TBPN Live - Merry Christmas, Hit The Dom Button, Yearly Tech Recap, 2025 and Beyond

Episode Date: December 29, 2024

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(00:00) - Merry Christmas (01:55) - John's Workout Challenge (05:00) - Yearly Reflection (41:28) - Reply Guy of the Week (43:45) - DM's (47:15) - The Timeline

Welcome to Technology Brothers, the most profitable podcast in the world. I am sick as a dog, but we're celebrating because we doubled and it's a Dom episode. You know what that means. We have a tradition around here. Straight into it. Every time we double, we pull out a bottle of Dom Perignon and enjoy it on the show live. And the timing with this actually worked well, given that it's Christmas Eve. It is.

Merry Christmas, everyone. It is. Yeah. Merry Christmas, everyone. We're having quite the morning. We got here, we woke up around 5, tried to get to the studio. Jesus. Jesus.

So bad. Oh, my God. I didn't realize I was so pumped up. Oh, that's great. That's fantastic. Anyways, strong start to the episode. Strong start. Jesus.

Merry Christmas, everyone. Merry Christmas. Anyway, so I'll just recap the morning briefly. So we got up around 5, wanted to get to the studio early because, of course, it's Christmas Eve and we have family stuff going on family in town but we weren't going to miss uh an opportunity to record a podcast it just would not be on brand for us um and uh so anyways we get here john is deathly ill from some uh from some

disease that his uh cousin picked up at school It's like a thousand times worse than COVID. It's yeah. So if you've ever had COVID, COVID is a thousand times worse than that. He keeps horrifically coughing. We just shot another video and it, it was tough to get through it. And then we had a huge sort of printer malfunction,

which was making this show difficult considering we rely on printed, quite a lot of printed content to get through it. Um, but we made it here. The worst part is that I got goaded into one of the hardest workouts of my life yesterday. So I heard a rumor that, uh, one of my buddies, who's a founder of a, I guess, deck of corn now, uh, is like the fittest man in tech. And so I hit him up and I was like, I hear, I hear you like the fittest man in tech. And so I hit him up and I was like, I hear, I hear you're the fittest man in tech. Like, I'm going to challenge you. Like, let's go. Like, I want to work out. Where do you work out? I want to work out with you and see, see how jacked you really are. And he sent me his, uh, his like just sent location. Yeah,

he did. Basically he was like meet here on the 23rd. He sent me his body fat percentage, 9.9%. Last time I tested, I was at 14.5%. Which is good. It's good. Healthy. But 9.9 is low. Yeah. And so I was like, okay, no whammy, no whammy.

I've been working out. Maybe I'll be down 9.8. If I beat him by even 0.1, I win. So I get down there, get on the body fat scanner, comes back 9.4 i'm like let's go i beat him he really you came in at 9.4 yeah he gets on 9.1 he dropped something it happened and so he smoked me but i i beat him on it gives you like this aggregate score for like muscle and fat combines everything complete like you know nonsense heuristic is this a dexa yeah exactly so it gives you like the score

and uh and i beat him on that so i was like this is what we were really focused on here of course not the body fat percentage of course and then he proceeded to destroy me he's been training for a you know the mirth yeah the mile and then the 100 push-ups 200 100 pull-ups 200 push-ups 300 squats or something and i just got destroyed yeah it was so bad but it was a lot of fun and uh yeah i was really close to bailing i said i was like i got this terrible cough i can't make it dude and he was like don't worry there's a bunch of doctors you can go to that are nearby and he sent me a screenshot just looking for gynecologists and i was like fuck i literally got in my car i was like i can't not go and i was just coughing up a lung dying the

entire time and it made me so you wouldn't have been invited back i i came back with a 103 degree fever it was so bad amazing so bad but i got through it and that's what matters so today we are recapping the big moments of 2024. Hey, before that, one big moment of the year was for us. We doubled again. Cheers. Well, we doubled in this podcast. Very grateful for it and the brothers.

It's been fantastic. It is an honor to record this. I genuinely feel like it is a platform for the community for us to highlight, you know, people doing great things, writing funny posts, um, and, uh, changing the world big and small. Yeah. Yeah. It's been, it's been a lot of fun evolving it from that first, first demo recording on zoom. That was just us like riffing about random stuff to a much more structured show, Being able to highlight funding rounds, the size gong and personnel news and people hire, job postings.

We have a lot more planned for 2025. We're going five days a week, people. Yeah, that's the big news. Subscribe right now. Five days. We're going to be in your feed constantly. It's a daily show starting in January. And just a little bit of a reminder,

we are 100% corporation supported. We never ask our listeners for a dime. We want to avoid audience capture. We ultimately don't care what the audience thinks. If the audience wants us to talk about some culture war thing or that'll go viral in the algorithm, that's not what matters. What matters is what do the corporations want us to talk about.

And so this Christmas, if you're looking for a Christmas gift for us, ask your boss. Maybe we could sponsor. Hit us up. Starting at six figures. Yeah. So.

Yeah. Just six figures a month. You can support Technology Brothers. It's less than what an average product manager makes at a fan company. So it's not much. Put us in the team. Awesome. Yeah.

So for context, we just recorded a whole bit on our on our annual awards which will be going out via x uh so you can look out for that but um in going through that process we spent the time to just kind of reflect on some of the bigger moments of the year and it felt like a monumental year in tech for many reasons. I think the election played into that because, uh, you know, tech was just so intertwined with the election this year, given the influence of X and Elon and big venture funds.

Um, and it was really more, you know, the election specifically was more of a, that more of a cultural battle, even within tech than we're now even remembering. Like I saw Based Baron posted yesterday screenshots of all the VCs that were VCs for Kamala

as a website. The website's since been scrubbed, but there were probably 100 names up there from many of the biggest firms in the Valley, people that were backing it. So I would, you know, since we don't talk about politics on this show or social issues i would just vote that we just try to forget about that list we just all move forward together we should just have a website called vcs yeah and

just any vc can can just you know put put your name on that site or like tech.com you just like if you support tech i'm just technology if you're a VC and you support technology, go to technology.com. You can put your name on the dotted line and you can support tech. Who's that guy with the NFL hat? Is that a meme where he's at the NFL game and he just supports the NFL? That's what you want to be? Just VCs for America. Pro tech, pro business.

You support both parties. Pro America. Pro America. vcs pro tech pro business support both parties pro america america yeah i think like the reason we're able to have a non-political show purely business is because we generally support everybody that is pro technology pro business pro america yeah that was one of my hot takes like the what what do you agree what do you believe that very few people will agree with you on is that every american president is amazing yeah i love them all and i would get beers with all of them and i think they're all incredible and i think that you don't rise to that station without having an incredible life story and every time i've dug into someone even someone who's been

controversial or i did a whole hour-long documentary about trump and then an hour-long documentary about uh biden and they both have like incredible stories and it's very easy to get lost in like or I did a whole hour long documentary about Trump and then an hour long documentary about Biden. And they both have like incredible stories and it's very easy to get lost in like, what's going on this week with this particular policy and is this policy good or bad? And there's certainly some politicians that benefited me or my worldview more,

but I think we've never had a Stalin or Hitler in America and every American, uh, president has been unique and interesting and someone that I would absolutely want to get dinner with if they were alive or available. Yep. That's a great point. Pro America.

Pro America. Uh, where do we even want to start here? There's a bunch of big moments, um, you know, everything from open AI to true social going public to an almost tiktok ban that didn't end up happening to software dying and being reborn to humanoid robots to open ai drama

to the open drama was mostly last year this year was kind of a rebuilding yeah but pivot to to product the talent exodus this year was intense that was intense but they were they were kind of two separate things the cadence of product releases is really really strong yeah that's really really good that's why i put it at the top is despite all the drama and the various yeah people high profile people co-founders leaving the product velocity was still insane. And it still throughout the year was this tug of war in terms of, you know, Anthropic versus OpenAI versus,

oh, what is Gemini doing? Like there was this- What's Ilya doing? Yeah, there was this sort of tug of war for attention and users throughout the year, but it felt like OpenAI seemed to be, stay mission oriented, even if their mission had changed year over year in some ways right shifting to for-profit yeah i mean it's a remarkable shift from a research organization with a very much more academic style of employee to product led

and like some of the earliest problems were like yeah it was an amazing magical like god box that you could talk to that defeated the Turing test like out of nowhere. But it couldn't keep you logged in. Do you remember that? Yeah. You have to keep logging in. And then for a while, it was like it's down or it would air out. And now they've gotten to a point where it's extremely reliable.

And yeah, it still goes down every once in a while. There's still rough edges. And there's weird things where like different models have different capabilities. What can generate images what can but clearly all of that product stuff is getting ironed out and they're going to have ads in the free version all the way up to a two thousand dollar a month tier or two hundred dollar a month tier like it's going to be like really really polished differentiated products so i'm excited i mean it's on my home row and nothing has been able to displace it for me yeah and we can't um understate how impactful it's been on this show.

We, on average, have been putting out 10 hours a week of content, something like that, and some of it is more or less scripted and then it's heavily researched and we write some of it out beforehand. And so ChatGPT has effectively replaced the need for a full-time writer, analyst type person. And then AI has worked its way into so many other things, just like transcripts are automated with AI, little captions that we put over videos, even just like some AI generated images and some of the short form stuff. Like it's just like weaseled its way into all these like nooks and crannies. It's not, it's very additive.

Yeah. It's not like replacing anything that we're doing, but it's been, it's been hugely beneficial. Yeah. Another one on the list, obviously, NVIDIA becoming the world's most valuable company. They were, Apple overtook them again afterward, but can't be understated how memorable that will be to me even 20 30 years from now like witnessing that jensen signing that woman's chest that was like he's a rock star now yeah after what 30 years of building that company more yeah more yeah and and and being i would say very benevolent you know positive yeah uh type of

character within the tech world right he's not doesn't have as ridiculous ego he's generally very humble uh the only the only critique i would have of him is he has gone on an interview to say i wouldn't do this again you know if i if i knew how hard it was going to be because we don't want to discourage any builders. But yeah, it really can't be. And it's so unclear right now. We're not in the business of making predictions around market caps or anything like that. But looking back at the dot-com era, comparable companies, evaluations they achieved, it would just be so interesting if this time it's different.

Well, they have over 100% of the profits in AI right now. Yeah. Because all the other companies are losing money and they're the one that's printing money. It's also fascinating to just compare them to their rivals, Intel, which is a complete mess. And Pat Gelsinger was ousted, and they just haven't been able to take advantage of the Chips Act at all. And then AMD, there was this bombshell report by Semi Analysis. Dylan Patel this week basically saying that even though they're hitting the benchmarks on the actual computing power, like price per flop, the software is so broken there's so many bugs that you cannot train models on it and george hotz is like chiming in and uh lisa sue the ceo

is uh kind of scrambling but she's done this before where she's like hopped on the phone with people and be like okay yeah we're gonna fix it and then nothing happens and so uh it's been a really crazy just kind of like a b test to see that NVIDIA just remains like so dominant, even when someone else can beat them on like the core metrics. Yeah. The, that CUDA, you know, polish really matters. Here we go. John's, John's, the coughing is going to begin. The other, the other thing to highlight under here that was seemingly monumental uh was tsmc actually starting to produce chips here in america yeah um i don't think that can be understated just given how important they are in in the in the value chain or supply chain

and uh yeah i'm excited to see where that goes in 2025. Um, but in, with the chaos of this year, um, it seemingly was not necessarily brushed under the rug, but it felt like it got like five hours of attention when maybe it was, you know, considering how dependent our entire way of life is on chips. It seems pretty cool that we were able to make those here. Yeah. Huge. Did you use an Apple vision pro this year? Yeah, that was, that was going to be the next one. Uh, I, uh, I'm going to sound like a Luddite for this, but I did not, you didn't even try it. I didn't even try it. One. Um, I had no interest in going to an Apple store and just like putting on something that hundreds of other people are putting on in a day. It's just a personal preference.

The other thing and the reason that I didn't actually buy it is I try to keep devices away from my head. So I use, you know, I use like wired headphones I have for years. I use my AirPods as like golf balls for practice in my backyard. Sure, sure, sure. They no longer work, obviously. But I didn't try it i haven't been um uh i i totally would it's not that i wouldn't but i just didn't happen to be in a situation um but it is it is fascinating that uh how quickly the swing happened from people going on x in the first couple weeks being like this is the next

computing platform amazing too i've never even facetimed a single friend that was like oh i'll call you from my apple vision pro and we have a bunch of friends that have these a lot of people use them for two weeks and then return that's exactly what i did yeah it was great for two weeks i watched a movie on a plane watched a movie at night but there were just too many bugs too expensive i would yeah you were seeing that like i remember i was at soho beach house like somebody's just sitting there with it on wow and it just looks you know it looks silly but it's also i don't just because something looks silly doesn't mean that it's bad it it was like it i i gen generally think that the apple vision pro is is very, I think it's cool.

I wish it was catching on more. I have a friend who's a super talented iOS developer. I was hoping that this would be a computing platform that he would be able to build on. I did try the Meta Ray-Bans, which I was pretty impressed with. But again, I don't really want to wear, I don't need to, it's not quite, you know, significant enough to want to wear them around all day. But yeah, I think there had been some reporting that Apple was going to discontinue the Vision Pro entirely. I think they, that would be such a, doing that, like there's almost seemingly some value to Apple in terms of maintaining their market cap to just continuing it. Because if they discontinue it, it's such a colossal defeat. And sort of like verifiable proof that you are an iPhone company.

And I think... I think they should keep it going. I think they should do another version with the same screen. Because the screen really is fantastic. Yeah. version with the same screen because the screen really is fantastic yeah it's just yeah and it's it's apple has the leverage to be able to say let's spend 20 billion dollars a year on this you know they legitimately could do that if they can for the next same screen get rid of the external screen switch it to plastic instead of metal so it's half the weight like and half the

price like they have a winning formula for sure. Just in terms of replacing a home movie theater. Because you can just put it on and watch a movie. But even then, the software was really bad. I would wear it in the dark, and my room is really dark, and it would get lost and put up an error saying, like, can't track. And it's like, I don't care if you can track.

I'm just trying to watch a movie right now. It was very, very frustrating. Not fully polished. And it felt like the people who were working on it weren't using it in a real way. track i'm just trying to watch a movie right now it's very interesting yeah uh not not fully polished and it felt like the people who were working on it weren't using it in a real way they were like using it in a demo at work and yeah not actually daily driving it in the way that steve jobs daily drove the iphone and realized like hey like it needs to be a glass screen because i put it in my pocket with my keys and that was like a fundamental insight that that hasn't happened with the apple vision pro yet it's been much like, of course you'd like a giant box

and like take it out and like put it on for 10 minutes and do a demo and then put it back. And like, it's like, no, that's not how people use these devices. Yeah. But yeah, people, people went from being like, Oh, what case should I get to my vision pro to the week, the next week being like, I'm going to return it next week being like i'm gonna return it yeah i'm gonna return it and yeah super bullish for meta because they're clearly going to release like something that's equivalent they just need to figure out the media uh like you should be able to just buy any movie on any meta headset and i don't know that they have a partnership yet but they need like a itunes store yeah because that's the first thing i did was i went in and i bought like avatar all the christopher nolan films all the movies that i would want to see in the

theater i bought all those like 3d movies and stuff and it was amazing all those were great but it was just too heavy and too expensive and too too janky but if if meta comes out with something that's the same resolution at half the price and has that feature just the movies are good i also do think that there's a big opportunity for uh starting to create content in 3d right now and then just waiting waiting waiting and then eventually it'll catch up i was thinking about doing that when when it launched if it was going to be big i was going to do the history of vr and i was going to do an interview series in in spatial video in 3d so you could watch it in the Apple Vision Pro.

And it would be like you were sitting right here and you could look to your right and see me. You could look to your left and see like Palmer or Carmack or whoever I would get to interview about the history of VR. And it would just be like these really chill, unedited interviews. And it would just be like, oh, I'm sitting at a table with these guys

having a conversation, learning a little bit. No edits, nothing. It's very natural. Anytime you're cutting, it gets really jarring. Anytime you're flying around, you don't want to move the camera at all. You just want to put the camera there and then let people look where they want to look. And that's really cool. But there's just no – they don't have a YouTube yet for Apple Vision.

Yeah. One thing that was interesting, they had that app that was native to the device it was you could see dinosaurs yeah and it reminded me of the original like iphone ipod touch apps when i was probably i forget exactly how old i was maybe like 10 years old 11 years old using these things where there was like the app where you could like play guitar or like drink you know you could pour a beer or whatever and it was just like these sort of i i don't remember anything outside of the dinosaur app that kind of broke through the noise um nothing which is really proof that there's no organic developer activity which is kind of a people are lazy when

it comes to vr like You want to just lay down. One of the best VR games I played on the meta quest was a shooter, but the whole conceit is that you're in a kayak and you're a sniper sneaking through waters and assassinating dudes on oil rigs, basically. So you're seated the whole time and you're just using your vr paddles to

go like this and then you reach down pull out your sniper rifle and then shoot and it's really that's a cool concept it's awesome because you don't need to be like running around and it's not very and it's good training for everyday life exactly yeah um he's great and so um it's uh like the figuring out how to break the like people just want like video games i feel like had a breakout moment during like final fantasy 7 where it was like a hundred hour game you could just sit there and it would be like lots of dialogue very chill sit on the couch middle gear solid like these long games and no one's figured out how to do that on vr because they're all so they're all so preoccupied like, well you could run around and you could throw fireballs.

But then it's like, okay, I'm just working out now. I'm like exhausted. Like I don't actually want to like climb a mountain to live like mission impossible too. Yeah. I want to watch it on my couch lazily.

Yeah. And so Palmer was actually talking about this one company that was building a VR headset that was meant to be used laying down. And so it had like a special pillow with like knobs on it. So you could like move your head very easily. So you can move your head, but then you'd lay down. It's like very kind of matrixy and like creepy. Like you could easily get like wire head and spend like eight hours in it. But if you look at like, imagine the brain rot. Oh, it's so brain rotty. But at the same time, it's like, how do people play Madden on Xbox when it comes out?

Or Call of Duty when that one comes out? They play for eight hours straight. Yeah. And so you're not going to displace the PS5 until you can give someone an experience that they can go and chill with for eight hours straight. Yep. And so you have to lean into that design, like elements, essentially. And you're competing with the phone, which is so good at capturing attention, right? Oh, yeah, totally.

And every single... And you can just do it anywhere. competing with the phone which is so good at capturing attention right and every single that's that's that's one of the big issues with vr is vr headsets aren't competing with each other they're competing with every other screen everyday life exactly so the competition is just intense especially when they're this expensive yeah um what else uh polymarket is worth bringing up like prediction markets finally breaking through. Seeing the picture of Shane working in 2020 or 2021, like in his bathroom on like a wicker basket. No one really cared. Very low volumes.

Slogging through the next few years and getting to this year. And I just thought that was cool because we've talked about this on the show before, but it had been this sort of thesis that every, so many smart investors had for so long that prediction markets would eventually go mainstream. And it kind of took this election cycle. It took PolyMarket's product getting to where it is. It took probably like sports betting, like blowing up and like becoming a lot more mainstream to be honest like kind of predated i would say like the the the the venn diagram of people that sports bet and people that like use prediction markets is like probably almost a circle you know um but just a breakout year and um yeah it was it was very i would say vindicating for believers and prediction markets based on how the election unfolded yeah it's been tricky figuring out

like what the next major polymarket segment will be because uh chad byers was in our replies talking about like oh we should do a polymarket for when these guys double the next time or like yeah 64k and my over at poly market was like the bots have entered the chat because yeah if there's money on the line somebody will just make a big bet and then 50k bots and just follow us and it'll just destroy everything and so it's very very tricky to figure out how to how to you know price that yeah the whole insider trading element is interesting too because people were betting on Brian Johnson's erections, like nighttime erection length.

Yeah, I saw that. But then I saw one of his employees was like, I just made XYZ money from insider training on my bosses, which is like, how do you really, like these are financial markets maybe they do need some you know element of regulation or you just need to go into them and treat it more as an entertainment product yeah like this is not a you know perfect security like it's sort of yeah you're almost you're almost trading on like how much insider trading will

there be like what what type of insider trading will there be like what what type of insider trading will there be like what are all the incentives that yeah or you could say like oh it's it's all priced in yeah everything's priced in um yeah it's like it's like if if we did the 64k thing would there be another army that's that's buying no shares and then actively reporting every bot that follows us to not hit that. Yeah, exactly. The levels. Bot war, essentially. It's all priced in.

Yeah. Anyway. What's next? SpaceX? Another one, SpaceX. Fantastic. Yeah. Landing with the chopsticks.

Landing with the chopsticks. So crazy. Crazy moment. Very, very well timed for Elon's brand. If that had not worked, there was so much attention on him, given the election cycle and everything. But it was just a very, it should be a really uniting moment, right? Like, hey, America is home to a company that can achieve a feat that is so unprecedented and feels like it's out of a science fiction movie.

The other thing that I thought was funny here was you had, for years, the meme has been, oh, you're an investor in spacex but you're an investor in this spv that invested in this spv that invested in like this fund that bought secondary at like 2x the pref value or whatever but i would imagine that most of those people are actually making money still for sure you know even if they overpaid or there was like you know a lot of fees bolted onto these different vehicles um but it's it's been a good year to be a spacex investor last it was i guess getting priced at 350 billion most recently um and uh yeah i'm interested to see i'm sure there's there must be an ipo on the on the horizon don't have any inside info there but you know I think the narrative has been they go and start up the Texas Stock Exchange

and can potentially bring a lot of... I don't know. Elon ran the AB test with Tesla and it was so brutal being a public company with Tesla. Why take it public if he can keep it private as long as possible? That's a good point. He'll probably just keep it private as long as possible.

Especially as long as there continues to be exceptional demand, then nobody can complain, oh, we need liquidity. It's like, well, you... But really, like, every single founder should buy a Starlink. Like, you should never be without solid internet. You should always travel with the Rome one. It's 50 bucks a month. Because I was in Yosemite and the internet was terrible and it was a big hassle. We recorded a video and John was like, okay, I'm uploading my side of it. Like 20 minutes of video took like four hours. It was miserable. I actually don't know what the upload speeds are on SpaceX and Starlink. I think it's probably similar as download, but either way,

like anything was better than what we had. It was brutal. Another one, I thought the F adobe breakup was pretty monumental yeah um last gambit her yeah cluster stand that was the most child what is it the child who is not embraced by the village will burn it will burn it down to feel its warmth yep that's um uh yeah i could have been so i felt this one i felt this one personally because almost immediately after it got announced i used figma quite a bit and the figma product uh they just rolled out these like absolutely horrific billing practices figma was always a very multiplayer product you could make something share it with somebody easily they could edit it comment whatever

and they immediately rolled out these like billing approaches that were one extremely confusing and basically meant that you if you wanted to share a file with anybody you were just going to get started you would start to get billed yeah for uh for them and it's so such a toxic way to do billing because if i'm creating something and i want to share it with you and you're not like logged in right at that moment, it's just very, very annoying. But they the 20 billion dollar acquisition was off. They realized they probably needed to two or three X, you know, ARR like immediately if they wanted to maintain this sort of narrative that they were a 10 to 20 billion dollar company. And so I get it from their side. But it was just extremely unfortunate.

I think a lot of those billing changes would have probably been rolled out with Adobe Cloud, but it was just like such, you know, Dylan, I think, I don't know. I think it's just the nature of becoming a big company. Like, I don't know. It's just the nature of becoming a big company. Like I don't know that there's any way to avoid that. Yeah. I think it's just like, you got to charge for, you got to capture all the value. Yep. I don't know. I think, uh,

I think it would have been, it would have been jankier with Adobe, which is probably the worst thing. Yep. But I don't know. But yeah. And, and again, there should be like an entire documentary on Ms. Khan's era because I don't know exactly, I don't think it's clear exactly why it happened. I wrote out a script once, I never published it.

But the fact that two companies like that can't just make love is tragic. Let's go to Defense Tech. Defense Tech. Yeah. I mean, this was the year of American dynamism. Credit, credit. As much as AI was a dominant part of the narrative,

American dynamism felt like it had an equal amount of attention from a meme standpoint. Coogan's Law. Coogan's law. Coogan's law and credit to Andreessen for, for not necessarily being the first and biggest investors in that category, but they certainly coined it and they get a lot of credit for the coinage. That's part of Coogan's law. And so,

yeah, I think you look at Teal and he co-founded Palantir. They did the SpaceX series a and and then incubated Anduril. And it's like, they didn't come up with the phrase though. Mogged. Genius. Mogged by coinage.

Genius. Yeah. I think, I think they're important. Yeah. So, so the three, the three major players,

Anduril, Palantir, SpaceX, all start, all, all started, you know, long before American dynamism was, um was a part of the vernacular.

I thought this was funny. This has been talked about before. A lot of the valuations over the last year have been pretty insane. I have portfolio companies that are getting multi-hundred million dollar valuations with no revenue yet. It's a little frothy um and can you imagine what how how much uh defense tech companies would have loved zero percent interest rates like zerp meets hard tech meets american dynamism would have been

iconic i hope we get that opportunity in the future yeah uh it'll be great um but yeah this is a good example like i think um trey I remember Trey went on a podcast and was basically talking about how the Founders Fund thesis is this sort of monopoly-driven power law. There's one company that captures a majority of the returns, and that's presumably in defense, hard tech defense that presumably is Anduril. Uh, I look at a lot of the defense tech happen, investing happening now. And, um, I think I've talked about this before, but the government actually wants vendor diversity. So there is some argument to say, Hey, we actually need more Andurils. There should be maybe one or two more.

Maybe they, maybe there's a power law dynamic there where if Anduril is like a hundred billion dollar company, we have another $10 billion dollar company so i think we do need this um andrew i think benefits dramatically from a lot of this because there's so many smart teams that are building cool tech that just won't figure out distribution and actually getting contracts andrew will acquire them and they've already proven that they can be in a m&a led product development org right where they're like we have the best distribution engine into the government. Now we're going to just bolt on other products to get in there. And then there's also just like the adjacencies that Anduril doesn't want to touch, like Flock Safety.

Yeah. Great company doing fantastically well, kind of like Anduril for police forces. Yeah, or Deterrence, the company that I helped get off the ground. Deterrence is doing something that doing high-volume energetics manufacturing is not a sexy program of record type opportunity. Yeah, yeah. It's not even necessarily. It's more deeper in the supply chain.

Yeah, it's deeper in the supply chain it's a critical you know critical part of our defense uh overall but um you know ultimately not not there there are still categories that that andrew is not going to go after you know immediately maybe they will long term who knows um but uh plenty of opportunities and and lots of great companies like Aon, Out Control Systems, not to just talk my own book. But going on, Meme of the Year, we covered this a little bit in our awards, Trend of the Year, Meme of the Year.

They're basically the same thing. I think PG had this one in the bag guys still got it um and uh founder mode founder mode coinage uh he took something that was kind of obvious this idea yeah but no one had ever put a phrase around it yep what is the difference between a founder-led company and a non-founder-led company yeah founder mode and it's a very vague post. It faced a lot of criticism when it was dropped because it didn't define it very precisely. And that's because it's different for every founder. Exactly.

It's very vague, but it's still useful. It's still a useful term. Yeah. One of my favorite moments, listeners will correct me if I'm wrong here, but Frank Slootman, I don't believe was the actual founder of Snowflake, was he?

He joined later as CEO, but very much amp it up. Like he is like brings that founder mode approach. And I just thought this was a crazy moment when he announced he was retiring from Snowflake. They lost 20% of their market cap in a day. They've been previously been a $76 billion company. And a month, roughly a month after the announcement,

they were a $50 billion company. So that's founder mode in effect. And yeah, it was cool to see people like Ryan Peterson, I think was it earlier this year, or maybe more than a year ago, but Ryan Peterson going back into Flexport. Now he's doing founder mode meets going direct. And that strategy is just like killer.

That's what Alex Karp is doing. That's what Palmer is doing, Trey are doing in many ways. So the winning formula is founder mode and going direct. And that's why if you try to make fun of a founder for posting a lot on x uh just check yourself because some of the best ever uh are doing that um and um yeah other otherwise like the the big one and um we were going to cover this later in the show was, um, just the foundation model wars. Like if this was this like constant, uh, game of, of tug and tug of war where literally felt like every other week, it was, Oh, and throw everybody loves anthropic next week,

everybody's back and, and, and what that kind of exposed. And you have to go back and look, there was at one point this year that everybody said the foundation all the value is going to accrue to the foundation models like ignore everything else apps don't matter and then people realize that people were just switching back and forth from models all the time and then everybody started to think okay well now the app layer is like what matters and you should focus on the app layer and then then it went back to oh like don't launch an app because open ai is just going to launch that as like part of chat gbt and

the next you know keynote search and so yeah search is a good example right um and all this just got so spicy with with um you know the perplexity ceo just like constantly posting against sam open ai uh and um you know el Elon feuding the whole time in this very public legal battle. And that's still ongoing. But yeah, we'll probably cover the O3 Pro launch later in the show. Another one, acquisition of the year for me had to be uh bridge oh yeah so i think that was we you know we talked about this but a non-token one billion dollar outcome in crypto yeah you know you had you had coinbase ipo obviously significant but uh to show that there was a second you know

sort of regulated crypto business that could be acquired by a non-crypto company, a tier one company, and Bridge showing that Bridge being... It doesn't even feel like they were that early to stable coins, but they were on time and very right. And they built the right product, a developer tooling that Stripe

clearly saw a lot of value in. So anyways, the last thing I would say is we kind of forgot about this, but in the Q1 of this year, TikTok was getting banned. Yeah. There is a law that's passed that should ban TikTok.

Yeah. It's happening unless Trump overrides it. And I thought this Trey post from february 21 was great he said concerned about losing our ai edge to china ban tiktok concerned about ai safety ban tiktok concerned about child mental health ban tiktok concerned about trade imbalance with china ban tiktok it's really not complicated so it's very unfortunate that that wave kind of died down i had bought bantalk.com at the time, which, um, you were going to do with that, but it's funny.

Well, the idea at one point was either to make like sort of a, like actually just build like a, like an organization to more. Um, and there had already been like internal stuff in Washington and pushing this stuff forward. But the funnier idea to use it, that domain, which I did sell it, somebody ended up paying, you know, buying it for something close to what I got it for. But if you did like a vampire attack on Tik TOK,

which was like sign in with your Tik TOK account, pull all of your followers and your following, and then we'll like rebuild the app. Sure. Somebody could probably pull that off. Yeah. Maybe Nikita, something like that. Um, and it's just interesting that the after that movement died down true social went public oh yeah and sword which you know maybe that was foreshadowing it certainly was foreshadowing the election it's crazy how much money flowed into that company it's just supporting Trump

it's wild yeah yeah I think it's time to probably get into some dms in the timeline but um yeah overall felt like i think this is a year that we're going to remember very long from now huge comeback year uh comeback year crypto's back defense is ripping ai is you know completely new paradigm and if you were a gp at a major uh venture fund you join the government that's true we'll be covering over the next year let's do a signal should we jump into some dm sure do you have them over there yeah yeah okay so first um before we get into that uh we got some reply guys to cover. Let's do it.

And this is the first ever double reply guy of the week award, and it's going to none other than Chris Amidon and Will – God, I'm going to butcher his name. Stromber? Stromber? Stromber? Stromber. Is there an R in there?

Will Stromber and Chris Amidon. Yep. And these two brothers, you know, not only exemplify brother behavior, but have managed to find the perfect balance between being highly active in the comments section and highly active building their business. They're building a maritime defense startup. Chris is the founder and Will recently joined to run Ops. And these guys are absolute dogs.

I will say that. Very patient, very focused. I can tell they don't have post notifications on, so it's not always immediate. But they're getting into the comments sections that matter. Very patient, very focused. I can tell they don't have post notifications on, so it's not always immediate, but they're getting into the comment sections that matter. They're engaging with other reply guys, other brothers. They are exemplifying the spirit of what it means to be a reply guy,

and they outwork the competition and patience. We've seen potential reply guy of the week come in, get super active for five days, not win, and taper off. And Chris and Will have just been at it, honestly, for months now. That's great. I'd love to see it. And so that's exactly what it takes to build a great company. And they have just proven time and time again that they, uh, they are very deserving of this award. So gentlemen, um, congratulations, uh, probably not the first time you guys win this award. I could see either of you winning brother

of the week at some point, potentially brother of the year, who knows we'll be doing the show for a long time. And, uh, you know, don't stop. We've talked about this before. You can win this award twice. That's great. What else do we have? Okay, funny meme. This wasn't in the DMs, but Andrea posted this and said,

Stimmies over demons. She said that psychedelic psychosis is out and cocaine-induced opulence is in. And we don't support hard drugs, but we do support stimulants. And we do endorse and live a loud, opulent lifestyle. And so loud opulence was, loud opulence I would say is going to be the trend of next year. I think so. So if going direct or founder mode was 2024,

loud opulence is the year of 2025. America's back. And thank you, Andrea, future brother of the week. I have no doubt. And thank you for meeting us. Yeah. We really appreciate it. Let's go to Sar.

He says, started listening to Tech Bros pod. In a sea of podcast, it has really nailed the vibes. A great blend of Twitch slash Discord style talk show. Very online discourse and podcast format. Thanks for giving us the shout out. We really appreciate this. I think Sar had to be one of the first hundred accounts I ever followed on X. He's not as active as he was maybe six years

ago, but when he posts, it breaks through. Thank you, Sar. Yeah. And I like that. Talk show is definitely what we go for. I don't know what he means by Discord style talk show. I haven't actually seen anything that's like Discord driven. I've seen a lot of Twitch content. And a lot of Twitch streamers will pull up tweets. There's a couple, like Ludwig is a YouTuber who is also a Twitch streamer or live streamer. And a lot of times what he'll do is he'll make a 10 minute video about something and he'll just have 10 tabs or 20

tabs open and just flip through them and it's a great format because it drives like okay he's on to the next thing he's on to the next thing and he can just do a one take or he does like a couple takes and it turns out to be a really polished product that feels like a scripted YouTube video but it's really natural maybe that's how we're going to do we're going to have to do

a remote episode this week so maybe that's how we're gonna do we're gonna have to do remote episode this week yep uh so maybe that's how we do it yeah just open you should just share the screen and then just go through and just be like this is the story we want to tell i really think that uh doing some more organization around our content would be great and being like yeah okay we're breaking down the chat gpt like open ai launch and here are 20 tweets that we're going to go through one at a time and take you through like a narrative on these, um, as opposed to kind of what we're doing right now, which is a little more random,

but we're always learning. We're always getting better. So thank you. Yeah. It's funny. We're, we're firm believers against building in public. Do not show your revenue online.

Let people clock it by the car you drive and the watch you wear. Yes. Uh, but content you're kind of forced to build in public because you're just sort of iterating live and everything's very public. So. Okay. Megan Niveold says what wine pairing, if you're in a warmer climate outside of Palo Alto, screaming Eagle seems a bit heavy for 75 in Miami. Thanks. What would you recommend? I mean, you can't go wrong with a refreshing glass of Dom pairing on cheers to that. Megan, enjoy some Dom. You can't go wrong with a refreshing glass of Dom Perignon. Cheers to that, Megan.

Enjoy some Dom. You can never go wrong. No, it's great. Almost every time I see Brother Megan's post, she's just somewhere like Aspen, Miami, just really indulging, the loud opulence lifestyle. So I think this is one of those questions where she already knows the answer and she's just kind of testing us. Sure.

Uh, so Megan, why don't you DM us, tell us what you settled on, what you settled on. Cause we know you already know the right answer. Let's go to base Baron. He's been on the show before we say,

if my second mistress is going to, my second mistress is going to be so mad at me. If I don't get at tech bros, pod reply guy of the week. Oh, that's a good point. Well, you missed it this

week but maybe next week keep up the replies based baron we love you uh yeah reply guy of the week is it it's it's a it's a duration meets quality yep and we value we value persistence quite a bit so for sure there's other people you know brody's been very consistent right trip trip has been um phenomenal like you know um and so it doesn't go unnoticed yep but um it does take some some sort of concerted long-term effort turner novak turner novak right it's great yeah let's go to julia black she says fascinating logic from tech bros pod i described the guys of el segundo in a way they would gladly describe themselves and that made it a hit piece. Humbled that John Coogan and Geordie Hayes would devote so much time to discussing such an untimely, uninteresting story.

Well. Got us dead to rights. She got us. We both enjoyed it and criticized it. Hypocrites. Yeah, yeah, yeah. She got us. I spoke with Julia after this.

You know, she's great. You know, we squ Julia after this. She's great. We squashed the beef. She was one of the nominees for Journalist of the Year. Technology Brothers Journalist of the Year Award nominee. She's close. Not by accident. Yeah. She's up there with some more merch.

She's going to be shifting to covering Doge starting basically now through next year. I'm actually very excited to see some of the basically now through next year. And so I'm actually very excited to see some of the articles that she starts writing. So if you're exposed to Doge and you have stuff that's sort of able to be shared, it's not going to be damaging

to the movement, go talk to Julia. Let's go to my tech CEO, Jason. He says, LA-based founder I backed has totally lost it and he replies every angel's worst nightmare tech bros pod would appreciate advice on this topic was this the jeff woo video uh i don't think that was a jeff woo video but but uh i think it was a founder uh that he was you know founder in his portfolio, moved to LA, set up shop, started taking acting

classes. And that is just a huge red flag. You'd almost rather a founder, you know, it's, it's, it's basically as bad as a founder saying that they're going to go to Peru and have a psychedelic experience. So acting classes, terrible sign. Very rough. Ayahuasca, terrible sign. I don't know. You might want to just write that one off jason good luck let's go to signal he says the eternal technological pursuit has always been about the commoditization of every single thing that includes the human body mind and soul i mean signal after the 03 launch just went on a generational posting run this is why he was one of the breakout posters of the year.

The guy went on an absolute terror. So this was in response to that. Uh, I think that, um, yeah, it was just like very, you know,

well put. Uh, I think that, um, commoditization of the human mind. I don't know if we have the human soul yet, but it feels like, uh, you know, talking with chat GBT doesn't feel far off from talking to a, a smart friend,

right? It's not perfect. It's not right on everything can make mistakes, but, um, so do, so do all of us. Yeah. Pretty wild here. Let go to Antonio Garcia Martinez. He says, I find the AGI alarmism amusing. We've been a post-scarcity society for decades. 95% of the work we do is invented, i.e. needless add-ons to food and shelter. We will always invent new forms of work, making new products and services to occupy ourselves and compete for status.

It's a good take. I mean, AI solved chess two decades decades ago and chess.com is a massive business and also chess masters like people like watching human chess players even though they are not the best compared to ai and i and i do think that there's uh something that will be you know around for a very long time around what can a human do in the ring or what or in this podcast or in this debate format even if an ai could win yeah yeah another another take here that was interesting sean frank was posting that it's like 60 60 million people still subscribe to cable so even though we have streaming says ai is going to replace everyone

brother 60 million people pay for cable we got time yeah it's going to take a long time to roll all this stuff out it takes time to implement AI into your organization there's a lot of risks a lot of legal risks like where does it sit I think the one thing that's obvious if you look at the cell phone

you know to compare it to something like cell phone adoption which was dramatic everybody suddenly having a smart you know, to compare it to something like cell phone adoption, which was dramatic, everybody suddenly having a smart, you know, computer and like a personal computer in their pocket. It took maybe 15 years to get to the point where everybody that was participating in modern society had this personal computer in their pocket. I don't think it's going to take that long at all now that everybody has that personal computer which is a distribution device for these ai products yep i basically think that i don't even know it might people are kind of already interacting with these generative ai products whether they know it or not and so um i think the adoption is is is already basically happened. And, but the impact is going to take decades.

Yep. Let's go to Mark Andreessen. He says, overheard in Silicon Valley, my biggest lesson from the last year is just, it's all going to happen. All my childhood dreams of space colonies and brain interfaces and AI and robots, everything is just going to happen.

And Yaxin says, not only is it going to happen, you'll get to play with it in your basement, 3D printers soldering the future rocks. I like that. It's very optimistic. Yeah, so I think we're able to say that now, that it's, quote unquote, just going to happen. I think it's a great quote, but we're able to say that because many it's quote unquote just going to happen like i think it's a great quote

but we're able to say that because many many years ago really smart hard-working visionary people decided to make it happen yep like you know elon musk's uh uh keynote the future should look like the future that was the culmination of two decades of work to say autonomous cars, humanoid robots, all these things. So our dreams are going to happen, but it's because a small number of people decided to make them happen. And so if you have a dream, don't just expect it to happen.

It's going to take 20 years. It's going to take 20 years and be persistent. But the best day to start is today. It's going to take 20 years. It's going to take 20 years and be persistent. But the best day to start is today. Let's go to Signal again. He says, it feels like instead of all of us stepping into the future together, a few of us are watching our world change

on a daily basis, while the remaining masses will one day have a startling realization that the world is radically different. This is true. I mean, there's a lot of people who are like, I talk to people who are outside of tech and they straight up say, oh yeah, that AI stuff. I gotta, I gotta look into that. It's like, wow. It's been like two years since ChatGPT dropped. And you're like, I'm, you could be benefiting from the co-pilot that is,

you know, chat GPT or whatever your preferred model is. It's, um, you know, my wife will be like, Oh, I have these four ingredients.

What should I make? It's like instantly like cool recipe that, that just like takes no real effort to find. It just comes out of, it's just purely magic. So, um, magic is here. it's just not evenly distributed this is a great take by pavel he says non-technical folks saying software engineers are cooked it's you guys who are cooked you're gonna have x software engineers competing

with everything you're doing now and they're gonna be ai AI turbocharged 7,000 likes. And then he follows it up and says, engineers were simply doing coding because it's, it was the highest leverage use of mental power. When that shifts, it's not going to all of, to all of a sudden shift the hierarchy. Couldn't agree more. Yeah. What's funny is I think he would also agree i got to i got to hang out uh with pavel for a bit at um holiday party um and uh he's got his takes are as good in person as they are online um but uh i would say that he's not necessarily speaking about the low agency

software engineer big tech company he's speaking about the low agency software engineer at big tech company. He's speaking about the high agency, you know, individual contributor engineer who's able to, you know, imagine something and then create it with code. And yeah, it does feel like right now, you know, somebody, this guy, is it Nat Eliason? You know that guy guy so he was talking and he was like he made some post about how think about how incredible the tools are that are public publicly available for 30 a month and you have to imagine that there's thousands of tools out there that have been

created by software engineers that are an order of magnitude more impactful and powerful than what's publicly available. And nobody knows about them yet. So if you're getting like a robo call and it sounds like a regular salesperson, you know, and then it's actually, you know, like some like, you know, model that person, the highest and best use of that tool might not be to sell it as a SAS product. It might just be to use it for their business. Yeah, I mean, for a long time, coding was just a proxy for intelligence. And there was a big difference between somebody who doesn't code

because they operate at a higher level abstraction, they hire software engineers, and someone who's just like, oh, I could never even fathom figuring out how to code. That's a very, very different type of person. It's a very different level. So Vittorio says, serious question, what should a CS student or any knowledge worker

for that matter do at this point? Even if the model is $2,000 a month, it's still cheaper than a graduate employee. What's the plan now? And so this is everyone freaking out about O3 and AI displacing software engineering. And yeah, I mean, we just discussed it. It's like, you know, operate at a higher level of abstraction.

Everyone kind of moves up the stack. This has been the story of programming for decades, going from machine code and writing like binary to compilers to, you know, higher level abstraction languages to, you know, autof abstraction languages to, um, you know, autofill and autocomplete on things to now being able to just, um, talk to Devin and figure out what you want to do. Well, yeah. So all of the AI coding startups are still hiring engineers. And so there's clearly plenty of value in what you do. The key is if I was, you know, if I was 21,

22 graduating with a computer science degree, I would be figuring key is if I was, you know, if I was 21, 22, graduating with a computer science degree, I would be figuring out how do I use AI better than any of my peers? And how do I do 20 times the work that a typical college grad would do? Because you now have no excuse not to write better code, more code,

faster, you know, do it all faster. And so if you can come in, if you can do the job today using these AI tools that a previous new grad engineer would have been able, you know, if you can do the job of 20 new grad engineers for you plus $2,000 a month, then you're going to be in a great spot. Let's go to Atlas Creatine Cycle.

He says, really worried about the amount of ex-software engineers turned DJs we're going to get out of this. Tasteless beep boop music too cheap to meter so funny Atlas uh Atlas has been on a tear as well he he said he's been bored uh staying you know with family for the holidays and he's just posting through it love it uh and uh yeah we love to it. There was already a big pipeline of people that were post-economic becoming DJs like Justin Kahn, um, plenty of other examples. Uh, and so, yeah, I think we're going to see a lot of, a lot of software engineers turn DJs, but, um, maybe it's beautiful things.

Maybe it's more parties, more human interaction, less time online, more excuse to go and be together. But yeah, I'd stick to the Metropolitan Opera and the L.A. Philharmonic instead. Good call. You know, learn to play piano. I think Atlas would as well. Yeah. And the violin. Learn a concerto.

Lulu says, building things will become easier and easier. Getting people to care will become harder and harder. And that makes sense. That was an entire thesis for OpenAI was that in a very crowded internet, a very noisy internet, they broke through and created a new consumer internet brand.

Yep. And to this day, you can stop someone on the street and say, where do you go to solve AI problems? And they'll just tell you chatgpt.com. Even though it's a ridiculous name with all these these acronyms it's not like the sexy name people still you know identify it and it became you know just baked into the zeitgeist and it's interesting there's like this entirely new behavior where i'm not necessarily wanting to search the

internet for something but i'm wanting to understand something, which people historically would use Google to say, I want to understand intermittent fasting. And now you can just go, I still, I oftentimes know I'm going to get a better answer by going to chat gpt.com. So I'll just go there,

type that what I'm looking for into the search bar. And so just the value of a new consumer behavior that I don't think people use, like I'm going to GPT it, but they kind of like, you kind of hear that. Yeah. I mean, this is a very different like pathway of like exploring things and asking follow-up questions and creating tables and ontologies. And it's just so nice to be able to say, Hey, break down this new concept in like different levels of abstraction, create a taxonomy for me.

Like what is the topology of this idea? Like what are the super categories, the subcategories? And you could get some of that on Wikipedia sometimes, but you can just do that for anything. It's been really, really fun to learn about all the new things. And I find myself just like dropping like weird arcane knowledge and getting up to speed, like way faster on things. Like there was somebody who was the first person like months ago to ever say

the word Vacheron Constantin to me. And, and within like three months I'm like clocking like, Oh, that's a cardio Santos because I've like literally gone from zero to 60 on that. And just like through YouTube and also through chat GPT but like understanding like okay break down all the watch brands break down all the

expensive ones all the oldest ones like just all of that it's just so much faster to learn and become an expert in some random category which is like crack yeah and you can be like what do you AP enthusiasts think about Vacheron exactly and then they you can get that summarized really quickly in a way that it's way harder on google yeah um signal john promoted i love ads we got a promoted post from our friend austin reef uh ceo of morning brew he says my favorite growth hack find business influencers and turn them into your biggest customers and advocate advocates it's awesome to see mr sharma genuinely loving the talent we source for him at ocean's talent

so austin's one of the owners of ocean's talent they help you find great offshore employees and he just screenshots hubspot here which was clearly like a you know know, MPS style survey. And, um, Nick says, uh, uh, XYZ employees, a beast he's taken on so much and has kept everything running so smooth. I'm never not going to not work with him in my life. He not only goes above and beyond what he's asked to do, but comes up with new ways to do those things better. Many times also just coming to me with these ideas in the first place before being asked. So, you need to hire remote employees, check out Ocean's Talent. Shoot Austin a DM. He's a pretty responsive guy.

And yeah, cool, cool example of the product in action. That's great. Let's go to O3. Mike Knoop, who runs ArcPrize, says O3 is really special, and everyone will need to update their intuition about what AI can slash cannot do. While these are still early days, the system shows a genuine increase in intelligence

canaried by Arc AGI, semi-private scores, and there's a fantastic chart here by Riley Goodside that shows the progress of the GPT models on ARC score, which are these like color-based block puzzles that any human can do, no problem, but AI has struggled with for a long time. O3 is a reasoning model. The high version costs $2,000 per solve. So it really thinks and it really puts tons and tons of cycles

and it's also been fine-tuned on these. And so some people were saying, oh, well, if it's fine-tuned, it's not that good. Rear-end's post was being like, they trained on the training set. Yeah, it is ridiculous. But at the same time,

there is a difference between a model being able to one-shot an eval and actually needing the training set. A human does not even need a training set to do ARK. You don't even need to study for it. And so there is something there, although Rune is basically correct. And that this is remarkable. And Dylan says, if you look at this graph and don't feel any existential concern for humanity,

you might need to revisit your priors. Yeah. So I was texting, I think it was Friday night, with Brother Madfess, Jared Madfess, who coined Too Big to Fail, which is his thesis that uh entrepreneur you know highly technical entrepreneurs that can that are in the thousand pound uh club and

weightlifting just are too big to fail uh so he says jared says basically tech used to be this great equalizing force but now this is a thing and so he's showing the uh the what we just discussed he says paying more doesn't lead to quote-unquote more tokens it now leads to quote-unquote higher quality tokens with costs being exponential but value generated per task also potentially being the same so if you're a billionaire with a device running ten thousand ten thousand times the inference cost now gives you a god compared to the average guy. And so right now $1,000 per task gets you basic human capabilities for most things and world-class capabilities for a lot more.

Within three years, that will be sub 10 cents per task. And what does the $1,000 per task model look like then? And he says, inference time compute means money energy and labor all converge on each other so uh and then uh he says today feels monumental to me go long feudalism or fascism al ai smells the end of democracy um so anyways bold bold take from uh brother brother madfess um but uh i think that there's a lot there interesting let's go to aaron So anyways, bold take from brother Matt Fez. But I think that there's a lot there. Interesting.

Let's go to Aaron Levy. He says, here's a way to think about the market size of AI. U.S. labor costs are $11 trillion. If a company pays 2% of headcount expense to get a 10% to 50% productivity gain with AI, that's $200 billion. The U.S. enterprise software market today is around $ that's $200 billion. The US enterprise software market today is around 150 to 2 billion. So AI doubles that just in the base case. Wow. Yeah, this is why this is why it's a good time to be bullish. The end of software was, you know, a meme in many ways yeah a lot of the best ai products or software meets gen ai yeah so don't be afraid to go built and here's a related post from adam d'angelo founder of quora and co-founder of facebook i believe uh wild that the o3 results are public and yet the market

still isn't pricing in agi and then Elon says, the markets claim to look ahead, but mostly look behind. AI will ultimately render money meaningless. Bold. By the way, I look forward to your deposition on quote unquote open AI. Wow. That's a heavy reply from the biggest,

most popular reply guy. I want them to be friends. Let's focus. Yeah. Let's focus on the task. But yeah, I mean, it's real. Yeah, O3, I think, is still underpriced. It's a big deal.

Yeah. Very, very big deal. Let's go to Aaron Slodov. He says, why do you, an adult technologist, believe that a nice programming benchmark score means AGI? Fascinating stuff. He's always been very anti the idea that AGI is here

because he works in the physical world and the AI has not been super useful to him in many ways and there's still a lot of tasks that he can't apply AI to. But I feel like this video, when a couple videos started going viral, that these $4,000 robots, like factory-grade robots, have started to seemingly get some amount of traction.

So yeah, I'm excited to see what the next year looks like for physical intelligence, which is a Lockheed Grooms company who's building foundation models for the physical world, presumably robots. And also just like Devin-esque technology for CAM programming and CAD in design and just what happens if you have a factory

that has some software and you're able to just speed up the development cycle time of every piece of software that you develop for every, every stream of tokens that you write. Um, hopefully they move a lot faster. Hopefully the delivery times are faster, which would be great. John, I got to jump in here with an exciting promoted posts from none other than Ben Heilack. Uh, he has two exciting pieces of news that I want to bring here. One is that his company, Don Analytics, is hiring again for a backend engineer. And this was posted, I guess this was posted before O3 launched.

So maybe double check to make sure they're still hiring that person. But he also says they will pay five thousand dollars to anyone who refers someone that we hire so brothers if you're hearing this go refer back an engineer to go work at dawn very cool company and ben is actually the designer that is responsible for the jaguar rebrand so that's true get to work with an iconic designer one of the best to ever do it led the jaguar rebrand was covered in you know the mainstream media for it and the opportunity to work with somebody that that can do that level of um that has that level of taste would just be uh you know a true opportunity of a lifetime yeah

exactly let's go to jeff bezos's wedding it was reported that he was spending 600 million dollars on his wedding uh and it was he was buying at a sushi restaurant and i was talking somebody about this and they were like 600 million dollars of sushi like that's all i and bezos debunked it he said furthermore this whole thing is completely false none of this is happening the old adage don't believe everything you read is even more true today than it ever has been fantastic yeah i think everybody saw the headline and started running the napkin math and was like what would you even if you were had having 600 guests yeah a million dollars a million dollars it's like impossible pretty

difficult to do i mean if anybody could figure it out we did yeah um but it'd be like a car one of one you know one of one thoroughbred you, you know, Kentucky Derby winning racehorse. Like it would really be the gift basket at the end. Exactly. Rack ran it up. Yeah. Yeah. Very hard to consume.

$600 million of food in a weekend. They just don't make it. You can, you can stay at a Mon for 50 grand a week. Yeah. You know, per room. Yeah. So you got two guests in a room and that's like basically all inclusive so yeah yeah impossible uh the math's not mathing and it's funny um uh you know jeff bezos owner of the washington post uh it was it wouldn't have

even have been effective for him to publish an article from the washington post saying hey by the way my wedding didn't cost this much he had to to go. He had to go direct. Yeah. Let's go to Oliver. Classic Lulu win. Yeah. Let's go to Oliver. He says, U.S. Department of M&A for Greenland, Panama Canal, and other strategic acquisitions. I love it. I'm so bullish on the Greenland and Panama Canal and moon stuff. I really hope at least we get one of them. It'd be so good. I know that's the best part of this is if this just only leads to one more American source of natural resources, economic dominance. I would love it.

Vacation destination. Let's turn Greenland into the world's largest Disneyland. Yeah. I can't wait. I mean, Ken Howery is now the ambassador to Denmark. And, you know, I think he reads Pirate Wires. Maybe he'll work on Greenland. By the way, brother Pat Blumenthal made a post that went quite viral and talking about how we were appointed as ambassadors to Monaco.

And all my least offline friends took it at completely. I posted the screenshot on Instagram, took it totally at face value, sending me congrats, like this is amazing, like I always believed in you. I got a bunch of those. So funny. So anyways, Pat. Thank you for creating the fake news cycle. You created an amazing fake for really you've created a lot of amazing fake news cycle i got a bunch of messages to go through time that i won't be moving to monaco

anytime unfortunately maybe the next cycle once jd's in that would be a great slot for us we could move the studio the families out to monaco i think we could do it um i have a good story i don't know if i ever told you this i my uh my, uh, my parents, uh, took us on like a 30 day road trip through Europe when I was 12. And, uh, maybe I was 10, somewhere in that range. And we drove through Monaco. We couldn't afford to spend the night there. It was too expensive, but we spent like three hours driving through it. And my mom said that at one point I stood in the Harbor and was like and i just said out loud these are my people these are my people and she's like what are you fucking talking about

our entire family's income wouldn't run one of these ships for an hour that's hilarious annual income uh so anyways our we have monaco in our future definitely this summer and hopefully the ambassadorship at some point let's do one more promoted post and we'll close with Darkash

this was a toss up but I'm going to go for a post from Hermes I don't know who it was but one of our listeners sent us

a saddle from Hermes Hermes you know started making you know very high-end leather saddles and um somebody sent this to us uh if you sent it thank you they said for the tech journalists listening to the show so if you're a tech journalist we know you love horses if you're shopping for a tech journalist today is christmas eve you literally have 12 hours to buy a gift. If you don't have something, go to your local Hermes store, get a saddle.

Pretty much, you know, gets it. Well, this was like a shot across the bow because we had publicly mentioned a YSL saddle. Yep. And Hermes has a much richer history in saddle making. Yeah. And so maybe this was Sarah Hess. I think so. And so this is a bit of mea culpa for us.

Yeah. I think we got it wrong. I think Hermes is the obvious choice here. Yep. And YSL is great, but if you're trying to convince a mainstream media journalist to stop writing a hit piece on you, you really should spare no expense.

You're going to expense it anyway. Go straight to the top. Just go straight to the top, the Hermes saddle. Let's close with Dwarkesh Patel. He says, we're way more patient in training human employees than AI employees. We will spend weeks onboarding a human employee and giving slow, detailed feedback,

but we won't spend just a couple of hours playing around with the prompt that might enable the LLM to do the exact same job, but more reliably and quickly than any human. I wonder if this partly explains why AI's economic impact has been relatively minor so far. So Dworkesh was spending hours tinkering with Gemini 2, Google's LLM, to effectively take his rough transcripts and turn them into really highly polished transcripts. And he finally needed to figure out that he needed to give it a sample of the input and the output and really write out. And so his prompt is big. It's really, really big, but he nailed it and it

just works super, super well. So anyone with a podcast out there, I highly recommend adopting Dorkeshi's script. It's such a perfect example though, right? Because when you hire a new employee, you're committing to 40 hours of training. Mentorship, weekly meetings. Sometimes I've had times where it's like, hey, let's meet every single day for your first two weeks just to fully ramp you.

And then constantly learning by osmosis and the LLM just doesn't get that. And i would say most people have spent less than an hour prompt engineering yep or like totally trying to dial any of this stuff in so this christmas eve i don't know if this will go live today might go live tomorrow might go live the next day uh but uh spend some time prompt engineering give give the gift of dialing in your models to yourself. Highly recommend it. And I think that's a great place to end. Merry Christmas, everyone.

Merry Christmas. Thanks for listening. Bye.