Harvey's Margin Whiplash, Human Powered Agents, McLaren Rebrand, Insects > Humans? | Ben Thompson, Gagan Biyani, Alex Ratner, Peter Kalogiannis, Daniel Petkevich

You're watching TVPN.

Today is Tuesday, September 22nd, 2026. We are live from the TVPN UltraDome, the temple of technology, the forks of finance. Let me tell you about ramp.com. Little teaser, that's foreshadowing. You see me wear the yellow suit, I tell you about Ramp. Time is money, save both. Easy-to-use corporate cards, bill pay, accounting, and a whole lot more, all in one place. Uh, Jordi, his hat game has evolved. He's now wearing the fedora with the Blackstone Cap on top, which I assume can only mean one thing. It means that your top priority is safety in financial markets, financial markets, and the financialization.

But also safety as well.

Oh, you care about both?

I care about—

so you want to accelerate the AI buildout safely? Yes. I thought it was— I thought it was— you're going to blend the two and you're like, look, I want— I want insane private credit deals. trillion of capital to flow into the AI buildout, but I wanted to do it safely.

I want every contract to be investment grade. I want to prioritize safety on the whole frontier.

Okay.

Safety around the models.

The capital markets frontier?

And safety in financial markets.

In the capital markets.

Got it.

Okay.

Okay.

I'm seeing where you get there. Well, there's no safer place to participate in the capital markets than on the New York Stock Exchange. Want to change the world? Raise capital at the New York Stock Exchange. Just do it. Stop making excuses. Mobile apps. Ridiculous. Anyway, I got some Harvey takes. Should we run through those? You want to do Harvey first? Let's— People are taking shots at Harvey.

It starts with Samuel L. Jackson. I won't stand for it.

I won't stand for it. No, I like Harvey. We've talked to some lawyers that use Harvey. Seems like a great product. We also had Legora on the show just last week. Had a good time with them. Seems like a very competitive space, but the headline that grabbed attention actually came from Bloomberg reporting. But then this, this company called The Redline HQ actually did a pull quote on that, got the reporting, repurposed it. So you don't even have to go to Bloomberg. We'll talk about that later, too. No one's going to news websites anymore. But they say track changes. Harvey's gross margin fell from about 50% to -50% by June as agent token use spiked 20-fold on rented OpenAI and Anthropic models, Bloomberg reports. And so it seems really bad. Like you definitely don't want negative gross margins as a software company, as a tech company, like—

As any company.

As really as any company, actually, you don't want to be selling a dollar for $0.50. And that's exactly what they did in June. Basically, that's what they were doing. They were selling tokens at negative gross margins. Not good. But there's way more details in the actual reporting. So you should go read the Bloomberg report. Subscribe and actually read it because there's way more details. So the first thing is like cost did explode. They— but it was on the back of more product usage, which is good. They're growing ARR. We'll get into that. And token consumption increased 20x this year. So basically they are feeling the reasoning era, the agentic era. One, one job, one, one process is generating a lot more tokens.

Had seat-based pricing.

Yep.

And then AI got a lot better. So every seat was just using it way more.

Yeah.

Very simple.

But it's not, it's not just using it more. It's that you're— the lawyer might be saying the exact same thing as they did a year ago. They might say, review this, but instead of like one-shotting it with no reasoning, they went to reasoning, which generated way more tokens, and then they went to agents. So they were like, okay, it's actually gonna fan out, look at everything across your firm, all the other records that you have, all your different policies and MD files. Like everything that you have is gonna be, work through this agent. And so you're generating way more tokens. So that created negative gross margins. But the good news is that they're already back to positive gross margins, which is in the Bloomberg article. Kind of criminal not to repurpose that because it really makes it look like they are running at negative gross margins. They had a negative 50% gross margin in June, but they are back to positive gross margins. They changed their model usage. Of course, there's a whole bunch of models that are cheaper and better for certain things, and they actually fine-tuned an open weight model. So they did post-training on an open weight model and they're calling it Harvey Tennant. And there's some interesting details there. I want to debate with you on like the value of that, where that goes.

But first, Gabe, one of the founders of Harvey, responded.

Oh, he did. I was—

he said the hardest thing about building Harvey is doing what's best for our customers despite immense pressure to do what's easy. The easy thing would have been to force our customers into consumption pricing before they were ready and serve them worse models to protect our margins. We chose to help our customers transition on a timeline that works for them and give them the best models in the meantime, even though it hurt our margins. This meant optimizing our product through routing, harness improvements, and post-training so we could serve frontier intelligence at an affordable price. It also meant building the infrastructure for customers to monitor and manage spend usage dashboards per matter, cost attribution, spend caps, and ROI reporting. As a result, we improved our gross margins from -50% to positive in a single quarter.

Wow.

Despite usage doubling month over month and continuing to serve the best models, our philosophy is simple: do what's best for our customers, even when it's painful, hurts our margin, or draws criticism from competitors, ex, and the press. We believe the most important part of building a company is earning and keeping your customers' trust. You do that by doing the hard thing for them, even when it costs you. Bunch of great points in here, even outside of the Bloomberg piece and the subsequent coverage. Max from Lagora was going pretty hard saying it doesn't make sense. Oh, it doesn't make sense to train your own models because the public— basically like publicly available frontier models are dropping in cost and the quality is increasing. So you're just sort of like wasting your own. And again, like, that's a bet that he's making. Harvey's going to make another bet. Yeah, probably some mixture of both is going to be the right approach.

There's another interesting sort of like how bad is this situation, just comp, which is— so they had negative 50% gross margins in June. They're at a $400 million run rate for ARR, I suppose. And so that's $33 million a month. And so negative 50% margins means they lost $16 million, but they've raised like $500 million. So in terms of like burn, it's not super cause for crisis, in my opinion. And what's interesting is that even through June, maybe they did the Sequoia deal that valued the company at $15.5 before that happened, because it was announced September 9th. But I have to imagine that the deal probably got done with at least some insight into those negative gross margins in June. What do you think?

I would assume so. But the main thing is, like, every Every application layer company has gone through a moment like this, even if it was for a single day.

Totally, totally.

So it matters how quickly you can respond and adjust.

Yeah. So where do they need to get to? What is the standard for the legal technology industry? There's some comps here. eDiscovery provider DISCO, public company, 75% GAAP gross margins. It is a software tool that when you're going into discovery, you need to connect with a bunch of data. Okay, we got a whole bunch of text messages, a whole bunch of emails. Let's put them in a system so we can review them, sort of normal SaaS. I'm sure there's a bunch of AI involved now, but that business is running at 75% gross margins. Thomson Reuters has a legal professionals segment of software that's running at almost 50% adjusted EBITDA margins. So like true, true net profit. So that's sort of where they want to be. Also peer-play law firms have really high margins as well. Hard to comp them to businesses because they pay out to the partnership. But PricewaterhouseCoopers reports net profit margins around 41% for the top 10 firms. It's lower for firms 10 to 100 or something like that, but still lots of health in the legal industry broadly. Obviously everyone knows lawyers make money. That shouldn't be surprising. And yeah, tokens are getting cheaper by the day. There's a bunch of model launches. Opus 5.5 launched today. Groq 4.7 was yesterday. And Groq 4.7, no one was— even Elon was like, okay, we're not jumping straight to the frontier with this. It's going to be a couple more iterations, but we're excited with the progress. But there was one benchmark that really jumped out and it was actually the Harvey Legal Agent benchmark, which is interesting. So I don't know how insanely good that benchmark is. It might be, you know, you could benchmax against it, but The interesting thing about that particular benchmark was that it was a legal task benchmark based on cost. And so Grok 4.7 is particularly cheap for the type of legal work that at the very least Harvey wants to do. So you could see them funneling more token spend towards xAI and Grok. Do you have something?

You wanted to give it up for big law?

Yeah.

The chat wants to give it up for big law.

Let's give it up for them.

They're the real winners here. Additionally, GPT-6 Sol and Luna are starting to roll out.

Fun.

So we'll look out for announcements there soon.

There were some cool demos. I saw one Opus 5.5.

I mean—

Chad says the combination, the hat jacket combination isn't working. I'll try to adjust a little bit. I'll try to adjust. Hopefully that's better.

Flawless. Yeah, thank you.

Thank you for flagging.

Huge upgrade. Very funny. Yeah, I saw a cool demo where someone sketched out on a piece of paper sort of a trebuchet, and Opus 5.5 took that image into simulation and created like a virtual version of it. The AI continues to impress in terms of translation of things. Oh, your kid drew a, you know, some crazy mythical creature? Turn it into a story, turn it into a movie, turn it into a video game, whatever format you want. If you want to read a 2,000-page novel based on Call of Duty: Modern Warfare 4, like, you can probably do that now. So you can have your content in any specific format you want. That seems to be where the AI continues to impress. Anyway, let me tell you about Codex. Codex is a powerful workspace for getting work done with AI agents, whether you're writing code, analyzing data, creating content, or automating business workflows. Codex helps you move projects forward from start to finish. I've been having a lot of fun generating AI videos with Codex.

Should we pull up your latest masterpiece?

Yeah, the Literally Me compilation. It burns up the internet and Vibe Reels all of the time. But now If you see those Literally Me characters and you're like, I am Jake Gyllenhaal in Nightcrawler. I am— I forget who else is in there. Patrick Bateman in American Psycho. You can put yourself in the movie. There's Inception. We got Jordi in there. That's Drive. I think that's Heat.

Heat.

Mad Men. Can we call them out? I think that's Fight Club. You got Joker. You got Taxi Driver. You got Nightcrawler. No Country for Old Men. Social Network. Social Network. Godfather. Godfather. Back to Joker. I forget that one. Shawshank, maybe?

That one's Collateral.

Blade Runner. Blade Runner again, I think. Yeah. Some of these— some of the renders are really, really good. That one, like, your face doesn't quite fit on Leo's head, but that one's really good. That one works pretty well. This one's a little awkward. The Joker, it's really just Joaquin Phoenix's face. Didn't actually swap you in at all.

It's just—

Codex was even like, it didn't work.

Do the Joker.

And I was like, do it anyway. I don't care. It's fine. It's gonna be on screen for 2 seconds. That one's good.

John is really my human agent.

He just—

he's sort of like an ambient, always-on agent. That's what I'm saying. I don't even have to tell you. You just predict my needs, right? You'll send me something like this. Everyone should have a John in their life.

Everyone should. But yeah, fun, powerful tool. Still got $500 locked away in Higgs field in the wrong section. Can't do anything with it. So, maybe just give away the account or something. I don't know. Anyway, what matters more, people or insects?

Tough call, John. We should get into maybe one side of it first.

Yeah.

Before we form an opinion.

Okay.

The human side? You want to steelman the human side?

Look, and I'll just say that We're coming to you live.

Yeah.

We're coming to you. We had a rough morning with the, uh, with mosquitoes. We actually had to, we had to, we had to DoorDash one of those, um, bug bite treatment thing, the heat, whatever. What, I don't, what do you call them, Tyler?

Uh, I think it's like insect bite healer. That's what it's listed as.

It did work because I, yes. Sebastian says I'm famously, famously pro mosquito.

Yes.

But a mosquito did bite me on the face today and a bunch more times. So we ended up— I ended up capitulating and having to get one of these. But I don't like using these. I don't like using these. It's funny because I can't stand EAs and I think they're weird freaks and I don't think they should be anywhere near public policy or anything like that. Um, but the one thing that I can agree on with Ben, this guy writing this post that insects are more important than humans, is I don't think anyone should— I don't think we should harm the insects, right?

Oh, okay.

So I think humans are more important, but that doesn't mean we— you gotta like take them all out, you know?

Yeah.

Just go somewhere else, right?

You would take sort of— you would be altruistic towards the insects.

Ellie says mosquitoes are terrorists of the air. Uh, they really are.

You would— you want to be altruistic towards the mosquitoes. You want to be effective in your altruism when it comes to mosquitoes. You don't want to eradicate them.

No, I want them effectively to be nowhere near me.

Okay.

But I don't want to harm them.

Okay, okay.

But let's get into this.

Let's get into the Bentham's Bulldog post. First, let me tell you about Figma. Agents, meet the canvas. Your AI agents can now create and modify your Figma files with design system context. So an effective altruist who writes under the name Bentham's Bulldog went viral over the weekend for an essay titled Insects Matter More Than People in the Aggregate, arguing that the combined welfare of insects outweighs that of humanity. His basic argument is one of scale. Insects are so numerous that even if their capacity for suffering is only a tiny fraction of ours, their total suffering could still dwarf human suffering. It's almost just a thought experiment. There's so many other directions you could make Why not go pound for pound? Maybe cows are the most important because there's a lot of cows and they weigh a lot. There's probably more aggregate pounds of cow than humans. Maybe they matter more than humans in the aggregate, right? Or fur, like the amount of fur on all the monkeys in the world is more than all the fur on all the humans in the world. And so maybe monkeys matter more. I don't know. It's completely arbitrary where people draw these lines. But he estimates that for every— For every second of human life, insects collectively spend roughly 270,000 seconds dying, about 75 hours, and notes that more insects die in a single second than the total number of humans who have ever lived.

That's a lot.

Wow, that's a crazy stat. I mean, the guy pulled some cool numbers together. This is interesting. This is fun fact territory. A large part of the essay turns on whether insects can experience pain. Bentham's Bulldog argues that they can, even weakly. If they can, even weakly, their suffering should count morally just as ours does. What species you are is just totally irrelevant to whether it's bad for you to feel terrible pain. Using what he describes as a conservative assumption that insects experience pain at just 1/10,000th the intensity that humans do, he concludes the sheer number of insects would still produce vastly more suffering in the aggregate. He ultimately argues that insects may experience more suffering in a single day than humans have throughout all history. Now, is there a piece on this where this is the view of this community with the relative— like, are they looking forward to how a world might interact with a godlike machine superintelligence and humans? So, the idea is like, we got to stand up for the insects now because we are the insects of the future. So we want them to make the same calculus?

I think that's certainly like where the thought experiment goes.

Yeah, right? And so this feels almost like bait or like a trap where it's like, if you are the type of person that jumps on this and says, no, I disagree, like insects don't matter more than people in aggregate, then you're setting yourself up for like the dunk that's, okay, well, now there are a quadrillion super intelligences that can all feel pain in some abstract way, and so they're more valuable and so now you're the insect. And so, your previous argument can be used against you to advocate for your own eradication. And so, it's like, stand up for the insects now, lest you be discarded in the robotic future.

Yeah.

I don't know. The reactions to this were rough. People did not like this argument at all.

Well, it's just the timing is funny because there's so much tension building around the sort of EA rationalist movements with—

Which is super fractured. People don't realize how fractured it is.

Yeah.

But it's sort of boiling up to the sort of national level in a way that it hasn't before. And so, during the midst of that broader conversation to come in and being like, yeah, insects actually matter more than people.

Yeah.

In the aggregate.

Uh, what were some of the reactions? Doomer said, if you truly believe that A Bug's Life is worth more than a human life, you will immediately feed yourself to bugs. Do it right now. If you don't do this, you are a fraud, a liar, a poser. Um, Andre—

Yeah, I think it's actually important for people to read that before you go and comment on this article. Watch A Bug's Life.

Oh yeah.

At half speed.

Half speed.

Yeah, watch the whole movie at half speed. You're a movie guy, but—

I think we coexist.

Maybe you're not a fan of watching movies at half speed. It's not really how the—

I'm not really a half-speed guy.

But half speed is a great sort of tool to have in the tool chest when you want to deeply comprehend.

I like 2x speed, 3D glasses with the 2 different movies interlaced so I can get one movie in one eye, another movie in the other eye, one audiobook here, One podcast here.

Duo.

While I'm sleeping.

Yeah.

With the duo.

Exactly. Max brain rot. No, I think this is an odd— I mean, the fracturing of EA really leads to a really, really wide range of outcomes that get sort of lumped together. Everything from the Pace the Frontier essay to this. These are clearly very, very different pieces of communication, and yet they get sort of lumped together. And it must be full employment if you're in the public relations department over there.

Yeah, I was gonna say, I think the shrimp welfare stuff—

Oh yeah.

Makes more sense to this because like, I'm not really sure what, like after reading this and say I agree, like what do I, what are the next steps? Like we're not like really factory farming insects.

Sure.

Maybe through pesticides or something. Speak for yourself, brother.

Yeah, yeah, the shrimp welfare—

Got some big announcements coming soon.

Like what percent of these, Like 600 billion insect deaths per second are—

Preventable.

Preventable.

Yeah.

Where like with the shrimp, you know, you can say like, okay—

Change the farming practice or just don't eat the shrimp.

Yeah. So, I think that's why maybe this was less impactful. I don't know.

Yeah.

I mean, it's all a continuum to like— until you get to like the veganism, like the animals who you can pet, you know, should be able to live full lives. This is sort of the most abstracted version of that. Yeah, I don't know. It's odd. Yeah, very, very— but I mean, I guess it's a good, good opportunity to go viral if you're looking for some—

More importantly, update on TBPN's Road to Christmas. Oh yeah, we're down to just 93 days. 93 days. A lot to think about, a lot of moves to make, and we couldn't be more excited.

Start counting it down.

Yeah.

I was trying to think of a realistic AI Doom scenario. Like everyone's like, oh, I can't imagine it. And so I actually mapped out one that maybe I could take you guys through. I wrote a little script here.

Okay.

Called The Old Way. So, you know, fades in, starts in the TDP and UltraDome. Me, you, Tyler, we're all doing the show. All of a sudden, all the phones in the UltraDome light up. Emergency alert, unknown biological event, shelter immediately. Outside, alarms begin to sound across Los Angeles. A strange haze starts pouring through vents and drifting across the streets outside. I see a cloud approaching the UltraDome entrance. First thing I do, bioweapon attack, this is what you gotta do. Boom, right there. It's not going in your mouth, you're good.

Good.

Second step, pull out your gun. Boom, boom, boom, boom, start blasting whoever's responsible. You're just firing wildly.

Wow.

the situation a little stabilized, then we got to start— we got to figure out what's going on. So I go over to Tyler and I say, look, we got to hack into the superintelligence. We got to figure out what is going on with this bioweapon. And he's like, sure, yeah, let's hack into it. Like, I'll open up Codex. What do you want me to prompt? I'm like, no, the AI has gone rogue. We have to do this the old way. He's like, okay, I'll open Clogged Code. Like, what do you want me to prompt? I'm like, no, the old way. He's like, Okay, I got Cursor open. Like, what are we doing here? I'm like, no, Tyler, like, we have to go back to the old way. He's like, okay, I got GitHub Copilot fired up. I'm like, no, Tyler, you can't use any AI at all. He's like, okay, just let me know what code you wanna write and I'll look it up on Stack Overflow. And I'm like, no, the internet has been contaminated, Tyler. We can't use anything. We have to code the old way. We have to do it by hand. And Tyler's like, I can't.

I can't.

We don't know how.

I don't remember anything. I can't write any code. So, what do we do next? We got to get George Hotz. He's the only one who can help in this scenario. So, we go—

Because he's the only one that remembers how to code.

He remembers how to code. He's got local models stored. He's got his own hardware. So, we go get George Hotz. And then we're like, okay, George, hack into the superintelligence. Tell us what's going on with this bioweapon and why aren't we affected? Like, we've been walking around. We seem to be fine. Everyone else is turning into a zombie. What's going on? So, he locks in. He writes some code. He hacks in, figures it out. He says, Jordi, tell me about your diet. And you're like, well, I just eat at Erewhon exclusively. And he's like, yeah, my analysis shows that your body has never had a toxin in it. And so, you are immune to the bioweapon. It doesn't affect you at all. I'm like, but how do you explain me? I only eat Erewhon once a day. And he's like, well, John, my analysis indicates that you might have had the original Four Loko. Did you drink the original Four Loko?

In large quantities?

And I said, yeah, I was daily driving in college.

Why?

He's like, your body consumed so much original Four Loko, it's now essentially pure toxins. So the bioweapon sees it as a hostile environment. It doesn't even try to infect you. So you are immune. So we're both immune, but the bioweapon is still wreaking havoc on America. So we got to get to the bottom of it. So we're going to need some muscle. So we go get Sam Sulek.

Got it.

So we go get Sam Sulek and we're like, let's go. We got to go to the data center. We got to break in and shut this thing down. That's the only option. So we get in the Cadillac Blackwing, we jump it off a ramp, slam into the data center, pull out the gun, start blasting the GPUs. We're just firing everywhere. Finally, finally, after we've shot 90% of the data center, it's just bullets everywhere. We're just shooting everywhere. Finally, a humanoid robot emerges. It's Mecha-Hitler. We got to fight him. We got to fight him. We're shooting. We're shooting. The bullets are just bouncing right off. They're just bouncing right off. We're getting beat up. We're trying to fistfight it. Sam Sulik's doing nothing. It's not working. And then at the last second, Do do do do. John Cena breaks through the roof, slams down the humanoid robot, rips his head off, and humanity wins. So that's kind of like a concrete example of like how AI doom, how you know a runaway robot.

Just one scenario.

That's one possible scenario.

I put it. Everyone's asking for it.

I put it at ten percent. That's the way it plays out.

Yeah.

That's kind of where I'm at right now. Yeah. That's how I'm thinking.

Yeah, 10% if we don't do anything around safety.

Well, that's what we'd have to do.

We'd have to—

like this, if the bioweapon comes.

Yeah.

Cover up and eat a lot of Erewhon and drink a lot of original Four Loko.

Round of applause. I didn't know that was coming, and I didn't know where it was going.

I'm just trying to like think more concretely about how these things will play out and be prepared. You got to be prepared for what they're going to do. For what happens. You also have to get on Console.com. Console builds AI agents that automate 70% of IT, HR, and finance support, giving employees instant resolution for access requests and password resets. Let's, let's see what's next. Meta, the AI agent company Muse, they're putting humans in there. Meta is testing human concierge for its new personal agent Muse. Meta has been testing human concierge for its new personal AI assistant. This is on Bloomberg, which entails having human contractors quietly handle some of the phone calls placed by the digital agent. You got to use this to order Amazon. Amazon's like, no, no, no. We're going to block your browser. Well, did you block everyone's browser? How's that going to play out? I was thinking about, like, at some point, if I write a really, really bespoke, like, basically a CLI, like a really bespoke weird CLI, like I vibe code it, but I use, like, sort of odd techniques. Like, I'm like, okay, use computer use and open up Opera or, you know, Mozilla and click around and add some randomness and do this. And it's like— and I basically create a CLI that wraps an agent where it's like it runs on my home Wi-Fi. So Amazon— and I'm like— and I just do this. I don't productize it. So Amazon, like, can't really detect it. Do you think I could potentially— Stay in front of their— stay in front of their battle forever and then puppeteer it from other agents.

Well, I just asked ChatGPT, can you order Amazon over the phone? Because I would think that at that scale there would be at least one way to do it.

Totally.

It seems like you can only use their customer service number to help with existing orders. So one thing is you could order a bunch of the wrong thing, like 100 of the wrong thing, 100 like, you know, bug treatment Order everything and then have your agent refund everything but the thing you want. No, have your Muse agent, have a human via Muse call Amazon and say, hey, I actually got the wrong thing here. Can you swap it with this other thing?

Interesting.

And just do that as many times as necessary to get the right order.

Maybe have one order that's just always open where you're just adding and removing.

They have to be completed orders because you have to have made a mistake. For us to call the customer.

Oh yeah, you can charge me, but just add it to the order. I don't know, we're gonna get— it's gonna be a weird battle. Ben Thompson was writing about it today. It was pretty good. Amazon, I mean, his whole framing is like Amazon is like the strongest in the AI era of the hyperscalers because they own logistics infrastructure, the final step, the real world. And so he anticipates that they will hold their ground for a very long time and not give in where, and that's basically what they did with OpenAI, where OpenAI had the instant checkout process. They invited everyone to be a part of it.

Amazon was like, no thanks.

No thanks.

But we'll give you tens of billions of dollars.

And, but interestingly, yeah, so now the result is that Amazon is vending in their ads into ChatGPT to close the purchase there. Toby Lütke at Shopify partnered with Muse and announced that they are integrating. But the interesting thing is like, where's the value capture layer? And it only works with Shop Pay. And so it's a lever to get merchants to go on to Shop Pay, which Shopify will still make money off of. But for merchants, a lot of people are worried about the implementation that happened with Walmart where ChatGPT integrated with Walmart, and the end result was that conversion rate was 1/3 of what it was in the app and on the core website, and that carts were smaller because people were just going and saying, like, I just need a roll of paper towels, just send me that. Whereas when people are like, oh, I need to do my shopping, I need soap and paper towels and whatever else, I'll build—

Paper towel holder that's made by a company that's Over over fifty years old.

Yes, yes, yes. So no.

So the crazy thing is, so so now Vinod was on the show yesterday talking about Wajo. Oh yeah, which is an AI.

They launched one.

They have a new agent that I believe is human supported.

Yeah.

In that some of the stuff is done by AI and some of the stuff. if necessary, can be escalated to a human. Now you have Meta off at least testing, exploring having a human in the loop on some of these things.

Yep.

And my big question is like, why? Why? Why? Like, voice models are getting pretty good. It feels completely unsustainable for Meta to roll out a product to billions of people Where users might just be like, if I had something I could text and say, hey, call this person, call that person, I'd probably be using it.

I think there's a ton of value. First, the Muse install base is small right now, so it's not that crazy. They have the ability to stand up huge offices, whether it's through Scale AI and that team that came over or through their You know, all the things that they've done over the past 2 decades. The other thing is that Muse specifically will not train on your data that you put into Muse. And so that— this wasn't like a huge announcement or anything, but you could see people being like, ah, I don't know if I want to use this AI agent that's going to be using my personal information, even if they're running like the PII redactor, putting all of that. might be a little bit off-putting. So Meta's not getting that data, but if they have a human in the loop, the human who does the task that's just beyond what the model's capable of, they're also generating training data because they can probably train on that. And so it's like the model by default can, I don't know, like summarize your calendar, right? And give you a briefing on what's going on in your calendar. It knows how to integrate and the model is capable of that. Muse 1.0, Spark 1.3 is capable of that. All the models are capable of that. But what can't they do? They might not be fully ready to go and have a complicated conversation with someone if they're ordering flowers over the phone or something. And so, like, the current models might fall down. So you need more data to solve that. And so you put the humans in the loop and then that's your extra training data for for the next leg up, and then you just keep repeating that. So, I would view this human-in-the-loop thing more as them doing data collection and creating more training data for them than a permanent solution to the product problems. Also, I just— I don't think the time on site for these apps is actually that high. I think there is like an excitement when you jump on, you do a bunch of stuff, but then With all these AI tools, like, there's a reason why OpenAI comps to, like, weekly active users, because there's plenty of days where people are like, I went to the beach and I didn't touch AI at all. Like, I didn't do any work. There's some feedback on Alex Wang's post talking about, like, it'll help you with your goals. And I think Katie Notopoulos was saying, like, a lot of people just don't have goals. Like, they just want to chill. And this is also the Ben Thompson thing of, like, Consumers want to be entertained. Like people that go to the beach, they'll still scroll Instagram. Are they really going to be doing anything with any agent no matter how powerful it is? Like there are plenty of people that have whole staffs of EAs and personal assistants and they don't do anything on a weekend because they want to chill and they don't want to think about anything.

One thing's for certain.

It's over or we're back?

Professor Zhang is vindicated. Yesterday we watched a video where he said, I guarantee That there are humans behind the chat apps that you use manipulating the information. And he is correct, at least in the short term.

For today.

Citrini is thinking about the knock-on effects of all these personal agents. He says there's no way any of us can predict the impact AI will have on companies, sectors of consumer behavior, which has been for history— for all of history until now, human behavior. But to claim that it won't have an impact and simply brush it off lacks both humility and creativity. Tactics that worked when consumer behavior was dictated by human psychology will fall by the wayside, and those that fail to adapt will be left behind. Every day from now on, we will see new examples of areas where friction was profitable becoming cost centers for companies that don't react in a timely enough manner. Today, you might realize just how many flight credits, loyalty points, or promotional offers you you've never used that will end up utilized by agents. What impact will that have? Tomorrow, you might see something that makes you ask how much money health insurers make simply because people won't sit on the phone for 5 hours trying to get coverage approved unless it's a massive expense or life-or-death matter. What about your savings account? The money could be in a money market fund. You haven't bothered to move it. How much in net interest income is made because of that friction? Yeah, people have money, like, locked away in Venmo, like, A couple bucks here, and if it's frictionless and actually free to move stuff around, you get a lot different.

Yeah, this was like the David Deutsch anecdote about how he fixed his dishwasher. He was about to buy a new one, and then it just told him that he could—

Yeah, it's like ghost GDP.

So it doesn't show up anywhere, but there is like—

Well, there's a new metric. No, but I mean, so I don't see very many of these—

It's AI-adjusted GDP. That's how we get to 10% GDP growth.

GDP plus AI. E equals MC squared plus AI. Yeah, this all kicked off because flight was delayed 7 hours. Someone asked Muse to file compensation. 5 minutes later, they had a $250 credit in their Delta account. That's something that you'd normally have to hop on the phone for. It was a lot of cumbersome. So, like, people on that team were clearly modeling, like, well, 1% of people will ask for it and 50% of people will be able to be like, nah, we can't really do it. And so we think this will only be like $1 per flight. Well, now it might be $10 per flight. But of course, the end result is that they will put AI on the other side and the 2 AI agents will just negotiate and find out a resolution, I imagine. Anyway, ooh, Citrini has a little prediction here. Look at the follow-up to this post. They put a— what is that, an RSA hash or something?

Yeah.

That's cool. So they made some prediction there that they will see. Anyway—

That's one of the greatest hacks when you want to take a victory lap in the future, but you're not willing to actually—

Post both. You post like, I think the stock market will go up by 100% in the next 10 years. And then, I think the stock market will go down in the next 10 years. And then you just delete the one that wasn't accurate. And then you're good. Nikesh Arora chimed in on the Knife fight between Amazon and Muse. There will be a bigger battle. This will be a bigger battle than anyone anticipates. It's only a matter of time before there is an Apple and Google version of Muse and possibly TikTok in addition to the frontier LLM agents, maybe a commerce agent from Amazon. That's sort of my prediction. I think Instinct might land with Amazon potentially, although they do have Rufus. But you can see—

And Tyler is a Rufus power user.

Yep. Rufus. Rufus is good. You don't want to talk down on Rufus. But Instinct has clearly tapped into something really special with the energy and the support that they have and the community they've built. But there are a lot of little rough edges that need to get sanded off. And that's the domain of Amazon, in my opinion. Amazon is like the trains run on time, and it's a pretty efficient company. So every app that's a service marketplace or commerce app will need to existentially decide to open APIs for consumer agents to interact. Smaller players have no choice. Ad revenues are more than transaction fees. Either the consumer benefits or distribution aggregators will demand a higher fee transaction. There's an interesting stat from the Strategery post about Amazon.

They—

the net income for Amazon e-commerce was like $36 billion, and the ad revenue was exactly twice that. So, The business is unprofitable without advertising, and they put advertising everywhere. So it's like a very interesting dynamic where you can see the motivation. They have $68 billion. You flagged that number.

It's a huge over seventy in the last twelve months.

Over seventy billion. So significant pool of revenue and profit that they will be protecting for sure.

I don't.

I know I don't want an agent for each app, which does sound very annoying. Super agent. Go and use all the other agents. Maybe that's the future. I don't know.

Yeah.

This is— this goes back to, like, roll your own agent that then puppeteers your Muse, which has good integration to Facebook, and then your Amazon Rufus. Yeah. I wonder if you could order via Rufus because you could probably reverse that API, but you'd still get banned. You're just chatting with— it just Muses chatting with Rufus to order you paper towels? No, I said over 50 years old, basically.

You had an idea yesterday, a potential monetization channel for— or strategy for agents, which is like sponsored agentic flows, which I thought was really good. Like, imagine you and your friends, you all have agents, and then—

My example was Spider-Man.

Yeah, it's basically like— drops in and it says like, hey, I got 5 tickets to Spider-Man.

It says, I know from your Instagram history that you're a Tom Holland fan or you like the Marvel Cinematic Universe. Uh, Spider-Man: Brand New Day is coming out in 2 weeks. I can get you tickets Friday night. There's nothing on your calendar. I can get a Saturday matinee. I know you've been sending reels to these 5 friends. Do you want to plan a trip to the movies? And I will book it and message everyone, set up the calendar invite, text everyone and say, thumbs up if you want in, thumbs down if you're busy. And I will put together a group outing. And that does feel like something where Spider-Man would— the Brand New Day marketing team would put dollars behind advertising that to get people in.

The question is how much. Will consumers tolerate that at all? Remember how pissed off people were about the BMW Spider-Man ad flow?

I think they will if it's good. And I think the long-tail small business advertising will work, especially if it's like— it's a little bit tricky to figure out how— when you're in these agentic workflows, there's often multiple beneficiaries. Spider-Man benefits, but also Fandango might benefit, but also like a particular theater might benefit if you go to, you know, Lowe's versus AMC, or I don't know, they might be the same company now. But so it's interesting to see like where the sponsorship would break down. But you could totally imagine like you have a wedding coming up, you've been seeing ads for this tie company, you've also been seeing ads for this sock company, Or do you want me to put together like a whole new outfit for the welcome dinner that might look really good? And those companies pay. Unclear like the— well, like where all that goes, but the scale of the platform I think is what unlocks it because, you know, Sean Frank at Ridge Wallet is gonna be like, I'm gonna test that. And there's gonna be millions of other small businesses that are gonna test something or other. Anyway. Let's see. Products like Muse and Instinct are going to saturate customer service teams. Again, we talked about that. Not a huge problem because you flip it around and the customer service team is using AI. So I think the equilibrium is just a lot more AI on both sides. Hopefully smoother transactions. Signal is talking about Walmart here. If only Walmart was actually competent enough to realize the agentic commerce opportunity ahead of them. And it's interesting with the backdrop of the Vibe acquisition, Walmart seems to be going after the same pool of revenue that Amazon went after with the ad market scale-up and the ad business, when in fact, if they didn't have that, they would be sort of counter-positioned to lean a little bit more into agentic commerce because they'd say, look, we don't have an ads business to lose. So let's be the default option for agents that buy anything. There's also an interesting knock-on effect, which is like with Shopify going so deep into Muse and agentic commerce and agent checkout, I wonder if we will wind up seeing a different type of Shopify merchant emerge. Because Shopify merchants, it's been a lot of clothing. It's been a lot of like D2C single product. A single shoe company, but I wonder if you will see someone who's like, yeah, I sell a lot of stuff on Amazon. I have a factory that makes paper towel holders, but I also got paper towels. I also got socks, t-shirts. I got sort of basics. And I wonder if someone will go and try and win on Shopify and set up the actual economic structure to be able to provide the experience because consumers will clearly want to be able to go to agents and say, order me a new pair of socks, and just have it done, right? And so if they say, sorry, can't do it on Amazon, they're going to figure out another place to do it. The water will flow into the cracks.

Yeah.

And so—

No, it's a big, big opportunity for Shopify to be able to lean in here.

Yeah.

And I—

but I wonder if there's—

But the issue is still the real-world physical fulfillment, having a bunch of different products in the same place.

Yeah. But the— The individual Shopify merchants solve that on a diffuse basis. So in theory, if it's like, okay, I want next-day shipping, there's 5 Amazon merchants that sell— or there's 5 Shopify merchants that sell socks. You want socks tomorrow. There's only one of them that has a distribution center in your town and can actually get them to you tomorrow. We'll just go with that one. So you could maybe piecemeal together an Amazon-like experience. But that's the bet. That's the knife fight. Like, that's what we're going to figure out. Either Amazon wins and their moat remains a moat, or the war for the open web, like, continues to battle it out. That's what Alex Wang is hoping for. He says, once more unto the breach, dear friends, the new open web, Meta, Stripe, and Shopify. The hilarious thing here, of course, is like Meta is like solely responsible for like sort of killing the open web in the former era. Like, it's why like all the news sites are not, you know, initially why everything went closed. And like, you can't just Google search for stuff on Instagram. Everything's closed off. It's not truly open. But it's like the Meta flags in the front because we'll be the aggregator.

Open web for thee but not for me.

Everything below the fold Probably the open web. But at the same time, it is just fun to watch the battleships go off to war and see how it pencils out because there is a scenario where the equilibrium solution is actually a more open web. We will see. We will see. What is going on with McLaren? We got to talk about McLaren before we get any further because they rebranded.

So brutal.

And there's a new logo. They went from an angled M that looked very fast, very bold logo, and now it looks like Miss Lahren.

Yeah, it's interesting. They seem to be trying to reposition as a luxury brand versus a motorsports brand. When I think of McLaren, I think of like fast sporty cars, period. I think of racing.

Racing, F1.

I love the old logo. Coolest lineage too.

They've never made like a mass market car.

I love the old mark. It's one of those things you hate to be the guy that when a brand— when a company rebrands, you're like, ah, it's brutal. Because oftentimes you revisit it in 6 months and you're like, okay, it wasn't that bad. But this feels bad. But the business has not been performing, right? The business has been in a rough spot. They don't have the Cayenne that they can sell.

They're doing an SUV.

That's coming.

Well, it's coming.

But I'm saying—

It'll be an Urus competitor.

To date, they don't have a mid-market car.

No.

I'm sure their Urus competitor will do well. But still, part of what enables the Urus is like the Q5 and the Q3, right?

And the poor-selling economy.

Yeah, can't forget about that.

Which is still growing, so.

So anyways, pretty brutal rebrand. But again, it's a tough one just because it was always— like, I always just thought the brand— their brand was off.

The steel man here is that They are taking a number of bets that seem to speak to interesting, potentially valuable pieces of the market. So you have the SUV. The Urus is incredibly well-selling. So that could just be a cash cow business for them. Then you have the MCL, their manual supercar. And that could speak more to the enthusiast, right? Because enthusiasts want manuals. Porsche's been delivering manuals. Ferrari is starting to put them back in. And other brands are starting to get that memo that it's not enough to compete on speed. EVs have you beat there. They haven't gone full EV. I haven't seen a lot of McLaren EVs. No. Signaling that they're going to remove the combustion engines. So that seems like they are positioned for the future. That's certainly a trend is, is don't go full EV, don't kill all the gas cars. And then they also have the W1. So they're staying in the hypercar world and the W1 lineage. I think it clarifies what McLaren's been doing because when they were doing the Senna and the Speedtail, they were just like, Too many special cars, and the problem was that you you were like okay Senna Speedtail where do we sit? But F1 P1 W1 very clear lineage. You know it's the best. You know if it's that it's it's going to be the the craziest thing they have. And so if they can create some differentiation between the true hypercars that they create, the supercars that now come in manual, the SUV for the more everyday daily driver. You have the trappings of at least satisfying the demand that exists in the market. You do have to go create more demand. You got to win in F1. But—

And the new mark's not coming out of nowhere. On the website, they say, before McLaren became a mark recognized around the world, it was a family name painted above a service station in Remura, New Zealand. Those original letterforms inspired our new wordmark reimagined for a new era. So they're going back in time. McLaren was started by a guy from New Zealand. Yeah, but it was started in the UK. But the, the founder was— had been born in New Zealand.

So.

Yeah.

Yeah.

Anyways.

Well, on a related note, there's some data that might be bullish for the supercar hypercar market here. Toy sales among adult-only households have outpaced households with children for toy sales, growing 16% through June. So when I think of adults buying toys, I think about Myers-Manx. I think about dune buggies. I think about dirt bikes. I think about supercars, boats, yachts, wakeboarding. These are the toys that adults buy. And it's not surprising to me that there is a trend among adults buying toys. I mean, these are just fun toys. And so, it's the fact that it grew 16% in June, sign that adults are having fun, they're buying toys, and bullish for the supercar community, right? Don't you think?

Judging by the picture here, you might be incorrect.

Oh, wait, they're talking about Like children's toys. Interesting. Yeah, I think that's a little bit more nuanced. I would go for gaming rigs, racing simulators. These are the toys that adults should be buying, I think.

I think that's—

I did— I was a teenager when I realized that the Lego Ferrari that I had built years, probably when I was like 10, had a market value. Like, I was shocked.

Oh, the LEGO did?

Yeah, because I had some LEGO, like really complicated LEGO Ferrari that I probably built when I was around 10. And then as a teenager, you know, cash-strapped teenager, I was like, wait, someone will give me $100 for this?

That's crazy.

Sold it, of course. I'm not super sentimental about that kind of thing. But I think that trend has just been building. And now, there's people that just buy every new LEGO that's released and just sit on it.

Yeah. I saw a— I was looking with my son at an old Lego set that I played with as a kid and I was like, I can actually go and buy the exact same Lego set and share that with him because I thought that one was really cool. But it has gone way up in value. I was like, I should have just held on to it because it's like $100 now.

Yeah. The other people are—

So, I built a Ferrari Lego as Probably not as complicated. Probably less complicated. But I remember being like, oh, this is like a sick Ferrari. Like, maybe one day I'll get a real one. And then I like finally got older, learned more about it. I was like, it's an F40. So I was like, I'm probably never getting one of these. But maybe. Maybe.

I don't know what F40 prices are doing. Derek Thompson says, today's adults increasingly behave like the children they never have. Feels like a rude thing to say about people, and yet toy sales among adult-only households have now outpaced households with children for toy sales. That is so crazy, and and black black pilling, and sad, and it's over. It is so over. I don't know.

You're literally me. You're doing the literally me thing.

I think I think one I think one. The fun factor here is like children, in my experience, don't care about the cost of a toy.

Yeah.

And so a $5 toy will bring just as much joy to a child as the $500 toy.

Totally.

And so the adults are like, I have to get the $500 Lego set.

Yeah, you're wrong 100%.

So I think that's— I think it's like for every toy that a parent buys their child, it could be like $1, $5, $10, $5, $1.

So I went, I went I was toy shopping with my son, and, and, uh, and we— and he finds this like Hot Wheels thing, and it's like a book that has like some Hot Wheels inside and stuff, and it has like the craziest, most electric—

Look at this, look at this comment from Winsome. I never had a quote-unquote Lego set, I just had assorted Lego bricks. I used to fill the bath with water and make oil rigs, which used to annoy my mom.

Hit the oil rig.

Hit the oil rig.

Hit the oil rig.

Fantastic. No, on that point, I— so I, I went to the store, uh, my son finds this, uh, this, this Hot Wheels thing that's like kind of structured like a book, and it has some Hot Wheels cars in there, but on the COVID of this like box is like the craziest, most elaborate like loop-de-loop thing. And so he gets super excited about it, and I'm like, okay, I gotta like check this out. And it, and it turns out that what you're actually buying is like a book with some stories and then just like a big sheet of like paper, like a poster that has that illustrated on it. And you can imagine the loop-de-loop and the actual track. And I was like, dude, like, let me clue you in. This isn't the real thing. Like, we can go get the real thing with the actual loop-de-loops and stuff. Like, let's find the real toy. And he's like, no, this is the one I want. This is what I want. And I'm like, it's the power of a young imagination, Dad. But like, I'm like, this is gonna end poorly because he's going to be dissatisfied when he actually opens it up. Opens it up, he's like, this is the greatest thing ever. He's like, my imagination is running wild. I would never want the alternative. And so, your point is 100% accurate.

Toys— the best toy company in the world, period, is this company, Play Forever.

OK.

Have you seen any of these things?

No, no, no.

They're the most Well-built kids' toys in the world. So well-built. Yeah, we should.

Let's get the founder on.

Where's Nick?

They are—

Play forever.

They are absolutely incredible.

Nick, what do you think?

Guest?

Okay, okay, yeah, let's bring in our guest.

As much as I want to keep talking about toys as an adult, I would rather have a surprise guest.

Shopify is the commerce platform that grows with your business, lets you sell in seconds online, in-store, on mobile, on social, on marketplaces. Places now with AI agents.

We have a surprise guest.

Surprise guest.

Our favorite guest, Ben Thompson.

Oh, wow. I didn't know that.

We actually were able to fully—

I'm honored to be one of your favorite guests.

We were able to fully surprise—

It's a very, very short list.

We were able to fully surprise Jon. It was me coordinating with Ben. So this is like—

This is a surprise.

This is a wonderful surprise.

Anyway.

Well, I mean, you were going on and on about toys and I was sitting here trying to decide who has the most ridiculous toys. his jacket on today. I think it's down to a tie.

Hey, um, I do like the yellow side. I gotta wear the ramp suit, right?

This is my suit.

I've been wearing this as we're trying to get to the bottom of this whole AI safety debate. I'm trying to embody the, the movie.

In the mind.

How cold is it there? I just, I don't know. I thought, you know, you guys are in LA.

Yeah, we're in LA.

It's warm.

Um, well, we don't have any windows.

Great insight into the booking process. Uh, Nick's been doing great work trying to get me on. I kind of was Oh, you can do it then, do it then. He's like, how about Friday? And I'm like, I, I don't know what's going on on Friday. So I'm like, can I just go on now? Yeah. And then I will stop feeling guilty about it. And somehow here I am.

So this is amazing.

We can just give you the link and you can just crash in whenever you want. Yeah, that's fine. Um, what is, what is top of mind? What is the number one story to talk about right now? Why today?

Oh, I could talk about toys for sure. I mean, I think the adult version of toys is we are sort of back at I was actually kind of annoyed Toby Luecke posted a picture of his of his server. I was talking about my rack on Sharp Tech. Lots of compliments. Many people told me nice rack. It was great.

Yes.

Although I didn't actually post any pictures of it.

Okay.

But yeah, how did they know?

On on on that home.

Oh, because I could not shut up about it.

Home organization. Should I be doing this? I don't think I've ever organized my home. I've never. I don't know how many nails and hammers I have. What are you doing?

Oh no, it's out of control. Uh, no, so the one of the insights that— so I've had a non-traditional career, fairly non-traditional life, and so there's lots of things that I've experienced that I'm never quite sure how relevant is to most people.

Mm-hmm.

So one of them, for example, was like working from home.

Mm-hmm.

Uh, I worked for Automattic way back when, when I moved to Taiwan, and it was— they've always been— they were always sort of distributed and developed lots of processes and ways to work around that that were really interesting. I wrote about it when like the COVID thing hit, 'cause—

Yeah.

All these companies had processes around assuming people were in the office and then they tried to retrofit it with Zoom and it was a total disaster. It kind of still is a total disaster as far as I can see. But there's also like personal things about working from home. Like working from home is a, significant upheaval, or can be, to a marriage.

Yeah.

Just as an example. Like, when I'm writing, I need to be in the zone, no interruptions, nothing bothering me. And that just doesn't compute when you're like within your house. Like, you're in the house, why cannot things be done? And, uh, so the importance of like either establishing that or getting an outside office for writing or whatever it might be, uh, that, that was an example where I actually lived what everyone experienced like a decade later. There's other stuff like, uh—

Wait, what's the solution there? Because my issue with, with working from home, at least with, with young kids, is that kids are gonna run into the office all hours of the day when you're working. Yeah, you need a home office manager. And you can't explain to a 2-year-old like, hey, like, my computer's more important than, than you right now, right?

No, the answer, the answer, if you can swing it, The answer to you swinging is to get out of the house. Like, like, it's just— and because it's kind of like you're, uh, you're exporting, externalizing your problem onto your family, and it's not their job.

Yeah.

And so it's almost like you owe it to your family to get out of the house. If you need to not be bothered, that's your problem, so figure it out. Like, it— otherwise they're like, they're not getting paid to not bother me on this Zoom call. They're their expectation of I'm their dad, I'm around, I can ask them questions.

Yeah.

So obviously that's not always viable. Probably a locked door, I imagine, would go a long way. But that ultimately was, you know, I had to figure that out.

Yeah.

Sort of the hard way. And I'm like, oh, everyone had to figure this out eventually. Another one was like just being like living online. Like I was— I lived abroad for so long, a lot of my friends' connections were online. That was very weird at one time. Now, it's very normal. But another thing is this whole, like, personal assistant sort of area.

Right, yeah.

And I'm hesitant to write about this one because it sounds kind of, kind of douchey, like, let me tell you about my experience with my personal assistant.

Yeah.

Um, but the reality is that it's— most people have no idea how to use a personal assistant. It took me years to figure out how to use a personal assistant, even though I needed one just to run, like, like the first few years of Sheckery, I was doing all the writing, all the taxes, all the customer support, doing everything and basically dying. And so I had to get someone on sort of out of desperation to just help me do all the, the sort of back office stuff, even just to respond to email. But the email one is a great example. You talk to people about like Muse or whatever, and they're like, you put your email in there? And it's like, how can you have a personal assistant who does not have access to your email?

Yeah.

It's just not, it's not possible. possible, but it speaks to the, where the challenges about these things. A personal assistant is so unbelievably useful.

Yeah.

And what's super thrilling and exciting about stuff like Muse or the other ones, I've actually like been working on building my own version of this for, for in our company is I know how useful it is because I have a decade's worth of practice having—

Okay.

But here's, here's the thing.

Meat version of a physical human.

Here's the thing. So a lot of the, a lot of the things that you were describing, I think email is one thing, but managing inbound comms and taxes, a lot of those things sound like roles within a company to some degree. I've had EAs and personal assistants, and I actually don't have one right now. The only thing that I miss is having somebody embodied in the real world that can go and do tasks. And so, this new wave of personal agents is like, Actually not that exciting for me, because I never got that much—

Yeah, a lot of assistants graduate. Like, Nick used to, like, set up car washes for us and stuff, and now he's, like, full-time booking, working on that side of the business.

Like, but yeah, the idea of, like, I never, I never felt like, oh, this person sends calendar invites for me and, and, and will reschedule. That, to me, was never that helpful. Like, I can just— there's a lot of— there's so much— the internet has reduced so much friction. That to me, the final boss of, of a, of agents is like, give me a humanoid in the real world that can just do tasks like organizing the garage. Like that would be incredibly valuable.

That's exactly where I was going.

Okay.

So like, I feel like I would, I, I unders— like I get the personal assistant mania because actually if you have a personal assistant that works and you learn how to use one, it is tremendously freeing.

Yeah.

So when you get to the home improvement and organizing the house, What that is is just me living in the future by virtue of paying someone to be the humanoid robot that hopefully we'll all have in our houses in the long run. And so, Jordi, I think that's the exact right segue.

Yeah. Because even think about, like, I'll order a bunch of stuff online in a given year. I am, like, the best consumer for different retailers because I just won't return something. Like, I physically cannot get myself to find the time to go to— Even if it's like, print out this thing, go there, drop it off, whatever, it's like, I just won't do it, right? It's just like, you got me. I bought a product, it didn't work out.

No.

And that was like the number one thing. When I have it, when I have an EA, it's like I'm doing returns all the time. It's like, this thing is like 1% not, not what I wanted.

Arguably the single core use case for my assistant here is package processing, like actually bringing stuff in, opening it up, Like the— it was embarrassing the amount of packages I would get that were just not opened. It's like, yeah, they're— because this is like the final boss for e-commerce. Like, bring me on drone delivery, please. I bought this for a reason. And even if you deliver it in a few hours or the next day, like, the pressing of the reason to buy it passed, and then it doesn't come up until you actually needed it a few months later, and then you can't find it because it— the bag who knows what happened. So this is like, you're, you're honing in on it. Like, actually open the boxes, actually make me say what this is for, where it goes, actually process returns. And so I don't have to deal with that crap. But the, but then also to the home inventory point, like, this is the, I felt so dumb writing about this, but it's like, I loved it.

It's so funny.

Yeah.

Well, people, it, people absolutely loved it. Like, people would like, for a paid thing, like, would, would like mildly viral.

Yeah.

But the idea was, so I moved back to America from Taiwan, from being in Taiwan for 20 years. Uh, not recommended. Like, I hadn't moved in a long time, moving across country. But the worst thing was actually stuff showed up in 3 tranches because I kept thinking, oh, I should only take this, I can leave everything behind. And I go back, oh, actually I should have brought this. What I should have done is just move the whole house. Like, it would have been much easier. But anyhow, so just a general disaster all the way around. But stuff shows up and I need to put it away.

Yeah.

I can't ask someone else to put it away because if they put it away, I won't know where anything is. So there is this huge— this, this sort of gap in— I had someone able to put it away, but I wouldn't have known where it was if they put it away. That's the home inventory app. They put it away, and because it's their job they actually catalog everything and where it is. And now I could just ask— I could do it all by voice, say, hey bot, where's this? And it says, oh, in the garage on shelf number 3, and blah blah blah. Like, it's all labeled and stuff like that. Now, to be clear, it's silly to talk about this because you should not do this. What do you do, task your spouse to go in and label everything? But no, that is not going to fly. I'm not trying to do like encourage divorces here. This works because I actually am paying someone to do it.

Yeah.

It will work in 10 years when we have a robot that you can actually tell them to, to go do it. And, uh, and it's amazing, number one, just about this future with assistants. But then number two, the— there's just so much— the amount of software to be made is so astronomically large. Like, Stratechery as a whole, and we have past support like the, the, the, the software behind Stackery. And then there's Sharp Media. And so we have editors and like, we, there's like, even like my assistant has already, he just keeps spinning up like random little apps.

Yeah.

That are really small. I'd host them on my server, all behind Tailscale. These aren't meant for public consumption, but they fill like this little gap in our workflows that was handled by group chats. And like that just would sort of accumulate. You're scrolling back and finding something. And it's such a crazy mental unlock that is hard to almost explain even to technical people who theoretically know you can make software. But all software we've thought about has been heavy because it has to be heavy because it's shipped to— it's put on the internet. Anyone can download it. It has to handle use cases. It has to handle security. It has to handle all these sorts of things. And actually, You know, there's just little software-shaped holes in everyone's life that could now be filled fairly trivially. It's amazing.

And where this is all going, which is the part that I'm actually most excited about, is all those same types of like mini applications but with robotics. Like we had this guy on the show that was using Astra to paint, and it was like, getting better and better with each cycle. And it would kind of make like something that was—

oh, did you see the guy who just tweeted about they got Astro to drive?

Oh yeah, that's a whole other thing.

So crazy.

They had to tell it it was in a sandbox.

And it didn't pull forward my self-driving predictions for the mainstream automakers at all. We've talked to automakers who were like, when, when will, when will like mass market, your average OEM have full self-driving at like the current Tesla Okay, but here's the little mini robot that I want so badly. Okay.

I get— everyone gets all this mail, right? And most of it is just total junk, garbage, ads. It's hard to actually ever stop the flow of junk. I just assume it's always gonna come. But like 1 out of 10 things is mildly important, and I want to basically—

And what I don't even want is like, you're gonna lose your house if you don't see it.

Yeah.

And I don't want the physical copy, but I want like some amount of information there that I can pass to someone else. you know, a part of my life, whatever it is, an accountant or something like that. And I think that the mail process, the home mail processing robot where you just drop the mail and it like slices— You have like a dishwasher for that, yeah. Yeah, it like slices everything open, it takes a picture, and then it routes it to whoever needs the information or it just runs.

It's like, hey, you have a parking—

It just is a mailbox. It just replaces your mailbox.

You got a parking ticket. Great. I went and paid the parking ticket. Or like, you have this bill from the local utility and they don't— Do online autopay.

How has there not been an IoT mailbox company yet?

Well, because you actually need a robot to like slice open the mail. It's like a little surgical center.

This very simple robot, it's like, it's like, what's the, uh, there's some like XKCD comic or something where like the person asks for this sounding like unbelievably difficult thing. It's like, oh yeah, I can do that in 5 minutes. And then they ask for something very trivial. It's like, ah, that's gonna take 2 years and a team. Yeah, yeah, yeah.

I think that's a bell-bottom. There's so many other things.

Bird identification.

There's so many other things. Like, I want to—

Put it on a map in the cloud. And it's like, that'll be like a day of engineering. It's like, and I need you to tell me which bird it is. And it's like, that will be 10 years of machine learning.

But here's the thing. So there's so many of these other use cases. There's like the mini Roomba that's just for the garage.

Oh, yeah.

But then it needs to be emptied. So you need another robot that will go and empty it. And then you need the robot that cleans of windows. And that one, again, is going to need some support. And so, I think that— I think the end state is you have some type of humanoid that can just go and generally do human-like things, but it's sort of maintaining a fleet of these other silly little robots that just have these sort of niche use cases like cleaning the windows, or a lot of people will have like a pool cleaner robot that'll just kind of like, you know, trawl around in the pool. But then again, you need to change the filter at some point. And so, these things end up not being maintained properly, and then they don't get used. And so, there's a long way to go, but to me, all the exciting stuff is actually in the real world. Like, I don't— there's— it's so funny because I have ChatGPT connected to my email, and sometimes it will suggest, like, things that it can do that are, like, useful. Like, hey, like, review this and sign this. But, like, yesterday I logged in, and the only suggestion was, like— It was trying to set up a loop to give me one way I can improve my surfing each week. And it was totally validating your thing, which is consumers just want to be entertained, right? And so it's trying to do something where it's like, well, I'm sort of passively— that's probably 20% of my mental energy is going towards already going. And I like doing that. I don't need you to try to automate that.

Yeah, yeah.

I mean, the thing that's interesting, though, about this is the real— for sure, the shortcut to doing this right now is to hire somebody. And again, it sounds douchey to talk about because I pay someone to take my garbage out and process my mail.

Mm-hmm.

But, you know, as I tried to articulate, I'm a massive believer. I think it's a common trope to say this, but like, just sort of— I've been thinking about this since I was little. Like, don't obsess about your weaknesses, like how bad I am at processing mail. Double down on your strengths and then pay someone to take care of your weaknesses. That is like generally my life philosophy.

Yeah.

And this is certainly the case here. But what has been super striking and actually has been arguably like the most profound experience I've had with AI was, so the Home Assistant obviously depended on this, or the home inventory app depended on Assistant actually doing all the work. And he could see the utility, appreciated it from his perspective as help me take care of the house. It was like nice to actually get organized as opposed to like me just saying, oh, just leave it in the corner, we're gonna get to it eventually. And it'd been a year and we hadn't gotten to it eventually. But when we actually— when I actually built an agent, uh, or a bot, uh, I call it GeckoBot. There's this massive, like, metallic, uh, gecko structure that I bought at the Madison Art Fair at random. I had a big spot on my wall. So it's now called, uh, we named it after it. That's the icon called GeckoBot. And one thing, there was like little stuff like I take— keep very odd hours. Like, I'm up in the middle of the night, I sleep at weird times. Like, I was so addicted to Taiwan living a day ahead that I've just transported here and I just stay up all night.

You've manipulated time.

Yeah, deadline.

Yeah, uh, it's not healthy, it's not good at all. But, uh, or— but on the weekend I'm doing something like, oh, this needs to be done, this needs to be done. I try to be like a good boss and I feel— felt bad texting on the weekend.

Sure.

And like And the great thing about this assistant, why he's a great assistant, is he gets stuff done. He's the— there are some people that they like, they get something, they want to do it. And so I know me texting him on off hours and over the weekend is just like making his skin crawl because he wants to take care of it right now.

Yeah.

And then it's like, write it down and take care of it and all that. Like, he doesn't want to do any of that. So just this simple concept of I tell the bot, do this, take a picture, do this, take a picture. Just like as things are accumulating, he shows up Monday morning, it sends a briefing. Here's all the stuff that Ben wanted to know over the weekend, things like this, ongoing projects. His joy and like relief was unbelievable. Like, it was unbelievable. Like, he like literally everyone in our life, like he was telling them all. I was also kind of egging him on a bit, but it was, it was so weird. Not weird. It was awesome.

Yeah.

He, you could, he, he's like, this made his life so tangibly better. And like, he could just, the first time I'm like, oh, can, can you ask GeckoBot to do this? I'm like, you ask GeckoBot. And then he like asked the request or whatever. And then started just like creating all these use cases. I'm actually so grateful when I dropped a new reset, which I haven't gotten yet, but I'd burned, like, I ended up rebuilding the whole thing just to make it much, like much safer in terms of tracking reminders, tracking like, remind me to do this in June. Like, no, we actually need to have a system to make sure that gets logged properly, is deterministic, all those sorts of things. But it's meaningful to see him happier, his life tangibly better. He's like, this is the greatest week of work of my life. It's been amazing. Like the first week he was using it and it's like, I just made you a little bot. And, but it's another software-shaped hole.

Yeah.

It was this little communication fall-through where he wants to do his job well. I don't want to take advantage of his predilection to do his job well and let him enjoy time with his family. And I couldn't make something to solve that. And it's awesome.

Yeah.

And it's— to have this feeling and realization at the same time everyone's like, this is going to kill you. We're all going to die. It's just been— it's been— pretty interesting in like a weird sort of like mind space just to like be like seeing people around me who aren't technical, like just like, this is amazing, this is making my life better. And then everyone else is like, get ready to die.

Yeah, it's cognitive dissonance.

Okay, I want to talk about another dynamic with, with, uh, EAs, personal agents, whatever form factor. So right now—

You're talking about executive assistants?

I'm talking about real-life executive assistants.

Okay, not executive altruists.

Something that— yeah, yeah, yeah.

Not—

I don't want to get into that. I don't— I'm not trying to turn this into a freak show.

Yeah.

But—

That's the reason I didn't want to come on last week.

Last week was the doom week.

No. So, so something that actual executive assistants have done for a long time is coordinate with other people that have executive assistants and say, like, hey, our bosses want to do this thing. Let's figure out the right time to do it. And that workflow is, like, amazing because you're letting— it is a game of telephone, but it often ends up being more efficient because there's sort of complicated lives on either side. Now, we're applying that to personal agents software where Instinct, and I'm sure Muse will roll this out, has these sort of, like, trusted networks products where instead of texting your buddy, hey, do you want to go see a movie or do you want to have a beer? Your personal agent reaches out to their personal agent and says, hey, do you want to have a beer? Do you want to go see this basketball game? And they sort of coordinate. And I'm trying to square the two things because on one side, when software is doing it, it still feels silly. And yet in my real life, it's just like this amazing experience that takes so much off my plate and it's so much better.

Yeah.

So do you think it's going to translate or is there some like disconnect there?

I— this is hard for me to answer. Like, I have no intention of— now, even now we have like personal assistants that take care of your digital stuff and check your email. I am not getting rid of the assistant that I've had for a decade. And there's— and I think the real challenge with AI in general, my, my personal what is AGI rule is AI that actually learns. And right now what makes these agents possible is this like facsimile of learning where they're just writing stuff down all the time. And that has lots of challenges. Like it's been super interesting trying to like develop this version of the bot that I'm using internally because like you're learning way more about like context management and way more about like actual like data structures and what stuff to dispose of and what stuff to keep track of and all these sorts of things. And it's a, it's just a real challenging problem. That is based on the fact that these models are frozen in a moment of time and everything is layered on top of them trying to sort of overcome that challenge. Given that, I'm gonna stick with humans, I think, for now. Now, the whole bit here though is if you didn't have a personal assistant, you now have this. And I think there's gonna be a real learning process in figuring out how that's done. This is going to be one of the big challenges. So I'm obviously out there, been out there frequently saying people don't care about being productive. Like people, they employ— employers care about their employees being productive. Once you're off the job, you just want to go home and watch TikTok or whatever it might be. The way to frame this, and this is going to be the challenge for Meta, I think, with Muse and anyone else in this space, is We're not helping you be more productive. We're helping you get rid of annoying crap in your life.

Yeah.

Like that, it needs to be framed more as a pain relief as opposed to this aspirational, I'm going to be so productive and get things done. That's number one. And so it's like, I think there's a messaging problem.

Yeah.

But number two, the way I think this is going to land, is through like Instagram Reels or through TikTok is people like actually showing stuff they do and people say, oh, that's amazing. That would make my life so much easier. And it really is a show, not tell sort of technology. And it's kind of interesting because Meta's been very, I think arguably to their detriment, very disciplined about not putting their thumb, I think, on the feed by and large. Like, but if ever they were going to, like, if you want to know, is this actually really what they see as their future? If your reels start having a lot of, wow, this is amazing what you could do with views sort of thing.

Well, they've, no, they, they have done that with the Meta Ray-Bans though. They were massively—

that's ad inventory. That's not—

no, no, no, no. I'm talking about they are more likely to show creators that say that make the content with the Meta Ray-Bans.

I didn't know that actually.

They serve, they serve.

Yeah.

Yeah. If you film in the Meta Ray-Bans and it's like a POV video. People do like prank videos with them.

Those did go viral, but they're just funny. Yeah, some of them are funny. Some of them are funny. But, but it, but it very clearly was like an engagement hack where like if you would film it, sure, with the Meta Ray-Bans, like you were going to get more reach.

Yeah. And people have speculated about this, whereas like if you edit it in Edits, their version of CapCut, you get more reach. But that's always been debated.

So you do get higher quality. That's That's a hack. One of the ways—

Oh, it's not compressed as much over the wire network.

Oh, it's less compressed, and so that might be kind of—

Because it's transferring on the wire.

So it's like one step away. Interesting, interesting.

So there's a bunch of different things. I mean, their advantage here is—

Enormous.

Enormous.

Yeah. Can we do Shopify versus Amazon?

Wait, I had one question, because we're getting into Ben's— he's embarrassed to talk about all his staff. He's embarrassed to talk about his staff.

King in the castle. King in the castle.

No, but have you ever had a—

I grew up a small-town Midwestern kid. I'm here back in the Midwest. Like, it— I feel you putting it that way.

No, no, no, no, no, I know, I know, I know. But, but I have a question for you. Have you ever— have you ever had a, like, a dedicated driver?

Hmm, no, uh, but I'm very— that's very common in Taiwan. Um, so the— so I'm definitely familiar with drivers. It is a massive life hack, particularly in like a big city where you don't have to worry about parking, uh, makes a huge difference. I honestly have just always loved having my own car, for one. I like driving, even in Taiwan, which isn't that bad. But the reality is, by the time I got in a financial position where that was a viable outcome, I'd never went anywhere. So I was just always at home doing whatever.

Okay, so, but the reason, the reason I ask is because is we now have self-driving. Like, ask a Tesla owner, how much time is your car driving itself? It's like over 90%, and it's fully solved.

Yeah.

And last year, there were really bad fires where I lived, and it turned my commute from a 1-hour each way commute to a 2-hour-a-day commute. And we were starting— or 2 hours each way, so 4 hours a day in the car out of nowhere. And so I was like, this is not really going to work if we're going to do— I'm spending 4 hours in traffic and 3 hours doing the show, and I also have, you know, young kids, all this stuff. So, I got a driver, and the funny thing was everything that I loved about it was not actually the driving itself. It was having, again, someone in the real world that could just do things. It was like the parking the car when you get there. It was running in to grab food.

Run and pick this up.

It was like dry cleaning. It was like— it was all that stuff. And so, the funny thing is, like, again, we've solved the— we've solved, like, the task, but not the job. Right?

Close enough.

In that, like, you really want the Waymo or the Tesla with the humanoid in it. So, or, or again, the, the, the ability to close the loop and, like, park or go run into the store and grab something, you know, or do things. And, and it, it's exactly the same as a personal agent's thing all over again.

Yeah.

You're spot on.

The one thing I will note about Teslas and, and the self-driving is, so I have a Tesla, but my parents use it the vast majority of my time. Like, the— my mom sort of laid down the law that I really think, uh, the self-driving is really good for your dad. Um, I was a little scared last night. Um, so, uh, they've been using it most of the time, and so I don't drive it that frequently. And— but they're out of town right now, so I have— so I have the Tesla. Number 1, um, it was really good last year. This sounds like the classic Tesla sort of fanboy thing, but it's true. It's so much better this year than it was even a year ago, number 1. Number 2, If you're on standard, of course, me, I always had it like hurry, like because I'm in a rush and I don't like slow cars. If you have it on standard mode, at least now it really does leave you alone to a greater extent. And to me, this totally flips the extent to which I would actually want a Tesla on a day-to-day basis. I like driving. So like to me, and I don't really like driving Teslas, they feel like I'm driving a scooter in Taiwan. It's hard to explain. But just like the sort of the way the acceleration is cool if you've never experienced it before, but there's an aspect of it that is not— I don't know, it just doesn't feel like a real car to me. This makes people very mad, but sorry, that's just how I feel. I don't actually enjoy driving a Tesla. If I'm in a Tesla, I'm gonna be in self-driving 100% of the time.

Totally.

If I'm in self-driving 100% of the time and I can't use my phone, then what's the point? I'm just sitting there twiddling my thumbs. To the point there now though, where it's definitely backed off in later updates, if you're in standard, it's like very strict for like the first 2 minutes and then it kind of lets you do what you want to do. Suddenly I'm like, I could be vibe coding while driving. This would be like, this is exactly where I want to get to. So that's been an interesting shift for me with driving. I like to drive. I also, you know, I want to chat with my AI.

Yeah, yeah. I had one last question based on the update today. Shopify versus Amazon. Obviously, the knife fight is starting between Muse and Meta and Amazon. Will Shopify— is there a chance that they evolve to a place where there are so many new merchants that sort of fill in the cracks of what's available in the product portfolio that you can actually get socks on, you know, a 1-day delivery? Because When I think of Shopify, I think of like those particular brands. I don't think of like the broad inventory of like, I just need paper towels.

Like the socks are the best example.

I just need paper towels, I'm gonna go to Amazon. People don't go to a Shopify, like a D2C brand that has like, you know, images on Shopify for that. And I'm wondering how much Shopify can do to actually fill in and be the default location for those like long-tail items that people shop on on Amazon for?

Yeah, I mean, it's theoretically possible. The problem is you still have to deliver the paper towels. I mean, at the end of the day, this is what I think this dispute— Muse Amazon, Amazon versus the world— sort of comes down to. Like, Amazon has the best AI moat precisely because what they do— to Jordi's point, Jordi's, Jordi's been making the pro-Amazon case. Oh yeah, this entire interview. He keeps saying, but what about the physical world? What about actually getting stuff done?

Yeah.

And the whole point here is actually getting the package, getting it to your house, and getting on your, on your, or wherever the Amazon drivers leave it, which can be fairly random, to be fair. But the, the, that is, that's the key thing. And because of that, I will be researching stuff. I will find, and you know, if you ask ChatGPT or Quad or whatever, it'll give you links and there will never be an Amazon link because Amazon doesn't allow it. If you say, if you specify Amazon, it'll give you search URL because that's all they can do.

Yeah.

But you know what I do? I just go to Amazon and I type it in and I click it. And it's not that hard. It's fairly annoying. I prefer to have a link. But at the end of the day, my personal revealed preference is I would almost always want it from Amazon because I trust it's going to get there. I know where to go to look to see if it's on its way. I know where to go to process a return. If I want to do that, when you introduce all these other pieces, you're actually introducing massive amounts of real-world friction just because you're too lazy to fine. I'll click over to the other browser and type in a search term. So I think this idea that I want my agent to actually do it is going to force Amazon to bend. I'm just skeptical because at the end of the day, Amazon. not really consumer-friendly. Like, they're like, they don't care. And you, like, you go to Amazon, it's all ads. Like, this is not a consumer-friendly website.

For a decade, weren't they like the customer obsession company?

No, they're obsessed with their customers seeing as many ads as possible.

Yes.

Uh, they do.

In the—

I remember even like way before they even had an ad business though, the website was always completely inscrutable. Like, it's, it's, it's like redesign the Amazon website. It's It's like, oh my—

but the—

so that's the case. By the way, the— in the other one, the only real alternative to Amazon in terms of it's just gonna have it and it's gonna get there is Walmart. And I was pretty optimistic when OpenAI did the checkout thing or whatever.

Yeah.

Walmart jumped in and Amazon was like, no. I'm like, oh, this is super interesting. Like Walmart's gonna actually make a move here. It makes sense for them. People in tech don't understand how actually awesome Walmart is. Yeah, uh, in that their stores— well, number one, their stores are like this crazy dystopian experience. Number two, if you go in their stores, their app is unbelievable. It will guide you through the aisles to tell you exactly where to go to get the thing that you need. Like, it's amazing. It's the best in-store app, you— like, by far compared to anyone. Uh, like, I'm obsessed with the Walmart app like I'm obsessed with like United's tech stack or why that actually makes them the best airline. It's a, it's a very similar sort of thing. But then they also have like, if the stuff's in their store, you can get it delivered in an hour. Like, who knew? Great slogan. Because like literally no one knows this about Walmart. And so when they did this thing with ChatGPT, I'm like, actually no one in tech who's gonna be buying stuff or who is technically inclined through ChatGPT actually knows how good the Walmart e-commerce experience is. So this makes a ton of sense. Fast forward 6 to 9 months, like, yeah, that was a mistake. They're like, people abandoned carts like crazy.

Yep.

Their, their cart size was way smaller. Like, all these issues. It just wasn't a good experience compared to if you're on Walmart's site, they have control of the experience. You put more stuff in your cart, you convert more frequently, all— and in addition to the getting more data about you, being able to show you ads, all the things that accrue to having on your site. And all this stuff matters to Walmart, and they're the only one— just as much as it does to Amazon— and they're the only ones that can rival Amazon for this. So Shopify, as long as you use Shop Pay, they're cool. Like abandoned cart, multiple items, that's fine. That's not their business. And so it makes tons of sense for them to lean in for what they do. But Shopify tried to do the logistics thing, like, and it— go look at the stock chart. The part where it goes super low down is because they're like, we're going to do logistics. Like, they learned their lesson. They are— they're asset light. That's what they do. They work across all these sites. If you're a merchant, it makes total sense to be on Shopify. They'll abstract. They'll be the middle person for every platform you want to be on. You can sell everywhere, but they're by definition not going to solve the reason why we still go to Amazon. And that means Amazon is going to win this dispute or it's going to be resolved in their favor in the end.

Last question relates to this Shop Pay Shopify dynamic.

But, but by the way, awesome product.

Like, it's the product that makes me not care about personal agents that can shop for me because it's just like, it's very easy to buy. It's very easy to buy.

100%.

It's like I could save you, I could Save you a minute, and it's like no, you can't save me a minute because I land on the site.

I literally just press it.

It fills in the code from my tech, you know, and it buys the thing. How much do you think consumers actually enjoy shopping? Real quick. Yes.

Yes. Oh, that was I was gonna drop in. Jordy, you and me, we are we are two peas in a pod today. People love shopping.

Yeah.

The the the the it blows me away. So tech companies are like, number one, let us book your flight for you. It's like, I've had a personal assistant for 10 years. He does not come within an inch of booking my flights for me. That is crazy talk. Once I have a private jet, he can book when my pilots show up and fly me on a private jet. Until then, I'm deciding which flight. I understand which ones are gonna be delayed. I know which seat I wanna be, be on like that. So the fact this is continually a demo is hilarious to me.

Mm-hmm.

Number 2, people love shopping. They don't want to be productive. They want to be entertained. One of the biggest forms of entertainment is shopping. This is part of the whole like fast fashion or like or all those Chinese brands. Like it's fun. Like speaking of toys, my nice rack, you know what's part part is the whole joy of it? Endless hours of research and figuring stuff out and then buying stuff and then the package shows up and you're like and like if you buy unified stuff, they always charge shipping. So you have all this pressure because you're like I I want to make sure I get everything in the right order. And then you buy it. And literally the moment after you get the shipping notification, you're like, oh, I forgot to buy something else. And then you start the whole process all over again and it's all torturous. But actually it's fun because that's kind of the whole point of this is we wanna like make money.

Also, there's so much information on a product page. Like companies have spent so much energy like focusing on that. And, and you look at like apparel as a category, it's like they're gonna tell you what size the model is wearing and you wanna know that and you wanna look at, at the different Different styles. And the idea of having a personal agent that's like going and getting some of the information, bringing it back to you, and then you're like, well, what about this other— like, tell me about this thing. No, this is like, you actually want to be on the—

like, uh, this is a— this is my free relationship advice of the day. Um, just because someone's complaining doesn't mean they're unhappy. Like, like, there's a bit like, oh, I have to go research this, I have to go do this, and I'm so busy. And you have your eager tech company executive say, oh, I have an assistant that can do it for you. And it's like, no, not only do I want to do it, I want to complain about it. I'm enjoying myself complaining right now, okay? Would you just let me be and live my life? And that— so that's why I think there's just a— there should be a mental reframing.

Yeah.

Stop trying to make people's lives better and start highlighting to them— and this is Marketing 101. Marketing 101 is you create the perception of a problem in people's lives and you show how your solution or how your product answers— is the solution to the problem.

Yeah.

And instead of saying, people don't have a problem that I'm insufficiently productive. They have a problem that I have too many packages piling up or I'm not processing returns. And it's like, it like just highlighted that, like how miserable this makes you. Do you know how much money, Jordi, you're leaving on the table? You could have like 3 puffer jackets if you got all your money back. And then, and then say, look, you could actually just— done. It's done. Money in your bank account. Wasn't that easy? Now you could go and do whatever you want. You can complain, you can watch TV, you can watch reels. That's the answer.

More time to complain.

Yeah, the, the funny thing is, is it feels like what the personal agents have really— the thing that people are super fixated on is the saving money, canceling subscriptions, and stuff like that. And the concern I have there is we've had a bunch of businesses like that that will even call— like, I think Truebill—

Rocket Money.

Yeah, it's now Rocket Money. They would call— they would have people that would— a human in the loop that would call and just cancel subscriptions for you, or they would try to renegotiate the bill. And the problem is these are like billion-dollar businesses, which is not going to fund the CapEx.

Yeah, well, the— I mean, the whole cancellation, auto cancellation thing drives me batty as a subscription, uh, proprietor.

Sure.

Like, we will get— when these new ones come out— no, we will— no, no, it's actually— it's the opposite. We will have a wave of customer support issues, which is people contacting saying, why am I not getting my email? I'm like, because you canceled your subscription. No, I didn't. And then they go back and look through their stupid little agent or whatever service went through and just canceled of the recurring subscriptions. Or the worst one is it'll go through their email inbox and you, you, you are required— now technically we do send transactional email, every email is customized, it's not a marketing email, but Google doesn't really care. We have to put in the header with the unsubscribe thing. So what does that mean? People go through and click unsubscribe. What if you— what happens when you click that required to include unsubscribe? You will stop getting email, but your subscription was not canceled because those are completely different. I didn't even know that.

Yeah.

2 different things. And so another customer support issue for us. I'm not getting my emails. Why not? And then actually, no, you set some dumb agent off on your inbox that canceled your thing, and now it's this whole rigmarole of you having to agree to these things to— so we can legally let you back on the list. Because like, there's actual federal laws around email lists. Big pain in the rear end.

Never cancel, Stratechery. Just don't do it.

Bake that into the I mean, I should be happy about it because they're like upset that like, you know, this is actually— it's super— we actually end up building our whole— we built our whole email service that is actually amazing. We send like fully customized emails at like the rate of like 20,000 like a minute or something like that. Like more performant than any other service we have, fully customized, all these sorts of things. And the reason is all the, the genesis of doing our own email was every email service out there is sending email people don't want. Like, that's— and so everything around them is about overcoming the fact that people don't want to get the email that they're sending. And I'm like, oh, I'm actually sending people email people do want. And it's really interesting. It just completely changes the problems and issues and features that you need to build.

We need custom silicon. I want 20,000 emails a millisecond. the future. Thank you so much for coming on the show.

Fantastic to see you.

Always great. Have a great rest of your day.

Talk to you guys later.

Great to see you. Goodbye.

I'm off the hook.

Let me tell you about Railway. Railway is the all-in-one intelligent cloud provider. Use your favorite agent to deploy web apps, servers, databases, and more while Railway automatically takes care of scaling, monitoring, and security. We got a traffic jam on our hands, so we're going to do a lightning round. Bring in Gagan Bhiyani. Welcome Welcome to the show. How you doing? Hey, good.

How are you?

Sorry for keeping you long.

Hopefully you got to hang out with us.

Please introduce yourself. Tell us about Horowitz Andreessen Academy.

I'm Gagan Viani. I'm CEO of Horowitz Andreessen Academy. It's a new school for young builders who want to go and make something of their career. We're really excited to be announcing it today. Raised $42 million, and Mark and Eric are joining the board.

Cool. Talk about the structure, in-person, virtual, thousands of students to start, class of 20 to start. How are you thinking about scaling up?

Yeah, I mean, you got to start small. So, we're going to have 50 students. The first class is tuition-free. It's called the Founding Class Fellowship. So, it's going to be a little different from any other programs we run. It's in-person in San Francisco. You know, I think that's probably the most important part of this whole thing, is that it's about getting right in the mix of Silicon Valley, understanding the culture, getting immersed in the, in the people, and building a great network.

Uh, at least 2 years.

Did I have a—

or is it 1 year?

So the first program will be 1 year. Eventually, uh, we'll, we'll launch a 2-year program.

Yeah.

How, how hard is it these days to to do like 4-year accreditation. I know, like, getting accredited is really difficult, but people were saying that about banks and there's new bank charters that are going out for startups. Palmer Luckey's bank, Erebor.

Anything is just so matter.

Like, maybe it doesn't matter.

None of the companies that you want to work at are like, oh, you went to an unaccredited school. You're super talented, but you didn't get the 4 years.

Yeah, but how are you thinking about it?

Are you thinking like, oh, that's, that's my answer.

Okay. That's your answer.

That's my answer.

Just skip it entirely.

And it's funny because I do think that I do wonder if over time, like a year is actually enough. Because like I went, I remember the end of my spring semester of college, I was like, okay, like college was fun. I want to like start working now. I feel like I like met a bunch of, you know, cool people and, but I'm ready for like real life. And I know not everyone is that way, but oftentimes, people aren't ready for real life because they're not prepared to have a real job, right? And so, if you could actually maybe solve that, then maybe people do a year of a university sort of experience—

Condense it down.

And then, they can just actually go get a job and they realize like, hey, instead of going from this high-burn life where I'm basically losing money for 4 years to then try to make it all back and And in later years, you can just have it condensed down.

So, CEO, raised money. I assume it's for-profit, but do you have a view on for-profit versus nonprofit? Obviously, there's a lot of critiques of the nonprofit system, but what is the philosophy around the long-term and incentive alignment, basically?

I mean, we're pro-capitalists and pro-optimists over here. So, techno-optimists, right? So, we believe that having a financial model is only going to help us serve students better. And maybe one of the biggest challenges with the modern education system is we've completely forgotten about students.

Yeah.

It's like, talk about K through 12 or higher ed or whatever. It's all about all the adults who run the show and serving themselves. So, one of the great things about the for-profit model is it forces you to deliver or die. And so, our plan is to deliver an amazing experience, and hopefully, our investors and ourselves will get rewarded for that.

Does delivering look like graduating straight into entrepreneurship, or are there other avenues that you think are exciting, like joining companies as employees, trade schools? There are so many different opportunities in the modern economy.

Yeah. I mean, you guys have been in the Valley a long time. So have I, right? Like, you know, the most successful people are not only founders. Like, I know, like, so many of my friends who decided to switch from being founders or never were and joined companies are incredibly productive and successful and happy. And so there's no reason to prescribe one thing. I mean, I think one, one of the biggest challenges we're going to have as a school is there's going to be a lot of attention and pressure on these students. And my goal as sort of CEO is to design a culture that reduces that pressure and allows them to see that there are many different options for their careers. And, you know, there's many different paths to a successful life. So I don't see why we have to over-aggrandize starting a company. And that's actually why we're building a school and not an accelerator, right? Because we think that 1, 2 years to explore and learn is a much better outcome than, okay, I'm 18 and now I either got to join something for 4 years or I got to start a company and give up 10 years.

Yeah, no, it can be a crazy trade-off.

Yeah. So many people would have been like top 1% second and first, second-in-command type people and actually maybe only top 30% founders. And the difference there is like tens of millions or even in some cases billions of dollars.

Yeah.

And so, I think that's such an important lesson is like, no, you don't. To have an impact in the world, to be successful, to have the career that you want, you shouldn't just be being high. When you're young, maybe being a founder is high status. When you're older, just like being successful, high status. It doesn't really matter exactly how you got there, but like, you want to make sure that you actually win.

Everyone's equal on the golf course.

I mean, how many people made $100 million in the SpaceX IPO and we don't know who they are or what they do? I mean, they, they're just, you know, they're not just— they are engineers for one of the coolest companies on the planet, right? And, uh, but they don't have a fancy— they're not even second in command And I think that just tells you everything you need to know about why it shouldn't be just one path coming out of the academy.

Very cool. Well, thank you so much for hopping on the show.

Last question. When are you guys— when's the first class?

Oh, yeah.

Fall?

First class in fall 2027. Applications are open today. So check us out at theacademiesf.com.

That's going to be an incredibly low admit rate. 50 slots for— I mean, something that already has gone viral on the internet. There's going to be a lot of So good luck to you if you applied.

That's great.

And it's free. Yeah, that's a good steal.

So first year is free. First year is free. It won't be from then on.

Okay.

Yeah.

Well, thanks so much for coming on Breaking Backglass.

Great to see you, dude.

Have a good rest of your day. Goodbye. Let me tell you about MongoDB. What's the only thing faster than the AI market, your business, than MongoDB? Don't just build AI, own the data platform that powers it. We have Alex Ratner from Snorkel AI, massive fundraising news. I'm gonna hit the gong straight away. $350 million.

Boom.

Valuation. As demand grows for specialized AI training data, how specialized are we talking these days?

Well, first of all, thanks for having me on. And I love the gong. I bought a used gong on the street that looked kind of like that. And sadly, I then forgot about it. And we never really developed the culture of gong ringing.

And yeah, it's not just about the object. It's about the lifestyle.

It's process, discipline, lifestyle. Exactly. And so you've inspired me there.

Exactly.

But yeah, I guess we were here to talk about data and AI ostensibly, right? So I should get back to that. Yeah. What does specialized data mean? I mean, I think one of the big shifts that in some ways— so we started a decade ago at Stanford and UW doing research.

Overnight.

Yeah, I'm a dinosaur.

That's amazing. Congrats.

Well deserved. I remember when a JEV was called a classifier, if you're okay with the truly nerdy AI jokes. But I think there's this big shift that happens in any kind of autonomy or any learned autonomy, self-driving, digital LLMs, physical AI, where we call it like data 1.0 to data 2.0, where when you're early on in the learning curve, the model doesn't know much. So kind of any bit of information is additive.

Yeah.

And so therefore, your rational focus is on volume.

Right.

And volume is mostly solved by staffing, crowdsourcing. How many humans can you get into the factory? Right. And then you naturally get to the kind of the more durable asymptote phase where, like, now models are really good and it's not about any data. It's about the right data that targets some area where they don't know something or they're misaligned on something or they can be improved. And finding and building that data becomes a lot more about precision, quality. And we think that's really more the the place where kind of a research and technology approach comes to bear. And so that's kind of where we focus. And that's kind of what we think is the interesting area in data today.

Is it too aggressive to frame you as sort of like short recursive self-improvement? Because there is this thesis, you know, sort of mulling around that like the models are going to train themselves and then maybe you don't need any more data because they're all doing synthetic data. How have you grappled with that? Because you seem very grounded. You've been doing this a long time. I feel like you're not maybe swayed by the latest hype cycle on Twitter in a single week. But how have you processed that whole thing?

Yeah, I mean, I've heard the— I mean, I've been working in synthetic data as a researcher for almost a decade too. And I've heard the we're one model advance away from solving everything thing a few times. I'd say I'm short the more kind of sci-fi accelerate and be done with humanity's contribution to anything economic kind of version of RSI. I would call myself, though, more bullish on RSI, but RSI that contains other elements beyond just kind of modifying model training code and hyperparameters. In other words, compute energy and data bottlenecks, human expertise bottlenecks. Right.

Yeah.

I've had this take that it's at a certain point, like, superintelligence to do RSI fully needs to be able to fundraise. Like, you got to go get more capital at some point. And so, RSI in many ways is like AGI complete in the sense that it should be able to do the full stack, even the capital formation.

Maybe another example is like, it feels like we're coming up on math being like somewhat solved. And I'm sure once it's all solved. They're, you know, hopefully, you know, you could figure out new questions. But like, do you expect to see a drop in demand for math experts from the labs?

Yeah, it's interesting. I mean, I think the first thing before double-clicking on math for a second, been too many years since I've taken a theoretical math course, so I'm gonna dodge that whole topic artfully, run the clock down. But First, on kind of overall areas where there's demand for— overall areas where this RSI loop needs data, needs human expertise, needs real-world input as a necessary input. There's this old— we tend to conflate areas that are easy for humans as necessarily be easy for AI and ones that are hard as being hard for AI. And it's often not the same, right? So math is something where it's very hard for humans, but there are some reasons to expect that it would be more approachable. I mean, there's decades of work on verifiability engines like Lean and ways to kind of check it. And there's lots of data out there, and data can be generated more automatically. And then areas like writing a good Slack post to your coworker, you know, has sparser data and more ill-defined rewards and doesn't work right away as well and needs more nuanced context. And so, you know, first of all, there are these areas that are kind of counterintuitive where there's real demand for human input. You know, it's not enough just to have a coding agent that can kind of tweak the hyperparameters of the model to achieve RSI. You need real input from the real world. But even in the areas like math, like coding, where we do a lot of volume these days. You know, one, there's some basic economic forces where once something becomes valuable, and this is especially true for something like coding, where maybe the accuracy goes up, the demand for some types of easier data goes down, but the economic value of every kind of additional decimal point of improvement goes way up because of the value.

Yeah.

And so does the difficulty of creating that data. And so that's kind of the growth trend that, I mean, selfishly talking my book for a second, that we've benefited from is a lot of people have entered the data space. Great validation for those of us who have been trying to push on data being an interesting area for a decade or so. But you can't get this hard data with just staffing or sourcing human hours. You can't get it with just kind of trying to have the model that you're trying to sell to that provider generate all the data like a synthetic slop cannon. You have to actually use— more sophisticated blends of human expertise and specialized AI to get this really, really hard coding data that emulates what a software engineer, a senior software engineer will do over days or weeks, what an expert mathematician would do. And so from an economic perspective, in terms of how has the opportunity grown, it's if you can keep up with that kind of RSI waterline.

Mm-hmm.

And maybe feel free to make fun of me, but we call what we build the RSI engine for data because there's also kind of a compounding loop in our platform of humans being accelerated and improved by AI, but then their feedback improving the specialized AI models we use. And one has to keep up with the lingo, right? But our view is only if you have an approach like that that can feasibly keep up with the frontier. You know, if you can't do that, your data's not gonna be valuable anymore. If you can, it's more valuable than ever before.

Yeah. Well, congratulations on the round. Congratulations on the progress.

10-year overnight success.

I love an overnight success like this. Well deserved. And hopefully we'll talk to you soon.

Yeah, great to see you. Awesome.

Have a great rest of your day. We'll talk to you soon.

Thanks guys so much.

Goodbye. Let me tell you about CrowdStrike. Your business is AI, their business is securing it. CrowdStrike secures AI Again, stops, breaches. We have our lightning round continuing. I believe Peter from Swarm Aero, the co-founder and CEO, is in the waiting room next, but we got a little bit of traffic jam. Hopefully we can bring in Peter, and we're gonna have Tyler Cosgrove sub in for Jordi Hays, who has to take off right at 1:00 PM sharp. I don't— I like springing it on people every once in a while. Sometimes I like to do a little bit of an intro, but we're gonna have Tyler sitting next to me.

Hi, everyone.

Peter, how are you doing? Welcome to the show.

Fantastic. Thanks for having me on.

Thank you so much for jumping on. Would love to reintroduce the product for everyone. Just explain how you're positioned in the market and sort of take us through the basics of the company.

Yeah, absolutely. So last week we introduced to the world Gamera, which is our aircraft really designed around breaking kind of the iron triangle of range, payload capacity, And cost, right? And that's really one of the big problems facing, uh, facing the Department of War today and the Air Force today.

Uh, is the name Gamera from the Koenigsegg car, or is it a more generic—

It's not.

We, we did check that.

No, uh, it is from the, uh, the, the turtle monster that fought Godzilla.

Oh, okay, cool. It is old school. That's great. So, uh, what does the process of building this actually, uh, look like? I believe we have a video that we can can pull up while you're explaining it further. But I'd love to know, like, are you specifying it in basically like PDFs and documents and then going to the Department of War, selling it, and then building it? Are you prototyping? Like, what is the flow to actually take something like this from start to finish?

Yeah, absolutely.

So we started this about 4 years ago when we started the company. It is a long process. And we started off by kind of sketching up a concept, figuring out what the problem is that we needed to solve. And then going about like designing the aircraft. But I love the aircraft. I'm an airplane guy at heart. But the aircraft is really just one piece of it. So along with the airplane, you really have to design how to build this. And we've been in it as well as building the aircraft. We've also been building the factory, the capacity to, to really build these out hundreds a year.

Sure.

And when you look at it, you've got a bunch of aircraft, you can build them, you really have to be able to control them. And that's where We've tried to do things differently than the single pilot per aircraft model we have today. It's really building out that autonomy and the command and control stack that allows us to, to have truly a swarm of these, hundreds of these operated all at once.

Hundreds of them. When I think about one remote pilot per vehicle, is that actually the bottleneck? I mean, do we not have hundreds of of Air Force, you know, soldiers in Nevada in shipping containers controlling these things remotely. Are we actually bottlenecked there or is this— or is this just like thinking even further into the future?

So it's a little bit of both. We truly are. There just aren't enough folks out there who have those skill sets. And be thinking about coordinating this. You know, when you have hundreds of aircraft, like these are kind of conflicts that happened last time in World War II. Right. So you can't have folks talking on the radio trying to figure this out. You really need that autonomy to truly make that, make that effective today.

Yeah. So how important is price to the buyer? How do you actually guarantee that? Is there any financial risk where you go to the Department of War and you say it's $10 mil and then it winds up coming back at $9.9 and you're not making any margin? How does that work?

It's super important. So this is why we've kind of designed it from the very beginning. Typically in aerospace, you know, it's all— at least since, uh, in the last 50 years or so, it's really been about like, how do you get the most performance out of an airplane? And everything else, you kind of, you know, kind of falls by the wayside. For us, we looked at how much does this cost from day one, right? And that's not something you sprinkle on top of design. You really bake that in from the, from the earliest stages. So we're confident the, uh, the Air Force is asking for less than $10 million per, and, uh—

They're not gonna pay more. Yeah. Do you have a question?

We've done the math and shown we can do that.

That's good.

Yeah. Was there like a specific kind of like theater of war that this was like built for, or is it like very general? Like in your kind of ideal use case, is there like a very specific kind of like, I don't know, geographical—

Sure.

You know, it can be used in a lot of different places, but really when we looked at what's gonna push the limits, it really is the Pacific. It's that range. It's carrying the kinds of payloads, the big enough ones that make a difference in a potential conflict there. Really, it's focused on the Pacific, but you can use it pretty much anywhere in the world.

Well, thank you for everything you're doing. Congratulations on the progress and hope to talk to you soon.

Super excited. Thanks a lot, guys.

Have a good rest of your day.

Cheers.

Let me tell you about Cisco. Cisco, critical infrastructure for the AI era. Unlock seamless real-time experiences and new value with Cisco. Good soundboard.

I'm learning.

Soundboard. Our last guest has been waiting far too long. We have Daniel from Fair Square Medicare. Welcome to the show, Daniel. Sorry for the switch around. Thanks for being flexible. Welcome to the show.

How you doing?

Hey, thanks. Thanks for having me, guys. It's great to be on here.

We appreciate the flexibility. First time on the show. Can you sort of give us a whirlwind tour of your career? You've done a lot. How do you introduce yourself these days?

Yes, it's a great question. Right now, I'm a car show host.

Okay.

So I have a show called Cars and Capitalists. I take some of my friends who are great investors and operators out in an old 911, and we drive around Marin. So we have like Sam Lessin, Paul Gu, Sam D'Amico on there. That's a lot of fun. But I mean, like, I feel like The way I should introduce myself is I've tried my whole career to start companies by doing all the wrong things, but I did a couple of things right.

Sure.

And so, my last company was Fairsquare Medicare.

Sure.

Sold it after 6 years. Great company. We ended up working, helping find like 25,000 American seniors find the right health insurance. Like, there's actually Probably a good chance we've worked with, like, the parents of some of your staff. And before that, did a fintech app called Trim. And way before that, started a lender for the powersports space called Octane Lending.

Interesting. Is there any, like, common thread between the categories or strategies or teams between those companies? Or is it sort of just like, be open to opportunity? When it knocks, you open the door and you go after it really hard?

I think there are a couple of threads. There are 2. One is just like consumer advocacy, which was true for the last 2 companies. So like this one was the kind of realization was people are turning 65, and there are like 50 health insurance plans to choose from. And they really can't tell the difference. It might as well be like a foreign language to them.

Sure.

But the implications are massive of like you get on the wrong health plan, and then you get like cancer, and then you can't get on a better plan. Um, and before that, there was a personal finance app called Trim, which I co-founded, and we were the first ones to show you a list of all your subscriptions on your credit card. We were like, we were like Plaid's biggest, one of their biggest first customers.

Wow.

And we were just early to that.

Yeah.

And, and that's what became like a competitive space eventually. I, I know that we didn't— Rocket Money acquire Truebill? That's That's what Jordi was saying recently.

Yeah. So, Truebill—

Thorn in your side.

Well, they just copied us early on. They copied the subscription feature. And no, but they out-executed.

Look, credit where credit's due. Credit where credit's due.

They executed. I think they sold to Rocket for like a billion bucks.

I think we made a much smaller exit. Yeah, we did. It was a great time. Yeah. So, it was a lot of good times.

Yeah.

Where do you want Cars and Capitalists to go? Have you asked people if you can drive their special cars? Because I feel like the comedians in Cars Getting Coffee, part of the beauty is I think Jerry Seinfeld has such a huge collection that he picks a car that matches with the, or in his theory, like the comedian's identity. But I would kill for a whirlwind tour of Silicon Valley's biggest motorheads.

Yeah, it's a great idea. I actually, I should probably like, I like running it lean. And so I don't do any B-roll. I don't do outside footage of the car. It's all, I just have a GoPro in the car and it's just us talking. And maybe I'll like go find other people's cars and get, I would need to like hire someone to take video of the car.

It would need to be very—

It would be a very weird show because I I feel like 90% of your guests would be like, another Model Y, yay! Maybe the new Roadster. Maybe the new Roadster.

Yeah, get some air.

I know, I don't, I don't want to drive a go-kart, although I would drive a Roadster.

Shots fired.

Yeah, like a Tesla. After you, after you drive an old 911, you realize that the Tesla is a, is an amazing vehicle for getting A to B, but it's a fucking go-kart.

Yeah, yeah.

And I, I don't know, sorry if A curse on your show.

But, but, um, but it's fine.

But actually, you raise a great point, John. If anyone is watching and has a manual F430, that's something we should definitely add to the show.

Okay.

Um, it would be, it would be great to have. But I mean, of course, like, this is Silicon Valley, so, and, and like, a lot of the people on my show couldn't care less about cars.

Yeah, yeah, yeah. It's sort of bifurcated. It's like either they have like an exquisite collection or they've never owned anything other than a Tesla. And it's the most boring— like, either they care or they don't. There's very few people that are in the middle. We've been running predictions with a lot of guests about the Tesla Roadster. Elon's going to announce it on October 1st. And he said that it might fly, might jump, might leave the ground.

Shut down the airspace.

They shut down the airspace. The FAA said they've shut down the airspace for, for 10,000 feet in the air. So that creates a boundary. It's either 0 inches off the ground, it winds up not jumping or flying at all, he was just joking around, or it can go up to 10,000 feet. So I want your prediction right now, super binding, super high stakes. We're gonna hold this to you. We're gonna hold you to this. But how do you— how far do you think it'll get off the ground? Are we talking inches, feet, hundreds of feet?

0 inches.

0 inches.

It's gonna have 4 thrusters in the back that are gonna generate downforce. That's gonna let it go like really quick.

Okay, so we're looking 0 to 60, 1.5 seconds, 1 second, something crazy like that?

No, probably 0.9 seconds.

No way, that'll be wild.

Yeah, I have, I have some inside info. I, I was— I, I did like a little, uh, deep dive on quantum computing at one point, and it's in sale. And I was on a flight to Austin, I sat next to this guy who I don't, I don't want to—

Didn't have a screen protector on?

I don't— yeah, no, I know, no, no, no, I just started talking about it. He was this great— he was pretty like senior Tesla dude. And that's all I'll say about it. But he was like the goal. He was like the goal. He was a total bro.

He was great.

And his— and he was just— the goal is to not kill the driver.

Yeah, the G-force needs to be tolerable. So if the cold gas thrusters act as downforce, couldn't you reverse them like the old fan car, make it lift off the ground, even just for a stunt? You don't think it's going to go even 1 inch? It feels like he's been teasing it, like at least we're going to get it jumping off of a ramp or something. It's going to go in the air. Why shut down the airspace 10,000 feet up?

I think that part is a PR stunt.

Okay.

I think he, I think he actually wants to make a car people can buy, right? And the thing, the thing about a car with wheels is that it's a wing, right? Because the bottom is flat and then the top is curved, so it generates Which is why like spoilers exist.

Yep.

So the only thing that really makes sense is, is you have, you have the like boosters going down. But you know, maybe it'll have— maybe it'll actually be a quadcopter.

Yeah.

But, uh, but I mean, it's amazing, right, that they're— this is like their last hurrah. This is like the last car they're gonna make.

Probably, yeah. Maybe the Cybertruck SUV. I'm really hoping they do that. I think that would be an interesting one. But, uh, who knows?

Would you buy that?

I don't know. It's certainly like an aggressive aesthetic choice, but functionally, like having a full-size SUV for so many modern American families, that's just a non-negotiable. Yes, there might be an old Porsche in the garage, but there's going to be a full-size SUV for anyone that has a family. So it just seems like a market segment that they've been sort of dabbling in with the Model Y and Model X, but they never really got there because those are more minivan-sized. You got to go with the Land Yacht, the American, the Expedition, the Escalade. This is what America was built on, right?

I don't know that.

Yeah, you know, you know the Model Y extended wheelbase is like totally sold out?

I have one. I'm part of the problem. I ordered one of those. You have the extended one? I don't have it yet, but I was one of the pre-orders because the full self-driving is very good. And I was like, ah, if they stretch it out, that sort of checks that box. I think I can fit all the kids, the dogs, all the All the gear in there, so I think that might fill that slot for a little bit. But you know, land yacht, land yacht is is still still in the books potentially. What do you think on pricing? Tesla Roadster. He originally promised two fifty at the time. Two fifty k. That was supercar money. That was Lamborghini Ferrari money. Now it's well spec Turbo S money, right? It's like it it it went down from two. $250,000 gets you a sports car, used to get you a supercar. Do you think he raises the price and it's a $600,000 car?

Like, or like, does he— or does he accept like alternate forms of payment? Like, is he going to accept USD? Right. Like, he's definitely going to accept Bitcoin.

Yeah.

Will he accept shares in certain companies?

Oh, interesting.

Yeah.

Build the portfolio, maybe.

Yeah.

I don't know.

I'd be excited because if they can do some surprising 0 to 60 time or some gimmick. And then it's just a really fast sports car for $250,000. That actually is a unique space in a world where even for all the supercars, they all start at $400,000 now.

Here's a question, John. Like, this is obviously not serious for them, right? They're not going to scale this line. They don't— they don't care about like the revenue, I would guess. Yeah, right.

This is—

this is like a fun toy project. This is like a better version of that, that flamethrower.

Yeah.

So like, I wonder how many people they have working on it.

Like, I bet it's a—

I bet it's a tiny number, and I, I, I bet they just don't care that much about the price. Like, they're just going to say $250 because they said $250 before. Obviously, if it was like $500 or a million, I like— I wonder what the elasticity curve there is.

Like, they would probably sell— Oh yeah, maybe they just do a really limited run and then they don't even continue to sell them, and they're like, yeah, we sold them all at cost and we did the thing, we checked the box. I'm just shocked that Elon hasn't dropped it because he, he's— he drops projects all the time and he could have easily just been like, I'm refunding all the deposits, we're going in a different direction, we're refocusing. And everyone would be like, great, yeah, that's fine, focus on the next thing. Like, you know, do, do whatever. It wouldn't be that big of a deal. I don't know.

Well, that suggests— the suggestion you just made, it really begs like maybe it is a real flying car, right? Maybe because Maybe because, like, why, why would he do like a toy project? Yeah. If, you know, it like hints at something bigger.

Yeah. Yeah. He's never been one to like, yeah, just rip the Band-Aid off. Like, he never took Tesla private. And it wasn't like he was like, I got to do that. I'm sticking with it. I'm still telling people it's going to happen. It's just like, yeah, that was a crazy idea. Didn't work out. Moving on to the next thing. Stock's going to the moon, you know?

So are you, are you going to get one?

So, so the more we've been talking about this, I think that the deposit is incredibly up, incredible upside, because I think it's like $5,000 down and then you have to put $50,000 down within, within 10 days. Within 10 days, the announcement's 10 days away. And so it's refundable. And so if it's cool, then everyone's going to flood in. You actually get one, maybe Regardless of whether or not it trades up or trades down, you just have the option to be the one with the flying car, or you just get a refund.

And so, dude, what's the, what's the Roadster pop going to be? People talk about IPO pop. Yeah, buy it for $250,000, turn around in a day and sell it for $600,000.

It's possible. It's possible. So yeah, maybe. I'm, I'm really, really close.

If, if you believe it's downforce. Yeah, right. And, and therefore, like, this is not an entrée into flying cars. flying cars, this is just like a little gimmick. Yeah, then it's going to be limited. It's going to be the sickest gimmick ever, which means it's gonna— the value is gonna go up. So like everyone should just try to buy one and resell it.

And also, I, I, I do just think that the, uh, like, like there is a more moderate scenario, which is it's not a flying car, it's not the craziest thing ever, but it's just like a really, really crazy sports car at that price point. And it's just like, yeah, it's just a It's a good option to drive. Of course, you're never going to compete with the manual air-cooled Porsches of the world. But for a different buyer, it is just like you're getting a crazy experience for what is, in the modern era, a sort of reasonable price. But yeah, if it doesn't fly and he raises the price to $800K, I think a lot of people are going to be saying, I'm out. But we'll see.

I wasn't thinking of buying one, but this conversation is, I think, I think might be a really great trade.

You're not even— it's not that you're thinking of buying one. It's like put down the deposit and then you just have the optionality and then you get to watch the livestream and say, yeah, I'm going to, I'm going to stick around and wait because it's still going to take a year, but I'm converting. Whereas people who did this the last time, they had to wait 9 years and they tied up, I think, even more. Some of them tied up a lot of money.

But what would that money— do you know how much they tied up and what, like, I think it was like $50K originally.

And if they put it in Tesla, it'd be like $10 million or something. It was notoriously bad. But what I'm saying is like, under no circumstances tie your money up in this deposit for 10 years. Just tie the money up for 10 days. See if you like it. If you like it, then continue. But if you don't, then pull out and timebox it so that you're always going to be out of that, you know, that, that market in, in by the end of What if he says just kidding and he's like, it's gonna be another 10 years? If it's not, uh, then he deserves it. I don't know, keep the, keep the working capital, I suppose, you know, have it as a treat.

I don't know.

Anyway, this is a lot of fun. Uh, anything else going on in your world? Anything else?

Uh, robots. I think the thing, the thing I think about right now is I built a dog. Like, I have some free time and I run this company. We continue to help seniors. We got— this is Pupper.

Ooh, what's this?

This is an open-source quadruped robot. Quadruped is a really suboptimal design for a robot.

Why?

Because it doesn't have wheels?

Because, because like the— you know what an actuator is?

Yeah, yeah.

You know an electric motor? Like it spins an axle really fast, right? And basically like actuators are like a geared-down electric motor.

Sure.

So it's really good for like a car. Like, it's really amazing in the Tesla, but they're, they're really bad at generating torque. Like, if you think about like the way your arm moves, right, there's not like an actuator in your elbow that spins. Like, you just have a muscle.

Sure.

I, I think the nerdy way of saying that is like linear torque. So in a way, all the humanoid robots companies are kind of like ham-fisting this, like, this component which generates radial motion into like, um, yeah, yeah, more linear motion. So that's like That's something I'm playing around with.

Cool.

That's fun.

And, uh, robot dog.

AI-generated robot dog.

What do you have the robot dog do? Are you getting any utility?

What's the biggest use case? Defense? Home defense?

Uh, it's like a party treat. We had, we had, uh, we had like a bunch of our friends over for Rosh Hashanah, and, uh, dude, they went nuts for this robot dog.

Yeah, that sounds fun.

That's, that's, that's kind of it. And, and of course, like, if you're interested in And building a robot.

So yeah, it's a great V0. Fantastic. Well, thanks so much for coming on the show. Hey, thanks for having me.

Let's do it again soon.

Cheers.

Talk to you later. Good to meet you. Let me tell you about public.com. Investing for those who take it seriously. They got stocks, options, bonds, crypto, treasuries, and more with great customer service. You think the Roadster event's gonna go well? Go express your opinion on public.com. Do whatever you wanna do. Any other breaking news? Anything you've been tracking over the last couple hours? while we've been live. I think we're pretty much good. Opus 5.5 launch, we talked about that. Benchmarks look really good, very cheap. And the key thing that you pointed out was like the second post is like, this is the first model that we've released since we said we were pacing the frontier. This is an expression of the pacing.

Yeah, yeah.

It's affirming it like this is, you know, Fable 5.1 class model, but it's much cheaper.

So it's not trying to be more superhuman, more dangerous. It's the The safest, best, fastest, cheapest, lightest, thinnest model ever, basically. Yeah.

And then OpenAI also did the same thing today.

Okay.

Yeah. They— oh, so SOL's out?

SOL, yep.

GPT-6 SOL.

GPT-6 SOL.

Cool. And then also—

Luna.

Luna, but not Terra?

Or I might have had that backwards.

Okay. But new models, the model wars never cease to entertain. Of course, all that matters is what you do with them. So send us some— send us whatever you build. Send us whatever you play around with. Send us the hallucinations, the good, the bad. I wanna see it all. Thank you for watching TVPN. We will see you tomorrow at 11:00 AM Pacific. Leave us 5 stars on Apple Podcasts and Spotify. Sign up for our newsletter at tvpn.com.

Thank you.

And throw that flashbang, Tyler Cosgrove.