TBPN Live - AI Sweatshops, Anthropic $3.5B Round, Tesla Takes Hit, Rivals Help Bybit, Mars Mission

Episode Date: February 25, 2025

(03:13) - AI Sweatshops (13:40) - Anthropic (33:12) - Tesla (34:25) - New EVs (46:27) - China in Africa (49:25) - Bybit Gets Help (56:20) - What Can We Send To Mars? (01:07:23) - NASA and Astroid Collision (01:08:31) - Claude Sennet 3.7 (01:14:03) - Default 2.0 Round (01:17:58) - Celebrrity Brands (01:25:58) - Timeline TBPN.com is made possible by:Ramp - https://ramp.comEight Sleep - https://eightsleep.com/tbpnWander - https://wander.com/tbpnPublic - https://public.comAdQuick - https://adquick.comBezel - https://getbezel.comFollow TBPN: https://TBPN.comhttps://x.com/tbpnhttps://podcasts.apple.com/us/podcast/technology-brothers/id1772360235https://youtube.com/@technologybrotherspod?si=lpk53xTE9WBEcIjV

You're watching TBPN, live from the shrine of shareholder value, the dojo of the dollar, the capital of capital. It is Tuesday, February 25th, 2025, and this show starts now. We got a great show for you guys today. We're talking about AI sweatshops. We're talking about the stock market crumbling. We're talking about anthropic raising, billions of dollars, and many more. We have tons of stories. I hope we can get to all of them because there's a ton here. But first up, Jordy, what's going on? How are you doing today? John, your intros get better every single time. It's incredible to watch. So I would

be in the studio today, but I had a little bit of a fire drill this morning about 15 minutes before my eight sleep was scheduled to buzz me awake uh i hear my son open his door like he he's not he's not very gentle with the door in the morning so he just always like slams it he yells something and i like come out i find him like he had made it to the kitchen he's just kind of like stumbling around because it's dark and uh he like can barely speak like in a very concerning way and he's just saying like i i swallowed i swallowed a glow stick i was just like at this point like i'm actually like very i'm like very concerned because i like he can't talk at all he's like clearly struggling to breathe and um keeps, keep saying that he swallowed a glow

stick. And, uh, so anyways, we're like freaking out. Uh, I'm not freaking out, but of course, uh, my wife is and, uh, call, call 911, get them over and, uh, end up having to go into ambulance cause he's not like breathing properly. Uh, and we get all the way there fortunately he didn't swallow a glow stick but he had this thing called croup which is like when a kid's esophagus like sort of swallows up or gets inflammation which I had never heard of and it's actually good to be aware of because yeah I somehow had never heard the term or was aware of it but the funniest part is uh by like i think like eight or nine a.m we were ready to go home they gave him like a steroid treatment or whatever and he was back to breathing normally and uh he was like genuinely like pissed that we

couldn't ride in the ambulance back home like like for him like a kid you wake up you get to hang with your dad all morning yeah you get to ride an ambulance there's firemen all around it's like for him it's like the best day ever so um anyways but uh good to be aware of croup uh for those parents and future parents but there was a lot of news there was a lot of news this morning, as there always is, seemingly. So I'm excited to get into it. Yeah. So yesterday, Y Combinator announced a new startup, Optify, and they posted a video.

They've been doing this thing, launch videos. Some of them have been amazing. Everyone's loved them. Some of them have been. We went over the Maritime Fusion company that people were people were saying oh hard tech like this will never work and you know yeah we we kind of defended that one uh this one was a little bit rougher uh the production value wasn't quite at the level of the amazing cgi in that maritime fusion uh fusion video uh and the response was pretty negative almost immediately i have a bunch of tactical uh ideas for how they could have done better uh we'll go to a couple posts here gabe uh people are just memeing this already uh gabe says absolutely sickening revolting imagine how

many sweatshop overseers will lose their jobs because of this uh and then uh peter comes in and kind of defends this he He says, the people dunking on this are idiots. This is literally scientific management that Frederick Winslow, Taylor, Deming, and even Ford implemented more than a hundred years ago, but with line form. And this is what makes Amazon warehouses and delivery trucks performant and what Foxconn uses to make your iPhones grow up. And that was my question was, I was, I was texting some people that work on factories and I was like, I get the pushback. Like the aesthetics were very bad. Like the launch video was like not empathetic. It didn't tell like a great story, but like, let's be honest,

like we, you, you really expect me to believe that China is ahead of America in manufacturing. We don't monitor the output of our factory workers in America. And somehow we're going to win if we don't do that. Like it didn't make sense. Like I couldn't really track all that. And so my take was I would have redone this video and instead focused on robotics instead of people and said, hey, look, there's going to be a lot of robotics in these factories

and we want to understand what's happening on the lines. And then instead of it being, so the video shows a negative scenario, someone who's underperforming, I would rather have it highlight someone who's a high performer. And because that's a more positive story. And then if you have to show something that's negative, show something that isn't, that is fixable. And so, yeah, not somebody that's like losing their job exactly imagine there's a robot on like it was like aisle 17 or something like work unit 17 they're underperforming you use their software to check on it oh turns out his wheels are squeaky or something you go put some oil in the joint and he's bad and he's happy and the robot's happy and

everyone's happy it's this very like almost pixar- moment. And that could have been really cool and positive. Instead, it really did just look like these guys on their laptops are now overseeing people that are working really hard in a factory and then just like firing them. So what was your take on it? How do you think this went? How do you think they could have done better?

Look, I think the product makes a lot of sense. There's clearly an opportunity for this. You can see the value. You can see the value also from my view for teams, if you're working on a factory floor factory line there it requires really close coordination even if you're just focused on a very specific lane and so my big issues with it was the the sort of um uh poor acting like i they're not they're clearly not actors they're founders they shouldn't have been been trying to make this sort of skit-like scenario to demo their product. It just set them up to be very easily dunked on.

And then generally, I would have focused it more around this sort of empowering of how are we going to empower teams at these factories to increase output and hit their goals at sort of like the factory level versus focusing in like of course like i think what you're saying is right show three people performing you know adequately two people over performing and then show one or two people underperforming and you know uh like there's so many ways to position that that are more like oriented around, like, how do we win as a team versus like this sort of slave driver kind of like approach of saying, like, you're underperforming. You've had a bad day, bad hour, bad month. Yeah. And anyways, just just the the other thing is it's very hard. You know, I looked it up and these guys are based in in india

it's very hard to sort of understand how this might be received really well in india yet you you you push it out to this like predominantly sort of american or international audience and it just falls uh flat and and to be honest like the the meme potential of this was just amazing it was insane yeah the shot of the guy like on the phone and i posted i posted the um i posted the the george bush one saying like sir uh work you know workspace 17 is underperforming um so anyways i uh rough day yc deleted it I think they should have doubled down. It's not, again, they also could have used, there's so many things they could have done differently. They could have made it feel more futuristic, to be honest. In many ways,

it felt like SaaS out of the, you know, it didn't necessarily, I'm sure that AI enables what they're doing, you know, to maybe build it with like a smaller team. But ultimately, there was just so many issues with it. And I don't know, I think that you can hold two things like, yes, they kind of set themselves up to get dunked on. But, you know, at the same time, there's like 20 things they could have done better. Yeah, like this has this type of software has to exist in factories all over the world already and yeah like i refuse to believe that there are major factories turning out iphones that where they don't know what's happening in unit 17 or whatever yeah but yeah should have been a lot more optimistic um it gets to this idea that you know we've been seeing the news with uh elon sending out those doge emails what did you get

on this week it's like the famous like his interaction with uh parag agarwal when he took over twitter what did you get done this week that became yeah that became his like calling card and i think that's been received with very mixed results like some people love it they're like yeah like everyone in the government should be able to say what they did this week but um what i think would make it a little bit more empathetic of a message would be if elon also answered that question and trump also answered that question because totally true people are doing that with we're going to talk about tesla later people are like oh what did you get done this week as soon as the stock falls or whatever and realistically like yeah elon

probably did a lot of stuff uh and that would be cool and i think there's something about like leading from the front there that says that, that just messages it differently. Um, there is something about the American worker in the factory. Have you seen American factory, that Obama documentary that went on Netflix? No, but I wonder if there's anything in what, what's that show that people brought up a couple of weeks ago about how it's made something. Oh yeah. Yeah. Yeah. So how it's made doesn't show any humans. It only shows the machinery typically. I mean, occasionally you'll see like the guy loads the machine or whatever, but it's not focused on the human stories of people working in the factory. American factory was very different in the sense that, uh, I think it was a glass manufacturing.

They make like windshields and all sorts of different custom glass installations. And, uh, it was, and this particular manufacturing company was bought by a chinese company and so they kind of brought over the chinese you know management processes and in there there are a bunch of really funny quotes about the chinese managers coming in and saying like the in america uh the american worker is like a donkey like you you you need to pet it with the fur going this way you can't pet it against the fur and so like a donkey. Like you, you, you need to pet it with the fur going this way. You can't pet it against the fur. And so like American workers respond to positive reinforcement, not negative reinforcement. And you know, I don't know, maybe that's our culture.

I think our, I think American culture is pretty great. Maybe that's a flaw, but I think you got to play the game on the field, right? Yeah, go and uh i don't think europeans would respond well to uh optimize i i don't even know how to say the name of the of the yc company but uh but yeah different different continents countries have like very distinct work cultures and that's totally okay right a lot of the approach of this sort of like manhattan finance culture would go over terrible in san francisco right yep uh this idea you know in san francisco you're supposed to get a pat on the back maybe if you're doing like 996 uh oftentimes in new york that's not even going to be like the bar right you're going to be like looked down upon for doing that like might lose your job over it so

these sort of workplace cultural differences are very real and i just don't think that uh so this company that the the annoying thing is i don't know how beneficial a yc hard post launch is going to be for them whereas for fundraising right yeah it's saying hey here's this demo bunch of people can see it even if they didn't make it to demo day even if they got lost in the clutter of demo day because there's a lot of companies now and i i think that's what those posts are for maybe hiring a little bit but uh certainly not customer focused yeah um but yeah uh it's got to go over well but who knows uh we'll see we'll be able to track them and see if they raise a seed round. And I think that

there's already a polymarket out there for will they pivot away from this or will they change? But I don't know. I mean, getting attention can be turned into a lot of different positive things if you work hard. And speaking of getting attention, if you want to launch your company, why don't you launch an ad quick billboard? There's a fantastic post here. Chris from AdQuick reposted it. Out of home advertising made easy and measurable. Say goodbye to the headaches of out of home advertising with AdQuick. Go to adquick.com. And there's a new billboard just dropped with Michael J. Fox for Calvin Klein and just an iconic image. And I'm sure that built the Calvin Klein brand wonderfully. And so sometimes simple aspirational is all you need. I just had an idea. We should take some of our

favorite reviews that include ads in them and run those as billboards. I love that. So we get some TV placement and then you get the actual ads under there. so we got to talk to i mean they they do that for movies now right like it'll be like you know five stars and and they have like the one they won the palm door it's the best picture of the year and what the critics are saying the critics say this we need to run that for us but it's what the critics are saying plus you know uh you know the ads that are that our listeners have dumped into the reviews uh anyway let's move on to anthropic claude is playing pokemon on twitch right now i saw about a thousand people tuning in uh very fun guys wake up says near claude's playing pokemon on twitch it's happening

and so uh let me get to that and so they did this and they simultaneously announced a new three and a half billion dollar fundraise is that correct no i do i don't think they have announced that i think the the fundraising round is leaking currently got it and they are finalizing the round we can get into that uh this has actually been done before uh there's been a twitch plays pokemon where the computer would just read the chat and everyone would spam up up up or a a a or b b b and then whatever letter was spammed the most was basically won the vote and then that would be the next move that the player made and it takes forever and it you can get stuck in crazy scenarios and i think when twitch played pokemon there's this one level where if you if you press down you fall down this like crack and you have to basically start over

this whole maze and and and people would come in and troll and like and so they spent like i think days maybe hours just trying to get through this play it's like pretty easy to navigate if you're just a player but it's very chaotic and so and then people would troll and be like oh like you know instead of let's attack let's just uh run away from this bad guy let's video games video games by committee probably not a good strategy but people had a lot of fun it's a lesson in democracy right and so now there are benchmarks around this people are people having fun with democracy like yeah yeah exactly um and so uh zeo fawn says we got

yo we got cool benches and it's uh clogged three versus 3.5 versus 3.5 Sonnet New versus 3.7 Sonnet. And it says milestone reach number of actions. And it shows how does Pokemon, how do you progress through the game? I've actually been playing Pokemon with my son because of the mod retro chromatic.

He actually finds it kind of boring. He's not that into it. He likes the other games, but it's pretty cool. The chromatic you can get. I got one card that has like 100 games on it. And then you can just like play whatever you want. So he always wants to play something new. So great. But it's pretty cool the chromatic you can you can get i got one card that has like 100 games on it and then you can just like play whatever you want so he always wants to play something so great um

but it's fun uh can you imagine if we had that can you imagine if we had that when we were kids like i distinctly remember you had your game boy yep which you'd be cruising around with and then i would have like a dop kit like shave kit that just had a bunch of yeah a bunch of cartridges yeah and you would have to save up you know games were like 50 you'd save for like three months to get enough money to buy one and and now it's like oh yeah just you know download them all for free uh yeah we live in the era of abundance ai we are post scarcity in terms of video games that is correct uh well let's go to vittorio he kind of sums things up the timeline in the span of three days people moved from uh openAI to Anthropic or to Grok, and then from Grok to Anthropic to Claude. Everyone is always obsessed with the new shiny object in AI,

especially online, especially in X. Everyone is always focused on the evals, focused on the latest and greatest models. People care about this stuff. But again, what's happening on the street, what's happening with the average person is what actually drives revenue at these companies and we'll get into what's going on at anthrop yeah and there's actually but there's this effect right now where there's you know in the same this the the critique of the whole uber versus lyft battle was that uber was like obviously the winner and by both companies being able to raise billions of dollars there was a bunch of capital incineration and Uber wasn't really able to recognize, you know, able to see the real value of the traction they had because prices were

very compressed. And same thing is happening right now, but it's actually much more deserving in many ways because we don't know how all these consumer products are going to shake out yet we want a lot of competition and i actually think that it's great for consumers because you can go on grok and get free intelligence uh that that you know a couple weeks ago you had to go to open ai pay money go to deep seek right yeah exactly so i think it's great for consumers and you know it's it's totally unsustainable yeah we're not going to have every every you know consumer is not going to have like 20 different uh llm uh yeah the fragmentation is unsustainable but i think the pricing will come down to the point where it is too cheap to meter i think these models will get baked down uh dylan patel was talking about this how already there are gpt4 level models that have been baked down to the

phone and so uh you know there's you know he doesn't even need open ai to open source that already because it's it is available and the optimizations are happening uh which is a lot of fun yeah uh anyway uh let's move on to uh uh chu jay jang who says are you kidding me dude talking about uh claude how many r's are in strawberry with a bunch of R's? This is like a classic, like a tricky prompt for these LLMs because they tokenize the words and they struggle to count letters, I guess. And it's hilarious because Claude recognizes

that this is a trick prompt and says, this appears to be an Easter egg prompt. The instructions specifically mention Easter egg. If the human asks how many R's are in the word strawberry, Claude says, let me check and creates an interactive mobile friendly React artifact that counts the letter R's in a fun and engaging way. And so it doesn't just count the letters.

It actually creates a whole website interface for you. It codes all of that up and it still gets it wrong which is hilarious yeah this is this is rough the reason that this is particularly rough is that they clearly were were like we're not going to get got here yeah exactly so they went and they they used humans to try to to try to inject themselves and make the model get it right and then still getting it wrong. It's so funny. Of course, the top comment is... Yeah, Eliezer, I'm cackling.

Yeah, Claude is in an odd place. I mean, we'll go through this with Evan Armstrong's post. But, you know, simultaneously, like a lot of people have a very strong affinity for Claude as a personality. Like they like talking to it. They think it's friendly and interesting and, and, and they like, they like the way it sounds that when they talk to it, but then simultaneously on the other side, it's like, it seems to be like the best programming model. And, and that's like, you know, completely utilitarian. You don't really care about the personality of the, of the LLM. And so they're clearly, you know completely utilitarian you don't really care about the the personality of the of the llm and so they're clearly you know finding little pockets of value here and there but isn't that armstrong isn't that fascinating isn't it just fascinating the way it's evolving because

it almost seems like it's it sort of feels like it's not fully in their control like that's like the like like there's a feeling of like yeah yeah, Anthropic was like, hey, we want our model to be like you can imagine them being very intentional about. We want our model to be sort of, you know, more fun, friendly, like more of like a companion. You know, you can develop like a relationship with it, whatever. But then do they have an off switch for that? Is there a way to make it go like, you know, sort of clawed hardcore where you just want to have this sort of like, you know, more simplified, linear, just direct relationship with the LLM? Because I haven't experienced that as a user. I mean, there's tons of different ways to do it. And maybe you could prompt engineer it.

Yeah, I mean, prompt engineering is one way. There's post-training. Even the data that goes into the pre-training, a lot of the reason that the models, the early models were like quote unquote woke was because they were trained more on Reddit and they kind of excluded 4chan. And so you have like a more left-wing perspective. And then just the internet broadly leans younger and leans more left-wing. And so you wind up with creating the average of that and get kind of more of a left wing viewpoint and then obviously xai has kind of it's kind of gone the other direction with mixed results honestly because it is a little bit of like you're trying to you know tame a wild dragon or something and like you don't really have fine grain control it's not just a slider yeah uh there's a lot that goes into actually you know evaluating these models and creating something that um is yeah it makes it one thing one one thing that seems obvious about where we're headed with all these apps is that you will be able to have like uh maybe like a four quadrant sort of graphic

where you can drag it up it's like do you want it to be more short and direct verse you know because think about it you know uh when you're working with people in a professional environment which a lot of these people you know most a lot of users are using them uh these products professionally you sometimes you want somebody who's just super direct like doesn't want to be your best friend just wants to do the work and you know get it done and then sometimes you want somebody to just talk to right yeah You can actually do that in open AI in the chat GPT app. There's a, uh, basically like a prompt that is like a permanent prompt where you can type in. Yeah. You can fill in a ton of text. You can say all sorts of things and people kind of trade like these, Oh, this is like the best prompt. You got to tell it to do all these different things. And

then the model updates and you kind of don't need to do that anymore. Um, but, uh, um, you can actually bake in, you know, those preferences and then it'll just come out, come through on every single prompt that you ask. Let's go to Evan Armstrong. He says, Anthropic is in a tough spot. OpenAI has 50X the brand and customer distribution. XAI has this platform captured. Google has an infinite money glitch going for them. How can they win good old fashioned marketing? They need to counter position, make everything less creepy than the open AI robot god energy.

This recent launch is a perfect example of their more human strategy. Handheld shot versus the creepy steadicam that makes everything feel like a webinar. A nice warm yellow orange color correction on the vids. Its videos will go with fun jazz or lo-fi beats that fills out the demos. No thinking about how the tools will take your job, lol. Intros that humanize the presenters with no C-suite

trumpeting accomplishments, just normal sounding team members. Most importantly, the models feel more soulful and are trained to have a softer, more human-like response to user queries. Yeah. So finding a little pocket of differentiation is really, really key. I was laughing about, you know, Google has this infinite money glitch. I was looking up Apple. Apple has returned, they've returned $928 billion to shareholders over the last 12 years or something like that. And they have, I think, $65 billion on their balance sheet. So they basically generated a trillion dollars in cash for their shareholders. And you think about like, what does a trillion dollars get you in the AI race?

Like, it's like clearly the best and like well beyond anything else. But they just didn't go down that fork in the tech tree. And so they just didn't pursue that. And I think the strategy will honestly probably work out because they're just going to act as the portal and take a toll on everything that happens. And they're happy with that. And they haven't had to duke it out in this game and put all that capital at risk and said they've been able to return it to shareholders. But it is very interesting. And these models are expensive. And that's why Anthropic is finalizing a $3.5 billion funding round. And so the company behind Claude Chatbot is valued at $61.5 billion after overcoming investor

fear sparked by the success of China's DeepSeek. Investors are eager to back the promising artificial intelligence companies despite their disruptive arrival of DeepSeek. They initially set out to raise $2 billion billion but they were able to increase that amount during talks with investors uh anthropic such a young company founded in 2021 by former open ai employees startup was previously valued at 18 billion and considered one of the few ai startups with enough talent and funding and to give to give sbf some credit I believe they definitely invested sub $10 billion. And that was from the FTX balance sheet investment, which is wild.

Yeah. And so I think that's part of what will wind up making the FTX depositors whole. Like if you lost money in FTX, you now have a stake on a small slice of Anthropic. And if they can liquidate that at some point, that's a really good return. I don't know exactly how much money is missing or how the math works out, but it is one of those weird narrative

violations. Yeah, it was highly illiquid, but potentially profitable. So who's in this round? It's Lightspeed Venture Partners, General Catalyst, Bessemer, Abu Dhabi. MGX is also in talks to participate. And I believe MGX was going into Stargate too.

Is that right? Everyone's in everything. Everyone needs a slice of everything. Everybody's getting conflicted. Getting conflicted. Anthropic trails. There's no worse feeling for a VC than, a massive market map appear and you're not conflicted out of it. It means you missed the boat. For sure. Anthropic Trails, market leader,

open AI, and among consumer users, although it's cloud chatbot has become popular among programmers and business clients, their annual revenue, annualized revenue recently hit $1.2 billion. And this is what's so funny it's like we're in this crazy ai bubble but you just compare it to the dot-com boom and yes everything could go really south and a lot of these companies could incinerate money but when we look back we're like okay yeah it was trading at 60x revenue that's still better than 10 000 times revenue which is like what was going on in the dot-com boom. These companies are generating real money and who knows where it'll pencil out once the models are fully distributed or if they really stagnate, they could become very cheap

and that revenue could decline. But so far they're doing pretty well. It's losing money and they'll use cash from the latest funding to support their efforts to train more powerful AI models. OpenAI, for comparison, in the October funding round said that it expected to generate $3.7 billion in revenue, so about four times the size. And of course, the valuation is about four times the size as well, or three times the size. Some Silicon Valley investors have been worried about the prospects of companies such as Anthropic since DeepSeek released a new model that rivals the most powerful in the US, but was made at a fraction of the cost and is free to use. And there's more information on that. So yeah, interesting to see. Everyone's piling in. Everyone's got to get a bet if you're not in a foundation model company and you're a VC? What are you doing? Yeah, we had seen an LP update recently,

of which I'm sure there are many like this. It basically was like, yeah, I haven't, not doing much AI stuff yet, but I'd like to at some point. Like, bro, bro, you're probably not actually late, but you should have been paying attention. Well, it depends on what happens tomorrow. NVIDIA is doing earnings tomorrow after the market closes.

And that will be something everyone is watching. Yeah. Very, very close. We need to time it so that we're actually streaming while it's happening. Yeah. It'll be after market close, but we'll figure it out. We'll find a way to cover it. Anyway, so you raise $3.5 billion. That's a lot of money going out the door. How are you going

to track it, Jordy? Great question. Before we get into the most important section of this piece, the other thing that was interesting is people have, the last thing I would say on this, people had sort of memed, does Anthropic buy Cursor? Does Cursor buy Anthropic, right? Being one of their most important enterprise customers, at least from a growth standpoint and a customer love standpoint. And, you know, anyway, so I wouldn't, you know, overall, the thing that was interesting out of this, they also launched a code editor. That's sort of, I think it's called claude code uh fitting name and so it'll be interesting to see do they end up you know it's not you know directly competitive yet but uh you can imagine all these things converging uh in the same direction and uh

yeah i'm excited to see that play out yeah they also have a a model that's trained to live in your command line and so it can interact with your computer at a very low level which makes a ton of sense and just gives it a lot more ability to interact everywhere it's been fascinating watching the ai companies figure out different ways to plug in the mac app for chat gpt uses the assisted device interface to basically automatically integrate with every single code editor, regardless of whether or not they want them integrated, because it's just sucking out the text and then injecting our text like it's any other screen reader.

So a bunch of different ways to kind of plant these models in. I used Operator yesterday for the first time, really looking for some office space and it was rough like i it was he kept coming back to me to like do captchas and stuff because it knows hey this is an ai that's going to this website and uh it wasn't you just become a fun part yeah exactly yeah yeah there's a there's a dark future where humans like primary work output is just filling out ever increasingly difficult captchas and that's the only time that you get hit up throughout the day is like getting it like somebody getting a door dash notification like oh new capture new capture and then you got like 40 of these things working for you at once

and it's just like every minute you're just sitting there you know addicted to your phone but it's just captures brutal i hope we can avoid that that sounds terrible anyway you got 3.5 billion in the bank you You want to track it. How are you going to do it? Ramp. Baby, let's go. Time is money. Say both.

Easy to use corporate cards, bill payments, accounting, and a whole lot more all in one place. Go to ramp.com to sign up. And we have a post from a clip of Eric Gleiman, the CEO of Ramp. And he was on a podcast or an interview. And there's a little breakdown of what he said. He said, when you put high agency people together, they ship faster. Uh, at one crucial point, a third of ramps team, 50 to 70 people were X founders. That's a crazy stat. Um, Eric explains these people understand that they, that the default state of a company is it doesn't work.

So they don't wait for permission or perfect conditions and they find out ways to make this happen. They've experienced firsthand how companies collapse without relentless effort. So they end up being independent and stubborn on what should be done to move things faster. But it also comes with an acceptance cost that these people might go on to start their own companies. And I think that's great. If you're looking for a job, I highly recommend starting a company, figuring out, building something, building your portfolio. Don't just hang out in the unemployment zone.

Build something. Put yourself in the cage. Do a PMF or die challenge. Try and build something. And then worst case, you find a soft landing somewhere as a job. You get a job offer. Because you did all of the things and you learned the 360 view of building something. Yeah, and I gave some advice to a listener yesterday

who has been a founder before, knows that he wants to build his career as an entrepreneur. And I give him the advice of sometimes it's great to try to use pattern matching to your advantage. And by going and working at a company, even if you were a founder, and then you go to work at a company like ramp, your your ease of raising additional capital from that point on, you know, will will will increase tremendously, right? Because VCs are just sort of trained at this point, hey, this person went and they worked at Ramp for two years. They got really deep on this one area of the business. And now they're starting this sort of adjacent business solving problems're just obsessed with company building, still should use stints at great companies to sort of accelerate their careers.

Yeah, 100%. Well, let's move on to Tesla. Having a rough month, down 23% in the last month, and people are now graffitiing superchargers i saw with the swastikas which is aggressive and solana had a great mike solana had a great post about like both nazis and people that hate nazis seem to both love like drawing swastikas everywhere very odd very weird i saw interesting uh wild because i i i just don't understand like who is feels the call to go and graffiti something like that uh i would say that uh you know the the the thing that we want to focus in on here is like what's actually happening to the um the stock because, uh, Jack Raines posted earlier and he said that basically the 20, uh, what did he say? If you bought Tesla in 2021, uh, you basically fully round trips. So his joke was

like, yeah, 10 K, 10 K. So yeah, it's a store of value. Uh, it's a pop to over a trillion and now it's back at a trillion and it kind of went up and down and up and down. But I want to know more about what's driving the market. You had a big question about are people moving to other EVs? And there's an interesting article in the first off. There's an interesting car that launched the Lee Auto Mega MPV. Have you seen this thing so it's a minivan uh but it looks like you know it has like like cushy seats that recline uh it has all of the all the niceties that these chinese evs tend to have uh and so stunning electric minivan from lee auto the mega mpv dubbed the highway bullet train due to its shape and ability to travel 311 miles in a 12-minute charge, while Musk's waste every second on toxic identity politics. The Chinese are crushing Tesla in innovation. Unclear if that's what's actually going on. We'll see how this actually performs on the road,

but they are moving quickly, and it's putting pressure on Porsche as well. There was an article in the New York Times why Porsche is no longer a premium sports car in China that I thought we might want to read through after decades of dominating China's market for high performance cars with precision engineering German automakers are losing out to Chinese rivals that have shifted the definition of a high-end car to one that is electric smart and affordable many new Chinese vehicles resemble their German rivals

like the wildly popular Xiaomi Su7 which is a phone company remember xiaomi has phones and now they make cars uh and this mimics porsche's taikon uh the su7 rivals the taikon in power and braking but it also includes integrated artificial intelligence that can for instance help with parking and greet drivers with their favorite song the cherry on top it sells for roughly half the price of a taikon yeah the problem is you, a lot of the wow factor elements of the Tesla early on were the software side. And the issue is like software is just so much easier to copy. So like not only do the Chinese are like the best in the world at, at manufacturing today, uh, at least like, you know, mass market consumer products, uh, they also can very quickly emulate.

And it's like, Oh, Tesla makes a fart sound. If you hit the right button, we can do that, but we'll like play your favorite song when you come in. And so exactly that ability to sort of spark, uh, joy or, or have some fun with, with the user is just so much easier to copy. Um, and not a durable note. Yeah yeah big driver here is that a lot of the chinese population is buying cars for the first time and so they're not like oh well i had a i had a poster of a career gt on my wall and then yeah i rented a you know a cayman and i know what you know a flat six sounds like. And so I really want a naturally aspirated V8, my next car. They're just like one car, please. And what is a car? Oh, it's electric

now. Great. Like that's fine. Like this is the best car. Great. And so the other thing that happens is that they're working. Is as Chinese consumer products and brands like started to work internationally. I do think that the Chinese consumer started to say, Chinese products are cool now. It's a small thing of like, oh, you're using TikTok overseas? Cool. Oh, you're using DGI overseas? Yep. Okay, why aren't we just buying our own brands? are i'm sure some of these brands are coveted and cool in the chinese market even though as american consumers like it means nothing like the su7 or whatever to me that sounds like i don't

want to knock off macan exactly but but to them they're like i that that might be like a super desirable vehicle yeah so the su7 sold a hundred thousand cars last year uh and porsche has been one of the hardest hit. They reported that last month its deliveries in China plunged 28% in 2024. Although Porsche's sales were up in every other region around the world, the decline in China was significant enough to pull down global deliveries for the year. For years, German automakers relied on the Chinese market to make up for weaker demand elsewhere, leading them to ignore deeper structural issues at home. Chief among them was the reluctance to adopt the technology that has come to manufacturers, namely the importance of electromobility and software-defined vehicles. And so, yeah, China is building their grid from scratch.

They're building their roads from scratch. So they're going to have a fresh start, greenfield opportunity. They can build their cities and roads for electric vehicles to put less stress on the grid and actually they can probably they can probably uh get more access to transformers to actually yep totally because they're they're making a lot of them yep yep and uh yeah and and and fundamentally people uh are are just fine with these electric cars the chinese are also doing something interesting where there's a few electric cars

that they're kind of a reverse hybrid. So instead of it being a gas powered engine with an electric motor that gives it a little extra boost or makes the range a little bit further or just improves the gas mileage, it's actually essentially an electric car, but with a really small, uh, petrol

engine or gasoline engine that can then just charge the battery. And that's all it does. And so you'll see these cars where it's, you know, a 300 mile battery charged, but then you get another 300 miles if you fill the gas tank. And so all of a sudden your range is like 600 miles and you can stop at a gas station if you have range anxiety so they've completely solved that problem of uh like oh if i get an electric car i can't go on the road trip um but now you can because you just fill back up uh but you're still getting all the benefits of the electric powertrain and all that and all the reliability and stuff so yeah it's interesting to see that hasn't happened in america because people have either been all in on ev it's got to be ev tesla doesn't touch internal combustion engines or you know oh we we

wouldn't put any electrical components in this hybrids are bad we need to be naturally aspirated v8s all the way uh china has not been as dogmatic about that yeah yeah the uh it's interesting to see uh just chinese they're not being uh they're not being necessarily super targeted with the international products that they're slowing down on purchasing right we saw caring groups uh sales were down 30 year over year uh and we're you know we're seeing that broadly in luxury goods and then the same thing uh they're not in the iphone sales too right so it's it's more you sort of utilitarian products like phones and cars and then also luxury goods. And this is probably, you know, there's probably a lot under the surface in terms of the CCP and the Chinese government broadly saying, hey, we need to switch from an export based economy, which they'll continue to be to some degree, but we also need to be consuming the products that we produce internally and having more of this sort of consumption based economy because they can't

be reliant on growing foreign exports to the same degree or having, you know, people move into cities, which also drives that consumption. So all of this will be interesting to play out. I hate I've been I went the last time I was at a mall, I went by there was there was one of these Chinese EV manufacturers that was like set up there. They had like, you know, two cars in a box and they were sort of marketing them. And I, I hope these products don't catch on, one, for our sort of local manufacturers. And then two, they're just like ugly and cheap, in my opinion. And I just don't want to see them on the road. I mean, there's already crazy tariffs.

So most of these cars are not viable to sell in the United States. And I think there's tons of pressure. And then, of course, Elon being so close with Trump, like there's going to be tons of pressure against this trend. So I wouldn't count on tons of pressure against this trend. So I wouldn't I wouldn't count on any of these products really going, you know. Yeah. But one one thing that's interesting is so as our in the chat mentioned, he says the U.S. will probably cut EV subsidies soon, which is scaring investors, which which we can see in the in the stock price. But the interesting thing with the Elon Trump dynamic is that you can imagine Trump would would ignoring the Elon relationship would just say, why are these EV companies getting all these subsidies?

Like, why are we paying? You know, why are we subsidizing the cost of these for, you know, let them compete on the free market. Right. But then he's boys with Trump and or sorry, he's boys with Elon. And like, that would be super damaging to Tesla. And so it's an interesting, uh, I'm interested to see how that dynamic. There's another interesting cultural thing, kind of going back to like the different factory worker, uh, you know, philosophies, uh,

the there's the, the Han, uh, I forget what, what company this is, but it's a Chinese sedan, electric sedan called the Han. And, and they have different trim levels. And And in America you want a bigger number. Everything's bigger is better in America. Everything's bigger in Texas. You want a bigger engine. So, you know, something like a BMW, you might get a three 28, I three 35. I, uh, same thing with a Mercedes E three 50 or, uh, you know, uh, what are the 6.3 liter V eights and the AMGs? And so you want a higher number, you want a bigger number, pretty much always. But they don't have that culture in China. And so the trim levels are just the zero to 60 time. And so you can buy like the Han 3.6, the Han 2.8. That's actually better. I love that system.

Yeah. And so you're driving behind a car and you're just like oh the 2.5 that goes 0 to 60 in 2.5 seconds which is yeah when we're when we're driving in the work to more in the morning and wondering who we should drag race yeah now it's a much easier way to figure that out yeah exactly i just thought it was hilarious culturally but it really goes so you know there are no priors there's no standards so it's just like yeah do whatever in uh sort of sort of adjacent to this in in the in the uae uh they people uh the license plates you have to like buy them and there's like a secondary market for them and so the most coveted license plates are are these like like the king will have like one and then if you see somebody with like seven like you know they're a big dog.

And so like, they're probably a part of the royal family or, you know, you know. Everyone has a million dollar car. Like that's not special. You need to add a couple million dollars in like license plate flexing. Yeah, and they'll trade for $20 million just for the plate.

Because if you have more money than you know what to do with, you might as well find out one more way to signal to the the other people on the road well if you're going to dubai or you're going to spend some time doing laps of the nurburg ring where are you staying jordy i'm staying at a wander let's go find your happy place book a wander with inspiring views hotel great amenities dreamy beds top tier cleaning and 24 7 concierge service. It's a vacation home, but better. And I didn't know this. We have a post here. Wander has 10 minute home tour videos on all the different properties. And you can see this one's 13 minutes as a full introduction, drone tours, cinematic shots. And so before you

make the decision to click purchase you can really get to know the property and it's a lot harder in the classic fake it with a couple wide angle photos when you're moving around with a with a video camera yeah the classic i when we were evacuating from the fires uh i was dealing with like airbnb drama i wanted to stay at a wander but um you know some of these homes are so coveted, they get booked out, uh, for too long. And, uh, I ended up booking two places that just were completely disingenuous with how they represented the property. And so that's like a pretty common experience with Airbnb. You see these amazing, uh, you know, uh, pictures and then you show up, but you can't fake a 13 minute video, uh, around the, uh, around the, the, the property. Uh, so,

so highly recommend it. Let's stay on China. Go to, uh, Bridget Harris, uh, my colleague over at founders fund. She says, by the way, China is lending money to African nations for infrastructure project like ports and energy, et cetera, via this infrastructure, making them more dependent so they can ship in Chinese goods more easily. And then if the loans are defaulted, they take on the land. Very rough. And so she says, yes, guys, it's not new, but still an interesting dynamic slash relevant. And yeah, this is the Belt and Road policy. And it is crazy and very controversial and has cost them a lot of money, but has earned them a lot of money, but has earned them a lot

of allies across the world. It's interesting. I was working with reluctant allies, different venture fund in in college and was looking through the cable gate leaks to hear what the embassies in different countries were saying about doing business there if you're an American. And overwhelmingly in some of these African countries, the sentiment was that you basically had two choices. You could do business with American companies or Chinese companies. But if you did business with the Chinese companies, they had the full backing of the CCP. And so if something happened where that company kind of stiffed you, you could go to the Chinese government and say, hey, make me whole. But you couldn't do that in an American country and American company. And that was there. There were a few oil spills that happened.

And American companies, there was, you know, they sue the company and they try and get the money from the company. But if the company is like bankrupt or something, there's really no more money to pay. Whereas China was kind of like, hey, we're backing up everything that we do in these countries. So yeah, lots of pressure. It'll be interesting. I don't think we're going to see an American Belt and Road initiative anytime soon. Well, the expensive, but some of the stuff that's happening in Panama, Greenland, like this is all related, right? Yeah, I mean, the the history u.s empire is basically one long belt and road initiative just wasn't branded in the same way but uh this is this is one of those things where

uh we like it when our team does it but uh when when our adversary goes and does it and yuck why how could they yeah meanwhile america we have military bases you know everywhere and we're sending money all over the globe for all sorts of reasons yeah so but some of that money high roi people people like to say oh we should yeah we're dying abroad with you i don't know about that yeah with ukraine we're sometimes uh you're giving us your mineral rights for uh yeah you're basically like our loans were backed by your mineral rights sometimes you gotta put your buddies on full scholarship and just help full scully full scully when they're when they're down you you spot them a couple bucks and then maybe

they help you out in the future it's uh one hand washes the other this is geopolitics uh anyway uh if you don't know bridget she's a fantastic crypto investor at Founders Fund. And let's move on to another crypto story, Bybit, which we mentioned earlier. They were hacked and they are now taking loans from rivals after a record crypto hack. Rival crypto exchanges are stepping in to shore up the finances of the world's third biggest exchange, Bybit. I had no idea it was this big. It must be Coinbase, then Binance, then Bybit. After hackers stole $1.4 billion from the firm last week, in doing so, the crypto

industry is hoping to calm investor worries that the hack could lead to another FTX-like implosion. Several exchanges, including BitGet and MEXC, have made emergency loans to Bybit to replenish tokens lost in the hack so Bybit's customers can withdraw funds without a problem. The loans are unusual given how quickly and easily they were arranged without prolonged due diligence and with few strings attached. And I think it actually makes sense because, you know, this was not, I mean, obviously they shouldn't have gotten hacked, but it's not Bybit was stealing the money, right? It wasn't fraud. It seems like it was like a mistake. And so, you know, the company management is probably more respected by other people in the industry and they're able to justify moving quickly.

They're able to say like, what would we want our competitors to do in this situation? And this is one of those things, if a $1.4 billion hole emerges in a crypto exchange, it is terrible for the industry. I mean, people were talking about this. People were talking about this last week. They were like, oh, this is actually bullish because Bybit has to buy back all the ETH, which would create upward pressure on the price. People will cope in so many ways yeah anyways so bit get a small arrival lent by bit for 40 000 ether worth more than 105 million dollars roughly five hours after the friday hack uh according to the ceo gracie chen uh the loan

has unusually borrower friendly terms since it charges no interest doesn't require by bit to post any collateral and doesn't require by bit to post any collateral and doesn't have a set timeline to pay it back it's more pure trust so i think everyone knows that like if there's a fracture it can go very badly for everyone in the ecosystem so it's worth it to just step in plug the hole and reassure the entire market and then sort things out i mean yeah and this was quickly a lot of the ftx stuff was triggered by luna and then what was three arrows capital those two major implosions yeah you catalyze this sort of uh but if but i mean you could play out a different scenario where all of

those all of those firms three arrows and and luna they all shored up the capital immediately and yeah you know coasted on although luna does seem like it was like a ticking time bomb and so the pure some serious problems but the pure irony here and i'm and i'm glad that big get is doing this uh you know i'm sure i'm sure they'll they'll it'll come back in other ways for them but the irony here saying the loan has unusually friendly borrowing bar terms since it charges no interest doesn't require them to pay back the collateral and doesn't have a set timeline. It's more of pure trust. The irony is that crypto is created to deliver trustless finance.

You didn't have to rely on trust. You could just trust the software and the code. So it's kind of interesting that they didn't do this. Reinventing everything, including bailouts and backroom deals between bank CEOs, basically. This would have been a really interesting way to flex the potential of crypto to be like, we raised a billion dollars on chain in this sort of trustless way in order to, you know, make our customers whole. And but, you know, I'm sure they had reasons not to do that. So Bybit said, we know each other well. We had common investors together.

We want to drive the industry forward and we want to fight our common enemy who are hackers, not each other. BitGet will not get any equity stake in Bybit as part of the loan. I'm sure all that stuff takes long time to paper to how much equity you're going to get. What are the terms? And so instead, it's just, hey, I'm going to front you and we'll figure it out later because otherwise it's going to bring us all down. It's so it's also just so crypto to be like, here, here's 100 mil quick. You know, the cool thing is that they were able to, I'm sure they were able to get that, the 40,000 ether over there in minutes in the same way that it took minutes to steal the 1.4 billion. Um, well, if you're looking for a little bit more of a normal finance app, uh, why don't

you check out public.com investing for those who take it seriously, multi-asset investing, they have industry leading yields and they're trusted by millions and we got a post from leaf abraham here scrappiness matters their team the public.com team produced a new tv ad entirely in-house no ad agency no production company just sweat and hustle from the team and a few grand of licensing fees start to finish 12 days and i love this i mean you can go see the ad but uh it's cool because like like people forget like it can go, you can deliver just an MP4 to the Super Bowl if you want to. You don't have to hire a team for that. Kanye does this. Yeah, if you're going to post it on Twitter or X, like you can just take that video and run it on TV.

And of course, they had a team internally that clearly knows their way around production. And so they were able to make something that looks very cinematic and looks great and you should go check it out but it it is cool to see a company that's as big as public like uh being still scrappy and and they probably had more fun and it was probably more aligned with their brand they just had less back and forth move faster uh and it's great to see and uh yeah i think that's an important lesson for for founders is like don't lose that scrappiness as you scale just because you can afford an ad agency doesn't mean you actually need one necessarily. You need to make sure that it's the right person, right along. Yeah, I'd like to see somebody do a company like Public do an ad that's just like a recording of a FaceTime call.

So it's like ringing. The CEO picks up, gives a quick spiel and like hangs up and like that's the entire ad unit. And it's just shot in one take screen recorded on their phone. Uh, cause you could imagine even just something novel, like having that, having that, the, you might have to license the sound from Apple, uh, but, but it would be a cool, uh, viewer experience. Yeah. Well, uh, let's move on to an interesting post from Casey Hanmer. Uh, I'm not sure if you're familiar with him. He was involved in the scrolls decoding project that Friedman did. He runs a very cool company here in L.A. out of the atmosphere and convert it into natural gas. And then they can sell that onto the grid as energy. Very cool company. Very cool. And he worked at JPL for a long time. He's brilliant. And he is asking the question of what should go on

the first Starship mission to Mars. And so the next window, so you can't just go to mars anytime it's not like the moon you actually have to the two planets have to be really close basically and it only happens every couple years and so uh we really got to hit that if we want to get on a cadence and we want there to be at least one mission during the next uh launch window and then we want like a hundred ready to go for the next window because you can't just, you can't just wait four years and be sitting on your hands. You got to test a lot. And so there are 601 days until October 7th, 2026, which is the mass optimal launch window to Mars and when that opens. So he says he doesn't have any privileged information, but it's fun to speculate about what SpaceX could choose to send on its first starship

flights to mars spoiler alert he thinks it's rods from god which is fun uh so over the side by the way uh calder in the chat says uh having to plan around the uh the optimal launch issue uh window is a skill issue yeah that's great uh and so uh over the next six or days spacex has a number of key technologies to demonstrate orbit reuse refill and chill so orbit uh this friday we're due for flight eight uh of starship which will be very cool to watch uh and this one may finally achieve orbit earlier flights technically had the performance necessary but deliberately targeted the ocean to prevent the possibility of a starship being stranded in orbit without propulsion capability and then undergoing an uncontrolled reentry and crashing. Reuse, this is the holy grail of rocketry. SpaceX has indicated that they may attempt to refly booster 14 on flight nine, which would establish booster reuse. Flight nine may also see the first attempt to catch the Starship upper stage. But in any case, the successful reuse of both stages of

Starship is necessary to fly to Mars. We got to be reusing these things if we're going to make it economical to get there and get back continually. Refilling. We can load Starship up with cargo from Mars, but it's not going to leave low Earth orbit unless Starship can be refilled with fresh fuel and oxidizer. SpaceX has been working on orbital refilling for some time, but we need to see it actually working. So you got to get a bunch of fuel up to space and then refill once you're in space, which is crazy because we're not sending something small. Can we also talk about how uncontrolled uncontrolled re-entry oh yeah the best uh that's just like crashing anywhere like crashing for just slamming back into the atmosphere yeah so it's it's high-stakes stuff they might have to shoot it out of the

world if it's a shoot out of the sky if it's coming down over a populated area uh and then chill by far the easiest of the four tasks, but long-term stability of Starship's cryogenic fuel will require that it be actively refrigerated, particularly in the challenging thermal environment of LEO or deep space. It's hard to make predictions, particularly about the future, but he's optimistic that SpaceX

will have multiple fully fueled Starships ready to go in October of next year, followed by a 10-month cruise and then either Mars orbital insertion. And so he broke this down in another article. He has a couple different things that he thinks that they should work on, but we'll just skip right to the spoiler alert. He says, Rod's from God. We know there's a high likelihood that mostly pure water exists within 10 to 20 meters of the surface across large swaths of the various prospective landing sites. Why not drop off a few dozen long steel or tungsten spears, guide them in while tracking them on radar, and then survey their impact craters with high rise as soon as the dust has cleared?

These rods will impact the surface at about eight kilometers per second. That's extremely fast. Penetrating many times their length and exposing the subsurface to our existing orbital instruments for the first time. The main attraction of this approach is that it requires essentially zero additional effort on top of the existing program, whereas the others require either a crash instrument development program or building and flying multiple intricate surface operations robots and landing them with an extremely untested EDL system. Rods from the gods merely requires dropping a few tons of steel in roughly the same area and then surveying the damage. It's also the

only method that can deliver enough energy to actually directly access the deep subsurface at scale and so uh uh jpl actually considered doing this of course they went with robots um but uh i i like that he's thinking outside the box and uh well this is the kind of this is the kind of citizen science that you've been asking elizabeth holmes to do right yes yes Like, hey, you're in prison, but like prove that you're legit by being at the cutting edge, making recommend, you know, I want to see Elizabeth Holmes, you know, writing an open letter to Pfizer and saying, hey, you're doing it all wrong. Johnson and Johnson. Yeah. This is how you're making this mistake. Yeah. You know, we actually had this figured out. Yeah. I mean, we'll get to this later in the show, but H51 the bird flu is going around what is elizabeth holmes take if she doesn't have one

you know keep her locked up uh but anyway uh uh he closes out by saying if all this goes well the technology situation in 2028 could look like this and that's i think the the next next transfer window uh easier technology uh spacex will have expertise in space solar photovoltaics mars air separation uh minor water prospectors and spacex will not be an expert in pressure structures and then the harder technology is orbital orbital refueling which we mentioned cryo fuel heat rejection mars edl space suits long duration life support uh surface life support and water miner and then uh spacex is not an expert in fuel plant rock miners construction robots nuclear reactors uh and this is uh this is i mean there's a lot of other companies that are working on this tech now um yeah yeah there's that uh what's the team that spun out of SpaceX that's doing nuclear stuff now? Not, it's the wife of the Farcaster founder.

Oh, I don't know. But there are a number of companies that have spun out of SpaceX to solve different problems that will come up as we colonize the moon and Mars. But what's really crazy is like, there's all this noise around Elon, all this noise around tech and stuff, but things do seem to be relatively on track for a private company to deliver something to Mars in just a few years, which is crazy and a huge milestone. And so I'm super excited for that and I'm excited to see where it goes. And I'm sure there'll be lots of other opportunities that crop up for businesses that are built on top of that plan because SpaceX isn't going to do everything. Yeah, I got a pitch.

I got a pitch last year. Maybe it was 2023 for a team building gas stations on the moon. They were like, we're Chevron for space. That was the thesis of like basically building, you know, fuel depots on the moon, which is a very cool thing to be thinking about. So Doug Burnhour, the founder of Radiant, another founder's fund, Andreessen Portfolio Company, he worked at SpaceX with Elon on special projects. Elon put him on a task. You know how the Falcon rocket has those legs that land

and that's how it lands? They took those off when they created the chopsticks that catch it. But those legs come down and they land the Falcon 9. Are they like LaFerrari doors? Basically, yeah. Oh, okay. Yeah, the LaFerrari of the rocket world, I'm sure.

And so he was working on that project. Then Elon was like, I like you, keep going, work on tunnels and boring company stuff for a while. And then he tasked him with, hey, when we get to Mars, what are we going to do for energy? And he ran the numbers on photovoltaics. And this is a little bit controversial. Not everyone agrees with his conclusion,

but basically because Mars is farther away, it gets less light. And so he was saying, if you need to set up solar panels to refuel a rocket to generate enough power, essentially, you'll need to bring tons and tons of like just so many solar panels. And the math doesn't quite work out in his mind. And so what he said was, you should bring a nuclear reactor, a smaller nuclear reactor, drop it down. It generates a lot of waste heat, which can be used to melt ice, creating water. And then you have a nuclear reactor that can then create the fuel out of the natural resources and then refuel the rocket and then come back. And it's a perfect, it's a perfect plot to a sci-fi movie now.

And it could be a, you know, good catastrophe documentary. It's like, Hey, we dropped this reactor and set off this chain of events that destroyed the universe. But, and now we have to say, we've got to save the universe. And so, yeah, I mean, they've, they've taken that. I mean, that's kind of like the long- And so, yeah, I mean, they've taken that. I mean, that's kind of like the long term inspiration for Radiant. But in the short term, they're building basically a nuclear reactor in a shipping container and you can drop it off in a military base. You can drop it off on an oil

and gas extraction site. You can drop it off at a hospital if they need, you know, a one megawatt diesel generator. You can just replace that with a one megawatt nuclear reactor. And, uh, it lasts for five years or 10 years instead of, uh, you know, like a couple of days until you need to refuel it with diesel. So in the military, you have to deliver diesel through these convoys and these trucking like over and over and over again, uh, radiant kind of, uh, eliminates that need. And so they're going into testing now and we're working on that. But when we get to Mars and you get sleepy, what kind of bed are you going to be sleeping on on mars eight sleep let's go we're putting an eight the first thing to go to mars is an eight sleep the first thing is a ramp corporate card for sure uh nights that fuel your best days turn any bed into the ultimate sleeping experience uh go to eight sleep.com

use code tbpn and uh send us your receipt and we'll uh we'll get your hat uh anyway let's stay on the topic wait i gotta look at my i gotta look at my sleep score last night before we jump in i i tried the warm-up thing i tried the warm-up thing and it's fantastic i woke up no problem at like 5 30 it was amazing uh yeah it's it's unbeatable all right i didn't do so hot i had an 84 but it was just because there you go it took my son uh needing to go to the hospital for you to beat me jordy texted me last night and says like i won't be out slapped yeah no one no i was serious about it i just only got six and a half hours i got six and a half hours too something something happened i think i think I got six and a half hours too. Something happened.

I think I set my alarm a little bit too early. Well, let's stay on the topic of space exploration, space travel, and NASA. NASA, you might have heard that there was an impact probability an asteroid was going to hit Earth. It was very risky. There was a polymarket all about it. But good news, the impact probability of asteroid 2024 year four has dropped to 0.004 it's expected to safely pass earth in 2032 let's go thank you to humanity undefeated and we have a post from uh sula here fell for it again award would you look at the time nothing ever happens uh so so anyways my my conspiracy i

don't have the tinfoil hat here unfortunately but my my conspiracy theory just said last week was that nasa really didn't want to get dozed and so as all the pressure you know they felt like the spotlight was going to shift to them they go oh oh there's an asteroid it's one percent more important than ever next day and they just were basically raising it like a point every couple days they just make us all afraid it's like well we better keep them funded because you know uh we don't want to we want to keep track of those there's a fun conspiracy i i wonder if it worked maybe maybe they got the email what are you working on and they said we're working on saving humanity guys and elon was like keep up the good work you know it's just like we're not cutting a dime. I don't know.

I mean, the NASA budget isn't that big. It seems reasonable. And it's fun. They produce inspiring stuff. That's important. All sorts of Americans all over the world. It's fantastic. Anyway, let's move on to Stratechery.

There's an article today about AI promise and chip precariousness. Anthropic released Claude's Sonnet 3.7. Dylan Patel had the joke of the day. He says Anthropic is also a Chinese AI company because of their aversion to the number four. And the point here is that, you know, Claude is just updating 3.5, 3.7. Why don't they just do Claude 4?

And it's because they see it as a GPT-4 level model in terms of total training size. That's the answer there. But it is funny that all of the different model companies are on slightly different numbering systems. And so GPT-4 is equivalent to Grok 3, which is equivalent to Claude 3 in terms of total scale. And this is all measured in flops. You kind of slipped there and you said guac guac guac which is a great name for a model grok has grok chocolate i want to say grok guac yeah uh

yeah for sure uh after this piece was published i was contacted by anthropic who told me that sonnet 3.7 would not be considered an E26 flop model and cost a few tens of millions of dollars to train, though future models will be bigger. So that's why they're fundraising. I updated the post with that information. The only significant change is that Claude 3 is now referred to as an advanced model, but not a Gen 3 model. It's all very, very confusing. I love Moloch's work, but reject his neutral naming scheme. Whoever gets to a generation first deserves the honor of the name. In other words, if Gen 2 models are GPT-4 class, then Gen 3 models are Grok 3 class. And whereas Sonnet 3.7 is an

evolution of Sonnet 3.5's fascinating mixture of personality and coding prowess, likely a result of some anthropic special sauce and post training grok 3 feels like a model that is the result of a step order increase in compute capacity with a much lighter layer of reinforcement learning with human feedback it answers are its answers are far more in depth and detailed model good but frequently becomes too verbose rl hf lacking it gets math problems problems, right? And that's, and that's referencing, that's referencing, uh, Grock, you know, saying that, uh, you know, who should somebody asked, uh, uh, if, if like one, one, what was the question? Who spreads the

most misinformation on accidents is Elon or, or who should get the death penalty? It says Trump. And it's like very controversial and not even aligned with like what Elon believes, obviously. And so clearly just an aberration of the model and it gets math problems, right? So the model's good, but its explanations are harder to follow because the RLHF is lacking. And that makes sense because it takes a long time to actually do RLHF, generate all the data and, uh, and buy new data. Uh, it's also much more willing to generate forbidden content from erotica to bomb recipes while having on the surface the political sensibilities of Tumblr with something more akin to 4chan under the surface. Grok 3 is also a reminder of how

much speed matters and by extension why base models are still important in a world of AIs that reason. Grok 3 is tangibly faster than the competition, which is a better user experience more generally. Conversation is the realm of quick wits, not deep thinkers. The latter is who I want to be doing research or other agentic type tasks. The former makes for a better consumer experience in a chat bot or voice interface. And so a lot of people are having this kind of paradigm shift where if they just want to chat and talk through an issue, they'll use Grok 3. And then if they really want a research report, they'll fire off OpenAI's chat GPT deep research, let it cook for 5, 10, 20 minutes, come back and have the full thing. And I found that there's this interesting collapsing process where deep research is great at pulling just every single possible source from the internet. And then I don't know what's going on here, but HDMI is iffy today. And, and then, and then,

so it'll generate 5,000 words, 10,000 words. And then I'll fire another prompt to say, Hey, summarize this in 10 bullet points because I trust that you've done enough research at this point to have all the good data. But so he goes on to say ChatGPT, meanwhile, still has the best product experience. Its Mac app in particular is dramatically better than Claude's

and it handles more consumer use cases like math homework in a much more user-friendly way. Deep research, meanwhile, is significantly better than all of its competitors, including Grok's DeepSearch, and for me, anyway, is the closest experience to AGI yet. OpenAI's biggest asset, however, is the ChatGPT brand and associated mindshare. COO Brad Lightcap just told CNBC that the service had surpassed 400 million weekly active users, a 33% increase in less than three months. OpenAI is, as I declared four months ago after the release of ChatGPT, the accidental consumer

company. Let me see if I take this out and put it back in. Let's see. Okay. Well, fun report from Ben Thompson over at Stratechery as always. Let's move on to some fundraising news. We got Nico from Default today. I'm incredibly excited to announce Default's new inbound orchestration platform and an additional $4 million in funding led by 8VC. We started with Default with the vision of building the best platform for revenue teams

to orchestrate their inbound workflows. And so default, if you're not familiar, is a tool for helping B2B SaaS companies deal with inbound requests and pipeline, basically lead qualification, pipeline them into trials, get them to convert into real customers. And so he says, what's that? We good? It's still staticky. I don't know what's going on here. You think it's the cable?

Okay, we'll see. We got a problem with you on the stream, Jordy. Good thing we're ironing this out. We're trying to figure out FaceTime today so that we can. I think it's the computer has now realized that I'm not wearing a suit. Yep. time today so that we can yep that makes sense let's see let me see if I can change this to audio video microphone okay can you hear me I can hear you. Okay. We're good. So he says the go-to-market landscape has changed more in the last five years than in the previous two decades. Teams can no longer afford bloated tech stacks, inefficient workflows, and fragmented

data. Winning in this new era requires AI-driven automation, seamless data orchestration, and intelligent routing. What is default 2.0? Embedded website intent data and conversion tracking. One of our biggest learnings since launching 14 months ago is just how much data is siloed in the inbound marketing funnel, especially on the website. Almost every revenue team we've met with has told us that they're flying blind on what's happening at the top of their marketing funnel with no visibility or control into website activity from target accounts. So they want to see, Hey, you're thinking about buying this product. What are you doing on our website? Are you a qualified lead or not? Are you worth the CEO getting on a call

with you to pitch you? Or should it be some SDR? Or should we just have you fill out more forms until you're actually ready? And so very cool to see him building. He's been building for a number of years and launched a new product. I mean, they've just been so focused. You know, they came out a few years ago at this point with the core idea of what they were going to do, but they've just been in the trenches every single day iterating on like, actually, what is the product? Like, what is the product actually do? Because people oftentimes come out with like, defaults plan was always to be this sort of like inbound lead processing, you know, sales enablement,

you know, product. And, you know, it's been amazing to watch them just sort of, you know, iterate and hit these milestones and, and acquire some very meaningful customers at this point. And I think he's been smart about sort of staggering out his, his capital raises. He raised, I think, a seed round from Kraft, adding more firepower from 8VC now. Awesome to see and bright, bright future. I mean, I think I just appreciate how Nico didn't leave anything to chance.

He's just been focused on actually making sure the product is meaningfully differentiated from the rest of the market and making it better every single day. It doesn't feel like one of those hyper bloated rounds where it's like, oh, they got $50 million out of nowhere. And who knows? Yeah. Like now it's $2 billion company. It's like he's putting one foot in front of the other, making sure the product's working, iterating, actually taking time to build this stuff. And going with tier ones instead of you know some frothy you know partnership i don't know how this was priced but i'm gonna guess that this was like a 10 percent deletion

round or potentially even less um seems great well uh you know if he keeps going he's gonna take a shot at salesforce salesforce scott's matthew mcconaughey on board should he get a celebrity to endorse saquon saquon yeah uh well Saquon's like the new Jeb Bush. It used to be that every founder wanted to raise from Jeb. Now Saquon's the next target. He's the kingmaker. Well, there's a report in the information about investors souring on celebrity brands. These days, it seems like every celebrity wants to start their own brand. Just look at Robert Downey Jr., who just raised venture money. I love that. Let's go. But investors say they're overwhelmed by pitches for undifferentiated or low quality

goods and are quickly souring on the sector. Four early stage investors who previously backed celebrity brands said they are shifting focus to promising products as opposed to celebrity buzz, especially after high profile flops in once trendy categories like beauty. This could be a rude awakening. There's a funny quote in here. One investor was recently pitched on a protein puck cookie like snack fronted by a rapper, but passed after the investor didn't like the taste. The firm was also reluctant to invest given the number of high protein products already on the market the investor said and wasn't convinced the rapper's fan base was enough to make the snack a breakthrough right now and that's true it's like for celebrities the ones that seem to do well in consumer have a they're

just like they're posting all the time. And so there's this idea that it's like, if you're just a celebrity and once a year you're taking over the world with some Academy award winning movie that doesn't give you permission to tell your audience about some promoted product every single day. Right. Like the Daniel Day-Lewis energy drink isn't going to make its appearance in Lincoln or there will be blood. And so unless you're following Daniel Day-Lewis's Instagram or podcast, which he doesn't have, you're just not going to hear about that outside of one press release. Now, there are other people you look at like the Mr. Beast case study or the Kim Kardashian, where they are posting their life every single day. They have so much content going out

that they can not only basically run a lot of actual direct response ads for their product on a very consistent basis, but they can also tell the story of building the company and building the product. And that's what makes it a little bit more um yeah there's there's there's a difference there's here's where i would separate it out there's a difference between fame and attention right you can be famous and have very little owned attention right so owned attention to me is jake paul yep you know absolutely grinding on content he's he's famous but he's also delivering content every single day for years and years and years and years. Logan Paul is the same way. Another, you know, another good example is like Kim Kardashian, right?

She's gone on these crazy television. And she has the sort of sustained attention of posting every single day. And I think that a lot of these celebrity brands have popped up. You know, Casamigos is a good example, right, of this sort of old style of celebrity brand where you just are relying on fame to drive acquisition. And I think that works in a category like beverage where a lot of the sales is happening offline. But for these like online, more like sort of digitally native brands brands like i think you need like actual distribution channels and so some brands that i've seen that that actually work well is like

product innovation of positioning a a like 200 milligram caffeine energy drink but positioning it as a sports drink and selling it to kids. And then like combining that with massive attention, better, uh, Jake Paul's betting, uh, and Joey's, uh, betting company. Like that makes sense because Jake Paul's posting about boxing and sports and training all day long. Like it's just a great, um, audience, but, uh, but ultimately, you know, I got pitched a company like a year ago that was like that Snoop Dogg was doing. And I just didn't really have, like, I was like, okay, you're the, your play here is to take a commodity CPG product, put Snoop Dogg's face on it. And, and that's going to drive more sell through in the grocery store. And, and I don't necessarily buy it, right. If it's truly a phenomenal product but the other thing is like you know another brand that i think has done well is

um kendall jenner has 818 tequila yep she's completely integrated that into her life so when she's hanging with her friends on the weekend they're drinking 818 when she's going to some show you know in vegas like they're serving 818 at the bar so it's hyper integrated into her life yep these other celebrities that are like oh yeah i'm gonna have my tequila brand and you know just um hope that my face moves bottle like i'm just super bearish on that and i also get overextended like you think about like snoop dogg or shack like they have so many different partnerships at this point it gets gets really, really difficult to know, you know, like Clooney was kind of right. And like Casamigos was his thing and you didn't really see him everywhere. Uh, with, with some of these other celebrities,

they're doing so many things. It's hard to track. Okay. What are you selling this day today or tomorrow? It might be something different. Who knows? Do they actually care about this product? Do they actually use it? yeah yeah i also think like ryan gosling applied the more traditional style of celebrity brand building which was just i'm famous i'm going to take a commodity product and and use my fame to just like drive more acquisition but that doesn't necessarily work in some of these more differentiated categories where consumers are discerning right they're not going to eat the they're not going to eat the snoop dogg pizza rolls every night unless it's actually like the

best pizza roll at the best price and you know um yeah and uh was it was it gosling or reynolds ryan reynolds was the one with the tequila and the sprint mobile or the the yeah ryan reynolds he he seemed really deeply integrated with those companies where like he was you know almost directing ads and like really seemed integrated with like his his brand he would like come into a company that was already kind of working mint mobile and then yeah really make it all about him for a while and kind of but do it in like this almost self-aware way that kind of broke through it didn't feel like he was just getting paid to endorse it it felt felt like he really had a stake. And I think he did when we looked at the numbers, he was making a lot of money and had a big equity portion. Yeah. And the last, the last

thing I would say here we've seen with, with, uh, skims, uh, um, Kim Kardashian's brand is that extreme to do bill. Uh, in the example, you know, in the skim situation, she has 5% of the company now after raising $700 million and skims is going to be a big company, but it required a huge amount of capital. So again, even if you have that attention, uh, you need brilliant execution and in skim's case, a brilliant product well so um well john i'm glad we're dealing with this technical issue while it's just me and not our guest tomorrow who uh is going to be uh you know is a busy guy yeah yeah we'll have to iron it out i think we'll be on zoom on a different computer running it from this laptop is just a mess um but uh let's do a

quick promoted post for bezel we love bezel here shop over 22 000 luxury watches fully authenticated in-house by bezel's team of experts jordan and i have been going back and forth on what the next watch should be uh we're gonna have to start doing some of these off the balance sheet it's gonna impact profitability but it's sort of you know siloed away from our household finances which get a little bit more scrutiny you know uh and uh what do we got today we gotta start we gotta start actually the ad is just us bidding on auctions yes yes yes uh they're highlighting a uh uh presidential rolex gold day date champagne diamond set president fifteen thousand dollars available highly recommended it's the watch from glengarry glenn ross look at this watch this watch costs more than your car it's great that's great uh yeah highly recommend bezel fantastic experience um anyway let's move on to the

timeline we got sully saying uh you have approximately one year before the normies catch on go build what is he talking about talking about he's just talking about just generally about ai okay i think i think this is one of those things you can tell the average person on the street that intelligence is free and they can create an army of robots to create wealth for them. Yeah. And they'll be like, cool. Like, did you see the new Netflix show? Yeah. You know, so I think this is one of those things like, you know, we the Internet was already like personal computing was amazing. The Internet was amazing. Mobile was amazing. All these major crypto is amazing. All

these, you know, tech trends, even when the average person sort of starts to experience them in their daily life, the lack of agency is what makes what makes life great for the entrepreneur is, yes, you know, millions of people, billions of people could use the same AI tools that everybody listening to this and, and the same tools that we use, but they're just not going to because they're fixated on, uh, the next, um, you know, the next, uh, TikTok video in their algorithm, the next, uh, Netflix special. And, uh, that's why, uh, there's never been a better time in history to be an entrepreneur. You can actually just lap, uh lap everyone over and over and over. And you absolutely should. Yeah, for sure. Let's move on to Sam Byers.

He says, in case you didn't have a sense of scale of Chinese overseas fishing fleets, strip mining the ocean life and leaving it bare for those who rely on it. Dozens of ships as far as the eye can see for months and years at a time argentines argentinia's military forces patrol argentine waters to keep chinese fishing fleets out china is running rampant on the ocean apparently i had no idea about yeah this this video is absolutely insane yeah uh it looks like a scene out of dune where you have the uh what do you call those those harvester things i forget yeah yeah i know exactly what you're talking about yeah the the sort of um you know america has gone back and forth sustainability is you know sustainability was i feel like when i was graduating uh like in like 2019 it was the hot thing it was anti-plastic

it was like you're gonna bring your uh you're gonna bring like glass to the grocery store and like put rice in it like it was like we sort of reached peak sustainability like pro planet there were consumer brands that said like their entire differentiation as a brand was you know we're eco we're green we're one percent for the planet that's why you should buy us and that's really faded in the mind of the consumer, right? I don't think that eco marketing even works anymore for a pretty broad. And this was like peak Allbirds in many ways, right? Yeah, totally.

And this, when you see videos like this, it makes all the sort of like ocean plastic brands, it makes them almost seem like, you know, what, what is even the point if, if, um, there are groups internationally that are going to, um, completely, uh, nuke, uh, this, you know, uh, all life in certain areas and and it doesn't you know i'm not a uh i'm not a maritime uh uh you know uh life expert but one one year of these sort of chinese fleets you know fishing like that off of your shores and you could set back a local um you know the local um fishing uh you know, the local, um, fishing, uh, you know, sort of supply or, or, or life back decades and decades and decades, potentially a hundred years. So what they're doing out there,

uh, is not, is certainly not sustainable. Uh, and we don't even know what the impacts are yet. Yeah. Uh, I mean, that's always been the trade-off is like in America where, you know, putting away the plastic straw as well. You know, Brazil is deciding whether or not they should burn down the entire or clear cut the Amazon rainforest or China is like firing up the next major coal plant. And if there's not some sort of global treaty, you know, you're just going to fall behind. I had a crazy story from that era of like hardcore eco-marketing. I had a buddy who I was still in college and he came to me and he was like, dude, I'm going to make this collapsible reusable straw that people are going to

put on their key chain. And he went all in on it, like took loans out from his parents and was like, he basically, you know, for him at the time, $30,000 was like a monstrous amount of money. And he was like, why don't you do this with me and i didn't i didn't see the vision at all to be honest like personally i wasn't i was like i'd rather not use a straw personally than just like

bring mine around and be like putting it on my keychain and stuff so i didn't get it but he saw the vision and he went for it and this is the beauty of consumer products is that you don't really know if something oftentimes like you have a strong feeling that something's going to hit, but you don't know until you start selling it. And so he spent 30 grand, you know, making this cool video and he launched this thing on Kickstarter and he did like just under $2 million in 30 days. Wow. And then he had a whole new set of, and I was like blown away. It was amazing. But then he had a whole new set of problems. It was like, one way it was amazing. But then he had a whole new set of problems. It was like, Oh, did we get our margins?

Right. Like, can we actually meet this man? Right. Um, absolutely insane rollercoaster. But I think that was, that was peak, um, peak. Go on the chat right now. Um, bring back plastic straw. Something like completely losing you.

Let's let's do a couple more posts. Jordy, you there? I'm there. I can hear you. Cool. I like this prompt. Save yourself $200 a month. Chat GPT 4.0.

Think of yourself as the premium model. Chat GPT pro plan. Got it. I'll provide responses at the highest level of detail, accuracy, and relevance optimized for your technical and business needs. Let me know how I can assist. Um, I don't know if that actually works, but it's a, it's a funny joke and, uh, never a dull day on AI, AI X, I guess is what we're saying these days. Never a dull day. Yeah. Never a dull day. Um, let's dull day. Let's go to Jordan Singer.

He's got some announcements. Say hello to Cobot Company. We're sharing an early preview of Cobot, an entirely new way for teams of people and AI agents to work together in a shared space. Jordan Singer says, we're so excited to share more about what we've been working on at Mainframe.

We're fascinated by agents because they're capable of doing real work and producing outcomes that can accelerate teams. We don't think the experience around them has been cracked yet. Enter Cobot Co. What's your take on Cobot Co? So one, I love what Jordan's doing here, which is he's created this sort of like parent brand, which is mainframe. He like launched that, he raised money for that. And then now he's shipping what you can think of as as you know products but he's almost operating more as a product studio yeah and uh at least from the outside i'm not i was um involved with his last company but but not in this one um and

uh ultimately uh you can imagine him like shipping more and more of these things as all the technology evolves quickly this sort of mainstream mainframe allows him to just like continue to ship and iterate uh without the without the same level of risk as when you start like lucy and then you ship lucy's first product and if it doesn't work or it doesn't scale as fast as you want it to you're you're stuck with it not that you experience that uh quite the opposite but um anyways i think this is fantastic. I really love the visuals here. It's so hard to actually stand out visually online anymore. You have to come up with something that's truly unique.

And I think that, uh, Jordan and the mainframe team actually crushed it here. And, uh, I like the name Cobot. Yeah. Uh, is, is fun. Well, speaking of, uh, robots, uh, let's go to Yaxin. He says prediction household robots are going to be the size of Dobby the elf. It needs to be small enough that you feel comfortable crushing it with your boots if it bugs out and starts attacking you. And yeah, we've seen a

couple of things. I think Reggie was talking about this, how, you know, you shouldn't have a computer that you can't throw out the window. And I agree with this. This is like the original Asimo. And once the Boston Dynamics robots start getting bigger and bigger, they start getting scarier. And unless it's an industrial application, a lot of things could be done with a smaller robot. It's cheaper and it's certainly more reassuring. Do you think that when robots like truly start to feel alive

that at night people will want to just go and put them like in a closet, like lock the door this feeling of like if something feels sentient even if you know that at the end of the day you can just kind of unplug it does that give you this sense of do you want to go to sleep when you've got three or four little dobby sized robots you know walking around uh dobby uh i don't know i mean if it's quiet i i know people put away the roombas uh there's been a lot of these features that have been built into these like uh you know the the nest the smart home the apple the the the alexis we typically have like a hardware level kill switch to let you know that if the ring is red it cannot listen to you at all uh facebook came out with a uh video conferencing product

called the portal uh you'd attach to your tv and it would have a very high quality video signal that would come through um and that had a slider that you could slide over the camera because facebook didn't want to be known well that was the thing wasn't on you didn't zuckerberg like have his he always would have his laptop like taped you remember that that came out that was a big thing in case you in case they want to be able to close the the the webcam entirely think about how many times zuck has been mogged by apple like he tries to uh like eventually he's like that that whole portal device like should have been a hit, but they launched it at the wrong time. And then like the technology itself was amazing. But then they were competing with FaceTime, which like everybody has their, you know, computers, iPads,

iPhone. FaceTime is like probably, you know, aside from the complications we're dealing with right now, probably one of the best products Apple's ever made. Yeah, it's not a FaceTime issue. It's an HDMI issue. We got this big, long HDMI cable and it keeps cutting out. You keep seeing the static on the stream. It's terrible. Well, let's go to David Holes. He says, crazy question.

How many kids would you be willing to have in exchange for paying zero taxes? And 26% of people said, I wouldn't change my plans. Only 13% of people said i would have 10 kids um i mean i guess it's only 13 i mean i don't know that's kind of crazy i mean it goes to this idea of like i i feel like kids are a fixed cost and uh taxes are a variable expense because they scale with your wealth so yeah if you're making like 10 million dollars like you're you're not going to wind up spending more on the incremental kid there's actually a lot of economies of scale to having lots of kids um yeah but it's more of a reflection of just like how many kids do you want to have i

think i love how you're you're you just said kids have economies of scale that's a totally that's a great post and it's real um yeah i think that one thing i saw this earlier um the you know how you can write i think it's like 6k a year you can write off for dependent support it's like child care basically and that number if you're a parent and you've like have to hire child care just feels like so like uh ridiculously low right uh i probably spend across like multiple nannies and babysitters, stuff like that, like around 100k a year on childcare. And so the fact that the government's like, okay, you can write off 6k. I'm like, okay, thank you for nothing, basically. But you know why it's 6k? Is it was that number was set in the 70s. they haven't adjusted it for inflation inflation so in the

70s it would have made sense like some babysitters making 6k a year something like that to to take care of your kids now it's like that gets you one month of child care in california and new york um and uh absolutely wild but um there are some countries that have done this type of stuff where they've put incentives in place to yes uh encourage more uh uh pronatalism and there's there's a bunch of other pronatalist uh features we should do like a whole deep dive on it yeah i mean the funny thing here is i do think the right answer is i wouldn't change my plans because you should be having kids based on your just broader ability to, and commitment to parent and provide for them in any scenario. Uh, and so, um, playing around that. Anyway, let's move on to crime junkie host.

Ashley flowers is building a $250 million podcasting empire. She hosts the second most popular show in the United States. Now she's brought on investors to take on YouTube, Hollywood, and her biggest live tour yet. Good afternoon from Los Angeles. It's warm, sunny, and great to be home. Thanks to those of you who attended our podcast business summit. This is from the New York Times. Let's break this down. This 36-year-old podcaster makes $45 million a year in profit. That's a lot of money. She's the creator of the popular Crime Junkie podcast, and she's still getting used to being on camera. I got into podcasting so no one would have to see my face.

But now, of course, she has to be on video. She built Crime Junkie into the second most popular podcast in the United States. Joe Rogan is first. The weekly show reaches about 6 million people, according to Edison Research. Over the last few months, Flowers has started recording video episodes of Crime Junkie to expand the show's audience. Her video setup is a work in progress.

The studio is an old gym on the ground floor of her company's offices, located in the broad, ripple neighborhood of Indianapolis. Soundproofing materials are scattered across the room, and a couple of her producers oversee a makeshift video bay. Not much, not much different than ours.

A couple of monitors on her table along one wall. One employee sits on the floor with her laptop, helping flowers when she stumbles over a pronunciation. The person in charge of the makeshift teleprompter sits to the right, scrolling the script up and down manually. Flowers is now constructing a new studio devoted to video, part of an expansion that will triple her office space to 30,000 feet and double her staff to almost 130 people. That's big. And so she raised $40 million from the churning group. Yeah, it sounds so insane. But when you're making $45 million in profit a year,

having 100 plus employees seems totally reasonable yeah definitely and it's cool to see so the churner group has been making you know it'll it'll take the fullness of time to understand how these bets play out but they were also in doug demure demure demuros uh doug demure cars and bids they did cars and bids and as part of that i believe the doug demure channel went into that holding company, essentially. Yeah. And they've also been in Barstool at various points and Reese Witherspoon's Hello Sunshine production company. And so they've been doing a lot of stuff in this.

Flowers could have funded the effort herself. Audio Chuck, named after her dog, turned a profit of about $45 million last year, according to uh who asked not to be identified discussing confidential if they were john let's just put this into perspective for uh the audience if uh they were getting figure uh ai's multiple they would be worth um 100 trillion dollars probably probably so just to give you some some context there and so she wants churnin's help expanding her business from a podcast network built around her to a media company that spans audio, video, merchandise, and live events. AudioChuck's founder already wears many hats at the company. She is host of two primary podcasts, Crime Junkie and The Deck, and serves as chief executive officer. She gives feedback on stories, closes deals with distributors, sells advertising, and tours the country, taping live episodes. She's also the mother to a three year old daughter. So she's got a lot on her plate and she wants help. And that's why she

called the churning group. Uh, crime junkie began as a side project. She loved crime stories since she was a kid reading Agatha Christie mysteries with her mom. After graduating from ASU, she got involved with local organization called crime stoppers that helps people report crimes anonymously. To help promote the organization, she began hosting a weekly local radio segment called Murder Monday. Flowers could tell listeners wanted more. Inspired by the true crime podcast serial, she and her childhood friend, Brit Prawat, started taping a show before and after Flowers day job at a software company. They released the first episode on December 18th, 2017. So,

you know, good run. I've never understood the, well, I've never personally been interested in these crime podcasts. I listened to one on Ross from the silk road that actually was cool. Uh, and helped us sort of,

uh, craft the story of, of what was sort of the business story around the Silk Road. But I think the story is amazing. There's so many more of these that you don't hear about. Right. And this is why David Senra felt for a long time. He felt like a crazy person. Right. Being like, why don't people understand that podcasts are basically have some of the best business models in history um why aren't more people like taking

these seriously so and then at the power law extreme it can look like a very very large business that's throwing off tons of this is a better business than cursor right like i mean right not not like not that it you know it can't scale to the same potential but yeah um you know it's uh and it's funny because we were actually we're messaging with sender about this and he's saying that it doesn't even make sense to look at these on an ebitda basis because there's this profit it's it's just every there we go you're back okay now you're back i mean rant around the day i know it's been terrible today.

Now we've been, you know, keeping the show going, um, back, but anyways, do you want to wrap up, John? Yeah, we should wrap up. We'll, we'll, we'll, we'll cover some of these tomorrow. Anything else? Um, but did you see the Sigma camera that launched i thought this was kind of cool i thought that was cool yeah so some people were saying this is what apple would launch if they built a photography camera so sigma just announced a brand new camera no sd card slot a single usb port 230 gigs of uh of storage that can support 2.5 hours of video or 4,300 raw images color modes 6k a log and so it's really just designed to be something that you just pick up point and shoot and it looks fantastic and i thought that was a interesting device we've been talking about like the desire for more gadgets more things that you could gift to someone at the holidays because we're definitely

in a we're definitely in a hardware renaissance in a hardware Renaissance. Yeah. Yeah. It's awesome. But what I, what I love about this is that it's not just beautiful, it's performant, right? So a lot of these new hardware products are sort of driving sales based on their sort of nostalgic side or super future facing, right? Like you saw with like the rabbit R one, uh, but seeing a product like this that's not only beautiful but highly performant uh is is fantastic and uh i already uh i already want to pick one up yeah

it seems like a good gift uh let's close with christoph he says movie idea saw but it's all developers and they need to solve leet codes to survive and we got of course tagged in this because we've been doing pmfr die and we highly recommend you head over there and watch that stream after you finish listening to this episode. And so good luck to the boys in the cage. They're making great progress. And every day the streams are improving. So thanks for listening. Love to see it. Please leave us reviews and stay tuned for the next one back in the studio. We'll see you back in the studio tomorrow. See you later. Thank you. Bye.