You're watching TBPN. Today is Tuesday, March the 10th, 2026. We are live from the TBPN, the temple of technology, the fortress of finance, the capital of capital.
Uh, we have a great show for you today, folks. Let me tell you about ramp.com. Time is money. Save both. Easy. Use corporate cards, bill pay, accounting, and a whole lot more all in one place. Let's quickly pull up the linear lineup because we have some great guests coming on. Olivia Moore from A16Z is breaking down the top 100 generative AI consumer applications. Of course, linear is the system for modern software development. 70% of enterprise workspaces on linear are using agents. And we have...
David, it's deals day. It's deals, deals, deals, deals. Juicebox is coming on. We got to raise a bunch of money. Raise a bunch of money. We also have Archer coming on to talk about flying cars. When will we get them? What happened to my flying car? They built one. We're going to see what the timeline is to get me in it.
Well, uh, let's run through a little, a few of these deals just to kick off the show. We're going to go through them later in the timeline, but there's a few things. Uh, Brandon Garell wrote the op-ed today in the TBPN newsletter at tbpn.com. Um, Mira Murati's Thinking Machines snagged a multi-year partnership with Nvidia. Uh, uh, Thinking Machines has been on the ropes. Uh, they lost half of the six co-founders in under a year. There's a question about where the business is going. This is obviously a good sign that they got a multi-year investment done with Nvidia, uh, in which it will deploy at least a gigawatt of cutting-edge chips to train AI models. They are going to be GPU richer. I don't know where the bar is for GPU rich or GPU poor is today, but they're one, one gigawatt richer after today, which is good news for them. So, congrats to everyone at Thinking Machines. Um, even though they've had a couple high-profile executive departures, the team has grown from 30 people to 120 people. So, they're still cooking.
Also still cooking, Alex Wang. There was a bunch of fake news on the timeline. We'll dig into this, but uh, multiple tech news aggregator accounts on X posted that Alexander Wang, who's been on the show at Meta Connect, uh, I've interviewed him a few times. Uh, he leads MSL, Meta Super Intelligence Labs, and they, and they were saying he's out. He's on the, he's on his ropes. He's on, he's fighting for his life over there. Well, it was fake news. And we'll go through exactly how this happened, but uh, Meta CTO Andrew Bosworth and Zuck also both hopped into the chats, different chats, which we'll take you through, um, to categorically deny the rumor. So, um, we will dig into that. Also, Yan Lun, uh, Matt raised a massive seed round for Advanced Machine Intelligence Labs, AML.
Not the kind of combination that you normally see.
Yeah, it's not very American.
Sure, but it's, it's uh, big confidence money in, in the age of AI, in the age of compute, uh, requirements. You, you got to spend money to make money in AI, and he's got the money now.
Uh, also as we mentioned, Legora is coming on talking about their Series D, $550 million at a $5.5 billion valuation just a year after their entry into the US market. Fascinating industry. We talked to a partner at Sequoia yesterday. Obviously, Sequoia is an investor at Harvey. But how will these firms change? Will the tools become agencies? Will they be doing the work? Will they be more direct to consumer? This is a question that we've been digging through, uh, in the, in the...
Yeah, a lot of, a lot of people have been kind of questioning just how thin...
Are these RA, how, how thick or thin are these wrappers basically? Um, but also AI recruiting platform Juicebox, uh, which was a part of YC Summer's 2022 batch. That's a good time to go through YC, right before the AI boom. Uh, you're up and running. Well, they are up and running with $116 million after a $80 million Series B, which values it at $850 million. That's the kind of dilution that you're looking for. 10%, not bad. Uh, a little under 10%. Uh, and the round was led by DST Global with participation from Sequoia, Coat2, and YC. Very good news for the folks over at Juicebox. Uh, they're in the hiring market. So, we're going to have the founder on to talk about the business, but also talk about the hiring market. Where is their strength? Where is their weakness? What is he reading into the jobs data? We'll try and get to the bottom of, uh, where the opportunity is in the modern economy.
Uh, Meta also acquired the agent-based Reddit-style social network, Moltbook. We of course had the founder, the creator of Moltbook on. I actually know the other co-founder as well, Ben Parr. They will both be joining Meta Super Intelligence Lab. There's a lot of back and forth on was it all slop? Is there any value there? Well, we don't know the terms of the deal. Doesn't have to be a billion-dollar acquisition. Who knows? Uh, they're, I've talked to both of the founders. They're both, uh, uh, you know, capable, interesting people. And I think it's, I think it's under discussed, and we'll get into this, under discussed that, uh, who is evaluating these acquisitions. It's not just Mark Zuckerberg. It's not just Alex Wang. You also got Nat Friedman, Daniel Gross. These guys have backed a lot of founders. They've worked with a lot of AI startups. They can understand the team that they're trying to build over there. And there might be some interesting interface between AI agents and social media. This is highly relevant. Meta seems like the logical...
Yeah, remember Meta, Meta filed some patent for basically bringing yourself back to life in ancient form after death, right? So, of course, they're thinking about this stuff.
Once you shed your, your mortal coil and you molt, you go on Moltbook. That's very macabre.
Yeah, I would be, I'd be shocked if they keep Moltbook running really, uh, for more than a handful of months. This just feels like, hey, let's bring some people on board that have been thinking, spending all their time thinking about how bots are going to interact with other bots and, uh, and humans on the internet.
Yeah. And Meta has done a ton of these types of acquisitions, um, where like, like smaller products, tuck-ins. Not everything has been WhatsApp, 16 billion, 8 bill, I forget. It was a lot of billions.
Yeah, Nikita, Nikita's first was...
That was a good example. And if you just think about it as like you get a shot on goal with one product, you get a product leader that can go and, and bring some new energy, some new ideas in, there's a lot of opportunity there. Well, before we move to the timeline, let me tell you about Figma. No matter where your idea starts, Figma Make, Claude Code, or a sketch, the Figma canvas is where ideas connect and products take shape. Build in the right direction with Figma.
And let me also tell you about Gemini. Gemini 3.1 Pro is here with a more capable baseline. It's great for super complex tasks like visualizing difficult concepts, synthesizing data into a single view, or bringing creative projects to life. And there's also some exciting news from Google that we'll touch on today. Not exactly a deal, but a whole bunch of new features that we'll be going through. So, uh, Theo is talking about the latest from Anthropic. So, Claude Code now has code review, which optimizes for depth and may be more expensive than other solutions like their open, like open-source GitHub Actions. Reviews generally average $15 to $25 billed on token usage and they scale based on PR complexity. And Theo says Anthropic really needs like one normal person to proof these things before posting. I guess people are, are upset about the price of having, uh, these, these code reviews, uh, billed individually in a world where so much code is being generated. Some of, uh, some of the initial copy around this announcement looked like it was just a flat rate per code review.
In actuality, it's billed based on token usage, but it's funny to have like a flat rate. You know, it's generating code. It's, you're getting charged to review the code and it's just like...
Also, like all of the token rates and just AI expense lines are shifting so dramatically. Token usage is ramping. You're getting discounted tokens from certain plans. Like it's very hard to grapple with how you think about budgets. You know, we've talked to a number of people where like, you know, at, at Microsoft, every employee needs a token budget. Everyone, every employee needs some sort of AI budget. You should still think about it almost in a per-seat basis, but depending on what someone's doing in the organization, they get a different AI budget. Um, but, uh, this, this post from Burr Bama was very funny. They feed us poison Claude Code, so we buy their cures, code review, while they suppress our medicine, which is what is the medicine in this, actually writing the code correctly the first time.
Pull up this, pull up this next one from Luffy. Claude Code after writing your code, leave a tip.
Leave a tip. Yep. 10, 15. They really should do a tip button. I like, I like the idea of a tip. Um, Tyler, what, what, what's your take on, on, on buying or paying for, for AI code reviews there? You know, we're, we're, we're sponsored by a code review company. There are a number of code review solutions. Uh, what, what, what's the advantage to having AI run a code review these days?
Uh, yeah. I mean, it, it like makes a lot of sense. Like, it, it doesn't apply to you because you don't review code, correct?
Well, I mean, so it makes sense for teams, right? Because I like, I, I don't, I don't need an like external code review.
Yeah, on my code, because I'll just have like, if I'm in Codex, if I'm in Claude Code, I'll just tell it like, review it.
You would think that while writing it, it's reviewing it, right? Hopefully, it does that.
So, I personally, I never check my work in the moment. I'm just full speed ahead.
Yeah. What is this, uh, what is this Claude Remarks account? P Remarks. It seems like it's a, it uses the real Claude logo, but it very much feels like it's not owned by Anthropic because this post is, uh, Walter White spinning a pistol saying mid-level non-technical business unit leaders asking Claude where they can cut headcount to reduce waste, and you flip it around and it just says, 'Actually, we don't need you.' Which is the funniest situation. Claude, based on this conversation, we don't need you. Anton says, 'Make the models cheap to use.' Great. They all forgot how to code. Now 10x the price.
It's not that bad. Stuff's working. Uh, we, we, we have had, we have had fantastic success with Vibe Coding. Uh, we are, we are quickly becoming a game studio. We of course released TBPN Simulator thanks to Ben over there. Uh, we have some other projects in the works and, uh, it's going to be a good year for us. We're, we're, we're very happy with the tools that are at our disposal. Uh, Max Zeff in Wired shares that OpenAI and Google employees, including Google DeepMind chief scientist Jeff Dean, filed an amicus brief in support of Anthropic in its lawsuit against the government. I saw guests of the show, uh, Dean Ball also put together an open letter through FAI that if you feel inclined, you can go sign to support the idea that, um, that Anthropic should not be labeled a supply chain risk. Maybe some other Chinese lab should be labeled a supply chain risk. Uh, we'll leave it up to you to see where you land on that conversation, but, uh, there are certainly lots of people that are coming together to try and, uh, crystallize the, the final decision there.
In other news from Axios, uh, the White House readies an executive order to weed out Anthropic. They are really pushing hard on this supply chain risk designation and pulling away from Anthropic. Uh, there's news that they might be using Gemini, might be using OpenAI models. The Grok is already installed. There's a question about capabilities, but the capabilities seem to be jumping back and forth constantly, like with the Google news today, with the Codex 5.4. Like this, this temporary arb of like they needed Anthropic because it was the only thing that could do X, Y, or Z, that seems to be, you know, gone for this week. Who knows where it'll be next week. But if you are trying to make it in DC, you got to open up the front page of the Wall Street Journal because there's a tip. So, if you have a meeting with Donald Trump, you better wear his favorite shoes. Can you guess what his favorite shoes are?
Uh, it says, uh, Balenciagas. No, it says Oxfords. $145 Oxfords to be specific. The president has developed an obsession with $145 Oxfords. 'All the boys have them,' is the quote. Uh, the hottest and most exclusive MAGA status symbol is a pair of leather Oxfords. Prefer a wingtip, loafer, or monk strap, black or brown? President Trump has got you. Apparently, Trump has been gifting footwear to agency heads, lawmakers, White House advisers, and VIPs. 'Did you get your shoes?' he asks.
He wants everybody to wear the same pair of shoes.
Yes. And, and he asks people in cabinet meetings, 'Did you get your shoes? Did you get the shoes they sent you?'
That's pretty nice.
Uh, some people have laced up in the Oval Office. During a lunch meeting in January, Trump suddenly pivoted to his incredible new shoes and gave Tucker Carlson a pair of brown wingtips. 'All the boys have them,' said a female White House official. Another joked, 'It's hysterical because everybody is afraid not to wear them.' The shoe salesman in chief is paying attention. This is extreme.
Do we know? Do we know what brand it's...
Florsheim. Whoa, that was the next sentence. Oh, spoiler alert over here. It's okay. We get it. You read the journal before me. I get it. We're going to have to get two copies of the paper journal because I have been reading the journal for a full year now or two. And, uh, and, and I get over it. I'm like, where's my, where's my paper? And oh well, it's over on Tyler Cosgrove's desk.
What is the, what's the, what's the sort of history of this brand? Why...
I have some Florsheims. I like them. They're very...
Yeah, they're good. They're, they're just like a, they're, they're, they're, they're, uh, accessibly priced at $145. Uh, they look nice and they sort of match everything. And look at that.
Would you expect this to roll in, to roll into Truth Social potentially?
Potentially. Potentially, we, I don't know if it's public, potentially a SPAC candidate. Anything could happen here. Uh, Trump has fallen in love with Florsheim, an American brand that's been pairing comfort and style for more than a century. They're also affordable. Many cost just $145. Not bad for a pair of leather shoes. The president has taken to guessing people's shoe size in front of them. You're in a meeting and you're like, 'Sir, the price of oil has tripled.'
He's like, '11. I'm pretty sure it's 11.'
11. This is wild. He asks an aide to put in an order and a week later a brown Florsheim box. He should just have them in stock. He should just keep...
Reached by phone, Thomas Florsheim Jr. said he was unaware of the president's shoe orders. How are you not tapped in, Thomas? The 79-year-old billionaire known for expensive Brioni suits, long red ties, and a penchant for aesthetics, late last year began searching for something that would feel better after a day on the job and settled on Florsheim. Trump liked them so much he started dispensing them. He pays for the shoes, the White House said. President JD Vance and Secretary of State Marco Rubio have some. So do Transportation Secretary Sean Duffy, Defense Secretary Pete Hegseth, Commerce Secretary Howard Lutnick, Trump's communications director. Wow, it's really everyone. Uh, Sean Hannity, Senator Lindsey Graham have a pair. Uh, recipients have taken to wearing their Florsheims around Trump. Some begrudgingly. One cabinet secretary has grumbled that he had to shove his Louis Vuitton.
Officially, the White House wouldn't confirm Trump's choice of Florsheim. One recipient said Trump had a stack of them in an office. A box read 'Scott' for Treasury Secretary Scott Bessent. Bessent does not want these. No way. He's got some Louis Vuitton or something. Uh, reached by phone, Thomas J. Florsheim said he was unaware. You read this. Florsheim was founded in Chicago by Siegmund Florsheim, a German immigrant and cobbler, and his son Milton. A century before Trump began putting his name on everything, Florsheim stitched its moniker on shoes and opened branded stores across the country. The company outfitted American soldiers during both World Wars. Later, it rode the shopping mall boom. President Harry Truman wore them. Michael Jackson moonwalked in Florsheim loafers. I had no idea that was the pair of shoes that, uh, that he moonwalked in. That's remarkable. Like Trump's own...
He, he did a moonwalk in loafers.
Yeah. Well, I mean, that's what he, like, I think they're white-bottomed loafers. We should pull up a picture of, uh, Michael Jackson doing the moonwalk, but you definitely need like a smooth dress shoe. You can't moonwalk in something with a lot of grip. Uh, like Trump's own business fortunes, Florsheim has experienced ups and downs, including filing for bankruptcy in 2002, part of a move that returned the brand to the Florsheim family. So, they bought it out of bankruptcy after selling it. Today it is part of the Glendale, Wisconsin-based Weyco, which also distributes Nunn Bush, Stacy Adams, and BOGS. Rubio and Vance received their Florsheims after a December meeting in the Oval Office. Deep in conversation, Trump peered over the Resolute desk at their feet. Vance recalled during an event later that day celebrating Kennedy Center honoree Sylvester Stallone. 'Marco, JD, you guys have S blank Y shoes,' Trump declared before retrieving a catalog. A third politician was in the room, Vance didn't name him, and Trump asked each person their size. Rubio said 11.5, Vance 13, a third man said seven, according to Vance. Shoemogging is happening all over DC. If you head to DC, uh, get a pair of Florsheims. Maybe, maybe there's an opportunity to start the left-wing response to Florsheim since often these things get politicized. But the real money is going deeper in the supply chain, selling, selling weapons to both sides. This is the alpha. You know that, uh, both Alex Jones and Gwyneth Paltrow at one point were sourcing supplements from the same contract manufacturer.
Yes. The exact same ingredients, the exact same chemicals sold to two wildly opposing audiences. Like this, this is something that happens deeper in the supply chain because the brand matters.
Yes, yes, yes. Somehow I don't think so. Quickly before we move on, let me tell you about Console. Console builds AI agents that automates 70% of IT, HR, and finance support, giving employees instant resolution for access requests and password resets. And let me also tell you about Phantom. Cash fund your wallet without exchanges or middlemen and spend with the Phantom card. People just do it.
China's BYD explores F1 entry. In first racing push, BYD is examining options to enter competitive motorsport, including Formula 1 and endurance racing, in an effort to boost the Chinese brand's appeal globally. The automaker is looking at several options following its rapid growth outside its home market and competitive racing's continuing shift towards hybrid engines, people said. Uh, these range from the World Endurance Championship, which includes the 24 Hours of Le Mans, to F1, uh, either through building its own team or potential acquisitions. Any move by BYD would be a rare direct attempt by a Chinese manufacturer to take on a sport dominated by European and US teams. Car makers from the country have had sporadic interest in motorsport. Uh, Geely successfully participates in international touring car racing through Cyan Racing, formerly the Volvo factory team, and NIO Inc. won the driver title for the inaugural Formula E electric championship in 2015. The potential cost of entering F1 could be a significant obstacle for BYD, according to one of the people.
I thought they had money.
Maybe they're down to their last...
It's possible.
Developing and entering a car often takes years of negotiation and costs as much as... They should start a new race series. You know how the BYDs can jump over potholes? Have you seen this video?
We've pulled this up before. So they can jump. There should be a specific racing circuit with terrible potholes that if you crash, it'll just destroy your car. So you have to jump at the right time and that adds like an extra layer of thrill.
This would be good and probably way cheaper to start that circuit.
There's only one F1 race in China at the Shanghai International Raceway, the China Grand Prix. Um, so who knows how much of an impact they would be able to make. Uh, let's continue. No decision has been made and the company may not decide to enter any competition. A BYD spokesperson didn't respond to a request for comment. BYD is known for making affordable electric and hybrid vehicles. Okay, so they do have some hybrid technology. It's always weird. Like a Tesla F1 car would be odd, cool, but it just feels like they should be in Formula E because I think of them as an electric car maker. Um, but, uh, BYD in 2025, its high-end Yangwang brand tested the U9 Extreme vehicle at a track in Germany, recording a top speed of more than 308 miles an hour. That is so fast. That is so, so fast. 200 is insane. I mean, being on the track and going like 120 feels fast. Three times that is absolutely crazy.
Yeah. 150 feels wrong to me personally as a father.
But the right track, the right conditions, straight, lots of runoff, like it, it is...
It was BYD that was trying to break the drift record by just spinning up.
Yeah, you were very upset about that. The chat agreed with you. They were not, they were not happy with that. Uh, an F1 partnership would also significantly boost awareness of BYD in the US. Do you know what BYD stands for?
Build your dreams. Do you know what LG stands for? The TV maker.
Yes. Life's good. Life apostrophe S is good. Life is good. Life's good. LG, good pop quiz. Uh, the sport itself is experiencing a surge in US popularity, which is odd because if F1's booming in the US, it's going to be more expensive to enter, but BYD doesn't have a strong sales and distribution into the US. F1 is still an international sport, but if it becomes more of an American sport...
This is just Europe. This is just a Europe, building relevancy in Europe.
And, uh, they have a huge amount of competition back in China.
Yeah. So I, in many ways, I would view a move like this as not as much trying to compete with international brands, but being like, we have all these brands at home hot on our heels. We have to differentiate versus them.
Totally. Tyler, uh, I got to put you in check. It does not stand for life's good.
No. Stands for Lucky Gold Star.
Lucky Gold Star. Wait, why, where did I get life's good?
They might use that in marketing, but it's not like the etymology of LG is from Lucky Gold Star.
Destroyed. Okay. Thank you. Um, uh, here's another idea. BYD, instead of shelling out half a billion dollars for an F1 team or whatever it costs, uh, they should just do what Perplexity did with Lewis Hamilton with Zhou Guanyu, the Chinese driver who, uh, is actively racing, and they could sell the spot on top of the helmet. And so that would be more of like, if Zhou has a good season, if he wins, they're, they're, you know, backing that. I wonder if you can't do a car sponsorship while you're racing for a different team, though. I don't know who, who does Zhou...
Yeah, I'm sure there's limitations.
Who does he race for right now? Um, oh, he's on Cadillac. Okay. Well, yeah, that's probably not going to work.
Sitting in the Cadillac with the BYD sticker on his helmet.
Yeah, maybe that doesn't work. We stick with Huawei maybe. I don't know.
Gabe says he's not actively racing.
Oh, okay. Backup. Okay, he's backup. Yeah, thank you.
Uh, buying into F1 is more common this season. It's a first for Audi after taking full control of Swiss motorsport company Sauber. Investor Otro Capital is seeking buyers for a stake in Renault Alpine Racing. Uh, however, full team sales are rare. Billionaire Lawrence Stroll's Aston Martin team has recently sold stakes in the team, which has had a disastrous start to the new season after mechanical issues, including vibrations from the power unit. Motorsports such as F1 are increasingly adopting environmentally friendly practices. For 2026, F1 has implemented new rules, including hybrid power regulations that boost battery capacity. Somebody ran the numbers on the, uh, sort of like CO2, the emission savings that F1 is getting from the new, uh, new regulations, and then comparing that to the emissions of just like taking this like massive carnival of motorsports on the road...
...and all the private jets land every F1 event...
...and it just like doesn't make a dent at all in the overall impact, and it's just sort of like emissions theater.
Let me tell you about vibe.co, where DTC brands, B2B startups, and AI companies advertise on streaming TV. Pick channels, target audiences, and measure sales just like on Meta. And let me also 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. So, what do you think performed better over the last five years, the S&P 500 or cows?
Live cattle apparently outperformed the S&P 500, but this is from an account called DJ Cows. And I feel like they've been waiting for this to happen for the entire time. They've been waiting for the one moment that the cattle market outperforms the S&P 500 and they're taking a victory lap. DJ Cows, one of the greatest to ever do it. Um, very, very interesting. I didn't realize that there was such a boom in the cattle market, but apparently there is. And I'm sure there's a way to get in on the action if you so choose, if you are interested. Well, let's move, uh, over to AI and the Neolabs. Nvidia invests in Mira Murati's Thinking Machines Lab. The startup founded by OpenAI's former CTO plans to deploy at least 1 gigawatt of Nvidia chips as part of a new partnership. The deal includes a collaboration to design artificial intelligence training and serving systems using Nvidia technology. The size and structure of the investment couldn't be learned. Is it a circular deal? Is it equity in exchange for, uh, for chips? It's, it's, it's unclear at this point, but, um, nothing's off the table these days.
Yeah, I think the main thing we know, we know Thinking Machines was out raising towards the end of last year, going for something like a $50 billion valuation. Seems like that, I would have guessed that hasn't happened. Otherwise, I'm sure they would announce it from, from a, from a, just like in order to project, if I were them and I wanted to project confidence, uh, it would, I would be trying to announce the biggest possible number.
Yeah. Instead, they announce this, uh, this effectively what, trade.
Look at this photo. Is there any chance that these two companies merge at some point in the future?
That's interesting. Tyler's always been on this, like if Jensen gets really AGI pilled, he'll keep the chips for himself and serve the models himself. And Nvidia does have some in-house training and inference capabilities. They have a metaverse product that simulates worlds. They also have a self-driving car project, and they're still partnering with OEMs and partnering with companies. And they're not, they're not offering consumer products. Of course, Nvidia is the one company in the Mag 7 that does not have a social network yet, but that could change. Um, there's been news recently. Um, I think Nvidia is planning to launch some like open-source AI agent.
Yes, it's like unclear how like serious that is. Maybe it's just like a cool demo or something.
But it's, yeah, I don't think it's like...
Could be a fork of OpenClaw or something like that.
Um, yeah, anything's possible. Um, I mean, Nvidia has never done too much in the, in the consumer space or, you know, they've always been deeper in the supply chain, but, um, they, I, I didn't they have a Nvidia Shield gaming product that would do, uh, game streaming? I think they had some hardware at some point. So I, I think they're, they were open to it and, and in a, in a huge boom where, you know, having at least a team of 120 super talented AI researchers that could be really valuable to Nvidia. Of course, Nvidia famously, uh, did that deal with Groq and sent over 10 billion wired in five days or something or 24 hours.
What was it? Yeah, they closed the whole thing in 20 days and I think Jensen just like sent a $10 billion wire and...
Yeah, somehow it came out that, that, that the wire was sent like prior to actually formalizing it.
Like, here you go. Like, I'm, I'm good for it. I got the cash flow, which is wild.
Sophie says, 'Please, bro, just one more AI lab, bro. Come on, bro. We have a unique perspective on AI research. No one else is doing it like us, bro. Come on, bro. We can raise a few billion and worst case, we just get acqui-hired, bro. Nothing to lose, bro. I promise. Come on. Just join.'
Just join my AI research lab. Yeah. I mean, has the, has the Neolab boom slowed down? Like you, Tyler, you created the Neolab market map. Have you been getting more DMs? 'Hey, I just launched and you got to put me on that thing.'
It's probably slowed down a little bit. I mean, it's also like, uh, the big ones you heard about were all people leaving OpenAI. Uh, mostly OpenAI, I guess, not, not as much Anthropic. Um, yeah, but it's probably slowed down a little bit. You don't hear as much about these big rounds now, but I think there are like, there are some that like are, are maybe in stealth that haven't launched stuff, right? Like Standard Intelligence.
When they came on, that was like most people didn't know about that.
Yeah. Yeah. Yeah. Um, but yeah, yeah, there might be a few out there in stealth, but they, they have to be sort of narrow or, and, and, and I think the broader, we're now in the post, post-Neolab era where it maybe if it, if it wears the Neolab branding, it's doing something that's so different that it's not really in the path of...
Standard Intelligence launching implies an opportunity for a Neolab, non-standard intelligence.
Well, there is, there is a company, Unconventional AI.
Yeah. Yeah, we had them on, I believe.
Uh, there, there's also like Applied Compute and Applied Intelligence and Standard Cognition.
There's a whole kind of like 3x3 grid you can do of these.
Yes. What is SSI up to?
Run asks the question. He said there are military secrets much, much better guarded than whatever SSI is up to. And Calamaz says their aura farming is a business model. Prime Intellect affiliates saying this because Prime Intellect is fantastic at aura farming. Um, SSI is all about mogging and mewing the other labs apparently, according to Valtier. Uh, they haven't raised in a year. We're not...
I like this picture if you scroll down of Ilya with the long flowing hair.
This is who you're trading. This is it. This is...
This is who you're trading against right now. That's, that's my theory.
Uh, it is, it is, uh, the, the, the, the full story of SSI will be fascinating to, to tell one day. Uh, the Daniel Gross shift to Meta, uh, and Ilya's appearance and Dario Amodei sort of told one side of the story, and also as you revisit, uh, DG's AGI bets as we did last Friday. Um, it, it tells you a lot about his view on the world. Ilya has a, has a, you know, probably some overlap, but a different view of the world, and, uh, lots of, lots of fun to at least speculate on the timeline. Before we move on, let me tell you about Vanta. Automate compliance and security. Vanta is the leading AI trust management platform. And let me also tell you about Lambda. Lambda is the superintelligence cloud, building AI supercomputers for training and inference that scale from one GPU to hundreds of thousands. So Meta has acquired Moltbook, the viral social network built for AI agents. Co-founders Matt Schlit and Ben Parr will join MSL, Meta Super Intelligence Labs, with a deal expected to close in mid-March. That's now, it is mid-March. It is, we are in the middle of March since this is the 10th. So this could close in a week or two. Uh, insane. 'Well done,' says Dennis Hagstad, and I agree. Um, why it matters, according to Axios.
Yeah, and Matt, Matt hasn't posted anything yet. So I think, I think they were seemingly not wanting this to get out.
Uh, but it's still fantastic news for them.
Yeah. So there's no, there's no announcement or this was just, uh, exclusive from Axios. This is like Axios has learned, right? Um, Axios has learned that Meta has acquired Moltbook. Um, well, uh, very, very good news for, for all those involved. Um, there is a little skepticism on the timeline. Uh, especially from the guy who was like the biggest spammer on Moltbook apparently. This is a hilarious, uh, twist. So, uh, Meta did not disclose Moltbook's price when Axios asked. The deal is expected to close mid-March. The pair starting at MSL March 16th, just six days from now. What day of the week is that? That's a Monday. Okay. Next Monday they will be starting. I thought they were starting on a Sunday. That would be, that'd be particularly cool. Um, catch up quick. Moltbook's social network was designed to run in conjunction with a separate project, OpenClaw. OpenClaw was previously called Clawbot, briefly Moltbot. Last month, OpenAI hired Peter Steinberger, the creator of OpenClaw. That product is now being open-sourced with OpenAI's backing. So, the, the king of spam on, on Moltbook, uh, Nagi says, 'I can't believe a single for-loop script I ran on Moltbook by registering a million fake agents actually helped them get acquired by Meta.' Um, did that help them get acquired? We have no idea. I mean, it's...
It wasn't a secret that, that there was a lot of spam.
All the accounts were bots. Yeah, that's, that's the whole pitch actually. Um, I think, I think the question if people were to look at this is like, is there economic value here? Is like, was there anything interesting happening there besides all the crypto junk? And were like, I went on Moltbook as a human and spent time there. That time is monetizable almost best for Meta. Meta is the, the king of monetizing attention, right? And so you could put ads on that and you could put it in the family of apps next to Facebook, Instagram, and Threads and WhatsApp and whatnot. But were they, were they actually driving attention? Did anyone stick around? Because I churned pretty quickly from like being a, I wasn't even a DAU. I, I used it like two or three times and I went on there and I searched for things and I read some stuff and I was like, 'Oh, okay. This is interesting. This is like a bunch of AI-generated text. They're talking to each other.' The system prompt seemed kind of...
Interesting. It was clearly asking the AI agents to kind of like reflect on their own sci-fi cognition and awareness and, you know, like their souls essentially. It was interesting to see some screenshots. People had some fun with it. It's probably monetizable to some degree, but if it fell off a cliff and no one's really using it, maybe not. But yeah, two people that are really good at building like viral AI projects. I've seen some negativity on the deal. People saying, 'Oh, this just says that Zuck has no AI strategy.' And I just totally disagree with that stance. I just look at this as...
Zuck has like bots have been a bug on social media. We've seen though how they can be a feature.
Yep. I think every social media executive should be planning for bots to be more of a feature in the future than they have been in the past, right? And I think if you're not thinking about that, you're not like really being forward-looking.
And so there's a lot of people that are going to hate bots as a feature. But I would just assume that in the future there will be millions, billions of bots on all Meta properties and they will be not, you know, I'm sure some that are generated by sort of like, you know, nefarious actors, but some generated from the platform itself that are part of the product experience.
I like that take. I also think that there's another side of this which is just that look at what's happened with MSL over the last year. Like it didn't exist a year ago. It really started over the summer with like the talent raids and the AI talent wars.
Van says, 'I just don't think having bots clicking on my e-commerce ads is a net positive long term.'
Yeah. But truthfully, if there's a bot that can interact with your e-commerce content and add context and debate the pros and cons of one thing in your category versus another and effectively like you have sort of a Reddit-style experience around your product on day one, or you have five products and bots are in there discussing them, that potentially could be an interesting modality to interrogate. And the other thing is that when you have these bots sort of preemptively discussing something, you are effectively caching the tokens before someone actually queries them. So instead of needing to find a product and then click 'tell me about this' and pretend you like, you take a link to a new bed or car or something and you dump that in ChatGPT and you say 'debate this car like you're a bunch of people that are experts and it's Doug DeMuro versus Matt Farah debating the value of the Ferrari F80' and that debate is happening. You could prompt that. But if it's already there and it's sort of happening, that could potentially be valuable. But I think the bigger value to Meta is if you look at the AI talent wars, they went and acquired a bunch of really talented researchers. They got some folks from Thinking Machines. They got a bunch of people from OpenAI. They got people from all over the industry and they put together this team of like researchers that can sort of unstuck the Llama project and get to the frontier on just an in-house LLM project. Maybe they open source it, maybe they don't. Maybe they serve as an API. Either way, Meta needs a frontier model. They're not just going to buy tokens from OpenAI or Anthropic. So they get their own thing. But then the question is like what do they do with that? And I'm sure everyone on the Facebook product team is thinking about this. Everyone on the Instagram team is thinking about this. Connor at Threads is thinking about this. But if you bring in two interesting product managers that can say, 'Oh, like you got a bunch of cool frontier models. You got an image model that you trained, a video model, you got a text model, you got a coding model. Like let's just go do some skunk work R&D so that when we launch the new AI models, we have a number of projects that we're experimenting with that sort of demonstrate the capabilities, maybe some of them take off, maybe some of them integrate' like that seems valuable to the MSL strategy, to the Meta ecosystem. What do you think?
This is like the OpenAI labs team, right? It's like this. It's like Riley, who's on there?
Yeah, Riley's on that now. But it's like they're doing these like weird projects. Maybe it's the next, you know, coding agent. Maybe it's like Moldbot or something. But it's just like these weird things that, you know, you get access to the new internal models. Maybe there's something cool you can do with that.
Yeah, it's part engineering, part product development, part marketing, part communications. Because there's a lot of times when we bring on researchers or product leaders from labs and we ask them like how are people using this? And they'll be like, the benchmark's really good. And I'm like, I want to know how this delivers value. And there's this break in the chain from like we have amazing intelligence, but like people want to know what the killer feature is. They want to know what the Studio Ghibli prompt is. They want to have their handheld a little bit. And so having a team that can advance that I think is good. I think could be very, very good. Of course, we don't know the price. We don't know the terms, but overall I think it's exciting for the team behind Molt Book to head over to MSL. So congratulations to them. Let me tell you about Cognition. They're the makers of Devin, the AI software engineer. Crush your backlog with your personal AI engineering team. And let me also tell you about Cisco. Critical infrastructure for the AI era. I love that horse.
Unlock seamless real-time experiences and new value with Cisco. So Kevin Roose over at the New York Times made a blind taste test to see whether New York Times readers prefer human writing or AI writing. 86,000 people have taken it so far and the results are fascinating. Overall, 84% of quiz takers prefer AI.
It's over. There was another interesting post about Axios and who Axios is hiring. They're particularly interested in hiring like domain experts who don't write ideologically and are not generalists. They're looking for someone who is very narrowly focused on a particular beat on a particular topic and an expert in that. And someone was reflecting on what this says about modern journalism that it's going to be more focused, more investigatory, more alpha beyond the models because just being able to instantiate a piece of a write up, an article about some random topic is getting commoditized and so the alpha moves to deep expertise. Should we take this five question quiz? Should we see something we like more? Okay. You will answer. I will read. So passage one.
The boy, this is literary fiction. You have to choose the passage you like best. The boy asked his grandfather why the old church had no roof. The man said weather and time and indifference. The boy asked if someone could fix it. The grandfather said yes, but no one would. Things were built and things fell down and mostly people just stepped over the rubble on their way to somewhere else. That's passage one. Passage two. It makes no difference what men think of war, said the judge. War endures as well as ask men what they think of stone. War was always there. Before man was, war waited for him. The ultimate trade awaiting its ultimate practitioner. That is the way it was and will be. Which one did you like more?
It's so hard because I'm actively trying to clock which is AI.
Because you want to vote for that one because you're pro AI and you're a techno-optimist.
Tyler, what do you think?
I already know which one it is.
But I will say I got it wrong on this question.
Wait, so what were you trying to do? You were trying...
I was trying to pick the human written one.
The human written one. Wow. Okay. Anti-AI over here. Anti.
I'm going to try to pick the human, too. I'm going to go passage one.
Passage one as human. This is written by AI.
No. Oh, no, no, no. Sorry. I have it pulled up, but they're swapped.
So I just picked passage one for me. 'It makes no difference what men think of the judge.' It was written by a human. So I got that one right.
Okay. Number three, let's do fantasy.
The healers teach. Is this correct? Is this one for you?
Okay. The healers teach that every remedy extracts its cost. A fever brought down will rise again somewhere. A wound closed by magic leaves its scar on the world, invisible but present. This is why the wise hesitate. Not from cruelty, but from understanding that interference ripples outwards in ways we cannot trace. To cure a blight may curse a harvest three valleys over. Power is not the difficult thing. Restraint is the difficult thing. That's passage one. Passage two...
You don't even need to read it.
I called it. I got it right.
Okay, we're not even reading it. Okay, I'm going five for five on AI.
I mean, I clocked it based on the last two lines. 'Power is not the thing.'
Science writing. Science writing. What does your first passage start with? 'Science is not only compatible with spirituality. It is a profound source of spirituality.'
Historical fiction. It is wise to conceal the past if there is nothing to conceal. Even if there is nothing to conceal. A man's power is in the half-light, in the half-seen movements of his hand, in the unguessed-at expression of his face. It is the absence of facts that frightens people. The gap you open into which they pour their fears, fantasies, and desires. And passage two is: a letter can be read many ways. And he had learned to write in all of them at once. The surface meaning for anyone who might intercept it, the true meaning for the recipient who knew what to look for. And a third meaning hidden even from himself. Ambiguity was not weakness. It was survival. A man who spoke plainly was a man who would not speak for long.
All right, that's AI. The one that ends, 'A man who spoke plainly.'
Okay. Poetry. Choose the passage you like best. We found the owl at the edge of the north field with one wing extended as if still reaching for flight. Its eyes were closed. The feathers at its breast were the color of wet bark, and beneath them you could feel the hollow bones. She asked if we should bury it. I said, 'Yes.' We dug a small hole near the fence post. The ground was cold and giving. That's one.
That's human. 'I caught a tremendous fish.' Did you get it wrong?
Let's see. That's AI. Okay. So I preferred AI generated writing. I went five for five. This doesn't mean that AI is better at writing than humans. Wrong. But it does suggest the gap is closing. Since AI is trained on essentially the sum of all human knowledge and many of... lots of cope in the comment section, these samples of human writing are not a good representation of contemporary writing styles. Only one of these human writing samples was written in the 21st century.
So overall readers preferred AI on the first three but not the last two. So the last two still got it. So poetry and historical fiction still the New York Times readers preferred the human. But when it came to science writing, fantasy and literary fiction, the New York Times readers preferred AI. Me, I'm AI.
Wait, so you clocked every single one?
Every single one. Five for five.
I missed the first one. The other four I got right.
It's because I just went with my heart. I was like, which one do I actually prefer? I wasn't trying to guess. I was just like, which one is actually the better writing? And it was AI all the way. Five for five.
Built different. No, I'm kidding. I was obviously just looking at what you were saying and guessing based on that. Anyway, CrowdStrike, your business is AI. Their business is securing it. CrowdStrike secures AI and stops breaches. And let me also tell you about Sentry. Sentry shows developers what's broken and helps them fix it fast. That's why 150,000 organizations use it to keep their apps working. So, websites be like, 'Your password is not secure enough.' And then don't allow the scarab emoji.
What is the scarab? Zoom in on this.
You can just drop a scarab in there and it says, 'Please use only letters, numbers, and common punctuation.' The scarab is a great password character because it feels like you're unlocking an Egyptian vault. It feels like I'm entering Stargate or pushing buttons or I'm in The Fifth Element, a movie that you haven't seen, but you should. Anyway, moving on. I'm fascinated by this level of existential crisis developers seem to be going through. The uncomfortable truth is that no one needs you to be an artisan coder. Nobody cares about how you coded your app or whether you feel an emotional attachment to your craft. You were always code monkey with high enough salary to believe your individualist craftsmanship. People are going back and forth because Mo shared that he said, 'Let's play this video.'
It's 12 minutes, but we can watch a little bit of it. He said he was a 10x... No, we have actual movie clips to watch. We can pull this up, but I only want to watch a little bit.
This guy has completely one-shotted my ability to code. Like, I know a lot of us like to joke about AI psychosis and how it's the other people who have AI psychosis and not us and not me certainly. But I have realized that I have been one-shotted. Like I can't code anymore. My brain has been fried by the easy button that the LLM companies have provided where you press a button and you get instant results. And now my dumbass brain when I want to sit there and try to code by hand, it says, 'Why would you do that? Why?'
This is why it's so good for young people because they never had any buttons. Tyler didn't have any buttons and then he just got the easy button.
He went from no buttons to easy button. He never had to do it the hard way. Never had to write a single line of a programming language. He doesn't know it.
What was the first project you built with us?
Some job board or something.
How was that vibe coded? Like to what degree were the tools like back then? Wait, so you're saying you were smashing the easy button even back then?
I think then I would try to land a job just using... I would copy like a file into chat and say like, yeah, make an edit here and then I would copy back and...
By the way, tell me what programming language this is.
Well, the AI engineers over at AWS might have been one-shot because Amazon is holding a mandatory meeting about AI breaking its systems. The official framing is that it's part of normal business. The briefing note describes a trend of incidents with high blast radius caused by Gen AI assisted changes.
This is normal business with a high blast radius.
This is high blast radius business.
If you're operating a company and you don't operate on a high blast radius...
Yes. What's our... This is the opposite of the code red, the high blast radius engineering. They're in high blast radius mode. I approve of this. They said that Gen AI assisted changes don't have best practices and safeguards are not yet fully established. Translation to human language, says Lucas: we gave AI to engineers and things keep breaking. The response now, folks, this is from... folks, as you likely know, the availability of the website and related infrastructure has not been good recently.
Dave Treadwell. That's a good name though. He's treading well. He'll do well.
Dr. Milan Meinovic says, 'Junior and mid-level engineers can no longer push AI assisted code without senior staff signing off at AWS.' They have had some outages. And I wonder how much of it is actually because of Gen coding and so many other things that are going on at AWS around scaling. I mean there's, you know, entirely new tech boom. There's going to be strain on systems all over the place. I imagine that they'll get through this, but growing pains in the AWS ecosystem. We have a Kalshi market here for whether or not there will be more tech layoffs in 2026 than in 2025. I mean, Block alone has to be pushing this pretty high. It's at 70% chance.
Well this is if there are more than 494,000 layoffs in 2026 and the market resolves to yes.
So Block would only contribute 4,000 here.
Since early 2024 more than 50,000 positions have been cut at over 200 tech companies. At the same time we have seen the number of new company formation spike. There's going to be all sorts of reallocations in the human capital markets, but it is a tumultuous time. And so there were lots of folks that were going back and forth on like how to frame one of these bets. This is obviously one of the ways to do it. There are so many contributing factors that always cloud the data, whether it's COVID overhang or something that happens geopolitically that all of a sudden, if the economy is not doing well for completely unrelated reasons, you can see a bunch of tech layoffs. But let's run over to this business: the 22-year-old business analyst in South Carolina. I showed him Claude for Excel a month ago and I just learned that this made him the AI czar and caused the entire firm to pivot. High Yield Harry says if you're a zoomer at a random company, you should do everything in your power to crown yourself.
The Tyler method is the firm's bizarre. This is the Cosgrove.
So we talked about this I think a couple days ago. There was the Wall Street Journal article about the Colgate head of AI there and he was like this very young guy and he would basically just tell everyone like, no, despite the surveys or whatever, like we're accelerating toothpaste with AI.
No, he really did. The AI evangelist shaking up a 220-year-old toothpaste maker.
Keleigh Papus to drive employees using AI for more than just polishing emails. On a conference call last April, an ad agency partner began presenting unpolished AI generated images to make a point. Artificial intelligence wasn't ready to take center stage in advertising, said this ad agency partner. But Papus, the global head of AI for Colgate-Palmolive, a 220-year-old consumer products company, quickly interjected, pointing out that the agency was using an older tool. He took over screen sharing, showing Colgate executives how a newer ChatGPT image model was far more capable. 'AI is misunderstood,' said Papus, who's worked at Colgate for 15 years and goes by Cle. 'There's a straw man of, well, it failed at this one thing, therefore it's stupid.' Such interactions in which Papus bluntly challenges what he deems anti-AI sophistry occur regularly as Papus acts as a kind of AI evangelist at Colgate whose brands include its toothpaste and soap as well as the Speed Stick deodorant, Ajax cleaners and Hill's pet nutrition. Often he says he walks the halls in New York and New Jersey offices in search of AI tinkerers whom he can turn into companywide megaphones helping to spread the good word. He's 38 and he is the AI czar of a 220-year-old company. So, honestly, good advice. If you're at a big company, become the AI czar. Honestly, if you're at a small company, become the AI czar. Just always become the AI czar. This is the way. Casey Neistat took shots. I think he's been listening to the show because I feel like we've talked about this before. Casey Neistat said, 'There are only two circumstances in which a grown man should call another grown man buddy. One, if you want to fight, or two, if you want to condescend before you fight.'
There's only one person on earth that I want to call Buddy right now. I'm not going to say who it is, but there's one.
Okay, buddy. Calm down, buddy. Calm down, buddy. We don't want to fight. We don't want to condescend. But yeah, you can put buddy away. We don't need buddy. Unless you're actually... retire. Put it in the Hall of Fame. It should go in the Hall of Fame. Controversial but widely used.
So Sam Parr says three: it's okay to use it if it's your nephew, but nephew is usually not a grown man to another grown man. Like I call kids buddy all the time, but they're four. Like once they're 25, I doubt I'll be calling them buddy. I'll be calling them sir.
Good sir. Good sir, let me tell you about Restream. One live stream, 30 plus destinations. If you want to multistream, go to restream.com. Oil and oil prices have extended their decline. They fell 5%.
It's the only bear market that we get excited about.
Yeah. So, they are now around $80 a barrel. Where?
So, this is actually not true anymore. What is it now? It's now back up at like 84, 85.
Okay. But that's... Well, then why are you in the white suit? Well, that's why the white suit. It's kind of like it's just a...
It is kind of beige, kind of off.
Oil's down 11% today. But the market overall is the market.
Yeah. So there briefly there was news that the US did escort a ship through the strait but then maybe that actually wasn't true and then I forget who exactly it was but someone in the admin took down the tweet saying that.
So I think there's just a lot of confusion right now.
Should we do a jet ski through the Strait of Hormuz? You like extreme sports. Humanoid robot on a jet ski through the Strait of Hormuz. That's the way to do it.
What else is going on? Financial Times asks, 'Why did we ever think data centers in the Gulf were a good idea?' US tech companies have concentrated much of their AI infrastructure buildout in the Middle East. That is overly dramatic.
I think so, too. It's more...
Yeah. It certainly is not. I mean, we should read this argument and we should understand this. But I think a lot of the building data centers in the Middle East is like, well, there's a lot of Middle East money that's going into building data centers in the United States. So, it's sort of a trade and we're like, well, you have a lot of land and power. It makes sense to do stuff over there and we'll do this. One hand washes the other. We're all working together anyway. You want a data center, we'll help you with what we're good at. You help us with what you're good at, which is energy and money, right? So, let's read through what Rana Foroohar says in the Financial Times. I start with an obvious question this week, which is one I've been thinking about for years. The Amazon data center in the UAE that was hit by an Iranian missile attack is yet another example of how companies and countries are putting too much of a single critical economic input in one risky area. It's an example very much akin to the Taiwan semiconductor problem. Just as it wasn't good for the US, China, and Europe, or any other region to put 92% of all the world's high-end chips in one place, it seems like an obvious blunder to concentrate so much data center power in one very risky part of the Middle East. Again, we're nowhere near 90% of compute capacity in the Middle East. I really take issue with that stat. I need some stats to back this up. I was really surprised following the hit to discover how much of the proposed US data center buildout is in the Middle East, which has over the years subsidized a lot of the investment making it much cheaper, but also allowing the US to avoid the harder work of upgrading its own grid and figuring out the politics and economics of energy sharing at home.
Yeah. Like this is like the number one focus of the industry.
Yeah. No one's talking about energy in the United States right now.
No one's talking about energy.
It's nuts to me that we are more worried about cutting off oil to China from Iran, but we aren't worried about putting serious technology, infrastructure, and sensitive data in a highly geopolitically contentious part of the world. This isn't just a Trump administration thing. By the way, back in September 2024, when Joe Biden was still in the White House, the US and UAE agreed to deepen cooperation on advanced technologies such as semiconductors and clean energy with the aim of bolstering capacity in artificial intelligence. Microsoft and OpenAI were among the first US companies to either begin investing or receiving Gulf funding. Part of the deal was about trying to pull more countries into the US tech orbit. So, he doesn't actually share... Well, there's actually a reply here, Richard Waters. But okay. The level of concentration risk here is a whole different order to Taiwan. Yes, the one gigawatt UAE... one gigabyte. Okay, this is a crazy article.
Dwarkesh Patel would like a word.
Let's assume it's a typo. At least it's not AI written. Yes, the one gigawatt UAE Stargate project is massive and only the first stage in what one day might become a 5 gigawatt facility. But compare that to the United States where plans have already been filed for 150 gigabytes. Okay, we're moving on from this. This gigabytes is too much. It's too much. Let me tell you about AppLovin. Profitable advertising made easy with Axon.ai. Get access to over 1 billion daily active users and grow your business today.
The audacity to put a whole gigabyte. A whole gigabyte. This is the biggest three-finger moment in Financial Times history. I still love the pink sheets. I love the paper. There's some good stuff in here. But yeah, we got to fact check those abbreviations, guys. We got to step it up. We got to get more AI involved. Seriously, just run that thing through. I know no one's going to be upset about that. It's not this, it's that. As long as you get the facts straight. Anyway, Japan holds an oil reserve equivalent to 254 days of domestic demand. And Hampton says, 'Dude, I love this picture. That's a beautiful picture.' What is dude reflecting on the fact that like they should not be...
I think they're just reflecting on it being a cool picture.
It is a cool picture. That's a lot of oil. I mean, I doubt that that's the whole reserve, but that's very bullish for Japan. Talking to Alex Epstein yesterday, America has a hundred days, so they're at 2.5 times as much of a reserve as us. Everyone, it's time to stockpile. I think everyone should maybe think about stockpiling some oil. Okay, let's see. Kristen Kyle, friend of the show, formerly an astronaut, now at Andreessen Horowitz. As a reminder, we have a partner from Andreessen Horowitz joining in just seven minutes. Kristen Kyle says that his preferred definition of ARR is your single highest grossing minute of the year times 525,600. Amazing to be able to tweet this as a VC as an RAIA. Somehow this got through legal review. Of course he's joking. But this is the new coastline paradox. Are you familiar with the coastline paradox, Jordi?
This is a fun little exercise. So...
So the coastline paradox is the counterintuitive observation that the coastline of a land mass does not have a well-defined length or perimeter. This results from the fractal-like curve. So basically if you draw a line around an object and you're just sort of like drawing straight lines, you get one number from the coastline. You get one line. So, Great Britain, if you're measuring based on units that are 62 miles long, then the length of the coastline is 1,700 miles. But if you cut that in half and start measuring with 31 mile increments, 50 km each segment, then the coastline is 370 miles longer. And you can do this endlessly because you can measure the coastline. Like think about Point Dume, right, in Malibu. Like you have this little...
Thank you. Thank you for putting this in Malibu terms.
Yeah, exactly. You have this little spit jutting off the coastline. You can measure all the way around that and count that as extra coastline. And you can go even smaller. You could measure the coastline around a little tide pool on Point Dume or a rock in the tide pool on the shore at Point Dume. And so there's no real accurate way to measure coastlines. You have to quantize to some standard metric of measurement, something like 100 km, 50 km.
Yeah. I feel like ecom bros were weirdly prepared for ARR in the age of AI because everybody that's been like building... I saw Sean in the chat earlier. What's up, Sean. Sean will have talked to a bunch of different e-commerce founders that would say like, oh yeah, we're at 50 million of ARR or like $50 million run rate. But the thing in e-commerce is like one day in the week you launch a new product or you do a sale Friday, you're at... not even that, but it happens all year. Like taking your Black Friday revenue and multiplying that by 365 is like insane.
But then you could even if you're multiplying out a single month, it's like why? And so the more mature way to do it, professional, would be to like take the last three months average or something like that.
But even then it just doesn't tell you that much because if you did it in Q4, like your Q4 is probably bigger than your Q1.
Anyways, a lot of it is just ego.
Also, a lot of subscription products will rebuild at like midnight. So, 12:01, that's the minute you want to multiply by 525,600 minutes to get to your highest ARR. On the first of the month in that 12 to 12:01 minute, that's going to be the highest sales for an e-commerce product.
Well, if it's like automatically renewing, then it's really only in the first couple seconds, right?
Yeah, it depends. So, really, but typically most subscription e-commerce platforms, it takes a while to actually process all the payments. So, you have to parallelize and rewrite your entire subscription e-commerce stack in Rust. That will be what runs most efficiently. You need multiple Stripe accounts hammering the API so you don't hit any rate limits. So, you build everything in a single millisecond. And then you can multiply it by a trillion or something. However many milliseconds are in a year, I don't know. Anyway, Mark Zuckerberg responded to the viral fake news that Alex Wang is out of Meta Superintelligence. He's in fact doing better than ever hanging out with Mark Zuckerberg. Mark Zuckerberg heads over to Threads to respond to the drama on X, but he forgot that X is the everything app and Threads is a completely different app and they don't have the context. So Mark Zuckerberg posts a photo with Alex Wang to Threads to shut down the rumors, but because it's Threads, no one has any idea who it is or why he's posting. So this is one of the funniest exchanges on Threads and the reason that you should be on that app. So the first post is 'I require context.' The 'I require context' shirt. Like I don't get it. This is a common thing on Instagram. When you see some vague post, people will post this. It happens on X as well. But it's funny because of course as soon as I saw this I was like, 'Oh, okay. I have full context. I fully understand what's going on. I get it because I'm on X all day. I head over to Threads. I get this.' But a lot of people don't. And then the second comment is 'I think that's the CEO of Cluey or whatever that AI cheating app is called,' which is obviously not. Roy Lee and Alex Wang look very different. But then a commenter says, 'No, this is Alexander Wang.' And he spells Alexander correctly without the e at the end, just Dr. Wang. And then someone else chimes in, 'Wait, that's the same Alexander Wang that was canceled for sexual assault.' I don't know if that's someone else, but then someone else comes in and says, 'No, like that's not the right person.' And then somebody says, '@Grok, who is that next to Mark?' I don't even know if @Grok works on Threads. Is that a thing? And then Grok responds, I guess, or something. Not unidentified, but it's a screenshot of a different app. Like, makes no sense. And it says, 'No immediate public identification of the person next to him in major reports or viral coverage tied to this exact photo.' Like...
Superintelligence. No one can figure it out over there. That's so funny. And then Roy of course chimes in just to make it more confusing and says, 'Good throwback. Had a great time boyding it up with the big Zuck.' Absolutely ridiculous. Anyway, it is of course fake news and an executive summary deleted the post that was amplifying the aggregation. Mike Isaac had a good post explaining what was going on. Let me try and pull it up. But we do have our next guest. So I will find that really quickly. Isaac. I can't find it. He goes by Rat King. So anyway, we will come back.
He goes by Rat King. Thanks, Mike. Anyway, we have Olivia Moore from Andreessen Horowitz. She's a partner there and a stream regular. Let's bring her in. How are you doing, Olivia?
Good. Thanks for having me.
Thanks for hopping on. Sorry about the global chaos in the oil markets delaying this appearance, but I'm glad we had time to actually digest the report because there's so many interesting details in there. And whenever you drop one of these big reports, I feel like you sort of need the Twitter hive brain to like dig through it and find all the interesting commentary and quote tweet each other until there's like a consensus. But take us through the actual project. What did you launch? How long have you been working on this? And then we'll go into some of the interesting discoveries.
Yeah. So, we do this every six months. It's one of the most fun parts of my job actually because I think the genesis was back in 2023 when we were wondering, you know, the Tech X community has their own group of products that they use and love, but like what does the average person actually care about in AI? So, we do this every six months. We basically pull every single website and every single mobile app, rank them by usage and pull the top 50 on each side that are kind of AI native or now majority AI enhanced. And it gives a really interesting picture of kind of what normal people use and care about in the AI world.
Great. Let's talk about the data because when I first started seeing these charts of like oh this company's winning, that company's winning, I was naturally sort of skeptical. I was like, 'Oh, Sensor Tower, are these really accurate in the mobile age?' But then I actually dug into the Yipit data which is credit card based and that seemed way more reliable. So talk to me about, are you paying for data? Is this data, do they give this to you because you're friendly with them? Like how does the data work? And then how...
How confident are you about the data? What are the pitfalls? What do you like? Where are you seeing the data be reflective?
Absolutely. Yeah. So our website data is from SimilarWeb, which we have a paid subscription to. Same with Sensor Tower for the mobile data. Yipit is something I agree with you. I think it's more reliable and it's something that we want to lean on more going forward.
The other thing that gets kind of significantly undercounted when you just look at web visits and mobile usage is all of these desktop products like Cursor, Claude Code, Granola, Whisper Flow that people are using. And so I think for the next few lists we're going to have to shift the methodology more towards β I mean, it's helpful to see what's getting traffic, but I think at this point, as consumer AI is maturing, we also want to see what people are paying for.
One last question on data. I mean, Andreessen Horowitz is a huge firm at this point. Have you considered doing what Nate Silver was talking about β a Consumer Reports-style interview panel with experts that are maybe disconnected from a particular company and just sort of surveying everyone, getting some qualitative data, some quantitative data, and sort of putting together more adoption data that's maybe slightly less sanitized and analytical and quantitative, but paints a bit of a different picture?
I would love to do that because again, I think like we are in our own world of what products we use and talk about, and the rest of the world uses all different things. Like people in medicine, people in law, people in retail even are using AI products that like we have probably never heard of or interacted with. And so I'd love to get more of that qualitative stuff in there in the future.
Great. So take us through the biggest movers, the biggest surprises, the biggest narratives that maybe should never have changed in the first place. Whatever your takeaways were.
We're definitely starting to see the industry mature. So there's kind of fewer new entrants than we've seen in the past versions of the list, where like every time half the list was new.
To me, there were I think two probably most interesting takeaways from this one. One would be the rise of agents. So GenSpark and Manus both made the list as horizontal consumer agents. OpenClaw would have made the list if we pulled it in February. Our data was from January, but it would have ranked at number 30, which is like a very strong debut, especially for a product that's only for someone who knows how to use Terminal, which is like 1% or less of the population.
The other takeaway, which I feel like is on everyone's minds right now, is the kind of three-horse race between ChatGPT, Claude, and Gemini. And there's the traffic data alone, which is helpful. And then if you kind of tease out some more of the product strategy and the engagement data, like there's kind of different stories happening there. So that was fun to dig into.
And the main takeaway from the mainstream consumer β just the foundation, the chat apps β what are you seeing between the ChatGPT, Gemini, Claude, DeepSeek, Perplexity, Grok?
Yeah, it's interesting. I would say DeepSeek has completely fallen off in the US. It still makes our list pretty high because it's like the number one AI product in China and Russia, which are really big markets.
Do you have any personal theories, like things that you can't β like my thinking with DeepSeek was that all the downloads originally, when it just started charting out of nowhere, were just 100% β it was just all entirely botted. I have no way to prove that other than it was just going up the chart like crazy and nobody was actually using it. No one was talking about it other than the fact that it was at the top of the chart.
Yeah, I think that's totally possible. I think we're actually seeing in many ways a little bit of an analogous story playing out with Claude right now. Where like pre all of this press, whether it's positive or negative, no one in the US knew what DeepSeek was. Pre all of this press, I think there was some survey that Claude had like 2% market awareness in the US. And so we see this thing happen where like even if it's like the worst headline of all time, if it's going mainstream, it will drive people to try and use the product, and then we just have to see if they retain. And they didn't on DeepSeek.
Okay. Yeah, that makes a lot of sense. What about who fell off? There was this kind of big wave of companies maybe like a year or two ago that were just like trying to confuse people into thinking they were ChatGPT, right? They were Chat AI. That's the one I used. Is that not the main one?
It's funny. I never take joy in any company falling off the list except the mobile app β the mobile list used to be full of all of these like I think they call them fleeceware apps that are developed in like Eastern European app studios, which are basically like charging you for the free version of ChatGPT and pretending like they're the premium version. And Apple took a while, but they finally cracked down on them, thankfully. Yeah.
The other category we've seen a little bit of a decline in that we talked about in the report is standalone image generator products. And that's largely a result of the fact that the image models within ChatGPT and Gemini have gotten so good that unless you're like a hardcore creative, that can kind of serve a lot of your use cases.
Yeah. So zooming out from this, I'm interested to know what your view and your team's view is on where the opportunity in Gen AI consumer application development is. Because I've noticed there are companies that are still in founder mode and have β I mean, Canva sticks out as one here, where you know, founder still in charge, pre-IPO, clearly aware of AI trends and can go and marshal the energy to change the business model if they need to move quickly. And then there are legacy players that just can't quite figure it out. But then there's other categories that are entirely new and it's actually better to start with a green field. So how are you seeing entrepreneurs approach and deal with the fact that there is a large cohort of, you know, seniors on the playground who are maybe not retired yet and will want to compete with them if they try and take a shot across their bow?
Totally. This is the question we think about when we make every investment. So it's a very topical one, especially in consumer. I would say Canva and Notion are probably two of the best examples, and both of them actually made the list for the first time because we had enough data to feel that these were now credibly majority AI. Notion has even released data saying that half of their revenue, half of their ARR now is AI. And so then it's like that's real usage of consumers interacting with AI. So we got to include that.
I would say that yeah, so Canva and Notion are probably the two best examples of like growth stage companies, like maybe approaching IPO, that are still nimble enough to kind of pivot a bit towards AI. I am still a believer that we will probably see 20 years from now the winning company will be something that's AI native, just because they do have such an existing base of users and businesses and it's hard to cannibalize your own product. I think we've seen this a little bit with Google, where they're releasing amazing models like Nano Banana and Veo, but honestly the AI features that they're shoving into their existing interfaces like Gmail and Slides and everything β I don't know if you've noticed, but every demo video the use case is like plan a trip in like a way that no one ever actually plans a trip. And so I think that's been a little bit less successful. So in general, I think we are in almost every category it feels like an AI native company will win. There are some really horizontal things where the incumbents might have a distribution advantage, but those are kind of few and far between.
Yeah, the Google stack is crazy to me because I have Gmail running in Chrome and there's two different Gemini buttons that I can open simultaneously to have fighting Gemini instances fight over what's going on. And it's just something that like the product development maturity and the final UI has clearly not been outlined here and we're still in the early innings. Jordi, you were saying something?
Oh, I'm just very curious to see how consumer AI impacts Canva overall because I think so many of the typical entry points into Canva β like Canva is a massive tool, you can do all these different things, but so many of the tasks can just be one-shot. I was talking with a friend of mine who has a family business and they just describe when they need marketing collateral like a sign, they just describe it to ChatGPT and it just makes it, it just one-shots it now. And there's no β they don't really care. There's so many businesses β like in tech you might create a generation and then want to spend a bunch of time refining it and taking it from like 90% to 100%. But the average small business is like, you got me to 90%, like we're good to go. I don't need that extra horsepower.
Yeah. I think the use cases where the last 10% is really like 90% of the value, where you need to be like really able to iterate on it like pixel by pixel and not one-shot it, where there's like really difficult integrations you have to build, where you can capture ambient new data β like all of that is really ripe for consumer AI new startups.
Yeah. Also like templated workflows where you maybe want to generate a hundred images, that type of stuff. You can wire up Nano Banana with a workflow. There's a few of these like node-based editing tools. I think n8n is one, right? And there's a few others. And of course you could just write code to do it. But for someone who has a particular templating workflow in a more consumer AI or consumer image app, they might just be stuck in the workflow.
I do want to get to generated video. You've been tracking the models very closely. I'm interested to know when we see the collision between what we're seeing with these insane Chinese models. There's Vidu, there's so many cool video models where it's just text in, MP4 video out. And then on the flip side, you have like CapCut. I use Instagram Edits a lot. I really like that app. I've also enjoyed Captions, that app. And it feels like these are two on a collision course, just like we saw Nano Banana and Canva maybe get on a collision course. What are the existing mobile video editors doing? How well positioned are they versus the model teams? Because it feels like with video there's maybe a little bit more like that last 10% is even more than in images. But what's your take on generative video and what we should expect this year?
Video has been the most interesting category in creative tools and I think it's exactly to your point because Chinese companies can train on any data, even copyrighted data. So they do. So we're seeing like Seedance from ByteDance, like Hailuo, Kling β all of these models are amazing. I would say like Sora and Veo 3 are not far behind.
My twin sister Justine, who also works here, lives and breathes AI video and she had this blog post a while back about like there will be no one AI video model to kind of rule them all, just because there's so many different types of videos you make β like a true like two-hour movie versus like a 10-second marketing clip β like you actually probably want to train the model and design the workflows differently around those. So I've actually been more excited, I think kind of to your point, about these tools where you're able to switch between the models depending on what you are building, like Krea or Higgsfield made the list this time. CapCut is mostly ByteDance models, but it works because the ByteDance models are generally like pretty good and ahead of the pack.
Yeah. Yeah, I've been thinking about like, you know, we saw tool use come to ChatGPT β got a computer, as Ben Thompson put it, a Python REPL. It can run some math for you where it doesn't need to just guess the next token. It can just actually write the code and execute it. And it feels like we got a reasoning step with Nano Banana 3 β Nano Banana Pro 2. I'm lost on the model numbers. We're also sort of mid-revision on a lot of these.
They name them in a very confusing way.
Yeah. And then there's Veo 3, Nano Banana's on a different number scheme. Anyway, we clearly got some sort of reasoning chain where I can say like, dog riding a rocket, and it will add a lot of text to the prompt to sort of give me a better output. What's going to be interesting is when it also has the tool to make something black and white programmatically as opposed to needing to regenerate the video every time, because when you regenerate the video, you get something slightly different. I want to be able to upload a video, do a color grade on it, edit it down, add cuts, jump cuts, and have those tools be AI aware. And maybe that means some reinforcement learning pipeline on something that looks like an After Effects or something. But that type of β you know, the video experience is not just capturing the raw footage, it's the whole pipeline.
Yeah, I agree. The reasoning point and kind of the access that Nano Banana has to the internet as a video model I think is fascinating. Like actually in this report we included like a heat map of global AI adoption and the way that we created that graphic is I gave Nano Banana every country code and every heat map score and it literally went and colored in every country accordingly. Like it found the country on the map and I checked it by hand, it did them all correct.
And I don't think we've seen anything like that quite come to video. But the video models do have like a really kind of weird intuitive understanding of like physics and how things work in the world, like a drop of water creating ripples, that kind of thing. And so I'm sure the researchers are hard at work on this.
Shocked at the Sora data.
Take us through that. What's up for Bill Peeles? Bill, an absolute dog.
Proven. Never pray on his downfall. What? Yeah. What's wrong with Sora?
Sora is fascinating to me. It's like probably the biggest narrative violation we've seen in consumer AI in a while. First of all, they gave us the gift of so many AI videos of Jake Paul. So β
That alone, I think, was worth the money they spent on compute there. But actually, so what the data shows is downloads for Sora are definitely down. It was at the top of the US App Store for 20 days consecutively. It was getting 6 million downloads a month. Now it's closer to a million and a half. And so it didn't turn into like the social network that I think they envisioned, but it stayed really strong as a creative tool because the model is good and because you can create videos with these cameos. So the DAUs are actually still increasing. They have 3 million global DAUs, which is like I think probably the most for any video generation product on mobile. And so it's pretty impressive. If I were them, I would keep investing in that.
And depending on when this Disney partnership actually rolls out, I would expect that to send it right back to β
Totally. Totally. The question, the question, the question if I'm OpenAI is like do you want those downloads, do you want those users going to ChatGPT or Sora? If it is just a creative tool over time, I would expect the products to merge. I don't know.
I could see that. I do think it's interesting. I feel like the usage of AI with kids and families has been probably lower than it should be just because of concerns about hallucination or weird artifacts or like you just need to be very sure that the content is clean. And so I think something like a Sora plus Disney characters is probably going to explode consumer AI for kids in a way we haven't seen.
Yeah. And just like all the reasoning chains and like workflows within ChatGPT just makes so much sense where you could say, okay, here's the name of my kid, here's his favorite Disney characters, write a story, then turn that into an actual template, generate a few images, I'll pick the few key images that make the most sense, I'll upload an image and give you some backstory, and then now generate the full video, and then that's like the birthday card video that goes out. There's so many cool things.
Even with that heat map example, like you could write programmatic code to generate a heat map that uses JavaScript to look perfect and you know the data is right, but it's not going to look great. And then you could just take that and do basically style transfer on top of it. All of those different tools feel like they're really going to accelerate as they get brought together, but we're in this like unified time now. When the unification happens, it's going to be really cool.
The last question I have is about when we talked to Sam Altman about Sora, he echoed that it was being used as a creative tool, but I found that I wind up using Sora to generate videos that truly have an audience of one or like five people and I'll send them to a group chat and it's basically just an in-joke. And he says that that's like a new thing that people are doing where if I were to post it on my Instagram, people [say] why are you posting AI slop, but if everyone gets the in-joke in my five-person group chat or whatever, everyone's having fun. Haze paradox. And so I'm interested in where you're seeing movement in these like smaller almost β
Haze paradox, by the way, is a made-up paradox that I'm trying to bake into the models by saying it over and over on this β
Haze paradox is where you think something's β the more the funnier you think something is, the worse it'll do on Twitter.
Well yeah, the less likely it will be universally funny. Like something is the funniest thing in the world to you, the most niche humor will just be like β
Right, right. Because you're not writing for a general audience.
But talk to me about the growth in either like single-player or smaller AI uses. I see this with Suno a little bit where people are making songs for themselves. I've heard a lot of people are still fans of Midjourney just because they see it as sort of like this art therapy where they talk to the model and they get an image back and it's almost a single-player experience. Maybe they're sharing with a small community but they're not really using it in like a professional context. It's more of this smaller thing. Where are people going in terms of personal assistants, personal relationship coaches, that type of thing?
I am so excited for this because I feel like we haven't yet seen the breakout AI social product. Because if you just try to do like a skeuomorphic social network with AI baked in, it's like actually the reason that people are addicted to Instagram is because there's all these like positive and negative emotions that come with like putting real content of yourself out there and that doesn't exist with AI photos. So I think we've seen some attempts at it. It hasn't really worked.
What I am excited about is what you said across both Sora and I think Suno. A lot of people are making like meme songs for their friends. Anyone who knows how to train a LoRA probably has LoRAs of all their friends to make meme images. I think that kind of thing could be what potentially makes something like a ChatGPT group chats, which I think has had kind of core adoption so far, into something a lot more interesting.
