TBPN - OpenAI Responds to Meta Offers, Meta's Secret List, Apple Scores Big With 'F1' | Will Brown, Will O'Brien, Kylie Robison, Joshua Meier, John Glasgow, Tom Lee

Episode Date: June 30, 2025

(01:28:55) - Will Brown discusses Meta's recent aggressive recruitment of top AI talent, including former OpenAI researchers, to bolster its AI capabilities. He highlights Meta's strategic investments, such as the $14.3 billion stake in Scale AI, and the formation of the 'Superintelligence' team led by Alexandr Wang, aiming to advance artificial general intelligence (AGI). Brown also touches on the competitive landscape, noting Meta's substantial financial incentives to attract leading AI researchers and the broader implications for the AI industry. (01:59:25) - Will O'Brien, co-founder of Ulysses Ecosystem Engineering, discusses his company's development of autonomous surface and underwater vehicles designed to perform critical ocean tasks, such as seagrass restoration, at a fraction of the traditional cost. He explains how these vehicles can autonomously plant seagrass seeds, significantly accelerating restoration efforts compared to manual methods. O'Brien also highlights the broader applications of their technology, including maritime security and infrastructure projects, emphasizing the need for affordable, scalable solutions to address various challenges in ocean operations. (02:21:13) - Kylie Robison is a senior correspondent at WIRED, covering the business of artificial intelligence; she previously reported for The Verge, Fortune, and Business Insider. In the conversation, she discusses her recent reporting on OpenAI's decision not to release a research paper defining AGI, allegedly due to concerns over its Microsoft partnership, and the tension surrounding Meta's recruitment of OpenAI researchers, including a memo from OpenAI executive Mark Chen accusing Meta of "robbing our home." (02:31:32) - Joshua Meier, a leader in AI-driven drug discovery, discusses Chai Discovery's new foundation model, Chi2, which significantly improves antibody sequence generation, achieving success rates of up to 20%, thereby reducing the need for extensive lab testing. He highlights the potential of Chi2 to expedite drug development and enable the creation of novel therapies, emphasizing the importance of agility in the rapidly evolving biotech landscape. Meier also envisions a future where AI models can design personalized medicines, though he acknowledges the challenges in predicting the pace of technological advancements. (02:40:09) - John Glasgow, CEO and Founder of Campfire, discusses his journey from experiencing the limitations of legacy ERP systems firsthand to founding Campfire, an AI-native ERP platform designed to modernize finance and accounting processes. He highlights the company's recent $35 million Series A funding led by Accel, emphasizing plans to enhance product development and expand market reach. Glasgow also elaborates on Campfire's innovative use of AI to automate tasks, improve financial insights, and streamline operations for businesses of various sizes. (02:51:37) - 02:55:35 Tom Lee, founder of Fundstrat and Chief Investment Officer of Fundstrat Capital, has been appointed Chairman of BitMine Immersion Technologies. In this role, he discusses BitMine's $250 million private placement aimed at adopting Ethereum (ETH) as its primary treasury reserve asset, highlighting the strategic importance of Ethereum's ecosystem, particularly in relation to the growth of stablecoins. Lee emphasizes that this move positions BitMine to leverage Ethereum's protocol-level activities, such as staking and decentralized finance mechanisms, to enhance the company's treasury operations. TBPN.com is made possible by: Ramp - https://ramp.comFigma - https://figma.comVanta - https://vanta.comLinear - https://linear.appEight Sleep - https://eightsleep.com/tbpnWander - https://wander.com/tbpnPublic - https://public.comAdQuick - https://adquick.comBezel - https://getbezel.com Numeral - https://www.numeralhq.comPolymarket - https://polymarket.comAttio - https://attio.com/tbpnFin - https://fin.ai/tbpnFollow TBPN: https://TBPN.comhttps://x.com/tbpnhttps://open.spotify.com/show/2L6WMqY3GUPCGBD0dX6p00?si=674252d53acf4231https://podcasts.apple.com/us/podcast/technology-brothers/id1772360235https://www.youtube.com/@TBPNLive

You're watching TBPN. Today is Monday, June 30th, 2025. We are live from the TBPN Ultradome, the Temple of Technology, the Fortress of Finance, the capital of capital. We got a great show for you today, folks. It's summer. Every venture capitalists is on vacation. They're celebrating the fourth in France. We're not. Yes. Literally everyone is abroad. But we're here in Hollywood, California, the future center of media and entertainment. If we play our cards, right? We're going to take you through the five most important stories in technology and business from our perspective, kicking it off with a dramatic. Yeah, get that gasp ready because there's a dramatic story in Wired. Open AI leadership responds to meta's offers someone has broken into our home.

That's the title of the article. You could hear it was written, but you could hear the pain in Mark Chen's voice. Friend of the show, I would call him a personal friend. He's a great guy. We've had people try to poach our team members with, I wouldn't say similarly sized offers, but meaningful offers. Meaningful offers. But yeah, I mean, certainly a high stakes situation. People think of Open AI as a $350 billion juggernaut, 90% market share and consumer AI.

They're doing great. They are the steamroller. But you forget that Zuck's built different. He has $100 billion in free cash flow every year to play with. And the vibe shift from two weeks ago where everyone was like, oh, okay, Zuck acquired scale. Yeah, yeah, yeah. And then, oh, he's got Nat and oh, he's got DG. Yep.

And then Sam's like, well, they're not recruiting any of our, they're not able to get any of our best people to, it feels like somebody has broken into our house and stolen something is very significant. So Bill Gurley has the post that defined the story. He says the company that was already planning to lose $7 billion this year, that's Open AI, is reevaluating comp to be more aggressive. Put that in perspective and he's got the Yehaw cowboy hat emoji on. I love it. I love post, you know, retirement bill.

Oh, yeah. Before he would be much more concerned about this, now he's got the cowboy hat on. He's just watching. Rodding into the sunset, you know. Riding in the Sunset talking about the drama. We're going to have a monsoon. Yeah, it'll be fun. I saw there was someone else posted that,

and I thought this was accurate, people forget that Zuck yeeded like $20 billion. He's willing to lose tens of billions of dollars. And he can lose. On this, the MetaQuest 3S, Xbox Edition. I don't think he would, I don't think he yeated 20 billion on this.

I think it was worth every penny. We're going to be putting this to the test on the show. Today, stay tuned. Tyler Cosgrove, our intern is going to be trying to beat the first level of Halo in virtual reality. But, anyway, OpenAI has incredible access to capital, but they are losing an obscene amount of money. Meta has incredible access to capital and can deploy it and still generate tens of billions of dollars of net income. Yes, and uniquely, Zuck has incredible control over the board and shareholders. And so he controls enough of the vote that he can't be. He's also just so big, but he can't be the subject of activist shareholder attacks saying,

hey, we want earnings per share to go up by a little bit more. We'd like our dividends to be a little bit bigger. He's in that early Bezos era where he has the full faith of the shareholders and also the legal control to actually go and spend tens of billions of dollars on a project that might prevent disruption, which is absolutely in the long term. best interest of the he's not thinking quarter by quarter exactly he's fulfilling his fiduciary duty just on a different timeline and so the only reason that an investor would be upset if they're like yeah well I just want the

stock to pop this quarter but he can tell those people that I'm not I'm not building with you in mind so yeah it's a it's getting spicy it's a mark on Mark violence Mark Chen is the Mark Zuckerberg pull it up I'm sure I'm sure half a point the market loves it No, clearly. What is that? Five billion? Yeah, something like that.

I mean, five billion, that'll pay for what? 50 researchers. That's exactly how many you need, right? Yeah. Who is telling us the pod? Perfectly priced in. Yeah. Yeah, perfectly priced in.

Nikita Beer. We got a trade deal, baby. Trade deal. Nikita Beer has joined X as the head of product. You probably know him from his viral social apps. He's been it. He sold one. to Discord. He's sold one to Facebook back in the day. Now, he didn't even have to sell one.

He's just going directly to X. He's joining as the head of product. He's cutting out the middleman, which were his startups. Basically. He says, X is the most important social network in the world. Fully agree. Fantastic app. I'm obsessed. It's where we built this company. It's where internet culture originates and where the world's most influential people convene. Finding my community and building an audience on X has impacted my life more than any single thing. It's online. friendships, professional opportunities, and it's even where I met my girlfriend. Well, I already spend every waking hour on this app.

I'll now be spending that time helping others unlock the same value. And we'll certainly be leveraging the power of GROC to create hyper-relevant timelines and help people understand everything that's happening. Thanks Elon Musk and the X team for inviting me to work on the product that I live in brief. Head of product, but also poster and residents, right? They lost the app scene.

Yes, they did. Or fired him. But they're bringing back a new poster and resident who residents who's going to be absolutely, you know, dominating the timeline. Yeah. And I think, I don't know. It's interesting, like, Nikitas had success iterating very quickly on product and, and developing viral apps.

He's also spent enough time in large tech companies like meta and Discord to like probably be able to play, you know, there's probably some overhead that still exist even in the leaner X. So I wouldn't be surprised that he's actually kind of the perfect person there. And then obviously the fact that he's so dominant on the platform and obsessed with it. And really, he's a power user. That's probably exactly what you want in the head of product. So I'm bullish. Are you ringing the gong for it? You know why I have this out? Why do you have that out? Because the rumor is that X backed up the truck. The truck? No way. This one. Okay. I'm going to hit the gong for Nikita.

Hit it hit the gong. Congratulations to Nikita beer. He's been on a generational run. And it continues as X's head of product. And I'm going to start texting him now when there's bugs. Yeah. Head of product. That means chief bugs squash. I actually don't have any experience that many bugs. You don't. Everyone says that there's a ton of bugs.

You got your head in the sand. I think it's a skill issue. I've been having great. But there are a bunch of things that could, you know, be improved with the product. And so I'm sure he'll come out with some interesting ideas. And it'll be a lot of fun to follow. Anyway, in other news, you got to also ring the legong for Apple. Our boy Tim Cook, he pulled in.

He paid for half of his annual salad. just by selling a sponsorship in Apple's F-1 movie for a fictional team APXGP, Expensify, paid an eight-figure sum to become the title sponsor. They also sold sponsorships, Shark Ninja, IWC, the Watch brand. They really built it out, and it makes it look better. I think it makes it look better than it's real brands instead of like fake brands. No, it looks super legit. Yeah, this was interesting.

I knew that I don't have a lot of insight into how entertainment companies sell sponsorships in shows. I know it's super prevalent, right? You'll see like, okay, why is this character driving a Hyundai? It doesn't actually make sense in the context of their life, but like clearly it's Hyundai. It's all over the place. So sometimes, like, when I was running Soilent,

there were people that would come to us and just like, they would just be like, we had it on the set that day and we wanted to throw it in the fridge. And so we did, so you got a spot in this. We were in a Super Bowl ad once because they needed to, the cars in the, Fast and Furious movie, we're driving through Time Square. And they needed a, they needed like some billboard in the background.

So they just put up an ad for our product. Yeah. Virtual billboard. So sometimes it's like we need something, we need a product that we can clear. Like, you know, Andrew Huberman is going to be fine with us, with us promoting, you know, mattaina Yerba Matae on here. But if we were at a really, really Hollywood level, like we would have, if we were going to drink like Monster or Red Bull.

Well, we're out of Hollywood level. I know, I know. But yes, yes, we are in Hollywood. feet are. But there is a level above where, where, where, like, you would have to get clearance to use the thing in the actual, in the actual film. And so a lot of times it's like, can we just get quick approval and not have to pay or not have to do any legal stuff? Other times it's more complicated where they're actually going out and selling it. It sounds like they sold this for this.

But anyway. And Polymarket of the day. They didn't just make money on the sponsorships. Yeah. The box office was at $140 million? They crushed it. $140 million. It blew it out.

Polymarket had it. If you were looking, they clocked it last week. It was, the market was, hasn't, I guess it's still in the process of resolving. I mean, the core predictions were around like 55 million, I guess, was kind of the middle band here, 50 million. And it blew that out. I mean, F1's popular. And it's so, have you seen the other F1 movies? Of course not.

Why am I even asking? Ford versus Ferrari, you haven't seen that. Drive, you haven't seen that. Right? No. No. But big news, I saw my first UFC event, so I'm catching up. George is going to see his first movie.

I'm going to see my first UFC event. It'll be great. John, yeah. I love it. We had a good crew. It was a lot of fun. UFC with Senra and Rob Moore. It's great.

And we had a great time. John asked, who's playing tonight on Saturday? I said, well, John, I don't think they're going to be playing. They're going to be fighting. Punching each other in the face. And the knockout on Saturday, was spectacular. My wife was like, oh, you know how like when guys go watch the football game,

they like toss around the football beforehand? Like are you going to be like punching each other before you watch the punching? And I was wondering if you guys are going to be like, yeah, let's just go like spar a little bit. Yeah, let's start. Let's start. Maybe we should do that. Next time.

Anyway, this is actually massive news. Toby Looky has a summed it up well. So Tesla has the first, the world's first autonomous delivery of a car. So this Tesla, I believe it's a Model Y, drove itself from the Gigafactory in Texas to its new owner's home 30 minutes away, crossing parking lots, highways, and the city to reach its new owner. And Toby, look, he says, put it on the front page of all the newspapers. Well, we printed it. It's in the front page of TBPN. What else do you need, Toby?

Come on the shop. No, this is significant. It's costly to get a car delivered. Obviously, if you're getting a car delivered from hundreds of miles away, the gas fees are going to be high. it costs money. The other thing, you just got a car recently, and the shipper damaged the car while it was being shipped. It should just dream itself. Now it needs to be repaired. Yes. And so I'm very interested in what this actually means for like structurally for the business of Tesla. We'll have to dig into this. But I wouldn't be surprised if like that like, like, so

Tesla is famously like cut out the middleman of like dealerships. And I think that's really improved their margins. And I imagine that delivery is not an insignificant line item for Tesla right now. Like you order them, they have to put them on trucks, they have to deliver them, there's all this extra, all this extra step. And I feel like delivery fees are around a thousand bucks maybe. It depends on how far you're going, but like you have to have multiple people securing the car, traveling with it. It's not just every time I've bought a car. Hundreds of dollars outside of, you know, a few hundred miles away. It ends up somewhere a thousand. A thousand bucks.

Yeah. And so, yeah, you assume that there's like a 50% margin tacked on that and it's like a, it's a high margin add on. The other thing here is it's effectively passing on the delivery cost of the consumer because it's really just mileage, right? Yeah, mileage and energy. And then you get a half-charged car and you're like, you're like, okay, I got to charge it up. But who cares about that?

Nobody cares about that. And so I wonder, I mean, this should be, you know, somewhat material. I just wonder how material. It might be like, okay, it drops their cogs by 0.01%. but it certainly is exciting, and you can imagine we're one step closer to, like, push a button and buy a car, and it just arrives, which is very, very cool. Yeah, finger hovering over the button. I went that one, I want to get one just. But I'll be waiting for the naturally aspirated V8 or V12, yeah, version.

Which is inevitably is coming. Yeah, it would be fantastic. Anyway, in other news, your naturally aspirated V12 Tesla is getting delivered. Yes. And it's just getting noise complaints along the way because AI is just joyriding it. Pulling over to do burnouts. I mean, have you heard Doug DeMiro's riff on electric cars? He says that they're technically naturally aspirated.

They don't have turbo chargers. They don't have superchargers. How are the batteries cooled? Naturally. All the air flows through the car, naturally. There's no turbochargers. Therefore, it is a naturally aspirated car. Yeah, Xiaomi is not far, right?

I posted a picture today. Their new car is just almost. a one-to-one copy of the Sunway. Yeah, the news is. Apparently, it's selling very well. They've got over 200,000 non-refundable orders. Not just over 200,000. This is the electric SUV, the new Xiaomi.

289,000, and that's not just so far. That's in the first hour. Okay. That's a lot. Well, just to show that we respect business, I'm going to hit the size gone club. Congratulations. to the folks over at Xiaomi for ripping off the Perosangue so successfully. That's on B3.

Anyway, the last top story we have today is from Luke Metro. He says, I know a 409 evaluator hates to see this coming because Robin Hood has tokenized stocks and everything else. And so now there are private company stock tokens and they're shouting out OpenAI and SpaceX. But Luke Metro works at Anderrol obviously has some stock options. and he's getting ready to trade them on Robin. potentially, who knows. In the pipeline, yeah. In the pipeline, I would imagine.

Yeah, this is interesting. I mean, we had Ken from Republican last week, also building his own private company. He must have known that this was coming or something, or maybe there's some underlying market structure bill, but this feels like it's a wave that everyone's doing it all at once, right? Part of it is that this administration has been extremely crypto-friendly. Yep.

Seems like everything, not everything. but most things are fair game now. And so I think a lot of companies saw this coming and have just been racing to deliver new product experiences. It is funny to think about, I guess the SpaceX and Open AI tokens are going to be available to European Robin Hood users only. And so it is funny to think about

when is Europe more liberal than America? Day trading open AI and SpaceX shares. I mean, could be the best place for them. This is apparently rolling out on the Robin Hood chain, so they're launching their own L2. Interesting. With Arbitrum. Yeah. And I'm interested to see this roll out.

Yeah, they cleared an exchange of Open AI tokens today or Open Aist stock. Like the first transaction, yeah, the first transaction's gone through. Luke Metro had another funny post that was like, I just hope that my and oral stock is valued like the Robin Hood Dijan retail trader valuation and not the price. that Founders should pay. This is very funny. That's good. Anyway, oh, sorry, we blocking the camera. Anyway, let's go through some of the deep dives.

I want to dig into that wired article that was burning up the timeline over the weekend. And then we'll read through some of the Wall Street Journal coverage of the list. AI's secret file of geniuses. But first, we have our first intern challenge of the day. The first. Yeah, there might be many. Even when you finish this, Tyler, you're not. So we were, we purchased and received a MetaQuest 3S Xbox edition.

Very nice foil. Look at that. This is a nice product. So this is the latest and greatest out of meta for their virtual reality headsets. And the challenge is to beat the first level of Halo complete edition, Halo C.E. This is Halo 1. Tyler, you've never played Halo at all? I've never played Halo.

Okay, but you've played first person shooters. Yeah, yeah. So you should be able to do it, you think? Yeah, 100%. So I asked ChatGPT. The Speed Run World Record for Pillar of Autumn, which is the first level in the first Halo game, is like three minutes, but you have to like cheat and go through the walls and stuff.

An experienced player can do it in like 15 minutes. First time person ever playing should be able to do it in between 20 and 30 minutes. So for this, let's give you an hour. If you can do it under- I was thinking like 12 minutes. Best of luck. Let's give you, let's give you 45 minutes. If you can do it in under 45 minutes,

you will get to keep this. Let's go. Okay. So can we get a timer? Yeah, start the timer. So it is 1120. Let's give you 40 minutes. So if you can do it before noon.

So go get it and start. Okay, time starts now. He's hustling. He's hustling for good reason. All right, have fun, Tyler. So the problem is that the level takes like five minutes, but unboxing and setting that thing up, I think he's going to take like 40 minutes.

So that's the real challenge. Godspeed, Tyler. But Tyler will be working on that during the show. And good luck to him. Anyways, back to the timeline. A poster by the name of Tuchan says, Nerds is quoting, on the left is Ronaldo, another post.

Real Madrid spent 80 million to sign him from Man United. On the right is Jahu you, which I'm struggling to pronounce. Meta paid $100 million to sign him from OpenAI. And Tuchan says, nerds have been complaining that athletes are ever paid for decades. Finally, the nine-figure nerd trading market has emerged. Take that. LeBron Jameson. LeBron Jameson.

He doesn't even know his name. It's fantastic. Well, you know, if you're on a hiring spurry, you've got to get everyone in your organization corporate cards head over ramp.com time is money save both easy to use corporate cards bill payments accounting and a whole lot more all in one place um anyway let's dig into this open a i leadership uh the open this was this is from was all this news was kind of coming out yeah over the weekend uh there's no there's no decency anymore you can't even take a summer you can't even send an email to a 3,000 person organization without it leaking to wired um mark chen the chief research

officer at OpenAI sent a forceful memo to staff on Saturday, promising to go head to head with the social giant in the war for top research talent. This memo, which was sent to Open AI employees in Slack and obtained by Wired, came days after META CEO, Mark Zuckerberg successfully recruited four senior researchers from the company to join META's superintelligence lab. I have a visceral feeling right now, as if someone has broken into our home and stolen something, Chen wrote. Please trust that we haven't been sitting idly by. I mean, this makes sense. You like, you need to, you need to tell the troops that, like, you know, we're prepared to go on the counterattack. Like, we got attacked,

but we're ready to go back. What's interesting is like, is like, I don't think that, I don't think that meta is necessarily going to try and go eat chat GPT's lunch. Like, I was revisiting that conversation that we had with Jeff Huber. And he was saying, like, you know, he was kind of reiterated like, never bet against Zuck. Like, he's working on an open source model. But a lot of it is, is more of this, like, uh, B, to be applied AI necessarily than trying to go and disrupt chat GPT because Google's already running that playbook. Mark can see that and see that like what happens when you're a trillion dollar company and you roll out a basically a direct clone of the chat GPT app. It's like yeah,

you can get a decent amount of users in a no in a notional terms. Like I'm pretty sure the Gemini app has like a hundred thousand user or a hundred million users and like lots of five star reviews, But it just doesn't have anywhere near the penetration of chat GBT. And so when you pull someone off the street and you say, what do you use for AI? They say chat. They don't even say chat GBT anymore. Yeah. It's so dominant.

And so I think Mark knows that he shouldn't necessarily try and go after that, but he should be going after the next thing and the next next thing and have a model that he can pipe into all sorts of different features within the meta ecosystem. But. Yeah, the crazy thing is there's not, it doesn't seem like there's clear press. for a raid of a talent raid of this magnitude right we covered Ken Griffin rating Enron but it was after host the collapse yep right yeah Apple heavily recruited out of Xerox Park in in 1979 in the early 80s but like it's kind of

down swing yeah it wasn't it wasn't yeah it wasn't anything of this magnitude nor nor was it hey we're gonna come in and just give you these like nine-figure off you know totally totally just and things like that yeah and it's very It's very common for a startup to be able to pull people from the big legacy company that's kind of sclerotic. And you know, we've seen a lot of people that were at Apple or Google or Microsoft or Xerox and then they went to Google and then they went to META and then they went to Open AI. I mean, Brett Taylor, who's the chairman of the board at OpenAI was the CTO of META. Right? And it's like, okay, he clearly just wants to be in like the hot thing and he's like a startup guy almost.

And I believe he was at Google before because he worked. Yeah, he redid Google Maps. And so like the story of Brett Taylor is really like the default, I think, for most like startup people. It's like he's at he's early at Google. He works with Paul Buchite on Gmail. He works on Google Maps. Then he goes over to Meta CTO there. Then he goes to Open AI, chairman of the board there.

And it's like he is a he's a startup guy. And so he's always riding like the new way. You usually can't poach them back into the older company. That's rare. Yeah. One other, some other precedent here in 2010 and 2011, Zuck and Facebook went super aggressive poaching from Google. So Facebook had crossed half a billion users.

They were in that period of transitioning from a super scrappy startup. And they went heavy into Google. So they went and got high performers from ads, search, mobile on the Android side, and some of the social product people from Google Plus and Zach did something similar not to the same scale but just went and with these sort of heavy heavy heavy comp packages and was very successful Google had to respond and increase comp in in a number of different do they

poach you designers any people that use figma.com any people that think bigger or build faster anyone that you know because figma helps design and development teams the web would be a different place but but fortunately you can go to figma.com you can think and you can get started for free Anyway, what do you think of the buzzword super intelligence? This whole, like they've really rebranded it. Is it a gold post-shifting thing? I think he poached the word super-intelligence.

It's mine now. It's mine now. Because, and he's not saying, and he's not saying the safe superintelligence team. He's just saying, super-intelligence. Reckless, the reckless-superintelligence implies reckless super-intelligence. I like that. But it's funny because, you know, everyone was focused on like AI as a buzzword. Then it was AGI.

Then it was now it's super intelligence. But like, you know, within our group, like we all think like super intelligence kind of the old thing. Like everyone like me and you like we know the secrets to this stuff. Like like it's kind of it's kind of already priced in. The real killers, they're focused on the next thing. You know what I'm talking about, right? Gig intelligence. That's the one that people are getting.

Really? Like the real killers are like, yeah, well, like, AGI is basically here. Super Intelligence is basically here. But Giga intelligence, that's the one we're working on. Because we're going to need a new buzzword because we're burning through buzzwords and increasing rate. And so we're just, we just steamrolled the touring test, completely blasted through AGI. No one talks about AGI anymore. AGI is here.

Yeah. But you got, you know, super intelligence. We're going to solve this buzzword. We're going to churn through this in a year or two. We've got to start working on giga intelligence now. Yeah. This is the way. Start adopting it.

Chen promised that he was working with Sam Altman, the CEO of OpenAI, and other leaders of the company around the clock to talk to those with offers, adding, we've been more proactive than ever before. We're recalibrating comp and we're scoping out creative ways to recognize and reward top talent. Still, even as Open AI leadership appears desperate to retain its staff, Chen said that he has high personal standards of fairness and wants to retain top top. talent with that in mind. While I'll fight for to keep every one of you, I won't do it at the price

of fairness to others, he wrote. It's such a brutal dynamic if you're Mark Chen, if you're Sam, because not everybody at Open AI is getting poached, but presumably Zuck is working with all the people he's poached already to identify who are the best people at Open AI, who do we really want. Yep. And then going to them and making the maxed out offers, meanwhile, it, you can't. You can't. You can't like the team dynamic if you have a few researchers on a team and somebody gets you know somebody's potentially getting poached by meta and you say to retain them you're like okay we're going to give you this sort of massive incremental grant and the other people on their team are going to be like wait well is that my market value I didn't go talk to Facebook or I didn't

I turned down I never took the meeting or whatever so you should pay me the same thing too yeah right I'm loyal shouldn't don't you want to pay I'm a missionary yeah I'm a missionary why are you paying the mercenaries more. You should pay, you should pay the missionaries the same price just because they're missionaries. Rough, for sure. Birdle. Anyway. Let me talk about. No, the other thing is it creates this like really nasty incentive to basically be like, oh yeah, you know, you maybe got hit up by a met a recruiter like many months ago and you can be like, yeah, I mean, they reached out to me and suddenly they're like hovering around you like being like, All right, what's it going to take?

We really don't want to lose you. Yeah. It is a... Yeah, I mean, it's crazy game theory. Like, every single conversation in a company is like this constant game theory. And then timing that with... Yeah. With this, like, summer break that I guess opening eye was saying, like, hey, everybody should take the week of the fourth off or not the week of the fourth.

But anyways, you wanted to say something more important. Well, I mean, you know, if you want to see who your high performers are, just look in the linear, right? Linear's purpose-built tool for planning and building products. Meet the system for modern software development, streamline issues, projects, and product roadmaps. If Zuck were able to gain access to Open AIs linear, it would be illegal, but it would also tell them, tell them a lot. Yeah, you don't want to do anything illegal, Zuck. You want to be on Vanta, automate compliance, manage risk, prove trust continuously. Vantta's trust management platform takes the manual work out of your security and compliance process and replaces it with continuous auto- whether you're pursuing your first framework or managing a complex program. So the news comes as competition for for top AI researchers heating up in Silicon Valley. Zuckerberg has been particularly aggressive in his approach, offering $100 million signing bonuses to some open AI staffers, according to comments Altman made on a podcast with his brother, Jack Altman. So this $100 million signing bonus, this just got baked into the lexicon.

but I was listening to Dylan Patel on Jordan Schneider's China Talk podcast. It's actually transistor radio, but it goes out on China Talk anyway. And Dylan Patel was saying that that might be a kind of a game of telephone a little bit in the sense that it's possible that you could maybe be making $100 million over a number of years in stock based on appreciation. But he was very skeptical that anyone was getting. a hundred million on day one without any sort of strings attached for like a true signing bonus. So I don't know how real that number is.

Yeah. And remember that the Zurich team came out and said, the Zurich team explicitly said we didn't get it. I know, I know. The way that he phrased that it could have been more. Clearly were getting. No, no, no. No, no.

No, I do think that the round numbers are easy to latch on to. Like, why is Mom Dani talking about billionaire specifically? He doesn't have a problem with 999 millionaires because like that's not easy to grasp. Like people latch on to round numbers. A hundred million is so much. A billionaire is so much. And it's easy to, it's easy to quantify.

And so the hundred million, if it was 80 million, it wouldn't be going nearly as viral as 100 million. And so I don't know, I don't know how real that is. Maybe that was thrown out as a total package and then it got kind of telephoned into, well, you know,

it's basically a signing bonus because, they got the deal on the, on the, on the, as soon as they signed, but they do have some sort of earn out or the shares are locked up in some way. But there's a bunch of creative accounting that can go into basically delivering someone what feels like a hundred million dollars of value, which is not far off from, okay, what, what would it take, you know, Johnny I've got, you know, a huge offer to go to open AI. And it was structured in an interesting way with like an acquisition and, you know, but he got, what, it was a few points.

Yeah, we were, it was a few, it was a few, it was a few. few percentage points. And when you think about the scale of Open AI and the scale of Johnny Ives legacy, like, yeah, like, what would it take to get someone like that's a founding designer? He's a founding designer, basically. So it's a couple, couple percentage points. So multiple sources of open AI with direct knowledge of the offers confirmed the number to wired. So I don't know, because it kind of benefits like both sides to have that number out there. It's kind of unclear who would, who would really want to confirm that and then what the nature of these things were. The other thing that's interesting though, the other thing that's interesting is if you're

working at Open AI and let's say their top researchers have on average a hundred million dollars in Open AI shares that they're like the truly top people and then you get an offer that's like a 20 million dollar signing bonus but it's you know it's kind of paid out over time and that or or they can claw it back if you don't stay past a certain date and you're kind of looking at it and be like okay I'm already a Senti and I'm getting an offer to like go to a new team that I don't know if I'm going to like and it's and it's unclear like the vision is not quite as clear at meta right I'm joining the super intelligence team it's going to be an incredible group of people with a ton of resources that's exciting but it is a big switch if you're

already happily investing out tens of millions of dollars. So it would make sense that these $20 million, you know, if it's $20 million, I can see that if it's $100 million for the, for the top, top, top people. It makes total sense. Yeah, I mean, like what was the, I mean, the deal to bring on Alex Wang was probably like way over $100 million to him personally based on his ownership in that company, which was acquired. And so, like, that number doesn't feel impossible. I would just be surprised if there's not some structure around it based on what Dylan was talking about. There's just enough precedent here that, again, going back to Zuck rating Google in 2010, 2011,

he was offering tens of millions of dollars in stock packages to mid-level engineers just because he was like, I need a really good ad product. I want to make my ad product great. Yeah. I mean, have you seen the, why am I even asking this? Have you seen the Blackberry movie? Of course not. But in the Blackberry movie, there's this whole story about they go and they try and poach someone.

And then it was controversial because they like backdated some stock options for someone to basically give them more like cash. But without having like the initial impact of that cash. It's kind of interesting. But yeah, I mean, to poach top talent, it's expensive. It kind of benefits both maybe. I mean, I don't know. Like on the meta side, do they want, does Zuck want the idea that he's willing to pay up to $100 million out there? Like maybe because that certainly leads to more people taking the phone call of like, oh, wow, Mark is really taking this seriously.

I should talk to meta's recruiting team. I should actually hear them out. And then if they wind up coming back to me with a $50 million offer or a $10 million offer, like, you know, know, we can discuss that and I can see it and they can evaluate how much I'm worth, they can make an offer, but at least it got the conversation started. But I don't know, there was also the theory that Sam Altman put that $100 million number out in order to kind of like poison the well because if, if Mark comes to you and says, hey, you're a top AI researcher, come over here.

And then the top of AI researcher says like, yeah, great. I hear $100 million at the going rate. And Mark's like, no, no, no. that's like fake news that Sam was spreading. Like we're actually paying 15. Then you're like, oh, no. I was, I was hoping it for a hundred. Sam is one of the most Machiavellian people.

Yeah, you see, it could be like 40 chess on both sides. I don't know. Yeah. Over the past month, Meta has been aggressively building out their new AI effort and has repeatedly and most unsuccessfully tried to recruit some of our strongest talent with comp focused packages Chen wrote on Slack. a source close to the efforts at Meta, confirmed the company has been significantly ramping up its research recruiting with a particular eye toward talent from Open AI, Google. Open AI and Google. Anthropic, while also a top rival, is thought to be less of a culture fit at Meta. One source tells Wired. They haven't necessarily expanded the band, but for top talent, the sky is the limit, which is funny because, yeah, of course, like, Anthropics are so missionary-driven. It's going to be very hard to go to Anthropic and research scientists and be like, ads. Better ads.

They're like, we're building God here. Yeah. Anyway. 100 million dollars is 100 million dollars. Not for, not for anthropic guys. Anthropic guys are true believers. They don't care.

What's $100 million in a post-scarcity world? It doesn't matter. You want to be on the, you want to be on the post-scarcity train? Chen's note. It is so, it is so telling, though, like how you,

like, sort of like, try to, you know, pinpoint like Zuck's views on AI. He's like, I'm not even going to bother with those people anthropic, not a culture fit. Two AGI pilled. Yeah, that's great. Senator, we sell ads. Anyway, if you're looking for sales tax superintelligence, go to numeralhq.com, spend less than five minutes per month on sales tax compliance with superintelligence for sales tax. The official sales tax automation provider for Lucy.

Yes, that's true. Chen's note included messages from seven other research leaders at the company where they wrote notes to staffers in an apparent effort to encourage them to stay One leader on the research team encouraged staff to reach out if they received an offer for meta If they pressure you or make ridiculous exploding offers just tell them to back off It's not nice to pressure people and potentially the most important decision WIRE does not naming the leader as they are not a C-suite executive This is weird where's the where's the end of this quote that's odd I'd like to be able to to talk you through it and I know all about their offers.

The remarks come as Open AI staff grapple with an intense workload that has many staffers grinding 80 hours a week. Open AI is largely shutting down next week as the company tries to give employees time to recharge. Let's give it up for Open AI's grind set. It's really fantastic. Also, extremely American to give a full week off right around July 4th. Like I'm normally pretty anti-taking time off, but if you're going to go really hard on a particular holiday, make it July 4th, you know, give a week off around July 4th.

I love it. I was posting, you know, the over, I need a polymarket on whether or not we get a Zuck American flag video on July 4th. Do you think it'll happen? I actually think they're 100% should be a polymarket on that. Right? It's kind of 50-50, right? Like he is definitely locked in, very busy, probably not a lot of time off.

He sees that, he knows from this that Open AI is down this week. So that's probably like really encouraging to be like, I want to go harder this week. They're off. Yeah, my advice for founders this week is one of the greatest pieces of content he puts out. Think about how hard Zuck is going right now. Go harder. The week of July 4th.

And just don't get outworked by him. Yes. Just don't get out work by Zuck this week. It's going to be tough. He's in year 20. What excuses do you have? You're three months into your startup. Come on.

Go harder. Go harder. I love it. Yeah. Meta knows we're taking this week to recharge and we'll take advantage of it to try and pressure you to make decisions fast and in isolation. Another leader at the company wrote, if you're feeling that pressure, don't be afraid to reach out. I and Mark are around and want to support you. While Open AIs leadership is taking Mehta's effort seriously, Chen also said that the company is getting too caught up in the cadence of regular product launches and in short-term comparison with the competition.

The sentiment is backed by a former OpenAI staffer who worked closely with Altman and said the CEO wanted to do. see buzzy announcements announcements every few months this was a really this was a key to their strategy was like always have a buzzy announcement ready if the competition does anything just drop a oh googleio is happening wouldn't it be crazy if we acquired a company called io the day before or something that like we're like oh the google's announcing the new gemini let's just steamroll them in the timeline the 24 hours before yeah i'd love to introduce you deep research I mean, it's masterful. And like, but that's the game you got to play.

That's the game on the field. Like I, I think he's being aggressive, but he's not doing anything wrong. He, he is up there with Elon in terms of creating and riding hype and constantly delivering enough real product value. Yes. To justify the hype that was being sold, you know, six months ago, basically. So like staying on the bleeding edge. And this is why they. anti-tech people hate him and Elon so much because they're like they're like it's overhyped it's not

useful and then like a year later it's like there's Tesla cars everywhere the rockets are landing and like 90% of people are eating Chachipati every day. Tens of billions of visits monthly. Okay so it was real but also he was over promising but he delivered at a at a higher level than anyone in human history and that's just like it's those two things are in in conflict. Someone said this like Elon is both Thomas Edison and Barnum and Bailey or whatever, P.T. Barnum. He's a circusman who can put on a show and put a robot in a human in a suit acting like a robot, but then he can actually go and deliver the thing, which is just, it just breaks people's brains.

Anyway, we need to remain focused on the real prize of finding ways to compute. A lot more supercomputers are coming on later this year into intelligence. Chen wrote, this is the main quest, and it's important to remember that skirmishes with meta are the side quest. Last but not least, I'll be around this week, recharged and ready to go pound for pound. DM me anytime. I love it. He's ready to go. He's fired up. Get him in the ring. I love talking with Mark. Yeah. That was a fun conversation. Also, I mean, guys in shape. Let's get him in the, let's get him in the octagon with Zuck. Mark on Mark

violence. Let's see it. I mean, if there was a pound for pound list for the top AI researcher in the world, he's in the conversation. He's in the conversation. It's amazing. It's really been amazing to watch Mark's leadership and integrity through this process, especially when he has had to make tough decisions Altman wrote on Slack in response to Chen's message. Very grateful to we have him as our leader. I mean, it can't be overstated making these decisions, right? Yeah. One of your best people comes to you and says, it's hard. Hey, I'm really happy here at Opening Eye. Yeah. I have 80 million dollars of stock. And I was just offered 150 over the next year and it's highly lick or over the next few years it's highly liquid. Yeah. Meaning that I can

market sell it quarterly forever. Yeah. And meanwhile, you're going through this for-profit conversion. Do you think the liquidity thing is real at all? I mean, obviously what are you, I mean like if you want to, let's say you want to buy 50 million dollar house and you have a hundred million dollars and like illiquid open AI. Like, Is that going to be hard at all? I feel like Goldman Sachs and Morgan Stanley, private wealth management, are going to be beating down your doors to give you a loan that's backed by the shares.

Sure. You're only going to have to put down 10%. Sure. But meanwhile, the company's failing to go through a for-profit conversion. Yeah, but I mean, so that appra- Satya's taken 20% off the top. I mean, that applies. No, I'm just saying.

That applies like a small haircut, I would say. I'm just saying there is always, there is always going to be attention. of somebody having massive wealth but not having green backs. Yeah. There's always going to be that tension. Yeah. And people seeing that meta is one of the most liquid stocks in the world. Yeah.

Can constantly trade in and out of it. And it's ultimately just, it's stable. Yeah. So. I mean, if you're interested in liquid stocks, go to public.com, investing for those to take it seriously. Multi-assad investing. Industry leading yields.

Good old-fashioned stocks. Trusted by millions. And who knows, maybe they'll have open AI soon. Anything's possible. You, yeah, we should, we should, you know, Robin Hood has, you know, the tokens around private companies. Maybe we should tokenize non-profits. I want to be able to trade, you know, PETA. You know, let me, let me go long PETA.

Let me get some, let me get some leverage on that. I want some levered options on PETA. Tyler right now. Let's get on the Tyler cam. How are you doing on time? I think I'm in the right mission. Oh, you did it? No way.

15 minutes left. Let's see it. Okay. I can't believe he got it set up so fast. That's amazing. Okay. That's honestly. He's sitting at 26 minutes so far.

All right, we'll have fun in there, Tyler. I just look over and I just didn't realize he was on VR. That's amazing. I think he's actually playing it. This is crazy. number one complaint with with the quest was that it just takes a long time to set up and I think that it well yeah you you intentionally did this challenge yes made it seem like oh I'm going to happily give away my my new my VR Xbox yeah thinking that it was there's no way he could

finish it in yeah yeah yeah but now now he's I think he's gonna do it running well pillar of autumn it's happening anyway uh swix has the story he's breaking it down yeah some good highlights here the mark you mark the mark war has resulted in going from oh the good people don't leave to someone has broken into our home and stolen something one week company shutdown question mark question mark I mean I was going back and forth here I wouldn't be surprised if the shutdown was somewhat strategic yeah in that hey people already need a break we shouldn't have them walking around the office with the way they were writing it it felt like this had been on

the table for months because the company has been working really hard launching a ton of stuff. And it seems like if it just seems like an odd, an odd choice to, if you're in the middle of this Mark v. Mark war, that that would be what you choose to do would be a one week company shutdown. But I don't know, maybe. I'm just, I'm not 100% that like the AI talent war has resulted in the shutdown. It's more just like the shutdown is happening at a weird time. Anyway, this third point is very interesting. Pivoting product launch calendar to AGI race, especially when Stargate comes online. Dylan Patel says December is when Stargate's coming online.

And that should be like the biggest cluster ever and unlock even more maybe pre-training scale. But there's a question about how important that is. And maybe you can do more reinforcement learning in a more distributed way. So I don't know. Also, like, it's not exactly like Zuck is GPU poor, so I don't know, I don't know how big of a deal it is, the Stargate thing. We'll see. But number four, for some reason, meta is ignoring Anthropic. Don't care about safety, if this is true, of course.

The Anthropic point was very funny. Anyway, there's other posts in here that we should cover. Satwick Singh says, what Sam never understood was that Zuckerberg torched 20 billion over VR headsets with Nintendo Wii graphics and didn't even flinch. Didn't even flinch. You know what's so funny? Like those, the, the Nintendo Wii graphics thing, that was like, that was like three months. And then after that, it was, he was doing that, that Lex interview in the Metaverse. Did you ever see that?

Where it's like, it is like photo real immediately. We got a wild demo at Meta HQ. Yeah. Of all the new products and we were blown away. I'm not even sure. I don't think we can say much. I think we can talk about it in like the abstract. But, but yeah, I mean, none of that was like Nintendo Wii level.

graphics but but but I mean that that was that one picture was like iconic and kind of like it felt like oh this is kind of a step back in terms of well many more people saw that picture than actually use the product yeah and so yeah because there were plenty of there were plenty of VR games that didn't have Nintendo Wii graphics like the hero game for the for the quest when it came out was Robo Recall which looked super sci-fi and just looked awesome and it was it was great and there were a bunch of other games but yes I mean the point is is that like uh Mark is totally ready to spend $20 billion on a big project if he thinks it's valuable. And here he clearly does.

And so he's going after it. Anyway, similar founder-led company that's building amazing AI stuff. Finn, the number one AI agent for customer service. We've had own on the show before. He's been on a narration run. Number one in performance benchmarks. Number one in bankoffs. Number one on G2.

Ranking on G2. Let's go. So if you're looking to do customer service, head over to Theron, I'm saying this. They're not saying this, but I'm going to say this is the official, the official AI CX tool of both AI safety and being AGI Pilt. Okay. Because Anthropic uses. Oh, really? No way.

There we go. Okay. Awesome. Yeah, I'm sure. We said you could put our logo on your website. We didn't say these technologies daily. These two podcasters could go way over the. their skis on this endorsement.

It's the official CX agent. Anyway, Rune had an interesting take here. Diving into why these trade deals are so, they're not really trade deals with these poachings, can be so damaging. And he's trying to put it in broader consumer terms. We're going to debate whether or not his point is well taken. He, of course, works at Open AI and has for a number of years.

Rune says, intellectual property by default is a market failure. Single well-informed talented defectors can walk away from organizations with billions of dollars of tacit value of knowing What works and what doesn't the athlete metaphor is somewhat wrong in industries where patents and such aren't viable market participants aren't viable market participants will drastically under invest in R&D efforts due to IP seepage just like the I Drink Your milkshake effect of oil seepage and land rights there is a compelling pro-competing consumer case for non-competes. Interesting. Very rare that you hear a tech person argue in favor of non-competes. Interesting. I mean, the immediate impact of non-competes in this context would be

dramatically lower wages for AI researchers like Roon, which is interesting. The other question is Yeah, and it's interesting you'd have to move AI out of California. Oh, people would then. Maybe. I'm just saying, I'm saying if you really want to use non-competes, go where they can be enforced, right? Yes. California has been historically impossible. Yeah, but then as an AI researcher, you'd say, I definitely don't want to leave California. I'm going to get a much better deal here. My question is, like, I've always been interested in, like, why can, you know, like, the consumer packaged goods industry, like, patent, like, a tab on a can for 25 years and, like, Walt Disney can only.

like the shape of mouse ears for like 80 years but like Google can't own the transformer for like even a minute like I get that it's like some sort of fundamental discovery but I've always been interested in how little IP works in technology so that's interesting is isn't is a byproduct of hacker culture sort of like positive some culture of Silicon Valley yeah but you think that would change if there were so many billions of dollars on the table yeah right Yeah, I mean, there's a culture of competition doesn't kill businesses. So I agree with you, and I'll give you an example. So a poll to refresh on the iPhone was created by a Twitter employee, like a Twitter UX, UI developer built that, built that user interaction.

And then Twitter had a patent on it. Lauren Brickter. Yes. American software developer best known for creating POLD to Refresh. Yeah. The poll to refresh interaction. So we created that design pattern, and then he had a patent on it. Twitter owned the patent, and Jack Dorsey and the Twitter crew were libertarian and said,

you know, we're not going to enforce this. And then it got baked into Instagram and it's now natively in iOS. And it's in all of, you know, it's basically like one button to like. If you're truly terminally online, you're using it hundreds of times a day. True. But now we've actually moved past that. We've moved into endless scroll. where you actually don't need,

like if you're on TikTok, you're never pulling to refresh. You just keep scrolling to the bottom. So we're actually post-pull to refresh, and now we're in the endless scroll. But endless scroll, I believe, was developed by Pinterest, and they had a patent on it for a while, I think.

It might be wrong here. But endless scroll was also something that was not able to be owned, which is just odd to me. And then there's the other question of, like, Coca-Cola's been able to keep the formula as a trade secret for like, years or something and it's just odd that that tech companies can't seem to do either they can't

keep the secrets you can have trade secrets yeah it's not like an open AI researcher can go over and say here's here's a bunch of code let's roll it out right yep it's basically they're the in the the sort of way of doing things that they've internalized from being at a company yeah for a certain amount of time tacit value of knowing what works and what doesn't and so the my my like naive read on this is that if you had had a top-tier open-a-eye researcher at meta during the Lama4 build-out, that researcher would have said, hey, don't focus on pre-training scale as much. We need to get a reasoning model out, and we need to focus on RL and post-training much more, much more. And when you talk to Dylan Patel, I was asking him about Deep Seek and how Deep Seek was doing all these innovations

and moving to floating point eight, like, smaller, basically small. our numbers storing the weights. And he was like, oh, yeah, but open AI was doing that like two years ago. DeepSeek just open sourced it. And so it's like the world found out, if you found out the best practices from Deepseek because they open sourced it, like you were not on the cutting edge and you should have been poaching AI researchers from the top labs earlier. So two things could be happening here.

One, Zuck is right to just go and overpay and try to recreate some of the magic and get a, you know, leading a leading model or set of models out of it that he can use in a bunch of different ways that we've discussed or none of this matters and Rune is is baiting baiting zuck into just overpaying the 40 chess is like oh yeah this is they're getting everything they're taking everything yeah yeah no no I don't I don't necessarily believe that because I did talk to Rune like years ago and I was like but like is this going to commoditize like is the foundation model layer going to get going to commoditize and he was like there like there are like the secret to open AI in many ways is like the tealian secrets like they understand things at a lower level that other

that other people just don't and so that will compound and stay at advantage and he hadn't even predicted the the the dominance of the consumer and like the aggregation theory that happens when you're the front end to consumer AI that's incredibly valuable but just on the research side just on the research side, like they can stay ahead just by understanding like the secrets of what works and what doesn't. But it's interesting. Anyway, let's tell you about Adio, customer relationship magic. Adio is the new is the AI native CRM that builds, scales, and grows your company to the next level.

Next level. Well, I mean, we have a light guest booking today. But, you know, we're spending an hour talking about one story because it's fun stuff. We can continue. We can do, go over to the Wall Street Journal because in the exchange section, there's a fantastic article related to this. It says, it's known as the list, and it's a secret file of geniuses. And I'm looking at this. It says, Mishka Belenco, Yu Zhang, Mark Zuckerberg's been reviewing this list. Lucas Byers on this, the new guy. It says Jordy Hayes, John Coogan, Tyler Cosgrove, Ben Kohler, a couple other folks.

But, you know, it's just the usual people that we know. Anyway, it's known as the list and it's a secret file of geniuses. Only select AI researchers have the skills for the hottest area in tech. Mark Zuckerberg and his rivals want to hire them, even if it costs ungodly sums. We love ungodly sums. We should be ringing the gong more to this episode. All over Silicon Valley, the brightest minds in AI are buzzing about the list. A compilation of the most talented researchers and engineers in artificial intelligence

that Mark Zuckerberg has spent months. putting together. Lucas Beyer works in multimodal vision research, which Tyler broke down for us. Can you imagine if Zuck spun up like basically like a hot or not style tool that's just like pins, pins to researchers together? Like he makes other researchers like just do the test and then just puts out a list of like these are the this is the top 50 people. Yep. This is so good. Um, uh, uh, Lucas Byer, describes himself as a scientist dedicated to the creation of awesomeness. Very redet-coded. But he's opposed to her on X as well.

Yu-Jong specializes in automatic speech recognition and barely has an online presence besides his influential papers. Misha Belenco is an expert in large-scale machine learning who also enjoys hiking and hiking and skiing, whereas he puts it on his website. applying hill climbing search and gradient descent algorithms to real world domains that's very cute the recruits on the list typically have PhDs from elite schools like berkeley and carnegie melon they also have experience at places like open ai in san francisco or google deep mind in london they're usually in their 20s and 30s and they all know each other they spend their days staring

at screens to solve the kinds of inscrutable problems that require spectacular amounts of computing power and their previously obscure talents have never been so highly valued the chief executives of tech Goliaths and heavyweight venture capitalists are cozying up with a few dozen nerdy researchers because their specialized knowledge is the key to cashing in on the artificial intelligence revolution. It is funny. Some of the losers out of this talent war are just VCs because how do you compete? How do you compete when, you know, historically you could have gone to some of these and really overpaid to invests? You could have come to them with a ridiculous. What would be very ridiculous for a VC, which is I will spin up a new company.

Not ridiculous, it happens, but it is kind of ridiculous in some ways. Spin up a new company. I'll give you 200 on a billion, and you can take 10 million personally and secondary immediately. And that is no longer even that compelling of an offer. If the odds of actually winning are low and you just want the most access to compute and be traded. Absolutely, absolute banger ripping. This was Tyler's handywork. Oh, Tyler, how we doing?

I'm in like some kind of maze right now. Oh, no, you lost it? No, I lost already? I think he lost. It's past noon. It's past noon. Ben, what you got on the timer? I've got four minutes still on the timer.

Four minutes still. Okay, well, let's get him another four minutes. You better finish this. I'm surprised this, but I bet you, I bet you if we roll the tape, it took him 30 minutes to set up. And he's actually on track. I bet he's going to finish the pillar of autumn, the level, in normal time, like 20 to 30 minutes. But the setup was a month because I only gave him 10 minutes for the setup.

Mogged. Anyway. This is the reality of pledging. Podcasting. Pledging to be a technology brother. Nobody is in this escalating arms race is chasing the prize recruits, quite like Zuckerberg, who has tried to raid Silicon Valley's top research labs, dangling $100 million pay packages. We're hearing it again to a select few superstars in hopes of poaching them.

The billionaire CEO of META wants them to join his company's new lab focused on superintelligence. One recruit who has spoken with Zuckerberg, who is personally courting his dream team of potential hires, describes the company's goal as nothing less than a transfusion from the country's top AI labs. It's all it is. It's a transfusion. It's blood transfusion. It is wild. Yeah.

Yeah, we got to talk to some more VCs about how they're coping. We got to get some coping VCs on here to hear them cope about. Yeah, yeah, it's actually a fine time to invest in AI. Hey, what do we got here, Tyler? I'm done, I'm done. You're done? No way. What's the clock?

What's the clock? Two minutes, 45 seconds. Wow. No way, he did it. Let's go. Oh, that's amazing. Let's go. Let's go.

Wow. Congratulations, Tyler. Did you even get to play your new? No, no. He had to unbox it. I wanted the unboxing to be part of the experience. Wow. He just ate your lunch, Sean.

He did. Congratulations, Tyler. Very good. That's amazing. Okay, so break it down. How long did it actually take you to get set up and get the game running? Yeah, well, okay, for the first, like, five minutes, there was just like a software update. Okay.

That's what I was worried about. Yeah, and then I had to connect to my phone. Okay. There's like the Meta Horizon app. Yep. And then it was probably like another couple minutes. I had to create a new Microsoft account. and then get Xbox game pass.

Yeah. And then I wasn't sure which Halo it was. It took me like a minute or two there. Yeah. And then the first like probably five or six minutes of playing the game was just cutscenes. It wouldn't let me skip it. It's like the tutorial of how to play the game. And then finally, and then I just crushed through it.

Okay. So how long do you think it actually took you to go from like I gave you the box to you were playing the game? I actually have no idea. I mean, I was pretty locked in. I don't know how long I was playing. Okay, we should have had splits because my big pushback on VR right now is that if you get an N64 or you get a like the old, if you if you got an Xbox for Christmas with Halo the DVD, like you would put that thing in, plug it in to the to the TV and you would be playing in two minutes. Like it was so fast to get up and running. And I've always been upset with how many things you have to you have to link it to your phone, take it off, log in, create a password, like all of.

that I've always been really annoyed with. I think that they should make you do the account creation after the fact, after they let you get hooked on playing and be like, this is fun. OK, cool. Yeah, now I want to create a social profile, because I'm having fun. Yeah, yeah.

I mean, the app seems very unnecessary. I don't know what the app is for. Odd. Because it wasn't like to sign in. I still to sign in on the, using the joysticks. I mean, at one point I had, I had four, I counted like six different passwords that were relevant to play Oculus games because I had, I had a pin code,

that was like you draw it. I had an Oculus password, but then Oculus got rolled into meta, and so I had a meta password. But then I also had a Facebook password that was different. And then I had like a pin for paying with things. And it was just like so many different things. And it's not that hard when you're on your phone and you get one password and you just kind of like auto fill passwords all day long and two factor codes. But when you're in VR, it's super hard to type everything out. So wow, I'm very, very impressed that you pull that off.

I thought that that was going to be a failure. So congratulations. Enjoy it. Great work. So break down the experience of actually playing. Were you playing on a big, big 2D screen, and you were sitting there in like a theater, basically? Yeah, yeah.

So the screen was probably like six feet in front of me. It was like in front of this light. It was like great experience. It was good experience. It was good experience. I didn't find it to be like, like the refresh rate was like totally fine. Great. I'm not motion sick at all.

No screen door effect or anything? No, it was totally fine. And then and then also break down. So you, Were you technically streaming the game over Wi-Fi? I believe so. I'll look, but I think it was like cloud gaming. Yeah.

So I think the actual Xbox, like frame rendering is happening in the cloud, and they're streaming that to you and then rendering in VR, and then the headset's just pinning that to the world, which is really, really cool. So I have this thesis, and you'll have to play with the headset more and let me know if it plays out that I think that, you know, Palmer Lucky has that take that's like soldiers will be wearing VR headsets on our daily basis before consumers are because the military can underwrite like a $5,000 VR headset.

And so he's betting on augmented reality glasses for the warfighter being the first real use case. I have this thesis that, you know, before people were using the iPhone for dedicated iPhone apps like TikTok and Uber, people were using to make phone calls. Like it was a phone first. And I think that that will be basically a TV first. And people will be playing traditional video games and watching traditional movies before we get VR GTA 6. People will just be playing the GTA 5. And so I think all the VR company should focus on making the movie watching experience and the traditional video game experience like super, super smooth.

So that you just get in, you can actually chill in that thing for two hours, actually beat Halo. So maybe, maybe, you know the whole game only takes three hours to, to beat. Maybe you got to be on the street now. Okay, gamers. Anyway, let's get back to the news. Let's get back to the sleep scores. How did you sleep last night, Rudy? I was absolutely mugged.

No. What happened? Get a pod five. 69 rough. I got a 92, baby. Let's go. Generational run. Hit me with the Ashen Hall.

You know the rule. If I beat you, I get that. sound effect. Ashton Hall. 92. Go to aidesleep.com slash TBPN. I'm worried to think about how I would have slept if I didn't have an eight sleep. And, you know, the only thing that I got penalized on was I went to bed too early.

9.30. And they're like, you don't usually go to bed at 9.30. And so I pulled it off. I was penalized by screaming children in the middle of the night. Brutal. Inside the secretive world. No, we have some breaking news. I'm going to dive into this. So this morning, Zuckerberg announced his super intelligence team.

We got 4,000 likes on this traded post. No way. Tyler. Oh, here on a generational run today. That is amazing. Okay. Wow. 4K likes. Okay. It's awesome. Back to the breaking news. Yeah, yeah, please.

So Zuckerberg sent a message this morning in a memo. introducing the new super intelligence team. So we were just covering his list. Yep. He got a seemingly a lot of people from his list. So Zuckerberg introduced Wang, who will be the company's chief AI officer and leader of MSL, as well as former GitHub CEO, Nat Friedman. Friedman will co-lead the new lab with Wang with a focus on AI products and applied research.

Here's the list of all the new hires as seen in Zuckerberg's, memo. It notably doesn't include the employees who joined from OpenAI Zurich office. So this is Kylie Robinson's reporting over at Wired. So Trappett-Bonsle pioneered RL on Chain of Thought and co-creator of O-Series models at OpenAI, cracked. Shoo Chow B, co-creator of GPD4 Voice Mode and O4Mini, previously led multimodal post-training at OpenAI. Hua Wen Chang, co-creator of GPD4O's image generation and previously invented Masket and Mews text to image architectures at Google Research. G. Lin helped build 0304 Mini, GPD 40, GPD 4.5, GPD 4.5, GPD 4-0,

ImageGen, and operator reasoning stack. Joel Pobar, inference at Anthropic, previously at Meta for 11 years at HVM, Hackflow, Redex, Performance Tooling and Machine Learning. Jack Ray, pre-training tech lead for Gemini and reasoning for Gemini 2.5, lead, gopher, and Chinchilla, early LLM efforts at DeepMind. Hong Yu-Ren, co-creator of GPD4-O,

and this just goes on and on and on. But the big takeaway here is that these seem to be some important people and not what Sam had originally messaged as none of our best people. Wow, that is a big... But obviously things have changed since. I think we need some more trade deal posters except it's not much of a trade deal

just full poaching. It's pretty pretty crazy. No, this there will whenever whoever writes the book on this era Ashley Vance. Obviously. We'll have at least a chapter dedicated to this talent rate. Yep, this is huge. Pretty unprecedented.

But yeah. I think the difference, the difference when when Facebook was rating Google in 2010. It was just spread out over a much longer period of time. And it wasn't... It wasn't... I mean, this is unique because

in that situation, Zuck was poaching from Google. Zuck poaching from a company that still is like many people view as a startup, right, even though it's worth hundreds of billions. So... Yeah, this is pretty remarkable?

Turner Novak said, is it true? They poached Matthew McConaughey from Salesforce. I like that one. Anyways, so we could go back to this list, but clearly he made a list. He checked it twice, and he got a very meaningful number. It's time for Zuck to head over to Bezl and get him something, get himself something to remember this moment by, because his Bezle concierge is available to source him any watch on the planet, seriously any watch.

He could get a new Richard Mill. He could get a new Vachron. a new Patak. He has a cubitist. Maybe he needs to fill out the stable a little bit after this. Pretty remarkable. Yeah, the other thing that's interesting here, Daniel Gross is not anywhere in this memo.

Oh, interesting. Who was rumored. Is the full memo listed, or is it just like they just didn't include this because it's just wired reporting and they don't realize how big of the deal, DG is? I think they would have included it. Yeah, maybe like the negotiation is still like ongoing.

Like it is it is a more complex situation because it's like a CEO leaving an active company. And so it's not it's not like a full like free agent or acquisition or, you know, going from W2 to W2. It's a little bit more complicated than that. So fascinating. Well, the AI talent wars are the current thing and we have been enjoying following them. It's a great time to be an AI researcher. Even better time to be a podcast. Yeah, incredible overnight success.

For all these researchers. It's not like they spent, you know, 10 to getting PhDs. 15 to 20 years, you know. Also in a field where people were like, oh, that will never happen. Like that, like this is crazy sci-fi. This is not like going to ever be real. Like you're just, you're doing like complex math. Cute chatbot.

Yeah. Cute. Does it work? No way. Yeah. Cute picture that makes a morphed hand. Yeah. Look at that hand with six fingers.

You idiot. You idiot. You bozo. Oh, you're working on image, Jen? Yeah. Why does that person have? Oh, deep dream, the thing that puts dogs in every picture? Oh, yeah.

That'll work. Yeah, right. Yeah. Cat detector. Hot, what is it? Hot dog? Not hot dog. Yeah, now it's the most valuable.

technology in the world. Yeah, I'm super, I'm super interested in seeing, like, what they build. Obviously, like, it's going to be, like, a good model. Like, we know that they're going to try and build, like, a big, good model. It's going to be open source. Maybe it might not be. Maybe not. There's, like, Mark hasn't guaranteed that he'll be open source forever.

Actually, the model is really good now. We're going to, we're going to charge for it. No, but he could. He could. I mean, like, Joe, open source, yeah. I mean, it could be closed source. There could be open source variations. It could be, you know, a chat-driven product that directly competes with chat GPT.

It could be something that just lives behind the scenes. He's going to, yeah, yeah. It could be a tab. It could be a new tab. You know what? I actually like, I would like to own a piece of search too. I got a bunch of DAUs. They're searching stuff all the time.

I kind of like that market. I think people might want to discover products that way. I think you could build a pretty good ad product on top of that. It is interesting, though. I mean, it feels like LLMs right now are so single player and meta products are so social and communication and entertainment oriented. Launching a consumer. And the social thing is like it's right around the corner.

It's right around the corner. Already I will, I've done this with you where I will go and generate an 03 pro response, let it cook for 10 minutes, have 4-0 generate some charts or graph. or images off of it, and I will just share the chat with you, and then you can trace through the whole thing as you like. And that workflow back and forth is amazing. Ben Thompson wrote about it, how he has a research assistant who isn't at the level of him in terms of analysis,

but can ask a bunch of questions, and then he can go back and see, okay, he asked this question slightly the wrong way, so he got slightly the wrong answer. Let me just continue chatting with the same interaction in GPT4 or 4O, and actually improve. and get the answer that he wants.

So it's fantastic. Are you about to hit the gong for something? Are you're just excited? I mean, there's like 20 researchers that just signed a meta. Do it. And I'm just going to do it. What is that? Why did you throw it?

We're just experimenting here. Yeah, it's funny. Dan Ratcliffe in the chat says, every researcher hire I see, I think, give Jordy's tweet on companies doing NBA-style trade deals, opening eye trading researchers for Anthropic, throwing in a couple of 20-26. Waterloo grads. The only thing here is it's very one way. So I called it, but I kind of didn't call it,

and that Zuck's just just absorbing, you know, he's not saying, he's not, turns out in tech, everyone's in free agency. Yeah. At all times. At all times. Yes, there are no binding contracts that can keep someone at a company for five years. It would be, yeah, it would be very interesting if sports shifted to that model where there was no, long-term contracts. And you could just mid-season be like, oh, I guess Tom Brady's going to the Eagles. Like, Sequin Barclay is being traded mid-game, like at halftime. He switches teams because there's a team that paid more. Like, that would be possible, I guess. And that's what we're going through here. I don't know. We'll have to get Tyler to give us a breakdown of these folks on

tomorrow's show. Maybe we could go through these folks and talk about. some of their history, what they've actually done, who's the most legit. I wonder how important the, what's that called? The Dunbar, is it Dunbar's number? No, I forget. There's a, there's a quantitative way to assess how powerful, how, why can I, jikster, Dixter's number? I forget.

Bench press. Yes, bench press. No, basically you look at the papers that they authored, and then you look at how many citations that paper has. And so if really great researchers are citing your research, then you are a great researcher, that type of thing. It's kind of like a six degrees of separation, like the Kevin Bacon thing.

Anyway, should we go to Apple F1? You know, it's also in the Wall Street Journal. Yeah, let's cover this journey, journal piece. Can Brad Pitt's F1 movie finally deliver Apple a big screen hit? It did. It did. This is old news. Also, it's funny because the paper version says, Apple is now trying to make people, make movies people actually watch. Do you see how brutal the written headline is compared to the online one?

So the online one is, can Brad Pitt's F1 movie finally deliver Apple a big screen hit? Like, okay, that's like kind of shade at Apple. Like they haven't really had a big screen hit. But then this one is just like, oh yeah, they haven't even been trying before. And now they're trying to make movies people actually watch. Grudal. Well, you probably, if you're familiar with F1, you're probably familiar with it,

some proofs from some billboard ads you saw. So if you're trying to run billboard ads, go to adquick.com, out-of-home advertising, make easy and measurable. Say goodbye to the headaches of out-of-home advertising. Only Ad-Quick combines technology, out-of-home expertise and data to enable

seamless, efficient ad buying across the globe. It's crazy how, so the, I was off earlier. Yeah. This is Apple's first box off at, hit. It's a $55.6 million debut. And the polymarket had it. There was under 43 million, 43 to 47, 47 to 51, 51 to 55, and then over 55 million. So

it had it at 701% just about a week ago. And that ended up being true by half a million. So close but accurate. Oh, interesting. I thought it was like a hundred and 40 million. What was that number? That's different? I think I... Okay, so they actually clocked it pretty well with the... Okay. Anyway, we'll get the details here. So, in the Wall Street Journal. It was, so, so it was domestic...

Oh, the domestic box office versus international? Yeah. Those in the numbers. Okay. Well, Ben Fritz and Joe Flint over at the Wall Street Journal have the story. They say, when the team behind this weekend's Brad Pitt movie F1 first pitch their idea around Hollywood, nearly every major entertainment. company wanted in and a bidding war quickly erupted. Then Apple bigfooted them all. Tech companies, they love big footing. The tech giant agreed to spend nearly $250 million, more than two AI researchers,

and more than any other studio, thought the drama about an aging driver's last shot at glory was worth, said people with knowledge of the matter. Pitt was paid more than $20 million standard, more than the $20 million standard for A-list movie stars, and we'll get a cut of the film's revenue if it's a hit. F-1 is one of Apple's biggest entertainment bets since it leapt into Hollywood in 2019 and embodies the unusually lavish and meticulous approach to the business, which has brought little commercial success. People who have worked at or with the company say Apple executives crave hits,

but only want to produce content they believe reflects the high-end aspirational qualities of the iPhone and MacBook. There's little chance, in other words, that Apple would make a critically reviled blockbuster like a Minecraft movie. Yeah, I talked to a very successful Hollywood producer about this dynamic with Apple that so much of being a movie studio is, it's a hits-driven business. You've got to slop it up sometimes. You got to do the kids movie like Shrek. You got to hallucinate a little bit. Yeah, Shrek is not is not an Oscar winning film, but it's a hugely valuable franchise. Franchise.

Yeah. And so he was basically saying that Apple had one hand tied behind its back because of the brand standard. And that it would only want to do things that had a polish and a prestige and that if you're expensive. Expensive. And if you're if you're only doing like half of the surface area of what a movie studio can do, you're just going to get eaten by Netflix. That's going to say, like, yeah, let's put the reality show on this show. Let's do this.

Let's try this. Let's try some prestige. Yeah, let's do Love Island. Yeah, let's do Love Island. Whatever. And Apple just wouldn't do that because it's not the brand. But it seems like they're figuring it out. And F1 is a good example of that.

So with sleek production values and strong reviews, F1, is exactly the kind of movie Apple wants. It was made by Top Gun Maverick, director Joseph Kosinski, and producer Jerry Brookheimer, and Apple executives have privately said they hope it will be the land-based version of the airborne blockbuster, which grossed $1.5 billion in 2022. Wow, you think you could get up there. That would be fantastic.

But pre-release surveys indicate F1 is struggling to generate interest among audiences beyond older men. Its success or failure will be a referendum on Apple's ability to meld carefully curated. And I got to go to B4. Carefully curated content with broad appeal after six years in which it hasn't released a single box office hit

To date worldwide box office of Apple original films Killers of the Flower Moon Did not see that one 2023 160 million Napoleon that was an Apple movie? Oh 221 million I saw that did you see that people didn't like it I thought it's pretty good that was long, but it was fun Argyle 96 million that one didn't do too well Fly me to the moon, 2024 did 42 million. You have not seen any of those, correct? No.

Fantastic. F1 will be available. It's okay. I haven't seen UFC 724. Who cares? To each their own. F1 will be available on Apple TV, the company's streaming service,

after its run on the big screen. Apple's TVs plus. You know, I knew this movie would be somewhat significant. Yep. I felt inclined to watch it. Yes. And so knowing that it was releasing, I thought it was releasing, I was like, okay, great, I'll watch it on Sunday.

I'll watch it on Sunday. Okay, and what happened? And it wasn't on Apple TV. It was at the movie theaters. Yep. And I didn't want to go to the movie. You're allergic to movie theaters. When was the last time you were in a movie theater physically?

When we saw Dune 2. With an absolute squad. Yeah, we did. This is the movie to take the boys to. We got to do it. It's hard to schedule, but we got to do it. This is a call to action for all the fans. Many podcasters are descending on Malibu for the month of July.

This is the thing. It's true. Maybe we'll get it done together. So, yeah, and Ben Thompson's been beating the drum on this, this idea of, like, Taylor Swift understands windowing better than, like, Netflix executives. Have you seen this thing? No. So basically, like, windowing is the concept that when you have a,

valuable media property, you should monetize it at increasing rates of cost based on time. And so when Taylor Swift comes out with a new tour, you can first go and see it live in person at SoFi Stadium and tickets are like $100, maybe more, like a couple hundred bucks. It's expensive to go see it. Then months later, you can see the movie version of her tour in movies. movie theaters, but it's not on streaming. And so then you're paying $20 instead of $200. Then after a couple months of it being exclusively in theaters,

then you can stream it for free if you're subscribed. So it's like, let's call it like an average cost of like $2. So you went from like $200 to see it live in person, $20 to see it in the theaters, $2 equivalent to like stream it on Netflix for like 10% of your Netflix spend that month or whatever. And then eventually it'll be like free on TV or whatever. free, it'll be even free, or maybe she'll put it on YouTube. And so it's this very, it's a very good way to price discriminate and get the maximum

amount of value out of a piece of content. And Hollywood had been very good at this for a very long time. Like, you would come out in the movie theaters, then there would be a big gap for like three months. You just couldn't see the movie at all, no matter what. It was like, you got to go see it in the theater because if you don't see it now, you won't be able to see it for six months. Yeah.

And then the DVD comes out or the Blu-ray comes out. And it's like exclusive, just and you have to buy it. And then, and then, it's, and then, it's, you know, after that it goes to Blockbuster you can rent it and then it goes on TV or HBO later and like it might not be it might be a year or two until you can actually go and see uh the movie for free and so Hollywood was really good at them they were printing money and then then you know the streamers came and disrupted all this and like the binge era started and you can just watch everything for free constantly um but ben thompson's point is that they kind of went too far and they started giving away

things that people would actually go see in theaters for free on streaming. And so even this, I think there's going to be a lot of people who are like, okay, it's an Apple TV Plus exclusive. Like, how long do I have to wait to just watch it at home? Oh, like two weeks? Like, okay, I'm not going to go see it in theaters. But if they window it hardcore and they say, look, yeah, it'll be on Apple TV Plus in seven months, then people will be like, well, it's worth it for me to go see it now.

Anyway. Well, we have some notes in the chat. One, Michael Ginsburg says, Community Note, Shrek won two Oscars. Oh, yeah. Owned. Owned. Onged. Onged.

Is not only a cult classic, but it's an award-winning film. Wait, did it win best animated? No way. Because that's actually hard one. Fact-checked that, Tyler. I don't know. Best C-GIS or Best Music or something. And then Blake, Blake Dreyer says, just got back from F1 IMAX, Incredible, as a fan of the sport.

Very well done. I think you got to see it. I think it's also the type of movie that you want to see. Turner says it wasn't a bad film. Cool. Nick Kotav says it was just too long. Have one? Too long?

How long is it? Over two hours? Got me 90 minutes. Give me in, get me out. Two hours, 36 minutes. That's so long. Anyone that would put out content, more than two hours a day is just completely out of line and unhinged.

We would never put out three to four hours a day. This is a pot call in the kettle black situation, but we're doing it. Yes, so Shrek, according to. to Michael says best animated feature and best adapted screen. Okay, okay, I say it corrected. Let's give it up for Shrek. Do we really think Shrek meets the brand standard of Apple? Like, no way, right?

My point holds. I don't think Apple. Tim Cook would kill to be able to put out a film of that Calvon. You think so? Well, the Wall Street Journal disagrees with you. They say that they wouldn't put out the Minecraft movie. anyway Apple TV Plus has earned a reputation

for content that excites critics more than audiences, including the Leonardo DiCaprio, Robert DeCaro. Content that excites critics. Yeah. Apple TV. There's a lot of examples of that in media. There's a lot of tech media that's just for the insider for the media folks.

It's journo on journal content. Killers the Flower Moon. Oscar's Best Picture winner, Coda, literary adaptation, Pachinko and the spy drama, Slow Horses. Yeah, I tried to watch Slow Horses. It's pretty good. It's few mainstream successes include sports comedy, Ted Lassow, sci-fi workplace drama severance,

and straight to streaming film The George or the Gorge. We're into entertainment to tell great stories, and we want it to be a great business as well, says Tim Cook. He told Variety on Apple TV plus spokeswoman declined to comment. Analysts have questioned why Apple spends billions annually on a business so far afield from its core consumer electronics operation, even if it's barely a blip for the company worth more than $3 trillion. The strategic value it brings is sufficiently mysterious for people not to talk about it very much. It says Craig Moffitt at Moffitt Nathanson. Good buddy of Ben Thompson's.

Apple TV Plus costs $10 months. And yeah, so a lot of people think that the Apple TV Plus production team is to like take pressure off of the like 30% Apple tax discussion. Anyway, we have our first guest of the show. Will is in the studio. Welcome, Be. back to TVPN. How are you doing? How's it going? Congratulations on the trade deal. So what supercar did you buy? You go Bugatti-Sheron or you go F-80. I mean, we hear that the numbers are huge right now for folks like you. I mean, I'm taking a plane more. So like that was. Yeah. Yeah, yeah. Yeah. People say, you know, we wanted flying cars. We got a hundred four-year-year-carters. that you would imply that Will drives himself is offensive.

I mean, yeah. Well, I'm Waymo for the first time recently. I'd like, I'd been to like SF, but like I didn't been around SF proper that much since the waymo craze took off. And so like that's been like the car experience really. Just like self-driving cars, damn. Yeah. So are you in SF full time now? Still back and forth, New York.

I'm like kind of slowly starting to spend more time than I saw. It's on brand. You're decentralized. You're decentralized. Exactly. It's a perfect fit. Perfect fit. Anyway.

Did you touch grass at all this weekend or were you locked into the timeline? A lot of stuff was happening on the timeline. I did, I mean, I'm in New York right now. I did some New York stuff on Saturday. But yeah, I mean, it's been, it's been good. Lots of drama, intrigue, as usual. It's a drama. The trade deals. I know I've been tuning into earlier. I know you guys have gone over all the major roster updates.

But yeah, it's exciting times. Open AI, open model soon, hopefully, but not this week. Yeah. So, yeah, I mean, the first place I want to start with is the actual trade deals. Do any of these people jump out to you as particularly interesting? Had they been on your radar? Have you met any of these folks? Can you give us more context on how academic are they? How product oriented are they? It feels like this is purely research, you know, beefing up of the research team

and meta kind of has products solved. Is that the right model to think about this? How can you characterize the shape of the team that Zucks building? Even there's thousands of really great software engineers. engineers that work in AI. Sure. And, but what, you know, even kind of getting a sense of like how many people was, was Zuck really going after?

Was it 250? Was it 100? The list that came out today is, you know, somewhere, somewhere, I don't know, there's like 20 people on it, something like that. Right, right. Yeah, I mean, it's a great list. It's a lot of like very senior, senior researchers who know, who have, like, been through some, like, real product cycles who, like, I think the goal is people who can bridge

the gap from research your product. Because like one thing meta has a strength of, but I think doesn't translate in like this era is they have a lot of really good like academic researchers. Like Yon-Lakoun's whole org has a lot of very talented, very capable researchers who write important papers, but it's the kind of stuff that doesn't really turn into exciting products.

It's like more betting five or 10 years out, how do we want to do things eventually, versus like going through a pre-training cycle for a Gemini, or a clot or an opening eye model, like that has a different set of considerations. And so this to me looks like they want like 10, 15 great people who have been through those research to product cycles in terms of the model being the product. And it seems like they probably they got Alex and then supposedly Nat and or Daniel to kind of hand, handle a little more of the, not the business side, but product-y business stuff.

Yeah, it's a management role. Yeah, yeah. Yeah, what is your take on Jan Lacoon right now? How should we think about his position in the AI world? There's, you know, I see these memes of like he's a non-believer. He doesn't believe in like super intelligence and like AI God. He's a little more practical. It seems like he's been, he's been.

right about some things in the sense that like we haven't had we we're not feeling a fast takeoff right now so I think you got to give me some credit for that at the same time like there's Ben Thompson's been criticizing on Lacoon a little bit for not driving the the product side of the business it's like it's one thing to say that hey we're not going to go straight shot to AGI God but then the the downstream implication of that is that if you believe that we're going to be living in like B2B SaaS world of AI implementation and productization, productization, then like, go do that and like make sure that the models work really well for business.

And so I don't know. What is your take? How would you describe Jan Lacoon in his history and kind of where he sits in the organization to like a layman? Sure. Yeah. So I think people overindex on thinking of Jan as the meta guy like too much. Sure.

I think it's a shame that they didn't necessarily have someone in the law. Lama org who was as visible as him because he was not involved with Lama at all. Like fair fundamental area research is like a whole other group that they mostly write papers and do very academic work. Jan Lekun's work is very academic. And it's really like it's I think he's like I think on one hand he like is right in the sense of like we don't have the full picture yet. Like we don't have full on AGI that can like do a human's job forever.

And I think that's kind of the drum he's beating. is like there's more work to do, there's more science to do. It's not just productize the current set of techniques in terms of like the end of AI forever. But his goal is not to do the productization. Like he, like the org he's in at meta is really isolated from product. He's there to kind of boost the reputation of the org academically as well as to be able to potentially advise on other stuff that's more productive, but his job is not to like go train a business model. That's a model that is going to be used by businesses. And like, I think people like that org fair,

the friends I have who are there, like, are there because they essentially wanted to be a professor, but they want tech company resources and not to not have to teach. And that is essentially what that org is. Yeah. Does that lead to better recruiting on campus essentially? Like you can go get research, because Jan Lacoon can come do a really powerful lecture? Maybe, but I think of it more as like the same way that Zuck did Metaverse stuff. Zuck is very willing to make 10-year bets. And so that org for meta is not about what product are released next quarter. It's what are our 10-year bets.

Yeah. Okay. What about, sorry. Yeah, you can go. So Meta shares just hit an all-time high. Really? So the stocks performing. I'm curious from your time at Morgan's staff.

Thank you for that, John. It's incredible. That is incredible. But generally up 5% for the last week. How did Wall Street generally sort of process the meta-AI story? Because now, obviously, at least the market seems generally excited. it may get to the point where we did with VR where people are like, hey, whoa, whoa, whoa, cool it on the Metaverse, you know?

Well, it's funny because like you add all this up, even if the $100 million offer number is real, it's like, okay, so now we're at like $1 billion in spend. It's 5% of the Metaverse hole that was burning. And it's like, you know, from a Wall Street perspective, you're like, I don't care about, I don't care about $1 billion. I care about $20 billion. Because that was what was weighing on the stock during the Metaverse.

business build-up. But this is like AI is so much more productizable. You can make so much more money from it up front. And it seems like even though it's a big number, $100 million offer, it feels like the spend is maybe less. But I don't know. What do you think, well? Right. Yeah. I mean, I think the, I don't know the context fully of the 100 million. I don't people have all these theories, but like the scale AI numbers we do know. Yeah. And like that that's a big chunk of their like, I don't know, they don't do those every day. Yeah. But they've done a handful in the past at that scale. Like WhatsApp was 20 billion or something like that. And I mean, it's hard to read into like the charts. Like I think people are generally like expecting that Meadow will take AI seriously and are kind of

of happy to see change. Whether that is like justified or not, I mean, we'll see. Yeah, it was interesting. George. The product side is going to be interesting because like they're not a B2B SaaS company. Well, so I want to push back on that. They're an entertainment company. Yeah.

I think that I don't, I don't think that people are fully understand the potential of Gen. A.I. around entertainment. Like it gets talked a lot around, oh, you're going to be to generate, you know, an entire movie or generate video games or things like that. And I think that we haven't seen, we've seen some fantastic examples so far, but nothing, nothing that is, I think, like, George Hots was on our show and he was talking about how, like, basically AI is going to be, like, having five CIA agents follow you around all the time, convincing you to buy

products. And, like, that is, like, one kind of dark bowl case for, dark bullcase, yeah. Dark bullcase for meta in the context of AI because it's like, it's possible that that Facebook is already the best advertising product in human history. Like period, hands down, there's nothing better. And then could you make it like two, three times better? It's very possible. Yeah. I want to push back on the B2B thing because yes, yes, they don't sell B2B software. But I was thinking about it in this terms. Like if you, if you were running a company where your entire you only had one client and that client was meta and you sold them CRM and infrastructure and LLMs to improve the back office and do censorship and and you know

reality checking and and looking for bad words and looking for improper posts and recommendation algorithms like how much would that company be worth and how much revenue would they be making just from that one client. And I think it's in the billions because it's just such a large organization that it's potentially like just the B2B applications of AI inside of meta are like, it's a multi-billion dollar cost center

or revenue driver or something like that. I don't know. I think a company at Meta Scale has already rewritten all that stuff themselves for the most. Like they probably have some services. Yeah. That they're using, but like they, like they could they rebuild their own cloud

because they use AWS for a lot of stuff. Yeah, yeah. Can they're not a compute company though. And in terms of like platform stuff. Yeah. I feel like those companies, once you hit a certain size and you just like have so many engineers that you want everything like Google rewrites all their own stuff. Yeah.

Microsoft rewrites all their own stuff. And can they, is there another like a tale of end of that has not been done yet that they can squeeze some more money out of her? Um, entertainment will be interesting. So one thing that just, I was just reminded of, have you seen the Italian brain rot videos? Oh yeah. Love them. Yeah.

Yeah. So, like, I know people, it's like silly, it's stupid. But there is something there in terms of this like communal narrative storytelling with a level of vibrancy that we haven't seen so far. It's like Looney Tunes. But people just like come up with these where you could imagine meta kind of eating a good chunk of the like video gen market. If they have an answer to like B-03 where that becomes part of the plot. form it ties into the whole metaverse thing of like creating stories and sharing these stories with your friends um but instead of like instagram stories it's like okay what's what's the evolution of

stories yeah have you have you seen higgs field no field i think it's an ex-snapchat team yeah their new image model is basically already at a point where it's creating images that oh yeah i've seen photo realistic, but in the scent, photo realistic, like some like some like full-time influencer generated it after spending hours trying to create it, and you can't tell that it's not real. And so it feels like, it feels like meta-integrating that kind of thing where it's like, it'll be interesting to see how this actually plays out because everybody will be able to be a super tasteful creator or like generate these sort of unique styles. and I'm sure at that point it'll switch everybody will be like well I only follow organic farm-to-table

influencers yeah yeah I feel like they're uh AI integration so far have all been kind of weird like there's these like Instagram accounts you can chat with and then there was the whole meta-a-i homepage disaster where old people were like leaking personal info without realizing it But I guess Facebook's always had that. Yeah, I think it'll, they'll have to thread the needle on product in a way that we'll have, they'll have to get creative because they haven't. Like, it's been a while since they had a product innovation that was homegrown that really stuck. Yeah, it is interesting that they haven't tried to create a studio Ghibli moment by your like going. I was saying this.

They should just pre-render everyone. profile picture as a Ghibli and just when you open the Instagram app it's just like here we did this for you it would be incredibly expensive from an inference cost but then it's just like you could share it and it's just more virality around it but I don't know and like even baking that down into a filter no it can be so like open AI branded now they have to go yeah yeah I'm not saying yeah I'm saying like an entirely new but yeah yeah basically like the the filters in the Instagram app should be like style transfer or like fully generative instead of just like color grades. That's clearly the next thing and they should just and they should just do that.

But what do you think of Sam Altman's ability to get through this? Like he's had a series of exodus and seem to continue to march on the anthropic departures, then SSI and thinking machines like XAI. Like this is not the first time. It's not his first rodeo. It's not his first radio. And so there's this whole narrative like never been against Zuck. He's been down on the Metaverse, came back.

He's been down on mobile with the HTML5 thing. And he came back from that, went native and super dominant and bought Instagram, WhatsApp, super dominant and mobile. And so never been against Zuck. But then also maybe never get bet against Sam. Maybe they both win. And maybe like the real loser is like, I don't know, some other company or something. I don't know. I think as long as Open AI doesn't totally drop the ball on product.

research, like they have the center of gravity for the AI world. Like, in the same way that no matter what Android ships, people are not going to switch from their iPhones, like, someone that have to go very wrong for people to not see Open AI as the winner, I think. Or there's the kind of the default. Yeah, I mean, the other thing is the average chat GPT users not waking up today being like, oh, I can't believe OpenAI lost a handful of its top researchers, right? They're just still using ChatGPT as a Google replacement or they're talking with it as a companion. So, yeah, I think the lead is still very, very real.

And I just, and again, it's going to create opportunities for people to say, hey, I want to go work on this product that hundreds of millions and soon billions of people use. every single day. Right. And I do think it'll get to a point, especially with like 20, 25 into 26, where there's a lot of product stuff that people seem like, I think we're early on products and we'll keep getting better versions of the same things in a way that's like kind of predictable. Like the image models will keep getting better.

The agents will mess up less. But like we can already use those things and we can start building a group of concepts around that. And the question is like, okay, what are the winning apps? apps. Like I imagine that like the real time, clearly kind of thing, opening eyes going to do their version of that at some point. Other people probably will too. Like does that become a modality that people really want like live overview on their screen, on their phone? Like

what's the what's the way that people are going to be interacting with AI generally? Because a lot of those use cases like it's they're already smart enough. It's not about making the models like way smarter at whatever. It's like how do you have the model be useful or fun engaging? There's a bunch of paths here going back to the Cluelly thing where one, Roy is like right about the U.S. And like wins the market. And the other option is like he's right about the UX but like doesn't win the market.

And then there's like, you know, it's just not the right form factor or whatever. but at least the first two outcomes are hilarious in that and that Roy Lee ushers in the new the new paradigm of for engaging with language models. I'm curious, is all the stuff around every B2B SaaS player descending on the sort of like single interface, like a chat interface that generates software? Was that predictable to you? Do you think that that's part of a multi-year trend or is that just FOMO? I mean like co-pilot was early. It was like 2020, 2001, like the first GPT3 co-pilot came out. And like that was already one of the early like LM applications that people were interested in at all.

And then it took until cursor for it to really like I think cursor plus like 35 summit was when it became a thing that was good enough that people were excited about it. And it really ushered in the trend because people were starting to find it more useful than a toy and like a thing that actually want to use day to day. And so I think that's one path is like and then the more background agent kind of things are like starting to take off now, which I imagine like those will get reliable enough that they're like useful for cranking stuff out. They already are kind of depending on what you're doing. Yeah. Like I think yeah. We're about halfway through the year. What are kind of like the big moments that you're kind of tracking?

The OpenA.I. Open source model could be one. Whatever meta launches next, right? I'm assuming they're going to be dark until this new team can really cook and bring something great. Maybe they don't do anything this year, but I would assume they come with something. What else are you tracking? Yeah, I mean, for me, the 03 release in chat CPT was like a pretty like game changer kind of thing where it was like, we saw with like deep research that like, okay, they kind of figured out how to make agents work, but it was also just like this one version of an agent. And oh three, you can kind of get it to be a pretty general agent where it can like do some pretty complex stuff that was kind of new to see from like the geogessor thing was crazy.

And having that as a like vision of like what. what AG and I starts to look like. I think it's pretty cool. Of course, from the like research open source world, there was a deep seek as like the RL craze taking off. But like, I mean, I work on RL so I like obsess over it and like think about it a lot. But I do think we're really starting to see these recipes,

at least in the broad strokes of like, okay, here's how the L-LM thing can go. We figure out what we want it to do. We give it some tools. We set up these environments. We figure out how to evaluate it. And then we can just kind of like let it go. and these things get better at doing those things via trial and error.

And so like I think that is one way to kind of forecast where things are going. It's just like, what are the plausible use cases that people want to use all this for? They want an agent to do XYZ. And then how do you make this a thing that you can help climb? And I think like you see a lot, like there's been a lot of news about opening I trying to like sell their RFT service. There's these startups popping up. There was a- Was the R-T service again?

reinforce and fine doing. Okay, yeah. And so, like, they're pushing this pretty hard. This is the, if you're spending over $10 million like Pallentier OpenAI, will also customize a model for you. So they have like a serious heavy, like a heavy paying customer one, but they also have like a more like in the thousands of dollars service. Sure.

Where you are getting to essentially do fine tuning on O4 Mini. Yep. And that is a little more self-serve. But they're also like doing consulting around that before deployed thing around that. Is that an important strategy for Open AI to kind of stay in the game against open source models like Lama? Yeah, but I think also against other like thinking machines, there is some report that this is a version of the strategy they're getting is like go to enterprises, talk to them about their problems, turn these problems into things where you can. create really good customized agent models. And if the, like, that's one potential roadmap towards like having more like AI everywhere.

It's just like having services that where whose job is to like come in, whether or not it's fine tuning or not, just like make them into agent shaped tasks where you can then optimize the model and have someone craft the model experience like for you. Because like I think a lot of enterprises are. still like in the spot where they don't really like there's there's not as you see by the like the talent war there's just not that many people in the world or in the market who really understand this stuff at a level where they can go make it happen and so like the open source thing on one hand it's cool that you in theory can go do it but it's also like there's not that many hands

who can who are equipped to like go make the thing happen actually fine-tune llama four yeah or whatever Exactly. Yeah, yeah, we actually talked to a startup that's doing basically that, like small models for specific business use cases. Like, okay, you just have a ton of CSVs that need to be turning to JSON and we build you an LLM that just does that or whatever. Interesting. How do you, after the scale news and then there's a bunch of players popping up or, you know, everyone from Mercore to label box to handshake. a lot of different people competing for that market. How do you see that market evolving over the next one to two years?

I think there were certain people saying scale got out at the perfect time in many ways. But clearly there's a lot of at least gross revenue up for grabs right now in the near term. Yeah, I mean, I think like the broader sphere of like creating stuff to train models on with humans in the loop to do that curation is going to be important, especially for these domain specific applications. I think it's going to be less like, oh, we just need more tokens. And it's more about we need curation of goals and objectives. Because like tokens, you kind of hit diminishing returns pretty quick in terms of just like more internet text or more like human, human, written math solution examples.

And it doesn't scale super well. But the nice thing about the past specification is you can kind of like pour in more compute without necessarily needing more data. It scales much better with compute. Whereas we don't really have like a great way of like spending 100x more compute per pre-chain token other than make a giant model. Yeah. Do you think Lama will stay open source for the foreseeable few? future?

We'll see. It wouldn't surprise me if they go kind of the Google route where like they are still, they still do open source. They still have the Lama brand, but it isn't like their flagship thing where they start doing whether it's for internal products or it's they really want to show that they're the winner releasing closed models via API or other services where. Yeah. Do they have the infrastructure to serve a closed model?

I mean, you mentioned that they use AWS. So it'd be kind of awkward, right? Yeah. I mean, I don't know that they really want to that much. I think that it'd be something that comes in a product. Sure. Yeah, that makes sense. Or maybe they partner with, like, you could see them partnering with someone like a Corby,

or Nvidia directly. Invidia is hoping to get into the inference serving game, it sounds like. Yeah. Do you think, so one of Swix's takeaways from the, the, article was that like one big development that's coming is when Stargate comes online opening I will have the largest single cluster for pre-training but it feels like we might be at the end of that game and we might be doing more compute intensive work in a distributed fashion and so maybe having it all in

one place is a little bit less relevant to just you know oh Trump card Stargate boom, I win. I have the best model by far. How do you think about the impact of Stargate on like the AGI race? I mean, I think the biggest experiment that they're definitely going to do is take a pretty big model. I don't know if it'll be like quite as big as big as like a 4.5, but like a big model and do way more RL on it than anyone's on the app. Like what does like an 05 level of RL look like? Okay. See how good that is.

See what new task is that unlock. though or could you do that just across a bunch of data centers? Because I've heard like a lot of RL is like you're generating, you're doing verifiable rewards, you're generating in a bunch of different data centers. You could do it completely distributed and you don't necessarily need a Stargate to do that. You can do it distributed. Like it is very inference heavy. There is still like a lot of weight updating. You have to sync the model across. You have to keep sending the training model to your inference workers. Okay. And so having it co-located certainly makes it easier. It's easier to do it in a distributed way if you want to.

is pre-training, but it's not like trivial. But I think like a lot of this is just data, people are building the data centers without necessarily knowing exactly what experiments they're going to run. They just know more stuff is coming, but also a lot of this is gonna be inference where like they wanna be able to do like meta has a ton of GPUs they just used for Instagram ads.

Recommendations. A lot of Stargate will probably be serving Ghibli. I mean, that makes a ton of sense. Like, I'm running into rate limits on... Except I haven't seen one in two weeks. Like, like, Google. It's like, how do they run out of TPUs? They spend 60 billion on GapEx every year, at least.

And O3 Pro, too. I mean, I get timeouts on this stuff. Last question. I'm guessing you weren't surprised to see OpenAI leveraging the TPU for inference or reporting to be starting to use it. Was that kind of? I mean, it makes sense that they are considering all options given the friction with Microsoft. Like, they don't have a core cloud that they're super close with.

They're not, like, there's, they need to both try to lower their costs for entrance as well. Like, that maybe was a factor behind the, because the O3 had a big price drop. That might have been TPP related. I don't know. Pure speculation. But like Google has TPUs and they can serve. Jim and I really cheap and really well, really fast. If Open Eye wants to compete at that level, they kind of have to consider all options.

Yep. Makes sense. Well, great day to have you pop on. Yeah, thanks so much for helping. Knowing how things are unfolding, we'll probably ping you to jump on again later this week. Stay ready. Good.

Great to have you. Cheers. Yeah. Bye. I'll be right back. And we got another Will coming in the studio. It's Will Day on TBPN. We got Will from Ulysses coming in.

Not talking about AI, talking about the ocean, talking about boats, talking about the gundo, military. Who knows what we'll talk about. But we will talk about the origin of his company. How you doing well? I'm good, John. How are you getting on? I'm great. Good to have you on the show finally.

Would you mind kicking us off with just an introduction on yourself and your company? Yeah, absolutely. My name's Will O'Brien, one of the founders. of Ulysses, we are building autonomous vehicles to solve the most critical tasks in the oceans. So in practicality, what that looks like, we build an autonomous surface vehicle. It's a 30-foot vessel that launches and recovers swarms of autonomous underwater vehicles. These underwater vehicles can read as well as right. They can sense.

They can take maps. They can take images and collect data, but they can also do physical manipulation as well, robotics. They can do kind of whatever you want, whatever you want to task them to. And we make both these vehicles probably about 10 times cheaper than anyone else in the world. So while some people are building maybe humanoid robots as a general purpose robotics platform for land, we're building the maritime equivalent to that. Humans obviously are not evolved for land. So we need something for that. So that's a bit about us. We're based in San Francisco as you can tell. I'm not an American. I'm Irish. and some of our founders

finding team are Irish and Scottish and yeah, that's a bit about us. How did you settle? Like, yeah, yeah, what was the inciting like problems set? Like, how did you land on this industry to work on? Like what got you in the industry? Yeah, good question.

So two years ago, one of my co-founders, a crazy Scottish man called Jamie Wedderburn was on a search trip. And he got talking to a, biologist working on this plant called seagrass. He'd never heard about seagrass before. I had never heard about seagrass really before despite growing up by the seaside.

And he became captivated by it. And then he, like, you know, like any good autistic engineer does, he goes down a rabbit hole learning about it and telling all his friends about it and telling us about it. Seagrass is this wonder plant in the oceans that supports 20% of the world's fish stocks.

It holds about 20% of the carbon in the ocean as well. Underneath it, it's 35 times. better than rainforest or removing carbon, but it wasn't in finding Nemo, so nobody knows about it, really. So we kind of discovered this and, you know, in learning about it, he realized like it's dying off all around the world and it's basically very time intensive and expensive to bring it back to restore it. And basically the methods people are doing to restore it and our manual, their divers going down with their hands and they're manually planting seeds and takes a lot of time, takes a lot of money. And it's not really working. And he kind of had, you know,

we kind of had this insight then that while we've mechanized, you know, nearly every kind of form of large scale outdoor work on land, that really hasn't happened in the oceans yet. Like nobody's built the tractor equivalent for the oceans. And that really like a lot of the kind of tasks in the ocean, whether it be in this, you know, narrow, weird, wonderful domain of seagrass restoration or in the world of defense or in the commercial industry, if you're building infrastructure, we're still sending down tons of divers to do manual tasks that really just make absolutely no sense. They're dangerous.

They're expensive. So we kind of started in this like where the sea grass got very obsessed with that and had a very successful kind of first year just doing autonomous seagrass restoration services for different governments around the world working with governments in Australia and the US. We did a million dollars in revenue last year and we're on track for over two this year. Congratulations. When I met when I met Will and I heard this like initial wedge, I was like, this is so fascinating, interesting, like, find, like, the most weird, obscure market that no one would ever think of and just absolutely dominate it.

It's awesome. And then use that to earn the right to do many other things. And I think I introduced Will to, like, 10 seed funds in, like, 30 minutes at one point. Without getting the opt-in from Will, it was like, people need to meet you. And then within a few days, people were committing. That's awesome. Yeah, it's good to see you on the show, Will. I learned to scuba dive off the coast of Catalina Island.

And when I was, I was like 15 when I was doing it and it was, had all this amazing tall California kelp. And I went back a couple of years ago. And you've heard this story. There's like an invasive kelp species that came over from Japan, I believe, on like a shipping container, shipping tanker. And it completely wiped it out and it looks completely different. And I don't know if it's particularly important to restore the original California kelp.

I imagine it probably is in some abstract sense. But it was just a bummer as a scuba diver to be like, It's not what I remember, which was like swimming through this beautiful forest. It's just kind of these tumbleweeds now. Yeah, it's, man, it's those fucking purple urchins. That's right. It's, that's it. What happens, a big, a big way of, like that invasive species are transmitted in the ocean.

It's like a ship will like, bam, you know, be in the, in Tokyo and in the port. And then it'll fill up its, it's like, it's ballets, so it'll pull in water. And when it's pulling in water, it'll take in fish or urchins and this. instance and philup and then we'll go out to sea and then it'll land in san diego or san francisco in the port and then it'll release them all so this happened as well with lionfish um in the Caribbean um in Miami and uh uh yeah those lionfish are nerdy they're just like chewing up the coral everywhere um i i went i went calling them once you go i was in the in the Cayman islands uh working there one summer and we were getting paid like spearfish these yes i've heard this

yeah i mean that is it tastes great yeah that is great that is the good side of the lionfish is that you can just hunt them relentlessly because no one cares about them because they're so invasive and they're like pretty fun to shoot and you shoot them and you can make leather wallets with them yeah yeah i'm making lionfish leather wallets and you're like i'm actually i'm hunting but i'm saving the world yeah yeah but these like pigs in texas yeah it's like pigs in texas i mean it's actually a problem and and it needs to be like really dealt with but as a as a consumer of hunting it's fantastic oh it's great but yeah these are these are good examples of past again that are just like still today it's like ours dude and spears going

down just like yeah like they're really realistically there are that many people that want to go spear fishing lion fishing all day every day forever because like the people like yeah a couple tourists doing it once the weekend is not going to do the job so yeah you yeah you need automation 100% yeah and like we've yeah so that's the kind of these are the types of this is like you know the same platform that we built for sea grass it's a service vehicle's underwater vehicle with a different robotic payload in a different model on board, fine tuned on finding urchins or finding line fish. You could, you know, we could have like a service that just like solving this problem for someone.

What's the origin of the name? Yeah, it was, there was kind of a few things. I mean, as most great ideas, it was conceived in the midst of drinking a rake of points. And we were, we were, me and my co-founders were out. We were just like thinking about different names for the business. And then we were thinking about it that. 1997 animated movie called Treasure Planet

where these people that go on this quest and the submarine to kind of save the world and we looked it up and it was called the submarine in that movie was called Ulysses and then interestingly when we got our first office we got our first office in Dublin just after we raised the precede in October

24 or 23 and we got our first office and to the left of us was the house that James Joyce grew up in. I was sorry, the Ulysses has the Irish connection as well, which is, you know, another reason we chose it. But to the left of us, there was a house that the James Joyce grew up in to the right of us was the house, or the Museum of Literary, Ireland, where the first copy of Ulysses is kept. And in the, in the park opposite of the street, just there's a big huge statue of James Joyce with a quote

from Ulysses there. And that also, like a bit a week later, my mom found a photo of me at home next to like a picture or next to like a James Joyce quote and there's a EEOC's quote underneath it. All this kind of happened in the same week. So it's kind of like, okay, we're keeping it now. Yeah, because the nominative determinism here is that, I mean, the story of Ulysses Odysseus is like a 10 year grind. Yeah.

And it's like absolutely miserable like facing the Cyclops and the getting enchanted by Searcy and the sirens. And like I imagine that if things, if nominative determinism holds. The sirens for. you or you won't be getting a quick buck in an aqua hire at meta anytime soon you're going to be at this for a very long time i'm rooting for you and i think i think you will ultimately ultimately be successful but uh buckle up you're going to go on an entrepreneurial journey if the name holds what uh who are your who are your biggest uh irish irish tech uh heroes what makes the irish so

infatuated with technology um i mean what makes irish people are so infatuated. I mean, you guys were well aware of the Hibernian conspiracy, of course. Of course. As two, I don't know what you guys have to say about your, the accusations that, you know, that Mr. Kugan and Mr. Hayes are at the core of this cabal of Irish people in technology or never beating the allegations. Never beating the allegations. What about Pachy McCormick? What about Molly O'Shea? What about all these people with Irish surnames? What are they doing? Why do they all suddenly have media programs. Why do they all follow each other?

Why does Apple domicile all of their cash there? It's a good question. Yeah. You think you think it's the Rothschild's, you know, running the money splight, actually policies. Exactly. So now she, Satoshi Nakamoto, Nakamoto is actually an Irish name. The old father on Nakamoto.

So there's all these connections here. It goes, it goes deeper. My favorite Irish technologist would probably be George Boole, the inventor of Boolean Algebra. Oh, there you go. foundation for modern computers. Yeah, I didn't realize he was that. Wow. Deep cut.

Yeah, and then modern, modern ones. Owen McCabe's a good friend as well. Big final one. I'm everything they're doing on Intercom. It's very, very inspiring to see, to see someone like that. And obviously, the Collisons as well. But I feel, I feel the Colossons,

everyone appreciates the calls on. I don't think, I know people appreciate the great business that the guys at Intercom are building. And I know you had him on the show recently. We just partner with him. Yeah, he's a sponsor now. Ring the bell Ring the bell

Yeah Yeah you can have in your tinfoil Hat conspiracy There's definitely a line between us and Own and then us in Intercom And own appeared on the show He just happens to pay us through Stribe, you know

I have I had a random question for you That I'm sure you have strong opinions on Every now and then I'll see a video That's off the like coast of Chile and there's just like just fleets of fishing vehicles as far as the eye

can see what's going on down there. Is there any solution to it? I'm sure you've gotten asked that. Is your guys tech something that could help address that? It is, yeah. And we're speaking to some partners about kind of supporting on this problem right now.

I think there's like two aspects to the problem. So yeah, like characterizing the problem. The Chinese are destroying many oceans around the world, particularly in countries where they basically just don't have the money to monitor their seas and enforce marine protected areas and these sorts of things. So up and down the coastlines of South America, this is like the single biggest thing that like South American navies are spending time on right now. West Africa as well is just being completely like the fisheries there, being completely decimated

in and around the Antarctic as well. There's massive krill populations and krill populations there. they're this kind of foundational species for a lot of other marine life in the ocean. You take away the krill, which is the kind of bottom of the food chain, you know, it knocks all the way up to the whales. And then you lose the whales as well, and nobody wants to lose the whales. So that's kind of like characters. The problem, okay, the reason the problem exists is, okay, two reasons. Natives that basically just can't cover enough area and get maritime domain awareness, they just can't, they don't have enough assets.

And, you know, that's part of the problem. And there's like a governance aspect of the problem as well, right? So how does China get away with it? Well, so, you know, the people responsible for kind of looking after, you know, law on the high seas quite often is what you'd call the flag state. So where the boat is registered, all these boats are registered Chinese. Chinese kind of like knows this is their kind of, they use this as what, you know, you'd call like gray warfare. And it's kind of similar with the critics on the sea infrastructure. You know, it's not really officially the Chinese state, but it may as well be because they're not like prosecuting on it.

And so on the technical side, I suppose like the problem here is like the, again, the paradigm of maritime operations is kind of, you know, is stuck and it's broken, right? It's, you know, you could probably characterize it. We call it this the heavy tech bottleneck, right? So basically any mission you want to do in the ocean kind of has like a few features around it. It like it generally, you know, relies on these expensive manned assets. So these could be airplanes like P3 or ions or they could be, you know, expensive national security cutters, these big host guard cutters, these 100 foot, 200 foot long vehicles. And these costs like tens of millions in capex. And then the running cost for them on a day is like tens of thousands, if not hundreds of thousands of dollars to run daily. And then you use these things as like a platform on which you put sensors or you use it like an unmanned system.

And that's kind of like what it relies on. So it's like small numbers of very expensive assets. Whereas the way we're seeing autonomy going around the world, whether it be like self-driving, transfer networks like self-driving cars or drones in Ukraine, is like we're moving away from these small networks of exclusive vehicles to large networks of small autonomous vehicles. Right. So it's like going from buses and trains to like a network of Waymo's or Tesla robo taxis. You know, on the battlefield, it's moving away from fighter jets to like a bajillion nearost drones. You know, it's like, it's like this is like just the way we're going, but that hasn't really

happened yet in the ocean. And the limiting factor we think is that basically drones are still like crazy expensive in the ocean, on-man systems are crazy expensive. Even if you want to buy, you know, the sexiest on-man surface vehicle, you're still paying hundreds of thousands of dollars underwater vehicles. You're spending hundreds and thousands, if not millions of dollars. But we haven't had like a DGI yet for the ocean, right? We haven't had someone that really completely collapsed costs to the degree that you can buy

fleets instead of single units. So again, that's kind of like what we're trying to position ourselves as is like this person who you can buy. If you're like the Argentinean Navy and your fisheries just getting obliterated, you can buy like a swarm of surface and subsea vehicles to kind of solve the problem for the fraction of a cost you purchase a mandass for. That works 24-7 with zero optics. How much? How much is the problem that these governments don't know something is happening? they know, but it's 30 boats that have showed up somewhere and they're just going to fish, you know, and basically decimate all life until they're maybe pushed away by a boat.

Like what is, I can see the technology angle of getting just like more coverage and knowing more about what's happening in these different areas, but it feels like there needs to be more will from the government as well to just. say like we're not going to let this happen. Yeah, I mean, the, the, the, the will is like certainly there. It's just they don't, they don't have the budget on current assets. Like, it, like, you know, like some of these, like, if you want to buy more of these, like, um, like boats that they would need or these planes that they need to like survey or get,

like awareness on the area, like these are like tens of millions, it's not hundreds of millions for some of these assets and the Argentine Navy just doesn't have that kind of money. Um, so they know that in general there's a problem there, but they just don't have enough assets out there. Um, I mean, it's crazy. that like most of the Argentinian Air Force's time is now spent on looking for Chinese ships over their waters. It's, it's, so it's like, they just, yeah, they don't have the money.

So, yeah, you really do any cheaper solutions for these types of folks. What other, what other stories are you tracking or kind of problems globally that that you think you're? I want to hear about like this big shipbuilding bill. Like, how does the new budget fit in? What are you tracking in D.C.? I know you're still an early stage company, but I imagine that, um, the long term has to be figure out a way to get to program of record, work with the DoD, but break it down.

How are you actually thinking about that? Because it feels like there's a lot of different applications. You've got to stay dual use. You've got to stay agile while you're building the underlying tech. But what are you looking at in DC? Yeah. Like for us on the kind of government DVD side, there's kind of three plates that were always spinning. There's like the DC place.

There's the like operator or warfighter end user. state and then there's like the acquisition teams it's met place so we're always just like in constant dialogue and with like you know different members of these three and these bodies so yeah the the most recent um you know appropriation for for defense is very interesting um there's many many billions going to maritime unmount systems particularly surface vehicles and underwater vehicles i think there's a general recognition um on the hill right now that like you know the you know the chinese are producing like hundreds of ships per year and you know US is producing one ship per year and that like trying to beat them on ship building is like probably not feasible we're not

going to ramp up that production that quickly we need to just play a different game and I think they're seeing like on man systems is like the the way to that and so yeah just this week it was announced again and another nine billion you know a few months ago as well there was another four billion allocated to this line at them so and yeah they're the things that that that that we're tracking, there's some existing program of records that are already set up that that were in the kind of running to kind of challenge for and get a position on. And there's also talk of some new ones being set up for some of these newer categories of vehicles that haven't been developed yet. But yeah, out of the three plates that we're spinning right now,

it's generally that you know, you do want to have like a presence in DC kind of like at this early stage. It's not the main focus. The main focus is we're, finding our product with the end user understanding what their needs are and building the link to spec. So that's kind of like our main focus right now and spoken to speak to some some folks in Navy about that and members of DHS and Department of Homeland Security and other other groups as well on this. So they're kind of our main focus is right now in that front. Probability P Atlantis. What's the probability that you guys find it? I think I think very high. I think it's probably very high. Very high. Very high.

You think that Atlantis exists? Yes, I think there's two very good candidates for Atlantis. There is one structure in Mauritania, in modern-day Mauritania, I call the recat structure. So the recat structure is basically three concentric rings in the middle of the desert with salt deposits between each of them. It looks man-made. You can see it from space. You can look up the recat structure, R-C-H-A-G. And the only description we have of Atlantis is actually from Plato.

Plato describes it as being this like walled city full of gold with, you know, built with the kind of city hall being at the middle of three concentric rings. The ring, you know, the rings were each filled with sea water. And Martin de Mauritania is also where the richest man that ever lived existed. So there was this Mauritania king who had. More gold he would have been the I think the you know the first trillionaire And yeah That's here for and and he and and and so that that's the gold part of it the recad structure the three rings of that part of it And then there's also these all these like salt stray actions all the way from the coastline west africa to this location so it's kind of the theory that there was this

And there was this apocalyptic event about 11,500 years ago called the on related to the comet hitting the earth and at Atlantis was there. There was a great flood and this city was lost. So that's one location. The other one is in the Atlantic Ocean, which then would lead to some of the first survivors potentially leaving the Atlantis and going to Ireland to set up the next modern-day civilization. Okay, okay, we're bringing back. This is fantastic. Next time we're getting to the bottom of the Bermuda Triangle. But thanks so much for stopping by. This is fantastic. We'll talk to you soon. Well, great having you on well. Thanks for us, last. Cheers.

Bye. Really quickly, let me, when we find Atlantis, you know what we're doing? The cure for male loneliness being called Lads. Lads. What are we doing? What's the first thing? Set up a Wander. That's right.

You got to find your happy place. Find your happy place. Book of Wander with inspiring views, hotel grand amenities, dreamy beds, top tier cleaning, and 24-7 concierge service. It's a vacation home about better. We have a video dropping with Wander this week. We do. We do. And we have a new guest dropping on the stream.

We are surrounded by. journalist. We're going to bring in Kylie. Kylie did the reporting that we covered for Wired, covering the leaked, the leaked memo. Very cool. So let's bring her on in. There she is. What's happening?

Welcome to the stream. Hello. Hello, my first time. Thanks for having me. First of many, hopefully. How's the day going? You've been scooping, scoop it up a storm? Yeah, it's been busy.

Just that this morning. and tweeting nonstop and now you guys. Fantastic. Awesome. So what did your weekend look like? What's like break down like a weekend of scoops? It's pushing a single messenger to the max. Well, it was Pride weekend here in San Francisco, so that wasn't fun to combine.

I helped a friend move, and then I got the first scoop and ran to Pride and Dolores, drink a bunch of white applause, ran home, slept for 12 hours. And my editor got to scoop. Then we're scooping again. Monday, first thing in the morning, scoop it again. So what, everything's been moving quickly? What was the first scoop and maybe like break that down? Yeah.

Yeah. Yeah. So I think that was Friday. I reported on a research paper that's not being released at OpenAI because it talks about the definition of AGI and sources have said that allegedly managers were saying that the Microsoft deal in those negotiations were a reason for not releasing that paper. And it's really, really tense as the journal and Reuters have reported those negotiations. Then Saturday for Open AI researchers went to Meta's new lab.

Sunday, Mark Chen, an executive at OpenAI, sent out a memo saying they're basically, meta's basically robbing our home. My editor published that one. And this morning, Mark Zuckerberg sent out a list of all his new hires. It's been fun. Huge. So we've got to ask, who's your source? No, I'm kidding. Not you guys. You're not going on the record as you guys. Yeah, yeah, yeah. No, no. So, yeah, I'm curious kind of what you expect next in this saga. Do you expect? I mean, it clearly, Open AI is not seeing Zuck announced his current list and being like, okay, we're safe now. We don't, you know, we don't have to worry. But at the same time, you know, it's very possible that the worst is behind. them. Yeah, my question I talked to Will about this was just like, were you surprised that the

memo didn't lean more on the fact that this has happened like three or four times at Open AI and they've gotten through it every time? Like I would be leaning on the fact that like, yeah, Dario left. Elon started XAI. Thinking machines exists. SSI exists. We still have 90% penetration in consumer AI, walk down the street, ask someone what they use for AI. They're not even going to say open AI or chat GPT. They're just going to say chat. And just leaning into that, that, hey, we are this compounding dominant force in consumer AI.

This is the place that you want to stay and be as opposed to getting, I think, a little bit more in the weeds of like the nature of the deal. It felt like an appeal maybe not restating the missionary element. It was very defensive. I don't know. And came across. I mean, they were clearly scared. Yeah.

Yeah. It seemed very focused on the actual dynamics of like how one of these offers comes together as opposed to just reiterating the strengths. Here you get to work on the consumer product in AI. Yeah, it was like a lot of messaging around, please talk to us if you're getting offers versus let's rally. Let's ride. Let's ride.

Yeah, I think it's, you know, a big part of it came over the weekends. These people are just, these executives are scrambling to tell staff, like, for the love of God, do not talk to anyone at Meta and sign any offers. And they're going on this like week-long break next week. So I think they were just trying to get ahead of that. But I think that your guys' argument, that's right. I think it's just like a very emotional, high stress time. They're trying to ship two very large models.

And Mark Zuckerberg's coming in, calling employees directly and offering millions and millions and millions of dollars. So I think like their response is emotional and defensive. because it's like just so high stress. But that would have been a good tactic and I don't doubt it's coming eventually. How solid do you think you are on the $100 million offer? That's been floated multiple times. And we were kind of digging through the game theory of it.

Like in one way it benefits Sam to get that number out because if someone, if Mark tries to poach someone for 50, they might be like, I've been hearing that everyone's getting 100. Like, come on. I thought I was one of your top guys. I thought I was near the top of your list. And then, you know, in some ways it benefits Mark at Meta because if you're hearing about these $100 million offers, you might be more willing

to take that preliminary phone call because you know that they are ready to put money where their mouth is. And so it kind of benefits both sides. Dylan Patel was on a podcast a couple days ago talking about how they're not really sending $100 million wire with cash on day one that there's earnouts to these structures and they can claw it back if you don't say. And so how much nuance do you think there is on that number specifically? Because it's certainly headline grabbing.

The Wall Street Journal, I mean, they put it in the special infographic right here. Heck, yeah. Because $100 million is a nice round number and it reads very well and it works in a headline. But I'm just wondering if you think it's 100% real or 50-50 or there's more nuance to it. Yeah. So my take is the Wall Street Journal in the New York Times. Both printed that number and they have really, really rigorous. editorial standards to get to print.

So they must have seen something, some high-level sources that proved them that that number is real. I, for one, have not seen that number. And for me, I would have to see an offer letter with that number. But knowing how tech salaries work, I don't think it's just like a wire transfer. It's, you know, going to be golden handcuffs. It's going to be over a certain amount of time. And, you know, the information had a really good story, breaking down, you know, researcher salaries.

And just imagine if you came to Open AI a couple years ago, at 80 billion, 40 billion valuation and what that is now. They don't really need all this crazy money from meta. But it is attention grabbing, attention grabbing for sure. It's also just funny. When you say $100 million for one person, it sounds like a lot. But if you were to say, we're going to spend $5 billion to put together a team that is super core to our entire business and win, it doesn't feel nearly as significant for a company

that is a multi-tralion dollar company. I'm looking at the WhatsApp acquisition price and number of employees. I'm pretty sure they paid more than $100 million per person because it was $19 billion, $4 billion in cash, $14 billion in Facebook shares, which are probably worth like $120 billion now because this is in 2014. And at the time, WhatsApp was a remarkably lean team with around 50 employees. So, I mean, not far off. I mean, what if they had?

Slightly different comp, though, just given that they were buying a product with... Totally. Yeah, yeah. The product's super valuable. But, you know, the researchers that can build the next model and bring Lama to the frontier makes a lot of sense. What else should we dig into here? I think that was it. Any, yeah, I don't know if you touched on this specifically, but any, any, like, specific predictions about what comes next?

or are you just waiting, waiting for one of your little birdies to send you a note? Well, if you're watching birdies, my signal is at Kylie.01. I will be impressed if they get someone of Ilias, Ilius subscribers capabilities. I think that the people who want to build superintelligence are true believers. And you're giving them a value proposition of, we will give you so much money, but you're making superintelligence for Mark Zuckerberg. And I think that some of the best brains that really believe in AGI's this prosperous future or, you know, perhaps the end of humanity don't necessarily want to do that for Mark Zuckerberg. So I will be impressed if he gets one of those minds.

He still has about 50 people he wants to hire for this team. So it's just going to keep coming. I think a lot of them are going to continue coming from the top labs. But what the vision is, what Mark's vision is, I'm waiting to see. So what about last question? What about Daniel Gross? It had been initially reported, but he wasn't on the list that that you released earlier. Yeah. Yeah, I noticed that neither was the team from Switzerland. I'm going to guess this is pure speculation for Switzerland. They have EU laws that make it strange to employ them right away. But for Daniel Gross, I'm not sure what happened. I saw that was reported by the information with Nat Friedman. I've tried to be in touch with anyone who could tell me if Daniel's going. But he was. wasn't on the list. So I guess we'll see. Yep. Very cool. Well, I wouldn't be surprised if you start getting some major offers. I could see the AI researcher space is heating up the AI journalist. It might be the next domino to fall. So yeah. Oh, God. We're here we're

here to help talk you through any, any sort of offers that you might get. Perfect. You guys can take a percent. We'll work it out. Fantastic. Awesome. Well, great, great having you on. And yeah, excited to follow the rest of your reporting this week. Thanks, y'all. Cheers. Bye. Nice. Next, we have Joshua.

From Chai Discovery with a big launch this week. It's always tough to launch when the biggest trade deal in tech history is going down. Not even a trade deal, the biggest poaching. We keep calling it a trade deal. It was non-consensual. Not much on the other side of the trade. It's a non-consensual trade deal. It's more of like a draft.

And stock is just drafting from the entire. entire league. I take 12 of your best people for nothing. Yeah, he's kind of just building an all-star roster. It's like an all-star weekend over at meta this week. Anyway, Joshua, welcome to the stream. How are you doing? I'm doing great. Thanks so much for having me. Excited to be here. Of course. It's great to have you. Great work, you know, launching on such a wild day in tech history. But we are here. Is the time run really getting like completely steam world? by this story? I don't know.

It feels like that. I feel like there's also people on vacation and stuff. But break it down for us. What is the big news? Well, the big news is we've just announced Chai 2, which is our latest foundation model at Chai Discovery. Chai 2 is actually allowing us to now generate antibody sequences with a success rate that's high enough for us to skip the usual high throughput lab testing.

This means that you can go take a target, put that into our model, go generate a set of potential protein sequence. and then test them in the lab, and they tend to work with extremely high hit rates. Previous methods were like 0.1% or so, so you'd have to screen, you know, hundreds of thousands to potentially find something. But here the hit rates are actually upwards of 15%, approaching 20% of success. Amazing. Break down some immediate applications of that tech across maybe different types of categories. You want to ask about making a more potent testosterone? Maybe we get to that next.

Okay. Yeah, maybe that's the right spin out. But yeah, no, it's really like that we said we would never incubate it. I think one thing that we're really passionate about is going after areas that are really hard to do today. So, you know, like this allows us to do things faster and cheaper potentially, but I think what's really going to be game changing is just opening up new kinds of biology that it wasn't possible before.

So we've got scientists that are like, you know, maybe working for years trying to enable a certain class of drug program. And if we can now use machine learning to jump to potential solutions, that could really just really kind of push up the bar of what people need to do. I think it's take a step back. Like look at what happened with LLMs. Everyone has to be an AI company now to compete, right? What's that going to look like in biotech where if you don't have access to a technology

like this and then you start to have these certain areas that are closed out to you? Like, what is that going to look like? So I think that that's something that we are really excited to figure out. So if benchmark it to the LLM market, do you see yourself taking more of an anthropic path or an open-eye path? And what I mean by that is, or do you develop the drugs? Do you wind up being the front end to this and kind of being full stack? Or do you go more like B2B and you have a bunch of pharma companies or, you know, traditional biotech companies that are buying a product or service from you? Yeah, it's a great question.

I think if we take any lesson from the LLM market, it's that the winners are going to be the ones who are most agile. The technology is evolving at a faster clip than we ever expected. Our company goal for this entire year was to get to a 1% success rate on this task. And we're halfway into the year and we're already past 15%. When you have progress like that, it is really hard to predict how the market is going to evolve. And that's why I think just like being flexible, making sure that we have the best technology, making sure that we have the best way to use that technology as well. You know, it's not just throwing a model over the wall and you get a drug.

We really have to make the models work well for the specific applications where they're most important. And we also have to make them much easier to use. We have to get them into the right people's hands. There's a lot of scientists that if we can bring their creativity and ingenuity together with these new models, that's where the magic's going to happen. What's the state of the art in, like, the actual impact of antibodies? Like the last time I heard antibodies in the news was when Joe Rogan was taking monoclonal antibodies for COVID infection. And that was one of those weird things where it felt like a very modern. Totally not controversial.

Yeah, it felt like a very modern piece of science. And the narrative around him was that he was not adopting the modern science, but he was kind of on the cutting edge of that fork of the tech tree. And so is our monoclonal antibodies like the most, like the latest and greatest? Or are there other applications that are particularly interesting? or just driving value, even if they're, even if there are indeed like the traditional way. Yeah, so about half of drug approvals these days are actually biologics, protein drugs, so antibodies. Onoclonia antibodies are actually just one type.

That's closer to what, you know, we've been exploring in this paper. But I actually think what's going to happen is that people have now started to push beyond that. They're basic, like, bi-specifics are all the rage right now. That's basically, you can think of it as two antibodies that are kind of stuck together. So you can actually go after multiple targets at the same time, or you can actually get like a more potent behavior for single target. You know, it's kind of just like two things in one. So if we can make, you know, the first, if we can make each one of those chains like easier to design, then, you know, bringing them together is going to be going to be even easier as well. So I think it just like raises the bar for the kinds of therapies that you have to work on because it becomes easier to make those drugs.

So that's a monoclon and antibodies are. They were really important in COVID, right? They can neutralize viruses. The way that people found them during COVID was actually one to patients who had COVID. COVID, they just like, you know, you would go and just like take someone's blood, for example, and try to find antibodies that are in there and then try to develop that into a therapy. But now we can just make it on the computer in 24 hours and then go test it in the lab and have some results in a few weeks.

So there's a lot more work to do to make sure that these molecules will actually work in the clinic and will work well in patients. We have early signs that they have drug-like properties, which is really nice to see. But we've got a lot of work ahead of us in order to really bring that to production, if you will. There's still some more engineering that you might want to do on the molecules that are coming out of the platform today. Yeah, what is your sci-fi vision for the state of drug development 50 years from now and how would the FDA have to adapt regulations to enable that? Because I can imagine my vision is like you come into some type of doctor clinic setting, you have some set

of conditions and they're just sort of like auto-generating drugs on the fly specifically to your current conditions, but that obviously wouldn't work in the current paradigm. Yeah, I mean, that would be the dream. The North Star that we have for this space is something that we think of as like zero shot drug candidates. Right now we have these zero shot proteins, right? So we specify target. We get some protein that finds it.

But the fact that this is possible makes us believe that someday we'll be able to actually just like have the models think about all the other properties that we need for a drug. And it's possible that the first sequences that come out of the model will actually be good enough to eventually move into patients And yeah, if you can do that then you know maybe the next stretch is like can we actually do Hyper personalized medicine, right? Maybe there's something where we take some sample from a patient We figure out what exact target we need to go after then we go and generate the drug I mean there's multiple leaps that need to happen

But I can't predict what's going to happen 50 years from I like honestly I failed and predicting what's going to happen in six months I thought you're going to be at one percent Do you use Alpha Fold or is it just a separate tech that is kind of complimentary but not in your direct supply chain? Yeah, the foundation models are built in house here. So we don't use Alpha Fold. We actually can't use Alpha Fold. The latest one isn't available to the company. So we built our own.

We have a state of the art model for predicting protein structure and its interactions with molecules. We actually open source that. So we started the company a little over a year ago and then six months in. And that was the first thing we did because you really need like an atomic level microscope if you're going to design drugs. That was our thesis when we got started. But we realized that, you know, that kind of level of the technology, we thought would get out there anyway.

So we open source that. It's being used in most of the major pharmaceuticals today. I think the community has gotten a lot of engagement around that. But now the models are becoming even more complicated. So that's why we're starting to invest in not just the model layer, but how we actually make those models useful. They're really starting to look more like pipelines. It's kind of like chat GPT-O-3.

It's not just a single inference. There's actually a bunch of complex tricks that you need to run if you're going to get the best results. So that's kind of how we're thinking about it right now. Very cool. Well, congrats on the launch. I'll have to have you back soon. This is fantastic.

Yeah, with the rate that you guys are moving, I'm sure you'll be back on in a couple weeks. Thank you. Congrats to you and the whole team. We'll talk you soon. Yeah, really appreciate it. Cheers. Thank you guys.

Up next, we have John. from Camp Fire coming on the show. Welcome to the stream. John, how are you doing? I missed this. You haven't been hitting the soundboard enough. I love it. Boom. Welcome to the stream. How you doing, John? Can you hear us? Yes, I can. Thank you so much for having me on today. Fantastic. Do you want to kick it off with an introduction on yourself and the company? Absolutely. I'm the founder, CEO at Camp Fire. We're based in San Francisco. And I was a finance exec for about 15 years and just lived the pain. Thank you. Notting software. Don't get enough credit. The unsung heroes. The unsung heroes. Never featured in the all hands, but we're working on changing that here.

And lived in spreadsheets my whole life and just helping, I don't think they're going anywhere anytime soon, but helping folks, I take all the transactional accounting, all the financial reporting, and get it into software. And so campfires the modern ERP for high growth companies to just help them move to an AI powered platform. And so that's how young were you when you realized he wanted to build the modern ERP? I would say it's not something you think of growing up. It's not like I'm sure kids these, I'm sure some ambitious kids these days think I'm going to build the next gen ERP.

You'd be surprised. So going into White Combinator, there's about 4,000 companies that have gone through it, but they probably hundreds of thousands apply. They're like, you know, no one came to us and said, we want to build a new ERP. They're like, you're like the only one excited about the category. So I would say this was in 2020, 2021, when I really was filling the pain at a mid-sized company,

that I'm like, this is the largest category in software. Why is nobody disrupting it? And that's, that was a critical moment for me. Well, it's interesting because it's not going to fit the pro, like, you need to have lived and breathed the problem to know what to build. And I think that YC has like the sort of default type of founder that does that. Maybe they have a few years as an engineer or Pallenteer or they were a product manager at Coinbase or something like that. It's very different than being in being in the C-suite actually living and breathing and having that ability to go and figure out what to build.

Give us the latest on the fundraising side. Yeah, we announced today Excel is leading our $35 million series ads. There we go. I love it. Thank you. I want to talk. You said high growth. I remember I was running a company and there was a big debate over like, oh, like, you

know, like thank goodness we don't need to implement Oracle. We get to implement NetSuite. And it was like still a lot of work. It was a big, big ERP. And so it seems like the playbook is start with high growth startups, be the first DRP and then scale into that. And then eventually you'll be able to service the really, really large Fortune 500 companies. But it's fair to say that you're not ready to go into a Fortune 500 company today. Is that right?

And how do you think about actually growing the business long term? You said it well. We're hiring on the sales team. So we would love to have you. But in all seriousness, we are not ready for the Fortune 500 yet. Yeah. We do have customers with hundreds of millions in ARR that have ripped out SAP. They've ripped out NETSuite.

They've ripped out everything below that. So we are filling the pull-up market quickly. And one of the reasons for the raise is to go make us enterprise ready. Sure. And our customers are saying, I don't want to be on an ERP for a few years, right? I want to be on it for 10 plus. I talked to the CFO at a 10,000 employee company still on NETSuite. They're really scaled into a $100 billion plus market cap on NETSuite.

and they bought it at 100 employees. And so it's like, how do we go, how do we get them at 100 and scale with them to 10,000? And that's really where we're going. Yeah, I mean, pretty sharp guy running the company over at Oracle. So it is a competitive market. But it does seem like there's lots of pockets of opportunity because the, I mean, I remember just like the implementation bill for NetSuite was going to be like half a million dollars in consulting or something like that.

It was pretty staggering. How do you think about? pricing and getting a growth stage company to actually match the value that you're delivering with the money that you make. It's obviously great if you get a big contract up front and they're locked in for a long time and they're paying this massive implementation bill. There's real cost to you if you have to implement stuff. How are you thinking about ramping a customer into a great profit center for you long term? Yeah. First off, like we rethought the whole

implementation process for the reasons you've stated. Like you tell your accounting team, we bought an ERP. Like everyone's ready to quit. Nobody wants to go through a one year implementation. Yep. And so we've made it where like you give us a log into your old system. We go do all the work because everybody says, I just don't have time. Our team is way too lean to do that. So a lot of the switching costs, to your point, a lot of it is just like dollars, but a lot of it is just time. So we have solved for both. But to your point on like how do we get them into a profit center, we're just very focused on getting them in the door and making people super happy because we know like customers drive customers. Yeah.

Like referenceable and one of your sponsors, Ramp, has done a great job at driving referenceability and customer love has really been, I think, the virality within the finance and accounting community. It's been a huge driver of their growth. And so we're focused on the same. How are you guys leveraging AI internally and then in the product itself? Never heard of it. Irrelevant. Don't need it. You can verify what it stands for.

I'm not familiar. Yeah, yeah. Don't need it. Build different. Yeah. We're doing it all by hand. Everything's statements over here. Yeah.

I know. You know, I found, I forgot what it was, but Excel's if statement limit. I like over. You know, you can. That's the final boss of the ERP. You have to build an ERP after you hit the limit on Excel. You just unlock a, like, like, Satchinadella just comes through the computer and says, you've broken out of the computer.

Matrix. Welcome to the big leagues. It writes you a check. Our very first version of Campfire was I built a three statement model that was I framed into like a login and it was me on the back end running a three statement

Google sheets for customers. You got to start somewhere. But yeah, we're super excited by AI. So obviously one thing about us relative to the incumbents is like we're investing in putting eye everywhere, but whether it's meta-winning the AI talent war or open AI, like,

we're always going to be on the latest models and we're always going to be multi-model. And so we're going to be inheriting all of the amazing work of all of these amazing AI engineers. And we're, of course, building out our own proprietary AI models. But the gap between us and the incumbents is growing every time a new model ships, irrespect of a vendor, because we're always putting in the latest models. A few examples, customers are uploading their board minutes in the last meeting and saying take all of our, every row in campfire of every financial data and come up with a story narrative for the next board meeting. And it's reading the board minutes, it's going through. And it's like, here's the five things you should focus on based on all of it.

On the accounting side, it's automating a lot of the accounting. But we've built like a chat GPT for your finance and accounting data. it'll write to the ledger, it'll report for you. Customers are saying like, kind of feels like we got a promotion. I'm not in there formatting my Y and X axis anymore. You know, the charts are coming out through AI. And that's been exciting to me because I live that pain.

Guys, I had a diligence report. I had to build the day on my wedding. Wow. That's brutal. Because we were in a bad ERP. So don't be me. You know, enjoy your wedding day. That's great.

enjoy your wedding day. Last question for me in the news. Zero acquired the U.S., the American B2B payments company, Melio, for $3 billion. I think this was announced last week. Two and a half. Who's counting at this point? Can you explain to me like what the rationale you think was behind that? And Zero feels like a different tier of product, not full ERP, more accounting suite for small businesses.

Why would they need a payments product? Are you thinking about payments as a potential revenue generator in the future, whether it's through Stripe or Stable Coins or maybe you're going to buy a payments firm at one point? I don't know. Yeah. So I can go really deep on this one. So as deep as you guys want to go. But at invoice to go, we were an air company, $250,000 businesses.

We were powering their invoicing globally in 150 countries. We competed with Melio. We sold the bill.com. Obviously, they're a Titan, an AP, a billion, an AP revenue. And so I know Zero and Melia well. I'll say Zero was really looking to be stronger in the U.S. market. They're very strong in Australia, New Zealand, very strong in Western Europe and other parts of the world. QuickBooks, unsurprisingly given their headquarters is here, is dominant in the U.S.

Melio brings a strong footprint. QuickBooks also launched their own AP product, and Zero needed a response to that. And so getting into the payments game, getting into APAR, in that solo entrepreneur category where i came from an invoice to go is where zero and melio have a lot of strength we don't compete with either at all yeah so we're really like a cut of bob so no competitive angle for for campfire um but i think one that actually no one has really covered is melio is very strong and indirect so if i serve and others and this is public um fyserv and others are are selling melio through their channel so

So Clover and other systems are out. And so this is a new go-to-market that's becoming really popular. We saw Brex launch Embedded, Bill is embedded. Embedded go-to-market is really a way to solve for really, like, efficient go-to-markets. And I think this is a new one for zero that is a great play. That's really helpful. Thank you for that context. I appreciate it.

I always like when people to break down on and their stuff. And congratulations. Yeah, congratulations. Thanks for having to you soon. Have a good one. Bye. Cheers, John. Up next we have Tom Lee from Bitmine, the first new treasury, crypto treasury company, this time on the New York Stock Exchange, not the NASDAQ.

We're going to talk to him about crypto. Really? Really an iconic stock exchange. It is. It's the New York of stock exchanges in many ways. Totally. The city that never sleeps, the Big Apple. Totally.

If it was a stock exchange, it would be the New York Stock Exchange. Very excited to talk to him. Yeah, so today they announced a $250 million private placement to implement buying Ethereum for their ETH treasury. Yes. The stock. ETH has been one of these. It's up 48% in the pre-market. Yes.

So welcome to the stream. How are you doing? Good. Good to meet you. Would you mind kicking us off with an introduction on yourself and the company and some of the news today? Glad to. My name is Tom Lee. I'm the founder of Fundstrat, which is a research firm, and many of those might know me from that role where I'm periodically appear on CNBC and talk about markets.

I'm also the chief investment officer of Fundstraat Capital, which is the asset manager, and we have an ETF called Granny Shots. That's 1.4 billion of assets now. And Fundstrat as a research firm has 300 hedge fund clients. and 10,000 RAAs. And today it was announced that I'm taking a role separately from Fundstrette as chairman of BitMine immersion technologies and the ticker is BMNR. And that's to help the company transform its existing treasury strategy to be one that is focused on Ethereum.

What was the decision to go with Ethereum? We've seen a lot of Bitcoin Treasury. story's happening over the past year, I would say. I mean, even going back further with micro strategy, but Ethereum is one that we haven't seen that much yet. Just before that I just would highlight that the bitmine is up 679% today. So the market is excited about what you're doing.

Congratulations. Back to John's question. Well, to me, one of the things that caught my eye is that stable coins are the chat GPT of Bintech right now because it's seen rapid adoption by consumers, merchants, and banks, and now visa and credit card providers. It's because it really does have a log function benefit over traditional dollars or over, you know, the credit card system. And it's $250 billion today. Treasury Secretary Besson estimated or said it's reasonable it could go to two trillion. That's a 10x market, you know, addressable market change, which means it's an

exponential opportunity. Stable coins have to transact on a layer one blockchain. And the, the blockchain that is considered the biggest and the most secure and the most commonly used for stable coins is Ethereum. So we saw an opportunity to say, listen, let's think about more tokenization, you know, that Chatt G2. Stable coins is the chat GBT, but it's a lot. It's opening doors to other ways to tokenize assets. And so we thought it would make sense to do a treasury strategy around Ethereum. And it plays a role that the bigger we become, the more important we are to the Ethereum ecosystem. So I think we're actually playing a role in stabilizing Ethereum because we essentially act as a staking entity.

And if we sort of think forward a few years, sorry if I'm rambling. No, that's interesting. Goldman and J.P. Morgan, as they get into it, to the stable coin business in Walmart and Amazon, they're gonna realize that they need to really secure the layer one that they're operating on. So they want it to be a big blockchain, a lot of validators, Ethereum is the biggest.

And then the opportunity is that we're kind of like the plumbing of what banks will look like in the future because they'll be staking Ethereum on their balance sheet or crypto in order to fund and operate their stable coin. Can you talk about the differences between the New York Stock Exchange and the NASDAQ and what it means to be listed on one or the other? Yes. If you look at our press release, what we announced and we put prominently in the headlines was New York Stock Exchange listed.

And we listed all of our partners. The reason for that is that most people know me as a traditional finance world person that also is trafficking in crypto. We've written about Bitcoin for eight years. We first recommended it when it was under 1,000 to our clients, but it's been a slow road. This project, the BitMind Emersions Treasury Strategy, is really a convergence between traditional finance and crypto, right? That's where stable coins are intersecting. And the partners we brought on, you know, Founders Fund,

Pantara, Cracken, Falcon X, Galaxy Digital, are companies that are really straddling the crossover world, and we want it to be on a prestigious exchange. Most of these Treasury strategies are trading on the NASDAQ CM, which is like a sort of the Junior League of NASDAQ. We want it to be on something prestigious because we want our traditional clients to come in and say, look, Tom, if we're going to be interested in this Treasury

and learning about why this might be a new bank model, we don't want to be buying stocks that are on a bulletin board. So it was really important to be on a prestigious exchange like the New York Stock Exchange. So I think that that's one of the advantages of BitMine immersion is that they invested the time to be uplisted and so that they are on the New York Stock Exchange. What is is liquidity like different on the stock exchange versus like I mean as as someone. who's younger and more, I guess, internet native, like the idea of needing to buy a stock and stock exchange is a little bit foreign compared to just, like, self-custodying crypto directly, and then you have access to buy the major cryptocurrencies on pretty much any app in the app store

these days. It feels very accessible, and yet in the past we've seen the ETF as a catalyst for price movements in the major L-1s. So talk to me about like what's actually going on in the market and and characterize some of the different buyers and market participants and why they choose one way to get exposure over another. Yes. Yeah. And so let's say I'll take it in two parts.

You know, the first is like in terms of a product to buy as New York Stock Exchange list, stock, which is what BMNR is, has market makers. It's almost like you have like this invisible hand protecting price and the exchange will halt the stock. If it feels one-sided, that's really important because it protects investors. And these treasury strategies need to get institutional investors to sort of back this and become, essentially created two-way market. So people can combine and sell these with little, price erosion. Now, the second part, though, is someone say, well, Tom, why didn't I just buy Ethereum, ETF? Or I'll just buy Layer 1. And I think the Treasury, the Ethereum Treasury

strategy, now there's another one, which is Sharplink, which was led by Joe Lubin and Consensus, and that tickers SBET. I think the Treasury strategies, especially for Ethereum, can actually deliver better returns for holders because let's say the metric for a treasury strategy is Ethereum per share. Now, if you buy an ETF, your Ethereum per share is going to always be fixed because you're just buying a unit of Ethereum. So you only benefit if Ethereum price goes up. Whereas a treasury strategy, let's say has an Ethereum.

per share value, then there's three ways that that price of that treasury stock goes up. One is that the company generates staking yield. So you're getting a little income. The second way that you benefit is if the company can capitalize on volatility, and Ethereum has twice the volatility of Bitcoin, which means the stock's twice as volatile. That means you're going to be able to fund transactions. like convertible instruments or sell stock in the open market, and people are buying it for the liquidity,

and then you can increase the net asset value of Ethereum per share. So you're kind of like growing. Someone's actually, if you own a share of a treasury stock, someone's working to get you more Ethereum per share for every share you hold. So they're actually like rowing it. And the third, of course, is the price of Ethereum goes up. And that's where I think it's really interesting because like Circle, which is the second largest stable coin issuer,

and it's probably the most successful IPO in several years, it trades at around 100 times enterprise value to EBITDA. So for every dollar of EBITDA, it's like $100 of market cap. Ethereum has a network yield of four. So in other words, like Ethereum's four times cheaper than Circle, even though Circle runs on Ethereum. So usually the more you get into the infrastructure layer, the higher the multiple is. That's why software has a higher multiple than a retailer.

So that means like it, but if you just said Ethereum should be a 1% yield, then Ethereum should be like 12,000 per unit. And it's only like, you know, $2,500 now. So I think that there's the third way is, of course, Ethereum could go up a lot. Yeah. What do you see as kind of the important milestones in Ethereum over the next few years? We went through kind of the NFT boom. We're now in the stable coin era. People have been still talking about store value. They're also starting to talk more about prediction markets. What is interesting to you about maybe crypto broadly over the next few years? Yeah, Gladys, well, you know, I'd say the story arcs that we see is one, the regulatory landscape has become very friendly for crypto. Huge deal. The second is that institutional fiat money, traditional, is getting into crypto. 95% of institutions don't own crypto. So that's a huge addressable market.

But then there's technological innovation, as you're pointing out, and there's a lot of great products. projects because of that. So Bitcoin, it's still our favorite layer one. And so, you know, but that to us is digital gold. You know, it's really the best way to measure trust. And if it ever traded at the value of golds above ground value, it's over a million dollars. So we still see a lot of upside. Ethereum unique feature is the smart contract. And we know that the majority of real world assets tokenized things, including tokenized dollars and tokenized real estate and equities, et cetera. They're actually being minted on the Ethereum or created on the Ethereum blockchain. So I think as the world starts to tokenize an experiment, remember, there's a lot more

equities and real world assets to tokenize onto layer ones than the other way as the traditional financial world is trying to equitize crypto, right, through ETI. There's more to move into tokenizing on layer ones. And the third, of course, is things like solving some of the risks of AI, you know, the whole idea of a decentralized blockchain. I mean, look, let's face it, there is a non-insignificant risk that AI is going to be malevolent. Like, there'll be a catastrophe because every innovation has always led to a catastrophe in the short term. Like consumer credit was a technical innovation for consumer access to money. And nearly every country that rolled out a credit card, whether it was Asia or Latin America

had a credit crisis. And then in 2008, we had a mortgage innovation with a securitization of mortgages, and it boomed for a while. But then we had the GFC. So I think AI is obviously going to be innovative, and there's a lot of opportunities, but it's inevitably going to lead to a catastrophe. And I think that's when decentralized blockchains are going to be one answer

to sort of managing the risk around AI. And so, I mean, I don't know if it tensor is a solution, but I can see something running off Ethereum and validators, you know, human validators is one way too. Interesting. Yeah. Talk to the background about that. Any reactions to there's been a couple announcements in the last week, Republic came out creating their mirror tokens products,

which are bringing private company stocks on chain, and then Robin Hood had an announcement, I believe it was this morning around bringing Open AI and SpaceX on chain. Any kind of predictions around private company assets coming on chain? I love all these experiments because they're trying to provide a hybrid form of liquidity to things that are illiquid, you know? And so it makes sense. And, you know, the answer to that is it is a big deal that Republic and then Robin Hood are doing this because you're going to create the probability of a more healthy market.

You know, this has to become a two-way market because if someone's just tokenizing assets but then very few people use it. And what I mean about use it, someone may want to own it, but someone might want to hedge something. So like an open AI token isn't just someone's long bet on open AI. They might be invested in, you know, GROC or something. And then they need to hedge it by shorting the token. Like if you can create a two-way market, then it's definitely going to work. But if the market is purely on like, let's just tokenize it, you buy it and hoddle it, then it's going to have the same liquidity problem that it originally had.

And then they might even trade it at discount. and then you create other risk because now you're putting it on a blockchain so now you have the actual risk of the physical asset being missing and then the blockchain being broken. So I think it'll be important for two-way markets to develop around it. Very cool. Well, congratulations. Yeah, congratulations on the announcement and we're excited to follow along. Hopefully you have you on again soon.

Yeah, fantastic. Thank you for having me. We'll talk to you soon. Cheers. Bye. What else is going on? We're done all of our ads. We are out of ads.

We need more sponsors. Put more on that soon because we have a new one coming. There's a story on Apple, Wang from Mark Gurman, Wang using Anthropic or Open AI to Power Siri in a major reversal. But let's save it for tomorrow. Save it for tomorrow. We do have some other news. Cursor is now available on your phone and on the web. Spin off dozens of agents and review them later in your editor.

Andrew says, I join Cursor to build the future. of coding and yes as of today you can code from anywhere with a coding assistant on every web browser and phone very very cool we'll have to see how i'll have to give tyler over there in the VR headset a challenge back in coding he's back in there he is what are you doing over there Tyler i'm just chilling in the metaverse oh yeah yeah but what's what specifically are you doing are you i'm not playing video games no you're grinding no yeah yeah how how is the uh now that you've had this headset for a few hours how well it's way better so I have the quest two at home and um I think it's just all the quest threes but they have this like back strap I don't know sure yeah yeah I see it's way more

comfortable it's way more like I'm not it's like totally comfortable to wear this right now that's awesome are you doing pass through are you using your computer I can see you guys you can see me yeah I'm using how many fingers or no way I can see it I'm using my fingers two oh wow you can see right through it's amazing I don't I didn't know the past through is that challenge don't take it off for 24 hours starts now go on a date with it use cluelie with it yes yeah I mean this is the clue this is the Cluelly device right here we need a clueli app in there ASAP for sure yeah you show up there why why why you take off your your meta quest 3 for the interview actually I'd prefer to leave it on anyway there's some other good

post we should go through. Apparently, Xfinity is allowing people to use Wi-Fi signals to detect motion in your home at no extra cost. And this five seconds account says, I apologize to all my local schizophrenic. Schizophrenics. I owe you an apology. I wasn't really familiar with your game. You remember there was a super, there was a video that came out. It was probably a year ago where somebody was showing how you can use Wi-Fi signals to detect people. Yeah, to get a very accurate map, like a 3D render of what's happening in space. Yeah. And it was 160,000 lives.

It was almost hard to believe. And then now X-Finity for free, no extra cost. Crazy, crazy. The concerning thing here is introducing this technology to more people. They're going to realize, oh, I can, if they're in an apartment building, they're like, oh, I can see when my neighbor's home. Or a thief could set up a Wi-Fi router outside. of somebody's home and start mapping it.

See what what's this person's schedule. When are we going to put this on an assault rifle and give the, give the special forces wall hacks? I bet they're no, no, somebody was, somebody was saying the CIA and other law enforcement groups use this already to understand when there's like an active shooter scenario, understand where they are. The positioning of everything.

We got to fly on the set. Yeah, this is that. She's going after you. But anyways. In other news, Christoph, Thuwe. lean futurists as I am developing an LLM benchmark test that I believe is truly the first of its kind very excited to launch this soon and hopefully partner with some of the frontier labs on it Congratulations we'd love to check it out when it goes live Christoph so thanks for the hard

work you're you're going up against Arc AGI so you know it's it's not easy easy the benchmark battle is hot but good luck to you and Delian reflects on the progress of AI He says, imagine gathering Alan Turing of 1942, Marvin Minsky of 1992, and Nick Bostrom of 2012, and explaining to them that we have passed the Turing test several years ago, and yes, there are some changes to society, but largely day-to-day life is the same. It would be a funny conversation. It is crazy. The soft singularity is real. Everyone's feeling the soft acceleration.

Anyway, any other. I mean, it's just like, yeah, there's just like, yeah, we beat the Turing test. And now we're on to super intelligence and then eventually giga intelligence. But that's a great place to end the show. Thanks for watching. We will see you tomorrow. Leave us five stars on Apple, podcasts, and Spotify. And have a great Monday.

See you soon. See up. Bye. Have a great evening.