, sign and track safes in a few clicks, according to Jeff. So very, very cool feature. You know, all the founders out there now, you're pretty much oftentimes by the time you're setting up the C corp, you're like actively trying to take money. So it makes sense too, extremely.

YC coded like Stripe, one of the earliest YC companies. The safe came out of YC and they're just like building more and more infrastructure for YC companies and then scales and taking a really long view because I'm sure some founder out there is going to raise a bunch of safes and then wind up building a billion dollar company and doing a lot of paintball.

I believe it's already happened multiple times.

Oh wait, like the top performing Stripe Atlas company is like a banger.

Oh, I'm sure.

I wouldn't be surprised.

They've made so many.

That's true. And it's been around for a while right now.

And yeah, I built this product at party round and it's very useful. It's not a great business, but it makes sense to tie it into the stripe ecosystem.

Do you think they're going to give this away for free? Is this all just onboarding into the Stripe ecosystem? Basically, because then you already have your Stripe account set up.

I mean they basically. Stripe Atlas is basically free. I think it's 3.99 to create a company.

$3.99.

No, stop that. $399. So yeah, I would assume this is free. Yeah, that's pretty cheap, which is great.

That's cool. Yeah, they probably need to put some sort of value there, just as a gate, so people aren't just like spamming a C corpse out. If it was free, you'd just be like codex. Set up 1000C Corps for me. That could get very, very odd. Well, let me tell you about turbopuffer search. Every byte, serverless vector and full text search build from first principles and object storage. Fast 10x cheaper and extremely scalable. Samo Berja, who we should have on the show at some time, at some point. I've had some good chats with him. He says the world is changed, subtle and obvious. For the first time in living memory, the scope of live player action expands. We shall see a new Napoleonic era, both liberating and terrible, with an ending that is yet to be written. Are you prepared? He has such a way with words. Do you know Samo? He follows me.

I mean, this is a great vague post.

This is a great vague.

Go work at OpenAI.

He might be talking about OpenAI, he might be talking about the US government, who knows? But he is, of course, the work. Yeah.

The framing of thinking of tech companies as nations. Right. Like I was talking to a buddy over the weekend, friend of the show and we were talking about Andy Jassy and how a lot of people are pissed off at Andy Jassy for a number of reasons. Reasons primarily kind of like in some ways, like not being on it with it. No, no, no, not aws. This was pre AWS going offline, but just kind of like not like fully missing AI, but not being in the game as much as people would like. And he was just saying, like Andy Jasson, Jassy may as well be like a Prime minister. Right. Even if you don't like him, he's still in office. Oh, totally and effectively has an unlimited term. And so if you start thinking of OpenAI in the context of a nation state, and the nation state is saying we're going to need trillions of dollars, we're going to need a huge amount of energy and yes, it's going to be distributed all over. It's not that crazy for a country to be saying we need to spend hundreds of billions of dollars on. It's like a digital nation state.

I always have this question about it still feels like like the US government is at the top of the stack and the tech companies answer to the US government. But I do wonder if there's some element of flipping that will happen at some point or has already happened. Did you know that Walmart spends more money than the United Nations? Isn't that remarkable? I mean, that sounds like it makes sense.

But doesn't the United nations just kind of like host some conferences?

No, no, they have like UN peacekeeping.

Troops, they're forward deployed, peacekeep, a lot of stuff.

It's the United Nations. They have massive buildings and operations all over the world. Health developments, all sorts of stuff.

It's not easy to prop up the details.

It is the closest thing, I think, that we have to a global government. It's who you answer to at the highest level potentially. I mean, obviously the US is much, much bigger than the UN and so we have this kind of flipped model. But anyway, we have our first guest of the show. We have Palmer Luckey in the restream waiting room. Let's bring him into the TVPN ultradome. Palmer, how are you doing? Welcome to the show. I'm here, he's here with the power of technology, the magic. How are you doing?

What's going on?

Doing really well. Although the thing I've been complaining about lately is I was recently on Joe Rogan and what you don't understand when you go on Joe Rogan is that it's like calling in a distributed denial of service attack on all your communications. Because your email, your voicemail, text messages, your Facebook messenger, your X DMs, everything becomes useless because you have so much incoming that even the important things like Homer, you need to go to your dentist appointment right now.

Boom.

Flooded by Brogan.

How many float tank operators have reached out and said, we gotta get you.

Have you used.

Hang on, let me think about this for a second. I haven't even caught up with everything, but at least three.

Okay. Have you actually gone in a float tank?

I mean, that, that hearing that, hearing that, the beginning of the interview and realizing that this guy you're listening to wants to do a float tank but hasn't yet. And you're like, I've been waiting for this moment my entire life.

It's funny too, because in talking about float tank tanks, Joe and I, we took a break during the show and when we took that break, we actually went to go look at his float tank.

Okay?

And it was so funny because you're right, I'm like one of the only people who has gone on the show knowing all about float tanks and the principle and the mechanism and the theory and the history, yet I've never actually managed to get into one. So it's a rare combination. I think Most people who haven't used float tank tanks don't really know much.

About them or care.

Yeah, I've actually looked a lot into float tanks, even beyond what we talked about on Rogan. Going back to my initial kind of push into this was in VR where I got into thinking about how do you do sensory management? How do you forget the senses that are at conflict with the virtual reality you want to present? Float tank is interesting, but then also it's a sci fi staple, is to fill your lungs up with some kind of oxygen bearing fluid and then go in your full fluid immersion pod so that you can achieve extremely high G forces during rapid acceleration. And what I've always wanted to do is build a dragster. Like build a car that can accelerate so fast that it would kill somebody if they weren't in a full fluid immersion G pod. And I think that if you could do that, it'd be some kind of of truly novel motorsports achievement where for the first time in any motorsport, the limit would truly be the ability of the person to survive it rather than the power of the machine that is involved. Wouldn't it be so cool? Like, oh, don't floor it unless you've got your lungs full or it will kill you.

Yeah, they talk about instantly be dead. Supercars that don't have traction control being death machines. Right.

But supercars that are accelerating over 1G, like that's crazy. I want to do an inverted fan car, basically a vacuum car that is accelerating at 80 or 90 GS.

Would you have to put rockets on the back too, or explosives or something? How do you actually, how do you max it out?

So my general opinion on these things is that there's an I'll know it when I see it element. If you're using rockets, you're not really a car. Look, I'm a huge fan. Like, like the NHR was the nhra, right? I forget. Whatever the sanctioning body was in the United States for drag races, they allowed rocket dragsters up until I think the late 1980s. There's a great video of a dragster called Vanishing Point and it was a rocket powered dragster. That was the last one before they banned them. It was doing like 400 miles per hour in the trap and its final run. There's a video on YouTube. I'll have to send it to you guys. Although if you look, look up Vanishing Point rocket dragster on YouTube. You'll find it. Maybe you guys can cut it into the video later. For people watching this not live, but Its final run, they ran it at true full throttle, went faster than everyone, and the camera pans up, and it blew out every single window in the control tower when it went. And the rumor is that it went so fast and so hard that the pilot passed out during the acceleration. And the parachute was actually pulled by a time. I don't know if that's true, but.

It'S a great story.

So, anyway, the point is, rocket dragsters don't count. You have to do it with wheels. I'll know it when I see it. If it's rockets, it's not a car. And there was a guy in the. I think, the Air Force who set the record for most GS sustained by a person. And he used a rocket sled to do that. Basically, it was a rocket sled that accelerated and then very rapidly decelerated. And they couldn't. He thought it would be unethical to have anybody else volunteer. So it was a pretty senior Air Force researcher who did the test himself. And I think he, like, blew out all of his.

The.

All the veins in his eyes and his sinuses, and he was bleeding from every hole on his head. But he survived and recovered, and he proved that a human can survive over 50 GS of relatively sustained 50 GS. Over 50 GS.

Brutal.

Wow.

But you're gonna have a very bad.

That's leadership. That's leadership.

That's true leadership. If the float tank's gonna get added potentially to your routine, what other biohacks are you using or a fan of or what else have you tried in sort of like the performance optimization?

Being on a mission. I feel like being on a mission is the ultimate biohack.

Yeah, yeah. I mean, it really is. I mean, I've always been very straight edge. Right. So no alcohol, no caffeine, no nicotine, no drugs, nothing. I've recently started drinking since having a child, and if we can get into that or not. But I'm very much high on life. I don't really want to do.

Doesn't Mountain Dew have caffeine in it?

So it does, and I will. But I generally will avoid Mountain Dew for that reason, at least as a dependency.

Yeah, sure.

I go to Taco Bell. I'm getting that Baja Blast, of course. And now that I've had a kid and I drink, I can go to the Taco Bell Cantina, of which there's only a hand. Are you familiar with the Taco Bell Cantina?

I'm not, no.

I've. I've. I'VE never been. It's in New York. Is there one in SF too?

I think there's one in New York City. I don't think sf. Sf. Nobody builds anything in sf.

But there's that Taco Bell right on the beach in Pacifica. You know what I'm talking about?

Cargo Cantina is this concept where it's like Taco Bell, but also they serve alcohol. So it's like a bar and a Taco Bell and there's one less than five minutes from my house in Newport beach. And so I can now go get Baja blasted on my. On my Baja Blast. You know.

Wait, but I have to ask. What about fatherhood made you start drinking?

You have kids?

Yeah. So we both have kids.

We actually passed the Palmer test. We have five between us.

Well, then you should understand. Understand?

Yes, I do.

You should understand.

Yeah, no, I. I understand a little bit. But, but part of it's like, like alcohol for me. I sleep terribly. Kids are also, you know, they make. They, they, they're very.

Oh, I see. No. Alcohol knocks. Helps. It's different for everybody. It knocks me. It knocks me out.

Yeah, yeah.

Maybe because I don't have a tolerance to it, but. So you asked about hacks. I've got one that I've been pondering that maybe we could talk about.

Yeah, please.

I'm really afraid that nicotine might be really, really good. You've probably seen this theory, right? Basically, America smoked its way to being the dominant hyperpower. It kept people focused, it kept people fit. It's an appetite suppressant. There's this really interesting health trade off theory that I'm not saying I buy into it fully yet. So for people who are watching and shouting and screaming, I don't buy into it fully yet yet. But I'm becoming more and more convinced that the health benefits of not smoking have not been properly traded against the health problems caused by the resulting eating. Not just in terms of appetite suppressant, but also just people filling cravings and filling the need for ritual rituals.

Well, I would tell you out of.

Experience with other things that are worse for you. I think the crazy version of this is we'd all be better off with lung cancer eventually, but fit until then. And it could be that smoking gets you there. So I don't know. I'm trying to figure this one out. But nicotine is near the top of my list of potential biohacks.

Sure. The other thing that it feels real is nicotine makes boring things pretty enjoyable. And it turns out that to do anything significant in life, like it's oftentimes very exciting early. Like I say you're working on a new product at Anduril. It's very exciting. You can make a 3D render, you know, internally. Not for marketing materials. I know you guys don't do that. No render policy, no renders. But let's say you make a render internally, you're like, this is really exciting. And then, and then it gets into the mud and you're like, okay, now we have to build this thing. And you're in the, in the trenches of product development and you get through many, many, many boring hours of struggle.

Yeah.

I'll tell you, a lot of the people in the trenches, they are smoking.

Yeah. The literal trenches also going back through like they were. That was how people used to be in rations.

Of course I was going to bring that up. Yeah. I mean, they didn't get rid of the, they didn't get rid of the cigarettes in MREs until pretty recently. I don't know what date it was, but it was not that long ago.

Yeah.

So that's the. And of course there's other ways to have nicotine without having to have the health impact of actually, you know, smoking. Atari substance. And so that's where it becomes interesting. Like you probably couldn't convince me that smoking is literally better than, you know, the food, but it's like, you know, is vaping worse than being fat? I don't know. I struggle. Based on the evidence I've seen.

Yeah, yeah, yeah. That's wild. Do you play video games before bed? You said that the alcohol helps you fall asleep, but do you ever get locked in and you like, are playing video games? Can't fall asleep because of that.

You know, I used to play games all night long. I mean, there was a time where I was spending maybe 14 hours a day on the computer if I could.

Same.

That's you.

It's amazing.

Wonderful.

The best times. I love them. I want to go back.

Was a good chunk of that. But also running online communities and other things. But the problem that I have now is the types of games I really like, especially multiplayer games. You kind of have to be on your game and if you play late at night when you're all ready, you're not even competitive enough for it to be fun. And I'm not saying I need to win or I, or I hate it. You know, I'm not, I don't hate skill based matchmaking that much. Yeah, but you want to know that you're playing to at least the upper limit of your own potential.

Yeah.

And I find so that that's been the struggle as I've gotten older. And I know, I know 33 doesn't sound that old, but there's a lot, there's a lot of things that are easier when you're 23 than when you are 33. And like, you're not supposed to talk about that. Particularly as an executive of a company in California high Department of justice, you're not supposed to acknowledge that perhaps it's possible. I'm not saying there are. It's possible that different age ranges have different strengths and weaknesses. Perhaps, possibly, but certainly nobody should ever make a hiring decision on that basis.

That's right. Yeah. I imagine in some ways being probably easier parenting could be easier for, for people in their early 20s. Right. Than their early 30s, just by.

What are you talking about, 20s? Look at this. Mainstream NPC. No, no, no. Kids should be having kids when they're in their teens. That's when they're supposed to have them. Now, you could argue that maybe there's a reason to stretch it out a little bit, but if we're just talking about physical ability and depth of, well, of energy, and let's not be politically correct, let's just admit you're supposed to be having kids. You're 16, 17, 18, and be done by the time in your 20s so that you can have your kids working on the farm by the time you're in your 40s. I mean, it's. I mean, I regret not having kids even at the age of 33. I regret not having them earlier because I'm like, geez, this would have been so much easier when I was younger. I've been with my wife since we were 15, though, so it's a lot easier for me. You would be like, oh, but Palmer, how do you know it's the right person? I'm like, geez, we should have just obviously had kids when we were 16. It would have worked out just fine.

Yeah, yeah, totally. Yeah. The whole world. You feel like when you're younger, everything's as complicated as it'll be, and it just gets more complicated every year and more busy every year. Take us through. Take us through.

Mod Rep. Well, this gets also to like the startup thing. I say all the time. And I'll take this, this moment to, to push it, especially for anyone young who's listening. People think that starting a company or taking a break from school is this very high risk move when you're young. They say, oh, how do you find the bravery to do that. What I try to push on people is it will never get better. When you are 18 or 19, you have no family, you have no real job, you don't have a career yet per se. That's the best time to start a company. You're not giving anything up. All you have to lose is a bit of time and all you need to do is, is better than you would have done otherwise on your resume like running a startup and failing it is probably going to look better on your resume than whatever you are going to be doing part time in school as a, as a, as a late teenager or early, early 20s.

Yeah.

And so yeah, the, the best time to start a company is when you're young because you don't need any bravery to do it. The people who start it later where they have a family and a mortgage and yeah, once, once these people have.

Acclimated to like a 250k base, it's like good luck building a hundred percent.

And also once you have a family, you have a sort of fiduciary duty to your family to maximize their quality of life. There's no life hack that says that you're allowed to degrade the quality of your child's childhood because daddy wanted to have that Grindset mindset.

Right.

It's irresponsible. At that point I couldn't imagine starting Oculus us where I am today. Like it would just be irresponsible.

Yeah. Take us through Mod Retro. How much of everyone's familiar with the Chromatic? We've talked a lot about it, a lot on the show. We unboxed one. We've had a lot of fun with them. How much of the roadmap is public now? Where's the company going? Where's the company now? Just kind of give us the general update and then we'll dive into some stuff.

Before I tell you where it's going, I'll go back to where it came from in the first place. A lot of people think that Mod Retro is this very recent thing for me, the first product we launched, the Mod retrochromatic which was kind of the ultimate Game Boy. This kind of heirloom grade tribute to the Nintendo Game Boy and Game Boy color, very high end hardware, magnesium aluminum alloy shell, a lab grown sapphire crystal screen lens rather than plastic or even glass. And the thing is, Mod Retro was actually the first company that I ever started, even before Oculus. So I started Mod Retro when I was about 15 years old. It was an online forum for.

It.

Was an online forum for people modifying game consoles, handhelds, portables, also home television consoles. The main kind of thrust of the community was this hobby called portableizing, which is taking home consoles and turning them into handhelds by combining them with modern technology like LCDs and modern high energy density batteries. And I mean, that actually was a pretty big deal for a while. I know in the modern day this doesn't mean much, but we were getting millions of unique viewers every single month because we were getting covered the projects on this site. We had thousands of active members. We're getting covered by Engadget and Gizmodo and Kotaku and kind of the whole group of outlets that back then liked technology. They're now anti tech. Anti tech.

Tate coverage.

But back when tech news was about tech and not anti tech, they were covering all of these things and again these days, millions of uniques. I mean, you can get that on TikTok with a good dance. But that really meant something in the mid-2000s on the Internet. And we actually started working on projects like what the ultimate Game Boy would be back then. There was a project that I did in 2008 called the PGB, the power game Boy. Power Game Boy was the ultimate Game Boy. It was a Game Boy Advance. It was Game Boy Advance SP hardware transplanted into an original Game Boy shell, but with an upgraded screen, upgraded controls, motion controls, an FM radio transmitter so you could broadcast the sound to your car so you could be playing and didn't have to tether in and could have all that sound if you're doing lsdj, Dells, DJ or some other chiptune software. And that actually won the 2008 portable Palooza. So very proud of the Power Game Boy. I was also the first person to ever backlight a Game Boy Pocket. The first person to ever led backlight an original Game Boy. So I mean, I am an OG to the max when it comes to this.

That was the original parental control was. I just remember being a kid and being like, if I turn the lights on to play Game Boy, my parents are gonna see the light under the door, they're gonna come take my Game Boy.

So that's.

I was like trying to. It was probably training my eyesight to try to.

Yeah, no, no, you're just maybe hurting your eyesight, but training your brain. Now you can see signal from the noise. You have ultimate night vision. Your Batman yourself molded by it, raised by it. Yeah, but so what happened is I started on this project all the way back then of what the Ultimate Game Boy would look like. I kept working on it at Oculus. Yeah, I kept working on it post Oculus. And then after about eight or nine years of not making any progress because I was just like, I made some progress. We had some prototypes and stuff. I finally said, you know what? I need to get more serious about paying someone else to do my hobby because I'm not going to be able to get this done on my own. And so I ended up hiring a few people from my past life and getting together with a bunch of people who believed in this kind of vision of building these tributes to what technology used to be the best parts of what gaming used to be, and building a company around that. And we ended up pulling together the chromatic, all the pieces and loose ends from 10 years of hacking and modding and turned it into a real product.

Walk me through the roadmap. The M64 is coming out.

That's right.

Do you have other ideas or ways you define, like, what sticks out? Obviously the Game Boy is iconic. The N64 is iconic. But how broad are you thinking? What reaches. Is it just games? Is it just games from the 90s, 2000s?

We did the Game Boy because it's the most important handheld console ever. I mean, it redefined the idea of gaming what it was, how you did it, how it fit into your life, how you could build games that you didn't sit down and be tethered to something, but instead could be a thing that you would share with friends, bring to places, meet new people playing them. That was really, really a wild idea at the time, and there's a lot of good in that. But I think that as the games industry has financialized and gotten much bigger, there's a lot of things that have gotten lost. I think the need to make more and more money has. Has taken things away from, I think, what were actually great product decisions in the 80s and 90s when people were just trying to make great games and great consoles. They weren't trying to figure out how to bump up their daily active users. They weren't trying to build live services. They weren't trying to build passive recurring revenue models that relied on a variety of modern marketing concepts to juice people and get them locked in.

Yeah, what do you like in the modern.

That's why I'd say just generally, Harvard, I want to go into hardware. There was like similar turning points. So the M64 is an obvious one. That was the beginning of 3D gaming, basically for everybody you can say, but Palmer, there were people who are playing 3D games on their PCs. Yes, but in terms of mainstream access, totally. I mean, that changed everything.

Totally.

I'm interested in going to all these points where there's things that we can learn and releasing basically the best version of that combined with modern technology that makes it more useful. So, so things like what would a modern Sony Walkman look like? Like, what was good about the Walkman experience versus modern streaming and kind of the digital morass that we've gotten into the algorithmic feed, like nobody's doing mixtapes anymore. Listening to music is not a conscious activity, it's not a curated activity. It's being a receptacle for a machine. People talk about just having AI slop fed to them without realizing that. The modern music industry has kind of been pioneering this for a long time. It's just the AI is behind the curtain, the computers are behind the curtain, and the algorithmic recommendation engine is behind the curtain.

Do you feel like there's some green shoots in the modern gaming economy? I'm thinking like, yes, there's a lot of mobile games that are hyper optimized and there's a lot of games that have gone free to play with all the micropayments and stuff. But then you do have folks like the team behind Last of Us where it's just a story and you don't really get sucked into any crazy micropayments and it feels like there's an artist behind it. But are there, are there other areas of the modern game industry that are sticking out to you as like, positives?

There certainly are. And there's lots of modern indie games that are, I think doing a lot of, like you look at like Silksong and a lot of. But where you're not seeing this, I think is on the hardware side.

Sure.

You're still living in an ecosystem. So like, let's say that you want to play a game that was built with these values. All right, so you're going to turn on your console, you're going to try to play it. Oh, there's a critical security update. Time for you to do your critical security update. Okay. Now you need to also update your console. All right. You're updating your console?

Nope.

Now that wiped out all of the cookies. It's time for you to log in again. You're going to log in. It wants you to do two factor authentication because fraud has become such a problem that you now need to do that. Oh shit, where's my phone? Let me see. I don't have my authentic. It's one of those terrible things where it adds so much friction. And that's before you then get railroaded into a big giant UI of advertisements that are auto playing the moment. The idea of that's just to get.

Permission to upgrade your skins.

I need to just be like, I'm going to take this.

Yeah. And just turn it on and that's it. And it's great.

Yeah, that is.

Yeah, yeah, yeah, yeah.

Modern games can't do that because of the platforms that they're on.

Yeah, yeah. I was comping like what it takes to play actually on the. One of the, one of the newer quests, I realized I had seven different passwords. Instagram, Facebook, Meta. I had all these different passwords, I had codes to different pieces to buy games and stuff. Versus I went and pulled up an old N64 where you turn the button and it just says Goldeneye. And it's just like amazing. And I was saying that like, that seems like an own goal. It seems like modern video game designers should at least know that you will see Lower Churn if you pre install one great game. And when the first time they unbox it on Christmas, you turn it on and it's Beat Saber or whatever, the piece of software or hardware that you're selling, like have it come with something great. Like why don't people do that? Are they just stupid or is it it like something more complex?

This is, it is complex. I mean, so you got to remember like, so first of all, there's a lot of people in the games industry who didn't necessarily get into it purely because they want to build great games.

Or they drink Starbucks and not Mountain Dew.

Yep.

You've heard my quote, right? And you have a lot of these people who it's. It's a lily pad, it's a way to make a check or it's a way to effects affect social change in the world through some larger companies. Budget, budget. And so that is part of the problem. But you're also at odds often. So you just talked about this specific idea of having a bundled title. So I've been on the other side of the negotiating table inside of Meta, formerly Facebook and formerly Oculus. What's going to happen is you're going to say, hey, this is going to be a great experience. And then you have someone who says, but if we have a bundled title, then that gets rid of the revenue opportunity to have another company pay us to be the bundled title. And also our developers are going to be really upset because we're going to be competing with them. And so for example, your holiday day one installs, they're mostly going to be of this new game. People are going to be less likely to do day one buys of all of these other games. If there's a game that you can already play, you say, well, which should we, we have to have something pre installed. And then you have these conversations. Well, it can't be a full game. It needs to be a thing that shows off the hardware. It can be like a fun little bite experience. But we then need to have people go into it. And then here's another thing they say if you have literally nothing, that means they have to download something. And the time where there's going to be the most impetus for them to put in their payment details is going to be that day one. So you don't want to do anything that degrades. You get the payment details. Because the moment you get their payment details, every purchase after that is an easy click. It's an easy dopamine hit. You don't have to go through all of the pain. And so it's one of those things where you have all of these intervariate money driven things and you have the platform holders so tightly intertwined with much of the content. This is the problem with when the hardware owners end up owning all of these studios. This is probably me getting a little too business political. But in the early days of Oculus we were very clear with developers to our business relations. We are not Nintendo, we are not going to compete with you and basically destroy your sales by preferentially treating our own developers better, our own studios. So we made things, but we really said our goal is to enable second and third party titles to fund lots of other people and we funded dozens of developers. As you become Nintendo, you make it harder for those third parties to exist. And the first party has their own goals that are not necessarily aligned with what you're talking about, which is turn on the game and play it and have a good time. They have other plans for you.

Is the answer just out compete them. I imagine that the libertarian in you is not saying we need to regulate this. So what's the answer?

Of course not. I think it's a combination of, of reminding people what they've lost, like what they had and how good it was. It's about reminding and showing off for the first time to a new generation how things used to be. I mean one of the things I love about chromatic and M64 is people show like we were just at the Portland Retro Gaming Expo and a lot of the people coming by and having the most funs were kids whose parents brought them over. Like, oh, check out this game. Like this is the Mario Kart I used to play play. And you can show these people, I think a lot about how things used to be, how it used to work. And I think you can prove that people still want that. There's people who don't believe any of this. You'll talk to people say, well that's not what people want from games anymore. They want live service games. They want games that are continuously updating with fresh new content. And they say this, I think because they're internal. People have to convince themselves that it's true because otherwise their business model of being a $100 billion company company doesn't pan out. And you contrast that with like the number, you know, how many people made the game. Boy, it was 12 people.

12.

I mean, it's just, it's. You contrast that with like the 60,000 people you have in some of these larger games. So look, I don't think it's regulation. I think it's reminding people, showing people. And then at some point someone needs to take the best of modern technology and they need to take these lessons from the past when people were trying to make, make things great without necessarily being concerned about the micro financials. And they need to build probably a new platform that learns combines the best of all of these things. Which is kind of the idea behind Mod Retro. You want to combine the best of the modern and the best of the retro. And I think at some point somebody's going to do that. Who knows, maybe you'll see something like Sega reenters the console war and kicks everybody's ass. But maybe it'll be in television. Maybe ColecoVision's going to come back, but at some point it'll happen.

Yeah.

Do you expect Mod Retro to continuously make decisions that sort of limit the tam and that's, you know, not going down the path of hyper financialization and you know, trying to create this open ecosystem. I just, I doubt you guys had any investors that, that passed on Mod Retro. But I imagine some of them would say, you know, the NPC VC that thinks it's contrarian to have kids in your 20s would say like, okay, I want to invest, but like how do you think about the tam? Right? Like how do you address that? Because I can see the obvious path of just selling a lot of really great hardware and building out millions of people in the world that just love the hardware. Buy the hardware and you have a fantastic business. But maybe it's not an anduril scale opportunity and you're probably totally okay with that.

I think the opportunity actually is huge. I think that the entire modern electronics industry is in a sort of prisoner's dilemma where nobody can stop doing these things that nobody wants to be doing because the first person to do it can't really do it alone. I mean, that's why your phone is full of crapware, that's why you buy a laptop and it's preloaded with all of this stuff. The companies know just how much consumers hate it, but they're all in this competing for each penny, competing for each dollar race against each other with largely undifferentiated offers offerings. And so they have to play that game. If you can hugely differentiate and if you can convince a consumer that it really is a differentiated thing, I actually think that you could end up where the TAM is. Look, here's what will actually happen. I think the TAM for doing things right is bigger than the TAM for things that do things wrong. The only problem is that the moment you prove that that's true, everyone is going to follow in your footsteps and become your competitors in this new post prisoners dilemma world. That's probably the form of success that I would be happy with. It's not let's do everything ourselves and nobody's going to follow us. Realistically, the moment you see success, everybody copies what you're doing. Changing the world is outcome people need to be okay with even if they don't own the whole market.

Would Mod Retro ever make a M1 Garand?

M1 Garand? I don't think so. I think it's going to stick to consumer electronics, like technology products.

We work a consumer electronics product that was created before you were born.

Oh yeah, for sure, for sure. I mean, we're working on a cassette tape player right now.

Okay, yeah.

So like that's one example. Like, what would the ultimate Sony Walkman look like? Like, if you were making something today with those principles in mind, what, what would it look like today? How would it work? And there's a lot of things, and I'm not even talking about the complicated things like algorithmic music feeds, just basic things. It was so easy to stop and start music or adjust the volume or know that it was playing in the first place. You have that haptic mechanical. You can literally feel the wheels turning. A Walkman on an armband is a much better experience than using your phone to play music. And I understand why we're using one and not the other. And I'm not saying that the cassette tape part is the part that made it good, but there's a lot of other parts that were good that I think are worth learning from. Another one is like an example would be modern televisions. I know I'm a libertarian and I'm not supposed to like regulation, but I sometimes flirt with the idea that smart TVs should be illegal. I hate smart TVs so much.

They're so bad.

See, you're clapping because everyone agrees.

Everyone agrees.

This is the danger of being a libertarian. You realize you can just adopt populist pro state dispositions and you can get voters. Oh, we're libertarian, except for this thing you hate. We're going to regulate that.

You can be, you can be trusted with total control.

I would trust you.

Exactly.

But you're not like the other one.

Like smart TVs create a lot of these same bad incentives where you have the hardware manufacturer not just trying to build a good tv, a good screen, good visual technology they're now dipping their toe into. Well, I actually need to be a services company. I need to be a software platform company. I need to run an app store. And it's a prisoner's dilemma. If you don't do those things and if you don't introduce advertising into your TV feeds and if you don't have a lock screen that's showing ads on your tv, how could you compete with the people who are, who charge just a few dollars less? But I think that there's value in saying, hey, one of things mod retro is looking at doing is doing a modern CRT display. And part of the value there is building something that actually has all those cool attributes of a crt. Especially when you're working with retro computer systems and retro consoles, that we're designing the art around the particular technicalities of how a CRT works in terms of having very high motion clarity, very, very good color gamut, relying on the persistence of the phosphors to enable certain visual effects, certain blending that happens on certain things, doing display interlacing techniques, but ignoring all that, there's also just the fact that a TV that you can plug something into and it just shows what you plugged into it is an incredibly novel idea. And I think bringing it back, reminding people of that. I wouldn't be surprised to see mod retro make a totally modern technology display that is just a TV that doesn't function. You like that's it. I think I, I mean, just, just think.

Just the, the, the. My biggest critique of all these smart TVs is how quickly the software degrades. Like it wasn't good in the beginning and then quickly there's like a, a small lag every time you push a button. And imagine a TV that was just super snappy and just had the three apps that you actually want to use. And I'm sure you go back and.

Use, go back and use like a, even like a flip phone. From the early 2000, the UIs were literally orders of magnitude faster and snappier than the UIs of these modern smart TVs. Like, you can buy a TV with 8 gigabytes of RAM and a 3 GHz processor and somehow you're right, it still manages to have that lag.

You know what, we're at a stage in the technology cycle where people make things that are different but not necessarily better. There's this pressure to do things that are new and different.

Horizontal movement.

Yeah. Yeah. And my biggest critique is the new iOS liquid glass. It's like, congratulations, congratulations, you made it different. I don't believe you made it better.

Yeah. What do you think? Do you think there's a world where the next turn of VR kind of displaces TVs? We talked to James Cameron about this and it seemed like he'd kind of seen the next iteration maybe and was kind of excited about the idea of being able to watch a 3D IMAX movie at home. And from my experience with the Apple Vision Pro, the screen's getting better and better. Meta's working on this stuff. We've talked to the big screen team. It feels like we might finally be at a point where we're thinking about replacing the home theater.

The way to think about VR headsets is not as a replacement for a tv, but as a replacement for a home theater. The reason that's important is it's not just a tv, a screen. It's a controlled environment optimized for watching that content. The thing I've always liked about VR is you can have all this cool new VR content, content, but it can potentially also simulate any previous form of any previous media experience before. So, like an example I've given when people said, well, what about like an ebook, you know, an ebook reader on VR, Are we really going to use that? And the point I make is, well, look, if you make this good enough, it's not just the book, right? The experience of reading a book is also your environment. Your surroundings, people will go places just to sit down and read a book in the right environment. Simulating that is interesting.

So.

And the same thing goes for music, right? It's not that you're not trying to simulate a speaker, you're trying to simulate, you know, a listening room. And I feel like that that's where things are going. I think that's why James Cameron is excited, because you spend a lot of money to control a viewing experience when you're doing it physically. Right. Like I'm building walls, I'm soundproofing a room, I'm painting the whole thing matte black so there's no reflections on my screen, by the way, I've done that. I have a basement home theater that I made myself and the walls are painted matte black and the ceiling is painted matte black. And I have custom matte black carpet that's mostly black except for a few of my favorite galaxies printed on it from NASA imagery, because I wanted to have some space carpet. Nothing's cooler than space carpet. But that was really expensive to do all that. And I think VR is going to actually surpass that experience in short order. The technological path to VR displays being better than 99% of people, people's home viewing environments is a single digit year problem. It's not decades, it's within 10 years, certainly.

Yeah. What's in the critical path? You've said that you liked pulling the battery out. Are there other things that you're pulling out? If you're building the next version of consumer like BRS home theater, what are the design considerations you're pulling out?

I mean, where do I even begin?

A second screen on the outside you'd have. You definitely have that if you want to be in the home theater, right?

No, no. I hear that was a Tim Cook special, but I can't confirm it. I don't work at Apple. But you need to relentlessly focus on the experience of the user of the device and you can't have features that are not adding to that. If your job is to try and make VR cool, which is largely what Apple was trying to do with the first generation Vision Pro Pro, you might do things like have a screen on the outside and like, if you're trying to make it cool and acceptable, but at some point you need to optimize towards the experience of it all. And luckily, in the long run, maybe even the medium run, these things are going to look like sunglasses, not like ski goggles. And so that's really where this is all going. I Think if you and me had James Cameron right here to grill, I don't think he would say, I think the future looks like a bunch of people wearing ski goggles. I think you'd say the glasses you wear to already like navigate and communicate that you're wearing all day will also simultaneously be the best home theater device you've ever seen. We're going to get there sooner rather than later.

Yeah.

Let's talk about Erebor. There's been some press hits lately. The press seems entirely fixated on Palmer Lucky's crypto bank. And obviously that is not the full story. The full story, nor is it the most, I think, exciting thing about it from anybody that that's been in the tech industry that lived through the SVB collapse. And so I wanted, I wanted to hear it directly from you.

Well, you know, it's a little early to start talking about it. We just got our conditional approval, which is fantastic. We're still waiting on fdic. Yep. We're still waiting on full approval. So it's a little early for me to really get into, into, you know, going out and marketing the bank.

Yeah, but why, but, but yeah, but.

What you're getting into. Yeah, no, I, I get it. Like the, I think there's this natural inclination for people to look at this and say, oh, it's like what SVB was. It's catering to the same customers. That must mean that they're going to be this really high risk bank that does all these high risk things. In reality, it's literally a reaction to that. It is an opposite. The reason that we got this approval is because we went in saying we are going to have the most conservative loan to deposit ratios of any bank in history. We are going to do this extremely novel service of taking money from people, holding it for them and then allowing them to have it back. We're going to be offering these services that banks used to offer. You can't get the service I just described in any bank. If you go and you say, I want you to just hold on to my assets, but I don't want you to loan against them. That's not what a bank is. What are you talking about? And of course I'm not saying that all assets should be held that way, but when you're a business in the business of making your business grow, you don't care nearly as much about getting an extra half a percent on your deposit deposits as you might about truly minimizing the risk that you're not going to have access to Any of your capital. I mean, if SVB would not have been bailed out by the government, it wasn't just the government's obligations. They were bailing out well beyond FDIC limits. If they wouldn't have done that, it probably would have wiped out half the tech industry in one fell swoop. That's crazy. How can the tech industry have become so subservient to the banking industry, to the finance bros, that not only can we not survive without them, but in working with them we sign our own death warrants. We can't see what the risk is, we can't control what is, and there's literally no other option. So I agree it's annoying to see everyone talking about how Erebor is this new bank in the vein of SVB that's going to be taking on higher risk bets. It's literally the opposite. It's do what SVB did for the people that SVB was otherwise successful in working with. So think like national security companies, hard tech, deep tech, biotech. These companies that just need to make sure they have someone who understands how their business works. And SVB did do a good job at that, but then also is going to make sure their money is actually there when they go to get it. That was the part that SVB screwed up.

Yeah.

How? How?

I guess I'll end with one bit here which is maybe worth noting. I'm not working on Erebor in an operational day to day capacity. I'm a board member. I founded it because I wanted this to exist. You got to remember that the finance industry is full of people who love finance for the sake of finance. They're truly in love with all the levers and the machines. They're in love with levers and what? That's right. And in addition they probably like what it enables. They like free flow of capital. They love the ability to leverage things like. I get that. Similarly, I'm a VR guy. I love VR for the sake of VR. I love what VR can do. I love that it can be a home theater. I love that it can be a classroom. I love that it can be a time machine. But I also strictly love it for the sake of the technology itself that you're presenting an artificial view to your peripheral nervous system that is sufficiently advanced as to convince a body that has developed for millions of years to discern what is and isn't real. That's incredible. I love it. I would love it from a tech perspective, even if it were useless from an everyday perspective. And I would Say, that is not how I am with finance. I don't love finance for the sake of finance. I'm not a finance bro. I want something like Erebor to exist because of and for the sake of my love for all of these other technologies. If something doesn't exist, a safe, reliable banking partner for people like me who care about tech for the sake of tech and what tech can do. If something like that doesn't exist, people are going to have a hard time. So it's a little weird. It's one of the few finance companies that started by somebody who wants it to be in service of all the other true interests they have, rather than the culmination of their own personal interest. I guarantee JP Morgan, Morgan, Goldman Sachs, these firms were not started by people who said, you know, I don't really care about baking, but, but, but I do. I do want to make this technology work.

Yeah, yeah, no, that makes a lot of sense. Speaking of other articles, there's this interesting article in Business Insider from six years ago. The US army wants mixed reality headsets that detect enemy fire, translate language, and see in the dark. Anduril, the startup founded by Oculus Palmer Luckey. Oculus's Palmer Luckey is on it. This was a six year project. What is the full story here? What was the genesis of this article and then how did the product come together?

Well, the thing that's so interesting, and I think there's a quote you should dig in there for. There's a quote from Brian Chimp.

Yeah, that's right. He says the real moonshot for us is this idea. You want to have every soldier, every operator be able to have total awareness of what's going on. They know everything they need to know to do their job. And all of this is available to them in a millisecond. And just the most critical information that they need.

I mean, you could take that quote and it's practically word for word what the army is saying about soldier borne mission command. I mean, it could be the press release for Eagle Eye, which we just started showing off publicly at ausa. I mean, it's interesting because a lot of people think that this move into announcing our augmented reality effort efforts is this new thing that we pivoted into, rather than the culmination of eight years of platform building, building the software that you need, building the data integration techniques you need, the radio relay and meshing systems that you need, there's so much you have to do to accomplish this dream of a soldier. Born heads up display that Shows you where the baddies are, where your buddies are, does your ballistics compensation and calculation. I mean, it takes so much work. And we've been working on this since the beginning of the company. And so I think that article is interesting because this was six, seven years ago and people are asking, what are you guys going to do? We said, oh, we're going to build combat heads of displays to integrate data from all these sources and put it right in the soldier's field of view. So it's available in a millisecond. And when we said that back then, people were kind of giving us side eye and they were like, okay, one, that sounds crazy. Two, the army just gave, like, to put this in context, that was like six months after Microsoft had won the IVAS contract. They're like, well, that's what Microsoft's doing.

Why?

Why would you be working on that? And even then, it's because we had a vision for what it needed to be that was somewhat divergent from what the army or Microsoft was doing. And so we kept investing, we kept building. And I mean, who would have bet that eight years later, that Microsoft contract, that $22 billion contract vehicle for the architecture of the Army's AR future would move over to Anduril, that we would be launching something like Eagle Eye, and that it would be an extremely. You know, that when we took over SBMC from Microsoft, that we integrated our UI and features that we've been building on Lattice in less than two weeks. We did in two weeks what the army had been working on for years, because we'd been building this back end for it and kind of preparing for the day where this might happen. So it feels very much like a culmination of destiny, a bet that paid off. I've been definitely talking to my investors and reminding them that they told us that this was a moonshot that probably was not going to pan out. Try telling someone, I think Microsoft is going to transfer their $22 billion contract to us at some point. It's just like, they're just like, Palmer, that's magical things thinking like, it's not like it's a. Literally, I've been told, Palmer, that's the type of thing you think when you just miss the boat and you just can't bear to.

Yeah, they're like, you're just coping. You're just coping.

You're just coping.

It's cope. It's pure cope.

I say, well, let's, let's give it up for a culmination of Destiny.

Yeah. Culmination of destiny.

I'm always a fan of people achieving. Achieving their destiny. I haven't gotten to mine yet. Although, you know, I won't do this. You know, I think you guys might have even asked me about this at one point. So I have this goatee that I've been growing ever since I was fired. And that was like coming up on nine years ago now. And I've told people who asked about it that I can't shave it until certain conditions are met, until I achieve certain, certain goals in my life. And I've just been cryptic about it and never said why. I don't plan on changing that. But I will let you know, I have achieved those life goals. I have achieved my destiny to end that I set for myself eight years ago.

Wait, so does that mean you just keeping the goatee because you like it? Is that what you mean?

I said I'm not going to shave it until. Until these conditions occur. And there were people who believed that those conditions will never occur. We'll have to talk about it another time.

Sure.

But it's not that I must shave it, it's that I now can shave it.

That's right. That's right.

I tweeted about it a few years ago and I said, look, I'm not gonna tell you what's going on in my personal life yet here, but suffice to say, when the beard comes off, shit's about to go to down.

We need it. We need a. We need a beard. We'll get a beard tracker on the website.

Well, we have Brian Armstrong joining the show. Thank you so much, Palmer, for joining. That was a fantastic conversation. We'd love to have you back. We'll talk to you soon. How you guys doing?

Good to see you, Palmer.

Yep.

We'll see you later, Palmer. Have a good one.

Awesome. Thanks for coming on. Cheers.

Live long and prosper.

Live long and prosperity. That is correct.

Brian, welcome to the show. Double Brian.

As you can see, I've already achieved all my life goals.

Yes. Yes. No hair whatsoever.

None left.

You look fantastic. Welcome back to the show. Thank you so much.

How's your 24 hours been? You guys have been dominating the timeline.

This acquisition of Echo out there, Kobe is just such a legend. He's an OG in crypto, so he's been giving us good advice for a long time, whether we wanted it or not. He was right every time he put out some stuff. So every time I'd call him. I DMed him on X a while back and was like, man, you keep lighting us up on X, but you're right every time. And he always had good suggestions, so I was trying to court him for a while to see if we could get him in the company.

Fantastic. I mean, I went on this little emotional roller coaster. Maybe it's just because I'm a little bit of an outsider, but when the story broke that you'd purchased the nft, everyone you know, I was seeing, oh, this is so much to spend on a podcast. This is so ridiculous. And I was kind of fallen into that trap. And then when the acquisition happened, it felt like a brilliant way to kind of actually draw more attention and even just share a little bit more of, like, the roadmap of what's happening here. They're obviously connected, but talk to me about, like, how you decided to parcel out the information and why. Why?

Yeah, well, I wish I could take credit for that, but we actually have some really great people on our team that are, let's say, a little more Internet native and in the. A little dj. And we've been, you know, we've been hiring some folks that really are following this. So, yeah, they actually came up with the idea of, like, let's bring his podcast back. That would be awesome. It had a cult following. I think everybody loved it. The Up Only podcast. So really, as part of this deal, we decided to have some fun with that. And the real news, of course, is we're acquiring the company. But if we can get this podcast back up and running, I think that would be really fun at the same time. So, yeah, and then I just. The broader story is just like, we're really excited about capital raising coming on Chain. I mean, that can make the whole process more efficient, more fair, more transparent. Every entrepreneur who I know finds the fundraising process to be pretty onerous. It usually takes like, two to three months where everything else that you're focused on has to stop. You go do tons of pitch meetings, you get told no, 19 out of 20 times. You get punched in the gut over and over with these smart people telling you that your idea sucks, then if you're lucky, you manage to get this thing over the finish line. And it's always a very tenuous process with tons of legal fees. And so, yeah, we really think that capital formation can be just much more efficient on Chain. And Echo is. Is leading this space. They've helped over, I think maybe two or 300 projects now raise money over $200 million. And so it's not just going to be like, token sales. Like, I Think other startups. What keyword did I hit for that?

200 million.

200 million. We got to ring the gong.

Yeah, get that gong. Okay, let's do it for the Echo.

T.

I love it.

Yeah. And I guess like dive into that a bit further. My understanding is like they're a platform but they have, there's some editorial approach there. Right? You're not. For now, it's not any company coming on Echo and spinning up or is it? It's community led.

Is that it?

It's like they have to have.

Yeah, yeah. The idea is you can raise from your community and they are curating it a bit. Right. Because I think if it's totally open ended, you could have an adverse selection problem. But yeah, they're bringing on really high quality projects and you can imagine that Coinbase has about half a trillion dollars in custody from our retail and institutional customers. These people want access to great assets and investments. So Coinbase is really building a network effect here. If we can have great builders come in who want to raise money and connect them with investors who have the money, we're the perfect platform to help accelerate this and give Echo even more distribution. So first when they come in, they're going to keep Echo standalone, just make sure it continues to operate and do great. But over time you can imagine us integrating this tokenization fundraising platform into Coinbase and really distributing it to all of our customers.

Is there. Can you walk me through some of the more like life cycle of a modern crypto startup? Like imagining if I incorporate raise money on Echo, but then is there a world where I'm in an ecosystem of like Coinbase B2B products or I'm building on top of base or I'm custodying my treasury with you. How many different products can I pull off the shelf if I'm building a crypto startup in 2025?

Yeah, that's exactly right. I mean you're pitching the vision better than I could. But yeah, let's say that it's the full life cycle. You're two kids with a laptop and a dream or whatever. You want to start something, you can go in there and open a Coinbase account for your startup startup. Maybe we even help you incorporate on chain at some point with like a DAO or these kind of things. And then you need to raise your first bit of capital like a seed round. Press the raise money button and put your pitch deck together, make a video like we'll distribute it to some of our early or some of our customers that are interested in these kind of assets, we help you raise money. Now you've got kind of a business bank account equivalent open where the funds just arrive instantly with USDC from people all over the world. You don't have to be tracking, tracking down these wire transfers and lawyers and all over the world we just kind of. It's all raised on chain in a smart contract. So your capital arrives immediately. Now you can actually start to generate revenue as well as you're building your product. Like crypto payments earn another button click away. Maybe financing is involved there and eventually someday you're going to want to go public.

Right.

And have your company be traded by every retail customer out there. We can help you do that on chain as well. So you can imagine this whole lifecycle coming on chain. And I think the goal is this will just increase economic freedom. It'll increase the number of companies who go raise capital and get started out there in the world.

Yeah.

What is the process for the private markets coming on chain? Like it seems like there's now enough momentum. Right. Echo is one version of this. Did you know on chain companies raising on chain. That makes sense. But how do we get to the point where companies that maybe aren't, aren't crypto native are able to utilize these rails and then particularly, you know, tokenize their equity and make use of, you know, crypto's full potential?

Yeah, so it's starting with token sales, more crypto companies. But you're right, we're eventually going to get this, I think to be every company just raising money this way. If it's more efficient, they shouldn't really even have to care if it's crypto underneath. They're they're raising dollars or stablecoins, whatever that they want to actually raise. There are currently some exemptions which allow this to happen under the sec, both for crowdfunding or having accredited investors raise it. But we're spending quite a lot of time with the SEC to try to get the next generation of this working. That would allow retail to participate under certain conditions. The accredited investor rules are there for good reason. They want to protect people, the unaccredited investors. But it also prohibits people who are not already wealthy from participating in these kinds of upsides. In many ways, the accredited investor rules are unfair. We're hoping that we can find the right balance of consumer protection and also making these available to retail. It'll level the playing field, democratize access. That's what crypto to is going at.

Do you. Do you. Has this happened yet for a company to go effectively go public on chain. Do you expect to see that in the near future? There's so much capital on chain that would love to invest in, you know, that loves to invest in all types of different opportunities. But do you see that?

Is that not just dropping a token? Like, do you not think of Ethereum as the first company?

Yeah, in some ways. But I'm talking about like a company that, let's say, you know, we had Figma IPO this year, they could leave an allocation, like an on chain allocation.

Yeah. Interesting.

And so some would go through traditional brokerages and I just feel like you'd be getting the first calls for serious companies that wanted to explore something like that.

Yeah, so there's been lots of precursors to this. You know, the whole ICO craze and everything. It just showed the amount of demand for this or like the Ethereum launch or anything.

Right.

But there hasn't been anybody who's done it yet with like a traditional ipo, like you said, with an allocation in the on chain. And actually when Coinbase went public in 2021, I really wanted to do this. We spent a bunch of time with the lawyers trying to figure it out and unfortunately the SEC at that time was not the kind of. It wasn't ready, let's put it that way. Now we have an SEC that's much more engaged and willing to innovate on this frontier and they really want the US to be the crypto capital of the world. I think that we will see hopefully in the next two, three years the first company go public on chain. It could be in parallel with a traditional ipo, but over time, I think you'll eventually see a marquee blue chip company go public entirely on chain. And that's the direction this is heading. Crypto is eating financial services and hopefully Coinbase can be the primary account for people in that new economy.

How much are stablecoins an important piece of this narrative? I imagine that even crypto founders who are super bullish on crypto are still not in the world where they see, say, I want 100% of my raise to be denominated in Bitcoin or Ethereum.

Because, well, I have a funny story from 2021. I built a company called Party Round back in the day and we had a fundraising software that some variety Web two Web three companies were using. And we had a big Web three company fundraise with our product and we had the functionality to raise in stablecoins. And they say we actually, actually don't want to offer this because it's really kind of complicated to try to cuss. Like, they weren't set up custody the usdc and so they were like, no, we'll just use Fiat Rails. Because, I mean, the variety, the tech's obviously come a long way since then, but.

But, yeah, but I'm wondering, like, are. Are modern crypto companies thinking 10% in crypto or Bitcoin and eth or something like that? Or are there companies that are like 5050 or are they like, I want to use crypto Rails, but it's got to be stablecoins. Is it all over the place? What's the current mood around. Just if you're building a startup and you want to build a product and you need to pay employees the same amount treasury management every month, how do people think about that?

I think most companies are going to raise using stablecoins just because it's stable, to price the terms in that, but then right away they're going to want to keep a percentage of their treasury in something that's more inflation resistant, like Bitcoin. We're seeing even large public companies like Coinbase and others use Bitcoin on their balance sheet. So it's becoming more acceptable to have Bitcoin and the treasury as part of treasury management. It's almost becoming like a best practice if you want to hedge against inflation, but just to actually get investment terms done. That's probably going to happen in USDC with people just clicking a button, you to sign the terms, agree to the safe, and then the money is instantly transferred as part of that.

Yeah, yeah, no, that makes a lot of sense. We're having Brian Chesky on in just a few minutes. I don't think we ever got your full take on founder mode like you've been through. So I feel like you're in like, the best example of, like, a founder who's been at the helm through so many cycles. How have you distilled your philosophy, your operating philosophy, on. On keeping the team aligned, keeping focus, agility, keeping new ideas flowing through the entire Coinbase journey.

Yeah, well, I actually started my career at Airbnb and I got a chance to work with Brian and Joe and Nate. I've been a big fan of Airbnb and everything I learned there before going on to Coinbase. But, yeah, I'm a big fan of founder mode as well. I think Brian struck a chord with that when he put it out.

I'll take.

I'll tell you one of the things actually we adopted from his Founder Mode Talk and subsequent content that came out is we're now doing product events twice a year where we go out and just package up everything we've been building and communicate it to the world. That's just one small example. But I think the core message of it is don't apologize for running the company how you want to run it. There's definitely times every week where I have to jump down into the weeds and be like, look, we're not hitting the bar here. Let's level it up. Let's go in this direction. And so you're constantly, as a founder, sort of making those edits in the room, editing. People's thinking, editing. Who's in charge of that team editing the roadmap to get it all moving in the direction you want. So, yeah, sometimes micromanagement is underrated and necessary.

Sometimes it's okay to be a dictator.

By Kobe DM as well. If he DMs you something, you got to bring it to the team. Well, I think we have Brian Chesky joining just now. Thank you, Brian Armstrong, for hopping on the stream.

Look, it's your old boss.

It's your boss.

Hey, Brian.

Hey.

Strong black T shirt game today. I like.

Yeah.

Looking good, guys.

You both look fantastic. Thank you so much for hopping on the stream.

Congratulations to the Coinbase and the Echo teams. Awesome moment.

Cheers.

Thanks, guys.

Bye.

Cheers.

And Brian Chesky from Airbnb.

Here we go.

Joining the show. How are you doing, Brian? We're doing great.

How are you doing?

Congratulations on all the progress. We've been hoping to get you on the stream this year. We're really excited to sync up.

We have to kick it off with the most important question.

This is what literally thousands of people ask us to ask you this question. We have.

We want to know your one rep max.

What's your bench press?

Oh, for bench press?

Yes. Maybe all time. What about all time?

Best three plates.

Three plates. 350.

Best 315. I was better at definitely deadlift. I deadlifted, I think 595.

Whoa.

Which I think is six plates now. I do like squats, thank you very much. I do 315 for sets of 10 or 12.

There you go.

Fantastic. Yeah. We were reviewing some of the more recent shifts in your strategy at Airbnb. I would love for you to give us just an update on this idea of Airbnb as a community, Airbnb as a platform. We were actually debating this earlier. I met my co founders in my first company on Airbnb. Like that's like, that's how I needed a place to stay. They were doing yc. I said, this will be a perfect fit. I want to build a company. We wound up putting the companies together, and I thought that Airbnb was actually a community on day one. And so when you say you want Airbnb to community, is that a return or is that an evolution? Is that an Act 2, or is that a return to Act 1? How are you thinking about community building?

Yeah, it's a great question. It's kind of both. I mean, I think Airbnb started very much as a community because people lived with each other. It was mostly bedrooms. And so in that sense, it was implicitly a community, and everyone left reviews for one another. Now, as we've grown, more and more people have traveled with families and groups, and they've rented entire homes, and there's still a community. I mean, a lot of people. Two out of three people who book an Airbnb leave a review.

You.

That's not true of most platforms. So people contribute, but I think that we can go so much further. So I think it's a little bit of returning to our roots, but then taking those routes and then taking a giant step forward. And so I'll give you a couple examples. You know, we launched earlier this year, Experiences, and one of the things we noticed was one of the biggest things we want to do with experiences is meet other people. So we basically allow you now to opt in. You can publish on a guest book. So when people book an experience, they can see who else is going to. You can then stay in touch with people after, and you can message them, because a lot of people were, like, staying in touch, but they would have to exchange WhatsApp numbers. It's really laborious. You can't communicate with everyone. So these are just small things. But we're going to be doing so much more. I mean, ultimately, you know, Airbnb is mostly a marketplace right now, and I would love it to be, first and foremost, a community. And it means that the atomic unit of Airbnb goes from a listing, a home, to a permanent person. In other words, the most important asset we have are all the people, not all the. Not all the inventory. That is a really, really big shift. And we've made a lot of investments in our community. For example, we have 200 million verified identities. There's only 100 million, 180 million passports in circulation, U.S. passports. So we've done a lot around profile having more of an ironclad identity system. Building out there. Really no social networks anymore. There's no profiles in the Internet, maybe other than like LinkedIn. And so we think there's a real opportunity to build something like a social network in the real world where you can travel and live anywhere and you can get a home, you can get a service, you can get experiences, you can meet people, you can discover interesting communities. And it could go even beyond travel because we're starting to see people using Airbnb in their own cities, especially for service experiences. And why are we doing this? Because I like to ask entrepreneurs, like, why do you deserve to exist? And the best generic answer I've ever been answered is, because if I don't do it, no one else will. And I kind of think the thing that we're working on, that I'm not saying no one else will, but the thing that's unique to us is the idea that we have a community. And I think it's a superpower. But we could do a lot more with it because we're built on trust and people deal with one of the most intimate things they can do, share their home with one another. So I feel like if we build the system of trust, we can do a lot more with it.

I was reading an article about run clubs and they've become really popular and they were kind of branding, at least in the thing that I, I was reading is a modern dating service or something. And I'm wondering how quickly does community building turn in? Does Airbnb become a dating app at some point? Am I crazy for thinking that people will meet each other on this app?

I think they will. I think it'd be more like a college party vibe. So less of a one to one matchmaking and more like what makes a college party unique. You kind of trust everyone because they went to the same college. It's not like a bar. So there's a filter. And so people, people's guards are a little down at a bar, people are a little bit. Guards are up. Or like a dinner party.

Yeah.

And at a dinner party, it's. You don't go to a college party necessarily just to meet people and hook up. Just like you don't go to a dinner party to hook up, you go to make friends. But sometimes you will meet someone that you can start dating. And so it's more like we're going to create a trusted environment for people to meet. There's no subtext of dating or anything like that. Yeah, but of course that happens. And, and so that's kind of how we think of it, we're not going to do like a matchmaking.

Yeah, but it makes sense. If you go to a city and you want to do an activity and you go with a bunch of other people, you're going to make friends, you're going to make business partners, but you're also maybe going to become romantically interested in someone. Yeah, all the above. That's the nature of these things.

Jordy, what's it like transitioning? And I'm sure you consult with founders that are maybe five, ten years behind in the journey, but what's it like transitioning from the venture world in terms of this sort of very mild, it's very milestone based. You're going from series A to B to C, it's very clear to ipo and then you're public and you're just out in the ocean. Right. And it's just sort of this endless sea of opportunity and paths and journeys. But I'm curious, like what mental shift was required to move into the public markets, especially now that you're five years in.

Yeah, I mean this might be surprising to say, but I felt like it was harder to be a late stage private company than a public company company because the late stage private company of all the disadvantages of being public, like your financials are like well audited. There's beat reporters covering you when you're of large scale. We found a lot of things leaking to the public market. So you know, we are mark to market. So we kind of have a lot of the disadvantage of being public where there's always a sense when you're a private company get really big, like you're hiding something or there's more to the story. And so this insatiable desire when you're a private company for people to quote, get to the truth, get to the bottom of it. And once you go public, everything is disclosed through the S1, the document you have to file with the SEC. And then I think people just assume like, like there's so much more transparency. I think it's harder and easier. It's a little less milestone based. I mean you literally, like I used to pay a really close attention to our valuation every round we did. And you're all, everything you do as a private company, it's often to get to the next fundraising and that can be helpful. But you can also make a lot of trade offs. Like I think we made just to our prior test topic. I think we made some tradeoffs around human connection in community, probably a period of time in the name of growth. And I think most companies, you live and die by that next round and that round being an up round, not a down round and making sure you have enough money. And we were one of the most profitable companies in tech. We raised not a lot of money relative to our valuation. By the time we went public, we had cumulatively burned I think $0 a cash or very little, little free cash flow flow and yet we were still pretty beholden to it. Now once you're public, it's almost like it's so omnipresent that you almost just, it becomes part of the background. Your valuation changes every second of every day of every business hour. And once you're a public, you realize this is you're going to have a stock price the rest of your life, even if you're not CEO anymore. You're going to live and die by it. Somehow I was able to just put it out of my mind a little bit and more. The quarterly earnings is kind of like instead of every year you have to explain yourself. The explain yourself every three months. But you know, I think you have to develop like thick skin. And I think the public markets teach you that you cannot like your valuation isn't your value. Do I mean by that, like if you associate your value to valuation, then your self esteem and the morale, the company's going to go up and down. And I think, you know, Jeff Bezos told me really early on this is more of a press thing. But he said, and this is when Airbnb was on magazine covers, when they had magazines and you know, 10 years ago and we were gone covers. And he said, beware, today's poster boy is tomorrow's pinata. And things are never as good as they seem. They're never as bad as they seem. And I think you have to have that like, don't get too high when your stock price is up. You know, when we in public, we pop from like we were marked at $18 billion. Five months, six months later we went 100 billion. And I'm like, we weren't as bad as when we were 18 billion billion. We're probably not as good as 100 billion. And today our stock price has been pretty flat. We're probably a lot better than people giving us credit for. So you just got to like, like not focus on that. And easier said than done. But the more you can like drown that out, the more you can focus on what really, really matters, which is creating a great product and eventually be rewarded for it.

Yeah, a lot of times when we talk to public company CEOs, they'll tell us that one of the benefits is that you have a public currency for M and A. How are you thinking about M and A? Airbnb doesn't feel like salesforce type company where you're just going to buy up a bunch of B2B SaaS companies every couple weeks. How are you thinking about other additions to the Airbnb ecosystem? What have you done? What have you looked at? What's your philosophy to that?

Our philosophy is to be extremely discerning around M and A. The reason why is we've taken an approach to run Airbnb that's kind of similar to Apple in this early days, where we are one app, one brand, and we're a functional organization and we have one customer. Right. So that that means that it's hard to take a really large company and bolt it on because we have a functional organization. For those watching, I assume, you know, a functionality, but there's like an engineering department, a marketing department. We don't have a division that was. We don't have division. So it doesn't preclude us from doing M and A if it's a really great opportunity. Opportunity. But for us to buy a big business, it just has to be extraordinarily compelling. Now, we're still looking at acquisitions and I do think, you know, we do billions of dollars of stock buybacks. We generate a lot of free cash flow. We 4 to 5 billion of free cash flow every year. I think, you know, we're a good stock to own. I thank you. I like to think that, you know, this is a good stock to own and that there's a huge amount of upside given our multiple right now, which is not super high. High. So we're very much looking at it. We're also especially interested in smaller companies because they integrate much more easily with Airbnb. And of course, we're definitely looking at like, talent acquisitions as well, especially companies related to AI. That'd be really interesting to us.

Yeah.

On running on a. On a quarterly cadence, there's been some talk this year and maybe interest from the admin to move to a sort of biannual reporting requirements as a public company. CEO, what do you think the impacts would be to your business and what are some of the sort of more broad impacts that you would expect out of a move like that? Because it's easy for the podcast class or people on X to talk about why it would be good or why it would be bad. But I'm curious from your view.

Yeah, it's a great, great question. I think it would probably be, from the company perspective, marginally better. I don't think it would be a complete life change or game change. Like, we don't spend, we spend. We spend a decent time preparing for earnings. You know, you want to show respect to the investors and be prepared at the same time. If I was an investor, I'd want to know that Airbnb management was mostly focused on growing the company and making the company more valuable. So, you know, I tried to spend a decent amount of time on earnings. And I do think some CEO spend a lot of time on earnings, and insofar they spend a lot of time at earnings, it's probably best for shareholders in the long run. Even though they want more information more frequently. I think there's probably just more upside in management being a little less distracted and being more heads down and not much changes, of course, of three months. I mean, that's the other thing I've noticed is so little changes every three months that you end up getting asked questions around, like foreign exchange, like currency. See, like, like the, the more frequent the meetings, the more tactical the conversations often are. And I wonder if a positive outcome would be less management distraction. But also the topics will be bigger, more strategic, and actually more fundamental. I mean, we get asked a lot of really good questions, but we get asked a lot of things because they're trying to predict what's going to happen next three months.

Sure.

And so the more frequent you have to report, the more tactical the tradies are and so the smaller the topics are.

Yeah. So great. A great CEO. You're thinking how, how can we be a more valuable company a year from now, five years from now? How do.

There's a.

There's a slight misalignment in timelines between the frequency of the reporting, the questions that get asked, and how I think about the business. Like, I'm not thinking about, like the impacts that much of tariffs or like whether the currency exchange between Europe and the U.S. u.S. Dollar, like, these are not things that you can optimize that well, anyway. And I don't think these are things that should determine whether to buy our stock and hold it or sell it. I think it's much bigger, more fundamental things. And so I think that would be, I think, on balance, probably good. I can imagine people that trade on information, want it more frequently, but I do think there's a cost to it. And so probably on net balance, it's probably good to go to every single six months. But make no mistake, it's not a huge problem for me. Like if it's every three months, we're fine.

How do you think about the long term opportunity in categories that previously struggled with disintermediation? Airbnb hasn't had that problem. But as I think about new products, whether, I mean, we actually went through this whole era of like Airbnb for dog walkers, Airbnb for house cleaners and the problem with those product, those platforms was that there's a lot of disintermediation. Yeah, you meet someone on the platform and then you say, hey, I'll pay you cash the next time you come by. Do you see any long term solution to those with just building a better payments platform, review platform, or do you want to stay out of markets that have a risk of disintermediation?

That's a really, really good question. And because historically we're a travel business, supply and demand are different cities and there's not a lot of repeat business. So there's a very low risk, a dis remediation. And because the transaction is fairly high risk to both parties, like, you know, you don't want to get scammed and you want to have recourse, you want to have customer service. Every reservation has $3 million of damage protection that is voided if you go outside the reservation. So in our case historically, because we're global network effect, it's a high, high consideration purchase with a lot of protections. We haven't had this problem, but we are now facing this because we just launched certain services now, travel services like we have. You can get a chef for your Airbnb, you got a big kitchen, why not have a chef come over? We're starting to now see people booking chefs in their own city. And so if you get a chef and you love them, you might order them again. So we are now entering this and it's really local recurring services. Whereas the risk of dismediation and the more it's local reoccurring services that are commodity where the differentiation is not big like a chef. You might want the same chef chef, but you might want a different cuisine. And so you might go to a different chef. And so, you know, you might not use the regular. My view on this is we should do what's best for the customer, not what's best for our business model in the end. And if it means that we need to have a lower commission for repeat business or it means we have a different business model, I think that's okay. You know, we are absolutely looking at loyalty programs, things like that, that could create incentives. But I do think that, you know, you got to align your incentives. We never want to have a commission structure. Structure where you have an incentive to go off the platform. So if it's a recurring business, we should not do taking 15%. Maybe we don't take anything. Maybe we take a really low percentage point, a low take rate. But maybe you also accrue some loyalty points, something like that.

Yeah, you turn, you start building, you know, business in a box.

Right.

Or starts the other way you can do it is membership. Right. You can have a paid subscription service where you just get access and then within that closed garden, you know, you can do whatever transactions you, you want. So I think it's just a matter of start with a customer, work backwards the business model. And I think there's a number of different business models here, but I do not think a 15% take rate on recurring business model. A recurring service works. So the margins would probably be lower because you're not providing as much value unless you're providing a lot of other protections.

Yeah.

Where do you think AI is overhyped and where do you think it's underhyped?

That's a great question.

I try to, try to answer this without pissing off any friends.

I might piss some people off right now. I don't think I will. Okay, here's my instinct. ChatGPT launched three years ago, more than this time. Three years ago, people were not talking about AI. October 2022, people were talking about Elon Musk buying Twitter. Like that was what people were talking about a little earlier, people talking about crypto and people were not talking about AI. And people that were, were like, this is five or 10 years away. And then ChatGPT launched late in November 2022. And in the last three years, that's all anyone's talking about. There's an old saying, people overestimate what they can do in a year and underestimate we can do in 10 years. I think that's true of AI. I think that people are wildly overestimating what AI will do to society in the next two, three, four years. And they're probably wildly underestimating the impact of society in 10 or 15 years. I think it's going to be slow. And then all of a sudden, and so three years later, after the launch of ChatGPT, daily life is not that much different for the average person. The top three apps in the App Store, Apps, ChatGPT Gemini and Sora apps 4 through 50, including ours or most of them are not AI native apps. Most of the of us have some AI in it. We have an AI customer service agent, we think it's really great. But we're not an AI app yet. And so I think the real question is when does AI change daily life for the average person? And the answer to that question is when do the top 50 apps, when do all those consumer apps become essentially AI apps? AI native apps. Think of AI intelligence as like a gold rush. And in this case the gold is intelligent intelligence. And so you have companies that are mining gold. Those would be like the large language model companies like OpenAI, Anthropic, Google and then you got a whole bunch of companies that are basically creating picks and shovels. Enterprise. There are very few companies using AI in the consumer space at scale. In fact, I'm on the board of Y Combinator. Almost all the startups we are seeing are enterprise. They're not a lot of companies doing consumer. There's a couple of reasons why. Why? Number one, I think some people are nervous about ChatGPT killing their startup. I think they're too worried. I think companies are too worried. I keep telling people, and I told this to Sam Altman, one of my best friends, that no one company can run the entire economy. First of all, governments won't allow that. But second of all, it's just too much bureaucracy in a company to do that. And there's a reason that when Apple created the iPhone, they didn't make every app in the app Store because can you imagine how big of a bureaucracy that would have had to be be for Apple to build Airbnb and Uber and Instacart and Instagram and blah blah blah blah blah. So there's going to be a whole series of companies, but they're going to take time. My prediction is that in the next three to five years, not in the next year, you're going to see a huge boom in the consumer space of AI. And to me, the entire economy is going to be built around the consumer adoption. You know, enterprise supports consumer. And I do think enterprise is being adopted quickly. Tools are becoming more efficient. But I do think the big question is when does it reach the consumer's day to day life? And looking at our timeline of development, I'm going to assume we're a little bit faster than the average company because we're really focused on this and we're still like it's going to take a few more years for us to really transform the company to become an AI company. And eventually we want to be every bit an AI company as the truly AI native companies, because I think every tech company is going to be an AI company or they're going to cease to exist at some point. Some point. So the question is, how long does this shift take? It's not a year, it's longer than that.

Wait, so you asked, you told Sam Altman he can't control the entire economy. Did he say, I'm going to try anyway?

No, he acknowledged that this was, you know, how this topic came up. It came up with the dev day where they had essentially those SDKs.

Yeah, yeah.

And we were debating like, yeah, the GBTS. And we're debating like, well, what would that be? And my very strong opinion was he said like, what's the role of apps in the world of ChatGPT? And I said, I don't know how much different it is than the App Store. It might be a little bit different, but I don't think that every app is just a mere data layer. And by the way, I think what you need to do, unless you want to build everything yourself, and again, I think you have a whole different set of problems you try to build yourself, is you need a really robust SDK. You really. Robust SDK.

SDK.

The thing they launched was a first version, but it wasn't a very robust SDK. We have, you know, we have a community. You have to have an account, you have to be a member of the community. There's a lot of things that precluded us from being able to have a really good integration with the current SDK as it is. But I think ChatGPT could be an incredible platform if there is a really robust SDK. Just like, I think, you know, like we can still have our app on the, on the App Store.

Yeah.

What kind of, what kind of conversations do you think management teams are having around integrating with alarms? We've seen OpenAI announce partnerships with Etsy, you know, Marketplace, Walmart, notably not Amazon yet, or eBay. I'm sure those conversations are happening. But from your view, running, you know, scaled Marketplace business, what kind of, what kind of conversation and kind of concerns or questions do you think these other players have starting to integrate around agentic commerce. Around. Yeah, agentic commerce. And specifically integrating with OpenAI when OpenAI's ambitions are obviously incredibly bold and they do want to own as much of the user experience, I think, as they can.

Yeah, I totally understand why a company want to do that. I think Most of these companies are thinking of this as an experiment at this stage. I don't think that most these companies think that the amount of traffic or the business they're going to get is meaningful. Yet this is really about learning and deciding whether you want to participate or not. I think a lot of companies going have to ask themselves do you want to be a destination or are you going to be or are you going to allow ChatGPT be the destination or a large language model to be the destination? And I have a, I guess I have a unique view on this. I'm not an AI maximalist insofar that I feel like a few companies are going to own the entire Internet. Here's why. Number one, the models that ChatGPT has that are, that are in Chargeboot are available to everyone via an API. And if you don't use their model, you can use open source models that are three or six months behind in the in. For most things, a consumer cannot discern the difference between a frontier model and a model three to six months behind it. Like if you need to have a travel concierge plan your trip, I think a model like three, four months behind, I don't think you'll notice a difference for the average person because the queries aren't complex enough. So imagine as a thought experiment we replaced the name. I know this is not a perfect analogy, it's a little bit flawed in some ways. But imagine replace AI with electricity and it was like 100 years ago and three companies had electricity. No other company had electricity. Suddenly these electric companies would have a huge advantage managed. But we have this mental model as if these companies are the only ones with electricity. Every company is going to have the access to all the same models. Unless companies start limiting their models only to their applications. But then other competing models would then get more widely adopted because they will have an API. So you have to make a choice. Do you want to limit your model or do you want to be like AWS and AWS? Amazon.com does not get much of an advantage advantage that they're part of the same company as they make a point about this by the way. And so I think you're going to see a huge change where on the one hand we all have to decide how to participate with platforms they ChatGPT and I think if they build a really great SDK and we can still own the customer relationship, there's probably not a huge problem. Has to just be integrated correctly. At the same time, you have to remember we're Also going to have nearly as good AI. AI, right. Via the fact that even if we don't produce our own models, there will be an entire economy that will allow those models to be accessible via APIs. There may be some advantages to the companies that build apps within. And so I think then, then it goes to the mental model. Do you want to go to one destination that then is like a macro agent that connects to all other agents or do you use different apps and those become different agents? So now we're starting to debate these mental models and there's a tradeoff. The trade off is the advantage going just to chat. CBT is now one agent can kind of cross pollinate and organize everything. But then if the SDK is limited, it will be not as powerful as going direct to the app. That is an app that can go really, really deep and do your job really well. And so this is the balance in where do I think this lands. Where I think it goes is I think think ChatGPT has to build the SDK that's really robust and it will be just a channel. That's my guess, but we'll see. And I might be wrong. And, but just remember that like all these companies are going to eventually have access to AI and we're in. So we're going through this whole electric electrification, so to speak period over the next three years of putting the latest models into our apps and for us to do that, we have to basically rebuild the, the apps from the ground up.

Yeah. To go back to Airbnb a little bit, I'd love your take on where culture is going. In the era of like online and offline content, there's this weird tension where everyone's brain rotted on TikTok watching five second videos. But the Taylor Swift eras tour is the biggest concert and everyone's talking about it. Or the sphere in Las Vegas is this place where people go to visit, run clubs are really popular. Yeah.

The other thing, there's some quote, I don't know who to attribute it to but like when, when an American has like a free week, they want to immediately go on vacation and experience somewhere new that feels really enduring.

And so I feel like you're in a unique position to kind of comment on like the runaway brain, brain rotification of American culture versus like touching grass basically.

Yeah, I think it's a great question. And when I came to Silicon Valley there was this thing called social networking. And social networking was literally, as you can recall, 20 years ago, a way to connect with your friends. Basically. People I cared about shared stuff they cared about. And social networking may be the most popular product of all time. That was invented and then uninvented. It was literally uninvented because around 2012 it became social media media. And the moment it became social media, your friends became your followers, you stopped connecting, you started performing. And social media is now becoming not social because pretty soon the feed became algorithmic, not your followers. And now with Sora the content is going to continually become AI. And so pretty soon the name social media is not even the appropriate name. I wouldn't, I'm not sure I'd call Sora social media. I don't know how social it is right now. Now it's really AI media if it's anything. And I think AI media is going to be where this goes. And I think the problem with it, maybe it's a problem, maybe it's not. But the name artificial intelligence, AI, the key word I think is the first word, not the second word, artificial. And I think what's going to happen is more and more what's on a screen will be artificial. Not to say it's bad, but it will be like a fantasy land.

Yeah.

And increasingly I like to say you want to ride a trend or ride the opposite, a trend. And so if we're basically creating this fantasy digital realm that is highly artificial, I think in reaction to that, people want what's real. And the way you're noticing this, if you look at Jen's Gen Alpha, you know, really young people, they're actually some of them adopting social media less. They're starting to see some of the adoption of social media go down and they are. There's an enamoration with the 80s. Kids born after the 80s are obsessed with the 80s now. A bygone era that they weren't a part of, part of before technology took over people's lives. Right there is like this nostalgia and we see it because we do these like, like pop up experiences that's based on nostalgic and it like young people are obsessed with them, especially eras before them and concerts are more popular than ever people now Americans go to Europe on vacation more than ever. Like how many of your friends do you know that go to Europe for vacation? And 10, 20 years ago they weren't doing that. So people are looking to get away and, and have experiences. And so when I automates everything or more and more things, I don't think is going to change how we go on vacation, the physical aspect of it that much. It might change how we book it, how we connect. But we want to be one of the companies that are getting people off their phone. If this was 20 years ago and you were like time traveling to today, you would remark how the world looks almost identical than 20 years ago physically. Except now people are looking at these pieces of glass all they long. What is in this piece of glass and they're just in this vortex. And I think increasingly, I think going to see a little bit of a reaction against that. This is not an anti phone like rant. This is not an anti AI thing. It's just about the fact that we need to have a balance. Do you ever notice that devices and screens aren't usually in your dreams? There's something about the digital realm that doesn't quite stick in your memory the way physical experiences do. And I think increasingly if I frees up more and more of our time, hopefully that time can be spent in the real world having meaningful experiences with people we care about. And to me that's what life is really going to be about. And I want to be a part of that. I want to be a part of getting people off their phones into the real world, meeting people, making the world feel smaller, having cool services, having cool experiences. And most these jobs that we're going to produce or, or like workforce is not going to be automated by AI anytime soon. I don't think you want a robot massaging you or pouring you wine. And so I just think there's AI is going to lead to this acceleration of really AI media which I think is going to push a lot of people also into the physical world. Just as you're saying.

I think people will slowly wake up that, that, that Airbnb is an AI bet, but not because you guys are going to use all the different forms of digital intelligence, but people are going to increasingly just want to log off. Last question if we have time. How do you see the role of the designer evolving with various gen AI tools? It feels in my personal experience the value of great designers is actually just going up. I want more of their time. I think they're commanding even higher and higher premiums. But I'm curious what your you is.

I think I would agree with you. You know, it's funny, just a quick story. When I started in Silicon Valley, I remember I pitched an investor and one of the investors, I give him credit for being brutally honest because the end of presentation he said I like everything but you and your idea. That's what he said. And I was, I want, I would, I Didn't think that was an insult. So I got home and I thought, well, what else is there? But he basically said, he was basically implying two things that, that didn't seem like they seem reasonable. Strangers won't live in each other's homes. Okay, that seemed like a reasonable conclusion. And designers don't start tech companies. And it seemed plausible. And there weren't really a lot of role models. The closest thing to a role model I had was Steve Jobs. I don't know if people thought of his designer. I kind of did. But then, you know, when he passed, there weren't really a lot of other iconic founders that came from that kind of work world. And I view design as a huge differentiator. And in a world where, you know, software, software program is a language and AI is really good at language and English to Spanish, English to a software language, I think the role of designers can be really important. And also I think everyone's going to be a designer whether they want to call themselves designers or not. Engineers, marketers, other people are essentially making design decisions, decisions. So then the question is, well, what is design? I don't think design is how something looks, it's fundamentally how something works. Design is an assembly. And I think the role of a designer is similar to the role of an architect of a building and engineers the ones building the building. But I think the architect is going to be one of the most important roles. And by the way, an engineer can design. I'm not saying they're only designers doing this. In fact, I think the tools will allow more people to be designers because you won't need to be be a craftsperson, you just need taste. And I think more people can learn taste than can get through the craft. And this allows engineers, designers and everyone to be designer. Now some people will be better designers and there really is going to be an expertise around design, but it's going to be about taste, it's going to be about intuition. And intuition is not this like, like just gut feel thing, like a vibe. I think intuition is based on your expertise. And one of the things I learned about great design is great design is simple. And simple isn't about removing something. Simple is about distilling understanding something so deeply you can distill it to its essence. And so I think great design is about making the complex simple, about caring about every detail, about having taste. Taste means having a sense of culture and history and where the world's going. And it's really about systems oriented things thinking. This sounds a lot like the skill sets you need in the age of AI. And so if we believe that, then I. Yes, I do believe that, like, suddenly more people can be designers because it's going to be much easier to build things in the age of AI. So I'm very, very bullish. And I thought when Apple rose, that would lead to the creation of all these designers as founders and designers being elevated in Silicon Valley. It kind of did for a moment, and then it kind of subsided. I am very optimistic that generative AI is going to really lift design in the world and make it one day equal to engineering. I don't think that's a crazy thing. One day.

Last question for me. We started with a bodybuilding question. What was it like being a bodybuilder at risd? Were you a fish out of water? It feels like bodybuilder, I expect, sec school or something. But was that a weird experience or was that just normal?

It was weird. And I was a weirdo at college. I'll be honest with you. The way I got into bodybuilding was I was an ice hockey player growing up, but I was very skinny. I was 100 pounds, freshman year of high school, and I went to a sports academy, like a prep school for ice hockey. And I was way too small. And my senior high school, I was 125 pounds and I broke my leg playing ice hockey. I had to do physical therapy. And I was kind of like an overachiever. And I decided I was going to start bodybuilding. And I got really, really into it, kind of obsessively into it. And I remember my friends were teasing me about how skinny I was. Was I said, I'm going to be one of the most muscular teenagers in the country. And I got into bodybuilding and I loved it. And the reason I loved it. And I ended up competing nationally as a bodybuilder. And I loved bodybuilding. And it was before anyone in tech was, like, working out or anything like that. It was kind of weird back then. Now a lot of tech founders have trainers and are really into longevity.

Creatine.

What?

Yeah, exactly. No one knew what that was. Back then, people thought creatine were steroids. Like, they couldn't discern the difference between the two. And I learned a couple lessons from bodybuilding, too, that I bring to Airbnb. The first lesson I learned is, you can change your body, you can change your life. And I was a kid. My parents are social workers, and I didn't grow up in an environment where you thought you could change your life. I Kind of grew up in an environment where a lot of kids didn't leave their hometown. And so to be a tech founder to kind of design the life you want to live, that didn't seem like anything that someone told me growing up. I think we take it for granted. But a lot of kids watch, watching probably come from hometowns where that doesn't seem possible. If you can change your body, you can change your life. It's almost the most tangible thing to change is your body. The second thing I learned from bodybuilding is you can't get in shape in one day. There's no one workout, I think, that gets you into shape. It's consistency over time. It's true of tech. Maybe you can have a flash of an idea, but you're not going to build a company in a day. It's better to just be consistent even if you have some bad days. And so you build your body one repetition at a time and you build your company one day at a time. So these are some of the things I learned. But at risd, I was totally a fish out of water because, you know, you have to eat a lot of protein. So I would walk around campus with like half a dozen hard boiled eggs, chicken breast. I'd pull them out at very weird times. I would keep sometimes like a steak in a Ziploc bag in my pocket and I whip it out and you'll appreciate. People thought it was pretty weird, but that dedication, I just didn't really care. And I'm glad I did it now that I'm 44, because it keeps you feeling young.

You'll appreciate my. I got quite into weightlifting in college. My hack was I would go to the school cafeteria, all you can eat, and they would let you bring sandwiches out. That was the one thing they allowed. And so I would take, take a sandwich and I'd make a hamburger puck of peanut butter and I would bring two between breakfast and lunch, I'd have two ham peanut butter sandwiches that were just like a full puck and then do the same thing between lunch and dinner. And then for dinner I'd have two more. And so I'd be happy.

Bulking. It's bulking.

Everybody has to know all those stories. Waking up in the middle of night to chug egg whites, all those kind of things.

There you go.

In my cafeteria you had this meal card and if you don't use your meal card credit, you lose it. And a lot of the women wouldn't eat as much and they would basically have all this meal credit they would lose in the year and so I made friends with them and they would give me their meal credit and I would. That was how I got my protein.

That's amazing. Do you follow the sport today? Are you a Sam Sulek fan, a Chris Bumstead fan, or are you too busy with Airbnb?

I'm pretty busy. I follow it a little bit. I like the classic bodybuilding more than the open class. It's just gotten kind of out of control. Opinion. I got the honor to meet Arnold Schwarzenegger and train with him at Gold Gym, Venice. That was kind of like a dream of mine, and that was really cool. But I don't follow the sport too much anymore. But my trainer was a former Mr. Olympia competitor.

Oh, no. Very cool. Well, thank you so much for hopping on the stream. This is a lot of fun.

Thank you, guys.

Lessons on bodybuilding company building life.

Yeah, we covered everything. We'll talk to you soon. Have a great rest of your day.

Thank you. Bye.

Before our next guest hops on, let me you tell you about Google AI Studio. The fastest way from prompt to production with Gemini. You can chat with models, vibe code, monitor usage, and more. And let me also tell you about Profound. Get your brand mentioned on ChatGPT. Reach millions of consumers who are using AI to discover new products and brands. We didn't even cover the Airbnb story of their brilliant SEO journey. During the Google era, Airbnb set up all these incredible landing pages for, you know, rentals in Tulsa, rentals in San Francisco. And so whenever you search for rental homes in a certain place, they were one of the greatest SEO beneficiaries during that boom. And you can be potentially one of the beneficiaries of the AI era with.

Profound Buco Capital bloke friend of the show says, God, I could kiss carpathy. SaaS has risen.

Sass has risen.

Of course, the first company I looked up, Salesforce, up three and a half percent today. Let's give it up for the chief mogging officer, Benioff. Back in the game.

Yeah, yeah. For a while, you've been. You've been like the a AGI bear AI sass bull where you've been like, it's all sass. It's all sass.

Every time I'm dropping it like agents, and it's like, it's SaaS. It's SaaS.

And you've been dropping it like a hot take, but you're not a doomer about it. And you're not like, it's nothing. You're like, it's the best. It's actually the best. It's a bunch of opportunity for entrepreneurs. There's a bunch of opportunity for business efficiencies.

I'm happy about it. As long as you're focusing on what matters. Proper enterprise workflow.

It's a bad day to be a manual workflow. That's for sure. That's for sure.

Did you see what happened? Did you see this picture? It looks like somebody found them filming the OpenAI.

Is this real or is this just someone dressed up as Ilya? So this is Chris. ChatGPT21 says this movie's gonna be sick. And it looks like.

It looks like Joseph Gordon Levitt.

Oh, you think so?

Supposedly playing Ilya, right?

Yeah, maybe. Oh, it does look like him. He's playing Ilya.

It would make sense that they would film in San Francisco.

Yeah, yeah, yeah. This is gonna be a wild movie. I can't believe they're making a movie so soon. It feels like the Facebook story kind of marinated for. I mean, I guess, what, Facebook came out in 2005. Yeah, it might be sloppy. Facebook came out in 2005 or 2004. I think they actually launched and then the movie came out in what, 2011?

2010.

Yeah, 2010. So Facebook was, you know, had five years to simmer, feels like. I mean, I guess OpenAI has been around longer, but really, narrative started with ChatGPT. Anyway, wild. Wild. Going to be a good movie. Hopefully it pumps some people up. We'll see. It's probably going to be a brutal hit piece.

I wonder if we're going to see any releases of who's playing Rune.

I. I would love to know who's playing Rune.

If they don't. If they don't include Rune, they should.

Yeah, they should just. He should be the animated character the whole time. You know how like in who Framed Roger Rabbit? It's. It's. You haven't seen this? You haven't seen who? Roger Rabbit. Who Framed Roger Rabbit is one of the first movies that was shot with film, but then they animated characters, cartoon characters over the film. And so the cartoon characters interact with the real humans and they, like, interact with each other in this funny, interesting way. The whole premise of the story is that the cartoon characters broke out of their cartoon and exist in the real world, something like that. And so I would love to see Rune played by an animated character character throughout the whole film. That would be pretty wacky. Should we go to Apple?

David sun says a supreme shape rotator can only rotate Shapes. But a supreme word cell can rotate shape rotators.

Yes. This is a quote of Mark Andreessen saying high IQ experts work for mid IQ generalists. What means? Yeah, I don't know if it's fair to call it Steve Jobs a mid IQ generalist Generalist. But yeah, Rune said this as well. The world is run by smart generalists. Smart generalists. There are no.

What's this Launch video from Apple. Let's pull it up before they probably.

Planned it on linear A purpose built tool for planning and building products. Meet the system for modern software development. Streamline issues, projects and product roadmaps. And if you're buying an iPhone, they better pay their tax on numeralhq.com sales tax on autopilot. Spend less than 5 minutes per month on sales tax tax compliance.

Okay, let's. Let's play the Apple video from.

We have our next guest already in the Restream waiting room. We'll bring in Stuart.

Every story you love.

Very simple.

Every. They're launching a browser that moves you.

They already have a browser they're launching.

The idea you wished was yours content.

You see that you can type whatever you want. They're not going to delete it if you go there.

They're all in just a flicker on a screen.

Who? Who?

Asking a intelligence is coming to pages.

This is. What do you see? This is Jane Goodall.

Is this a new Mac MacBook Pro? Is that what we're getting?

I don't know. It seems just like a pro Apple creativity vibe reel. People were saying it's kind of like a return to the roots. More like anti AI. Great idea. Start here. It's just like.

It's a Mac ad, Joe.

It's a Mac ad. But was that even a Mac? That looked like an iPad to me. Was that a Mac? I don't know.

It was a MacBook.

But it hits extremely emotionally because that ad is voiced by Jane Goodall who passed away on October 1st of this year. She was, of course, the British primatologist. She studied monkeys, studied chimpanzees. A very noble, noble cause.

Does she own a bored ape?

I don't know if she ever got into the Bored ape gang game. I'll have to figure that out.

Jane Goodall would have loved.

Extremely disrespectful to her like legacy. Let me tell you.

How is that disrespectful?

She's one of the most revered, some.

Of the most valuable apes in the.

World and you're trying to drag her down into the D gen mud you're.

Tarnishing legacy on chain art.

Stop. Before we bring in our next guest, let me tell you about Fin AI, the number one AI agent for customer service. Number one in performance benchmarks, number one in competitive bake offs. Number one ranking on G2.

And before. Before, before we bring purchased the new MacBook Pros available starting tomorrow.

Yes. I'm going to buy one M5 which is almost as fast as an M3 Pro and almost as fast as an M1 Ultra. It has this weird like thing but.

It'S not better than it is the.

Best MacBook you can buy right now I think.

Okay.

I'm not sure but there's some like plateauing going on in the M world. We'll have to dig into it. Anyway, we have Stuart Landsberg from Senior in the restream waiting room. Let's bring in Stuart. Welcome to the stream. How are you doing?

Welcome.

Doing great. Thanks for having me.

Great to have you. Thank you all for joining us.

We are both excited for this one. Yes, we hate fire and we love water. We love water. I live in Malibu, John lives in Pasadena. Beginning of this year was super chaotic. Fortunately our homes didn't burn down but we had a crazy start to the year and we are looking and excited to see more companies working in the space. That's great to meet.

Yeah. I would love for you to kick off with an introduction on yourself and the company and then we can get in the news.

Sure. Well, thanks for having me and sorry for you and for all the folks who live in Southern California about what happened earlier this year. I think it's a story we've heard too many times in California and that's a big part of why I'm here and why, why Seneca exists. I've spent the last decade and a half as an entrepreneur and doing, doing other things in technology. But over the last few years it's become I think increasingly apparent that building physical things that can solve real world problems. We need more entrepreneurs like doing this kind of stuff and maybe fewer people building, I don't know the, the things that are fast to generate financial returns. Exactly.

The slot machine, the trough. Yes. So anyway, where are you on the farm? You're helping put out the fires on the farm. We're big into farming analogies. Anyway, give us the news. What happened?

So started this company, started working on it about a year and a half ago doing research, doing ride alongs with fire agencies and was amazed at the quality of individuals that we have in the fire service and amazed that most of these people are doing their jobs. And remember, these are people who literally run into burning buildings. They're doing their job with technology that in many cases from the 60s and 70s.

Right.

It's for people before that.

When did we invent the hose?

Hundreds of years ago.

We invented buckets and hoses. And that's still, you know, key to the, to the fight.

Yeah, yeah. It's amazing. And so you look at what we give to the war fighter, right. The other people who we rely on to keep our communities safe. And then you realize that aerial attack, which is so necessary.

Right.

And you saw in the Palisades and in paradise where you couldn't get aircraft up in the air, how bad that was still largely using manned helicopters on platforms from the 60s and 70s that cost tens of millions of dollars. And we all know what's possible in autonomous aviation. So you put those two together and it became really clear there was an opportunity to solve one of the biggest problems in fire, which is how do you get to the fire before it becomes too big?

Yeah.

So put together a good team. Announced yesterday that we'd raised $60 million in startup capital. Appreciate the gong. Looking forward to that moment.

Raymond Thomas. Well deserved caffeinated capital.

Thank you gentlemen.

Let's.

I want to get into the specifics of the early product and kind of how you're thinking about, about. Obviously wildfires are a big problem. You can't solve everything to do with them. But you know why, why drone based system and kind of where, where's the early focus?

So when I, when I first looked at the problem, I assumed that we would go structure by structure. But when you talk to folks who are really deep, two things become clear. The first is almost no structure is safe. If you have a fire like the Palisades, right. The, the flame lengths are too long long. It's just too dangerous. The second thing is the holy grail of fire is really how do you get to fires before they become big? And a lot of times these fires start in places where it's like windy roads. They're not near a fire station, they're up in the wildland. And so the only way to get there is through the air. And so then you say, okay, well what we did, we built a fire model and said, what's a 5% risk fire? How quickly would you need to get there? How much suppression payload would you need to carry and do physics the work. And once we got convinced the physics worked, we're like, well, do the economics work. And the answer in both questions is absolutely. So we fly small fleets of autonomous suppression copters, as we call them, and they carry about 500 pounds per trip. They sort of can go round and round and round, load and refill from an engine. And the goal is to a lot of use cases, but the best is to stop an incident before it becomes something like the palace.

Yeah. I don't know if you know the story of the anduril firefighting tank. Are you familiar with this? One of the first projects they worked on was an autonomous tank that would fight fires. And apparently they ran into a ton of pushback from firefighting communities. That said, this is job displacement. We're worried about that. How have you framed this technology as, like, fitting in and augmenting the firefighter as opposed to replacement, replacing them?

So I think Andrew does better than anyone understand the job is not to replace the war fighter, but to give them superpowers and use their market copy. Our specific philosophy statement is to build advanced technology to support firefighters in situations that were previously unsafe, inefficient, or impossible.

Right.

And so we think about all the utility lines out there. It is literally impossible today to be able to get to those quickly. And so you need technology, technology to do that. You talk to firefighters today like, okay, well, there's a start on the ridge. You know, you're joking about, like, buckets and. But literally, they're hiking up in the heat with backpacks that have, like, a couple of gallons of water and an ax or this tool called a Pulaski. Sort of like half ax hacks, half shovel.

Yeah.

And so we're not. We're not trying to prevent those guys from getting up there. We're just trying to make sure that if you've got an aircraft that you can take off instead of by the time those folks get there, hopefully there's something that's under control enough that they'll be able to do their jobs.

Yeah.

Are you guys focused on detection at all? Because, I mean, when I think about Malibu specifically, it's the area that I've obviously spent more time thinking about than any other area with wildfire risk. It's like you have homes basically around pch, and then you have all this wildlife land up in the mountains that nobody's around. So, like, theoretically, a fire could start or could start on an electrical due to an electrical line falling. But how do you actually, like, how does the discovery of a fire happen so that you guys can bring the response?

There's a lot of good work that's gone on in detection, there's satellite based detection, there's sensors, there's cameras, there's people with cell phones.

Phones.

So the detection problem, it's not solved, but it's getting there. The big challenge is that it can take 30 minutes to 60, 120. Sometimes it takes days to get to a fire if it's really inaccessible and at that point can be too late. So we're really trying to fill that gap between. Okay. Detection is, is a problem that we sort of know how to solve fast response is something where there's only one solution. There is no solution today. And that's, that's really where we exist to stop. And I will say, say, you know a situation like Palisades, the difference in getting there from like you know, one minute to 20 minutes. Fire growth is exponential. Right. It could be 100x.

Yeah. How do you think about making the drones fireproof? I remember watching videos of like those huge like 747s dumping like tons of fire retardant and it feels like they actually don't really need to fly through smoke. But do you need to harden the system or can you kind of of use off the shelf componentry while there's a fire truck pulling in right behind you? That's it. Now you know, it's real background.

That's a great point right now.

Yeah, you timed this up perfectly. But yeah, yeah, just hardening for heat smoke. Is that relevant or can you just fly above it and it's not a problem.

So yes, it's relevant number one. And number two, if you go out and test on the field with firefighters, you can see in this video you got to be really careful. Right. If you put a helicopter above a fire.

Yeah.

You're going to put a lot more air on the fire. And anyone who's ever blown on a campfire.

Oh yeah, that's crazy.

Air does to a fire.

Yeah.

So if you want to build a drone based system. Yeah, we built a, we built a proprietary. Gotta be super lightweight because you're flying with it. Super high pressure pump. Like a water cannon. Yeah, that, that sprays foam and we spray it out of the rotor wash so it can go 40, 50ft from the air and hits the fire with pretty good precision decision. So anyway, that's like one of the core design elements. You end up having to deal with smoke and obstacle avoidance. And it turns out that some of the like IR stuff and it can get tricky but in practice, you know, hopefully if you're doing your job, you're in and out relatively quickly.

Yeah.

What about wind conditions? I'm sure you've tested it in a range of, you know, different conditions.

That was the big problem with the Palisades fire, was there was so much wind all that week. It was crazy. Crazy.

Wind is, is going to make any fire hard when it comes to aerial response. Just always like should start by acknowledging that.

Yeah.

That said, the risk profile when you've got a $30 million helicopter with two firefighter pilots in it versus a couple hundred thousand dollars piece of hardware that's autonomous, I mean, they're not on the same level.

Right.

So where you, you have to have a 99.999% chance of success in one mission. You know, could you take a risk in higher winds with a robot? Absolutely.

Sure.

The first thing, the second thing I'd say, if you look at our aircraft, right. A lot of drones are sort of agnostic about front and back. The whole thing is meant to be super aerodynamic. We've got 100 acres in Sonoma. That's our test range. And you know, when we fly up there, wind blows 30, wind blows 40. Sometimes wind blows more than that. Aircraft dead stable in the air. Doesn't mean that we're going to Recommend it for 40, 50 mile an hour winds. But we absolutely think hard about, you know, this is valuable a lot of the times, but it's most valuable when literally nothing else will do the job.

Yeah.

Okay. Last crazy sci fi idea for me because I've spent a bunch of time thinking about this. So I get like a small discount on my fire insurance because I have good fire sprinklers in my home, like a modern system. And the home that I own previously, before I owned it, burned to the ground because a single tiny ember had like flown and gotten caught in the roof in this one area and nobody was around. So the house burned down even though there wasn't like a wildfire in the neighborhood. And so I was thinking at some point, and I'm sure you've thought about this, I'm curious how you think about, about it. You could just have a drone based system that effectively had a fire extinguisher attached to it that would just monitor your property if there were wildfire conditions. Why? Why or why not? Is that a terrible idea?

D2C instead of B. Yeah.

Where like, I would go to Seneca and be like, okay, I want to buy one of your drones and park it on my roof so that if there's a fire and I have to evacuate it, it will monitor the Property.

We're hiring on the sales team. Do you want.

I love it.

In another. In another life.

Yeah. One of the really big challenges. Everyone understands.

Right.

90% of structure loss is ember driven. Which are for those who don't know, it's like a little spark that's floating in the wind from house to house basically. And aerial suppression is the best way to get those because they often hit the roof. Also an obvious. But you're really limited in terms of air resources. The Seneca systems are made to be stationed remotely. So if you're part of a community that's in a high risk area and you say, hey, I want to think about how do I defend my community. We're not meant to defend house by house. Right. That's. That's the wrong way to do it if your neighbor's house is ripping. You know, it's going to be hard for even the best systems in the world to keep your home safe. But if you can protect the whole community, that has real promise. And so the systems are built to be able to be up in the air. It's relatively easy with an IR camera to spot like a tiny start on the gutter and the roof, whatever.

Right.

Gets caught in these little, little mesh pockets or on the side of a deck. It's really easy to spot those. And then, you know, class A foam is not exactly. Which is what we shoot. It's not exactly what's in a fire extinguisher.

Sure.

But it rhymes.

Yeah.

So absolutely. We think about structured defense just like all the firefighters do. Right. This is a core part of how we make sure that wildfires don't. Don't prevent us from living in the American West.

Right.

We have to have to be able to protect our communities, we from fire.

Well, thank you for everything that you're doing. Congratulations.

So happy that you're building this company.

Yeah.

This is amazing. I think we, we probably talked about this exact so many times idea back in January and I'm really.

Congratulations. Good luck and we'll talk to you soon. Have a great rest of your day.

Great to meet you.

Thank you, gentlemen. Be well.

See you soon. We have our next guest in the Restream rating room. While we bring him in, let me tell you about Adio Customer relationship magic. Adio is the AI native CRM that builds. Builds scales and grows your company to the next level. We have Daniel Glassman from Samsung. Imagine putting Adeo the CRM on your tv on your Samsung smart tv.

That would be beautiful.

That would be a beautiful. How Are you doing?

What's going on?

Welcome to the show.

Great, thanks for having me.

Please kick us off with an introduction on yourself and the news today. We'd love to get the update.

Yeah.

Dan Glassman. I lead a new business development team at Samsung Cross Device, both TV and mobile. So working across AI gaming, gaming art, a host of other different initiatives. And today we had an exciting announcement. We announced that Perplexity is now live on Samsung TVs hit the gong.

Hit the gong. Congratulations. How does that work? I feel like Perplexity. I gotta type a bunch, like how do I.

But here's my thesis on why this makes sense. A lot of people want to have hardware, AI. Hardware devices.

Oh yeah, that's right.

You guys are sitting there.

Yeah.

We got these and millions of homes. Why not integrate it so that people, I imagine is the workflow. You can just ask a question to your remote effectively and it'll bring up whatever you want on the screen.

You nailed it. You nailed it.

Yeah.

So we have a new 2025 devices. So our freshest lineup. We have a new AI button. Hit the AI button and there's an AI home. We have a first party agent. We have Perplexity, we have Copilot. And really I think there are a host of endemic use cases. Right. Like what to watch tonight. I think like cavemen and cave women were arguing over what to put on the wall in their cave and we face that same, that same issue. And I think search has been broken for so long on TV and so this is maybe one step to help improve search.

Yeah. I mean, just the other day I went to ChatGPT actually and I asked like, what are the. What are the best documentaries in the last 10 years? Maybe true crime, maybe business focused. But I don't want anything that's too depressing. And I wanted to have over 70% on Metacritic and this and I had like seven different filters and it wound up pulling up like 20 different options. And that's something that is just so natural for an AI system, for an LLM. And why not just have that all baked into one system? So yeah, exciting. That's really cool.

Yeah, absolutely. And I think there are also a lot of orthogonal use cases non media. Right. Like I'm constantly on the couch trading my phone back with my wife around, like home renovations or trips we're planning. And so being able to query and saying like I want to trip an hour away and have that come out on the TV with hotels and options and then take action, I think it really Goes from this one to one use case of querying a mobile agent to more of a communal, let's call it like couch adjacent AI experience, which you know is phones down, which I think is really neat.

Yeah. How do you think about.

Oh, sorry, yeah, I was just going to say how are you thinking about partnering with other companies at the AI application layer?

Yeah.

So we, so we launched Perplexity. We also have Copilot integrated into our TVs as well. I think we want to bring users choice. We have our own companion first party, Samsung developed people have affinities to various different agents and we want them to be able to translate those affinities to the tv. And they all have their own strengths as well.

Yeah.

How are you thinking about the trade off around features which people want every feature individually. Sounds awesome. Adds a bunch of value versus product bloat and just adding so much that it becomes overwhelming and surfacing the right product at the right time, that feels like a really big challenge. But how is the trade off? How do you think about that internally?

Yeah, that's a great, that's a great point. I mean we've seen this as we've kind of expanded the aperture of use cases on the TV. So 10 years ago we launched Art Mode on the TVs and Art Store on Frame TVs and we're trying to redefine the black screen on your wall. And then we launched cloud gaming with Xbox and GFN and Amazon Luna. So console list cloud gaming and each step there we've had to figure out how do we create like the fastest path to engagement. And it's a work in progress how you balance choice and feature overload with getting people into really frictionless experiences. So I think it's a work in progress. Having a button on your remote is helpful. Yeah, just to get right in there and then to be able to query directly with voice.

But yeah, yeah, I was about to say I have a Samsung Frame TV, I have a PS5 plugged into it and I and I open up the frame TV and. And it was like do you want to play the game that you had in mind on your PS5 but cloud stream it? And I was like, I know I don't have an account. So yeah, having an AI button that's just like actually just switch the source to PS5 because I know I have the game installed already. I can speak more naturally and that makes a ton of sense. Well, congratulations on the partnership. Thank you so much for stopping by the show.

It's great to meet you.

We'll talk to you soon.

Thanks, guys. Appreciate it.

Have a great rest of your day. Jordy, get ready to play the biggest sound cue you've ever hit because I got a 99 on my eight sleep last night.

99.

Eight hours, 19 minutes. It's incredible. 100% on quality. 98. Oh, we're on roll. We're on a roll. I don't know if we've ever gotten that close to both getting 100, that is.

Go to 8sleep.com, by the way, after the AWS outage. Back in the game. Yeah, yeah, back in the game.

It was just training your body to work harder, to sleep harder. It was great. Anyway, we have our next guest in the restream waiting room. Harrison from LangChain. Let's bring him into the TBPN Ultradrome. Harrison, how are you doing?

Welcome.

I'm doing well. Thank you guys for having me. Excited to be here.

Thanks so much for hopping on the show. We'd love to get a brief update on where the company is. I think most people are familiar loosely, but I'd love for you to just break it down for us and then we can go into the news today.

Yeah, absolutely. So we started as like SO companies. LangChain. We started as an open source project about three years ago almost this Friday is the three year birthday. So we'll do something fun for that. And then started as a company a few months after that. Right. In the whole hype of the whole ChatGPT gen AI cycle and have evolved from an open source package into a whole company. We have Python and typescript packages and a platform now that's used by a number of sponsors here, including Vanta and Fin. I saw the commercial for them earlier.

Thank you.

And yeah, we've grown up into a company.

Awesome. And give us the news today. I want to ring the gong.

Get that gong ready.

Yeah.

We're excited to announce a funding round of 125 million at a 1.25 billion valuation.

Amazing. Let's get into the specifics of what kind of the core product focus today and how companies are getting value.

Yeah, I'm super interested in comparing how your approach is working relative to. There are like no code agent builders. There's all these different building blocks. And then if you go to yc, you can kind of get an agent that's pre built about anything as a SaaS product. So walk me through some of the customer use cases like how are people actually getting value out of Langcheng and how have they Evolved the way they work with LangChain.

Yeah. Okay, so core thesis. We think LLMs are great. We think they're going to transform what applications look like. We think they're way more interesting when you connect them to data and APIs and build these systems around them. And that's what we in the whole industry call agents now. And we think it's quite hard to build these agents. I think there was Karpathy did a great interview the other week where he talked about the decade of agents and how we're not super close to AGI. I think that's totally true and I think it's quite hard to build these agents. And so we've really focused on going almost down the stack and providing more low level tooling to build these mission critical reliable agents. So comparing to some of the no code solutions, which I would argue are mostly used for internal productivity things, we're much more focused on external facing or mission critical things. So there's a lot of, I mean Fin is a great example of a customer support bot. Like it's external facing, you're going to have engineering team building it. The interesting thing we've seen is that again we're primarily like a pro code platform, but a lot of the evals, a lot of the debugging, a lot of the prompting happens from product managers and designers and subject matter experts. And so that's really where Langsmith, which is the platform we're building, kind of comes in to bring all those worlds together.

Lang Smith, what does your token consumption look like? I imagine that I should think about you as like a SaaS company with great margins and stuff. You don't have to go into exact details but it seems like you're not in the token reselling business, the buy from the token factory resell intelligence. It feels like you're building more of like the proper shovel for the gold rush.

Yeah, that's exactly right. Some fun news though. One of the things we announced today is an agent in our product actually kind of like agent in the product and basically what it does is it will look through all of your agents logs. Basically one of the things that we saw is that people will put you put these chat boxes in front of people and they ask anything under the sun. You have no idea how folks are using it and so people want in. A big part of making these agents work better is understanding how people are using it and then bringing in the right tools, bringing in the right data for those use cases that people are trying to actually use it for. And so those types of insights typically done by humans, we now have an agent in the product that will help with that. But yes, generally our token consumption is very low.

Yeah.

What are you seeing what's coming down the pipeline on the consumer side? Obviously engineers are using coding agents. You have agents like Deep Research that are search focused. But what, what, what's exciting to you on, on that front?

Yeah, so you mentioned I'm coming from.

The framework of like the decade of agents and that like what we were sort of maybe promised last year that, that people said would get delivered this year is maybe just taking a little bit longer.

So one of the, one of the areas that I'm most interested in is this idea of what, what we call deep agents and basically build upon the idea of Deep research. Claude Code Manus, all of these general purpose agents that do run for an extended period of time and are actually, they're quite simple algorithms under the hood, but there's a lot of prompt engineering and context engineering that goes in. And so Deep Agents is kind of like this agent harness that will help power a lot of these more longer running things. And so to answer your question, I do think things like Deep Research, we see everyone building some version of Deep Research. Like it's just such a good kind of like product fit. Not only because the agents, especially for some of the search things, they're good at it. But in terms of the UI ux, I think it's a quite natural fit because basically it runs for an extended period of time, but it creates like this first draft of something. And if you think about AI in general, it's always struggled with like the last mile problem. Like it's easy to get to 80%, 85%. And so if you can come up with, with like products where AI will do still a sizable amount of work, but you have a human in the loop when it gets to the end and it can review it and it's basically creating what we call like a first draft. I think those types of products are really well suited to be kind of like the next gen of what we.

See come out with the raise. I imagine growth has been fantastic. Institutional venture partners, they're not ripping checks on a whim. But what's been the secret to growth? Has it just been a bunch of great customers that have been scaling up consumption and their LangChain bill is going up and so you're making more money? Have you been moving up market to the Fortune 500? Have you been just onboarding tons and tons of new companies Maybe a combination but how should I understand the way the shape of the business is changing?

Yeah, it's a combination. I mean everyone's doing stuff with Genai. I, you know we have, we have gen native companies that, that are better customers. We have big enterprises that are big customers. Everyone's doing stuff there. I'd say you know of our, of our revenue about like 30 to 40% comes from our self serve to maybe give some kind of idea there just a split. So it's a pretty even split. Like I think we, and so we see everyone kind of doing things. I'd say in the past, past it has been a lot of the consumer facing companies. There's kind of this interesting dichotomy of use cases. So consumer facing use cases are like really high volume but they're generally like shorter interactions because latency is still really high. It matters a lot totally. And so we, and so we see like a lot of our usage coming from these consumer facing companies. But then on the other end you have these more B2B companies where the, where the agentic workflows are just much more valuable. Like they're doing way more work and so there's maybe like fewer of them but the ROI that they're providing is like much higher. And so the way that, the way that we charge is basically usage based and so a lot of it comes more from the consumer side of things but we see that there's a ton of value on the B2B side as well.

How are you thinking about the open source strategy? What are your role model companies there? I'm always fascinated by this. I know the story of Red Hat Linux a little bit bit GitLab. There's a bunch of other data bricks and it feels like open source is sometimes just like a tool that people pull off the shelf as marketing. Other times it's like the company would not exist without the open source community. How do you see that playing a role in evolving over the next couple of years?

It's a huge part of our company and what we do. The way that we think about things in terms of the lifecycle of building agents is there's like a build phase and then there's like test run and manage and everything in the build phase we try to do in the open source. And so this is LangChain and Lang graph and so I think like Vercel and Databricks are two of the companies that I kind of think are good analogies there. And yeah, we want to build the platform for agent engineering Just like Vercel is building, kind of like the platform for front end engineering.

Yeah, but so like pitch me if I'm using the open source, the open source repo, I'm happy, I'm scaling and you know that I'm running this at scale and you'd love for me to be a customer. Are you just saying, like the downtime will be less or you'll get extra enterprise features or you'll have some sort of forward deployed engineer that can come and jump in, do a sprint for a little bit and help me level up. Like what's the shape of partnering with you if I'm already big and scaled on the open source repo?

Yeah. So it kind of goes in the story of build test run manage and we've kind of built out our product suite in that order. So people get started building like that's where they enter in. And so when they come to us on the commercial side, it's mostly for the testing phase. So the biggest blocker that we see that people run into is that their agents just aren't reliable enough. It's really hard to get them working well. So we have best in class debugging and evaluation solutions that work with or without are open source. Actually that's one decision we made early on is like the test run manage is going to be separate from the open source. Kind of like you can run more than just next js. So that testing and evaluation and debugging is the first thing that people typically come for. Then we have the run phase. And so this is really running agents at scale. We didn't do deployment for lanechain in the early days because to be honest, a lot of things that people were building in March of 2023 were really simple compared to what they are now. And so there just wasn't that much new infrastructure that was needed as we start going into like longer running and more stateful agents. Like that's really what our runtime is aimed at. And then the managed aspect of it is something that we're just starting to do more of. This is where like our insights agent comes in. So now you have like millions of traces going into your agent. How do you manage that at scale? And so our product development has kind of followed this, you know, this life cycle. I think we, I think we try to stay pretty grounded in where the industry is and what's needed at the time.

Do you think reliability will always be one of the key challenges to agents?

I think making them reliably good? Yeah.

Because right now it's like, okay, it's like 80% success rate for enterprise agents for a lot of use cases isn't good enough. And then it's hard to get it to 95 and then it's probably even harder to get it to 99 and then the last 1%. And depending on the use case, the reliability maybe matters more or less, but it feels like the age of agents will kick off when we have highly valuable reliable agents in multiple categories. But I'm wondering if it will just be kind of like a perpetual challenge because of the nature of LLMs and constantly being in the business of predicting the next token.

I think reliability is definitely the number one blocker that we see people focused on today. We did a survey about a year ago and like twice the amount of people said that they were worried about reliability compared to like cost or latency. And yeah, I mean I think one of the taglines we use on our website is the platform for building reliable agents. Like we think it's by far the biggest blocker out there. I think, you know, as an industry we're learning like techniques and practices to get better at it. The models are improving as well. That helps. I do think it will continue to be a challenge and I do think that the best agents that we see out there innovate a lot on UI UX to help overcome some of those reliability issues and put the human more in the loop. Like cursor for example. I would say like they're superpower in my mind. Like they've just nailed the UX of how you with interact, interact with these models and what it's like. And I think a lot of the ambient coding agents are taken off because the UX is great because it creates this first draft. And so that first draft paradigm goes hand in hand with this reliability thing. So we're approaching it from different.

Karpathy had a great take on this. He was like, yeah, you can get a self driving car to 99% accuracy and it's like you'll only crash every 100 miles. 99 of the miles will be safe. And then the last mile miles you get in a crash. And so he was saying that like it's not just going from 80 to 90 to 95% it's like you need to add a 9. So it's 99.9999. He really did the meme. I was wondering about the Carpathy. You mentioned it earlier. You clearly watched Carpathia. And Dora Keshe, what was your reaction to the Vibe shift around AGI timelines. There's one world where you're like, oh, maybe the froth in the market will cool off. Maybe it'll be harder to fundraise something like that is one possible reaction. Another one is like breathing a sigh of relief and that, like, I'm going to have a job for a long time because I'm building something that's really durable and useful. Like, how did you process that? Or did it up? Did. Did it update you at all? Maybe you were already thinking all the same things.

I think I probably have a boring answer here, which is like, somewhere in the middle. I mean, I think one of the first things Karpathy said in the first five minutes was, was that right now he's really focused on what these LLMs and agents can do for us and how they can be applied. And that's what we really focused on as well. How can we take these LLMs and do things? And I'm personally not an AGI maximalist. I still think the LLMs and agents will transform what applications look like. And I don't know. I think that's a reasonable middle ground to take. So I don't know if, if. I don't know if that interview updated a lot of my beliefs on things. I think one of the memes on Twitter recently has been the idea of that we're in a bubble with a lot of this AI stuff. And I think, again for that, I have kind of a middle take as well. I think we're probably in a bubble in the sense that there's a lot of interest and a lot of hype here, but I think there also will be a ton of value created. And I think both those things can be true at the same time. And we obviously aim to be one of the companies that creates a lot of that value.

Well, it looks like you're well on the way. Congratulations and thanks so much for hopping on the show.

Great to meet you. Great to be back on again soon.

We'll talk to you soon.

Thank you for having me.

Let me tell you about adquick.com, out of home advertising made easy and measurable. Say goodbye to the headaches of out of home advertising Only Ad Quick combines technology, out of home expertise and data to enable efficient, seamless ad buying across the globe. I would love to see a Lang Change campaign on the 101. Let's make it happen. Back to the timeline. Warren Buffett missed out on $50 billion in profits by selling Apple too early. Barron's calculated that he sold 650 million shares at an estimated price of $185. Now they're at $263. His cost basis was an estimated $34. One of the greatest trades I think he made. He might have made more money than any investment manager ever with that trade. And it was sort of a, sort of a narrative violation because he's often bought, you know, these low PE ratio stocks and not stayed out of technology and just generally stayed out of technology. Of course, Apple kind of violated them both. But it was a fantastic crazy.

Because I remember when the trade war hit, the tariffs, liberation day, everyone, everyone was like, oh, Buffett is a, is a genius, right? He looks so smart for selling when he did. And then of course ripped back up because nothing ever happened.

Nothing ever happens.

Roblox.

Roblox is in trouble.

Criminal Attorney general has said it has become a breeding ground for predators to gain access to our kits.

Yeah, I'm very interested to see where this goes. We've talked to some folks at Roblox. Very cool piece of software. I've always said that. Yes, whenever you have kids online with chat rooms, there's risk of this stuff happening. But hopefully AI should be able to do even better filtering. Previously you filtered for keywords. If somebody's saying a certain keyword, that might be a flag for an administrator to go in. And with LLMs, you should be even better equipped to screen messages, just like Elon said. GROK is going to read every post now to decide where it goes in the algorithm. Roblox should be running every single message. I think no matter what the cost, it's worth paying to go through an LLM and say, is this indecent? Is this a problem? Let's flag this to a moderator. They also probably need to hire a lot of human human beings because we've seen the concurrent numbers of users on Roblox. It's what, 50 million people? You need a huge, huge task force to control that. You need moderators. It's not all public, so you can't have a community notes system. You can have some sort of tattletale system, I suppose, but when the stakes are this high.

Calling Buffet paper hands.

Paper hands. Yes. Well, Jacob Rintamaki is having fun on the timeline. He says, excited to do more writing. Introducing agents and demons. Our first place is on AI goulash, the gems one finds in AI slop. And just asking ChatGPT5, is this something cursor for serial killers? Stab, stab, stab. Well, you can go check out the latest substack AI gulag Bosch at angel and Demon AgentsandDemons AI and that's Daemons with a Dae.

And on that note, we'll see you tomorrow.

On that note, head over to getbezel.com your bezel concierge is available now to.

Source you head over to get Bezel.

Seriously, any watch give you some time. And Wander also dropped an update video. You can go find it on Kyle Tibbett's X account. So much exciting stuff coming coming at Wander. We've been happy to partner with them and you can of course find your happy place and you can book a Wander with inspiring views, hotel gray minis, dreamy beds, top tier cleaning and 247 concierge service. It's a vacation home, but better. Thank you for watching.

Very fun day. We had ambitious goals for the year but I wouldn't have predicted that we would get Brian Armstrong, Brian Chesky and Palmer on the very same show.

Yeah, we definitely in early be like it'd be cool to interview a billionaire on this show, let alone three in one day. Remarkable. Absolutely remarkable day. Well, thank you to everyone in the chat. Thank you to everyone listening and supporting us.

Can't wait for tomorrow.

We will see you tomorrow.

Have a great evening.

Goodbye.

Cheers.