Alex (3:00)

Yeah. First of all, we have two bits of IPO news here from the United States in the same week, Jason, which is fantastic and very exciting. The biggest one is that Discord, the consumer gaming chat service that everyone likes to use, has filed confidentially to go public. Long awaited, long expected. I think actually on Monday, Jason, we went through a bunch of the polymarket odds for who's going to go public, and we both thought that Discord was pretty high up there. Well, here we are. So it's a filing that's confidential. So we don't have all the numbers yet. We don't have a per share price. We don't know much that's new, Jason. Other than that this is coming, though, I do want to provide a little context for folks. If you recall, a couple of years back, Microsoft tried to buy discord for between 10 and $12 billion. The company said no, went out and raised capital at a $15 billion valuation. And now a couple of years later, according to secondary trading, the company's worth something a little bit more like 7 to 8 billion dollars. So it's the 2021 era. Valuations were a good time for Discord. Still a very valuable business. Still great to see it go out. Probably a little disappointing for the last money in, Jason, but I think a lot of folks will still make a lot of money.

Chris (4:02)

Yeah. And we need to get more public companies out there. The public doesn't get to participate very early on because of the state. Private, longer movement. Discord's been around for a Long time. I'm trying to remember when I first heard about the startup. You can tell us when they launched their product.

Alex (4:18)

Always a really interesting May 13, 2015.

Chris (4:21)

2015. Okay, so it's only 10. I thought it was more on the 15, but what an amazing story. This is 200 million global monthly actives, I believe, which is, you know, that's a pretty big footprint folks. If there's 3 or 4 billion people online, you know, that means 20%, whatever, 10, 10%, maybe 5, 10% are using this product every month and they make money, I believe through subscriptions. Right. That is this Nitro product. And then they may have some advertising on the margins, but I'm not sure exactly how that works because it started as a GAM platform but then people took the platform and started using it for everything, right?

Alex (5:00)

Absolutely. I'm in like 15 different discords for different things. It's great fun. Everything from like fantasy authors to individual video games. In early access, it's super popular. Nitro is the consumer subscription side of this, Jason. I think individual servers or kind of instances of Discord can charge for access and The Discord takes 10% of that. They did roll out a number of new advertising formats in 2025 after they swapped CEOs, I think clearly gearing up to juice revenue ahead of an ipo. And also they have a new product called Orbs, which is essentially a kind of like a virtual currency that you can earn by doing certain advertiser driven tasks that they're expanding. So they're really trying, I think have a full court press on the advertising side to try to probably one augment growth and then de risk them from just having a single main revenue line.

Chris (5:43)

I'm looking at my discord right now. We have a this week in startups one we're actually hiring a community manager here because we have so many disparate communities. We have angel investors at the syndicate, we have founders, we have Founder University. So we have the founders we've invested in in one Slack. We have Founder University. Now Global. We're using for Saudi WhatsApp and we're using the same thing in Japan, a WhatsApp instance. Because in those cultures that's predominant here.

Alex (6:07)

That's what everyone uses.

Chris (6:09)

Yeah. And then here, you know, I'm in my Knicks fan TV Discord. So when they have a live show and they go over what happened in the Knicks game, it's kind of like a call in radio program. They actually take the calls for Knicks fan TV through Discord. Why they used to have a phone number, etc. But it always was low quality. According to my friend cp the fanchies over there and they just push everybody through Discord. Leo Laporte is very successfully running something called Club Twit, which I'm a member of. Everybody should join. It's like eight, ten bucks a month. But he's got, from what I understand, a decent audience there and his business, you know, has moved I think into subscriptions as a big denominator for him. And that's just great because it becomes a foundation and then you become less reliant client on advertising and audience size and then you can stay true to your mission. In Leo's case, I think true to the mission means just that. OG a Silicon Valley Bay area look at tech and just people who just love the products and services, passionate about it. Whereas you know, you, it's hard to compete now with so many podcasts out there. You really have to carve a niche for yourself. And I think people have now made Discord substack, Beehive, Patreon. You know, people pick which one of those they want to make their central hub. But it's, it's a juggernaut of a program up and a platform and I think it has great upside. Also it seems to have used the tried and true Reddit and siphoning off Reddit users to sort of create a new experience. Reddit didn't have chat, they kind of have a kludgy chat on it and doesn't seem to be app based. So that has also become a big win. I saw our friend Kevin Rose is now doing communities in his new Dig reboot. So he's, he's back in the game and people love a good community, right?

Alex (7:50)

Yeah. The interesting thing though in the case of Discord is I'm not sure how quickly it's user base is growing when we think about Reddit right now. Jason, to go back to that example, they're seeing a boom right now in usage, actives and so forth. In the case of Discord, they've been saying this 200 million plus monthly active user number for at least a year. I think I saw a data point from like 2024 that was the same number. So my question is how good are they at driving increased monetization from a, let's say roughly static user base? And if they're not very good at that, then I'm not sure what the growth story here is. So this is an IPO that I don't think we have all of our thinking done ahead of time. I think we're gonna have to read this document when it comes out and learn quite a lot about the business and figure out how good its 2025 monetization efforts were and what kind of hope that gives us for growth down the road because Discord is doing. You know, I think the last number I heard was like $600 million revenue a couple years back. So it's at scale, but, you know, as we can see from box, if you're not growing your private sales, multiple drops to the low single digits pretty quick, quickly, and that's brutal.

Chris (8:49)

Yeah, Reddit doesn't report their mouse, but they do report their daily active users, which are 444 million, which is crazy. And then people estimate Reddit is at 850 million to a billion. So community you can make money on. It has to be a platform, it has to be user driven. It's very hard to build like these community websites without the audience really driving it with these subreddits, these sub communities. If you want to be our community manager and you're passionate about startups, looking for a bar raiser for people who don't know the term bar raiser that comes from Amazon, this was one of Bezos's big innovations. Hire somebody who would raise the bar of everybody on the team. So community Launch co, if you want to apply for that job, we need somebody who's been doing community for five or 10 years and can tell us what we should do and how to run communities. We, we have ideas, but this is always the tension. Alex, when you are running a business, you could hire maybe a young person or an enthusiastic person and kind of train them up on the job. But if you've, you're not an expert on it, it's going to be slow going. Right? Whereas, you know, we're experts on podcasting. Launching a new podcast, not very difficult for us. But I need somebody who can teach me about community if you're that person. Community outlaunch co. One of the core challenges facing every single early stage company is getting your first 100 customers. And I say this all the time here on this Week in startups, to get there, sometimes you have to do creative things, even things that don't scale. So what is that actually mean? And what are some tried and true techniques that have worked for founders in the past? And which one of those can you tap for inspiration? To find out, you could join an accelerator, or you could just check out this free resource online at HubSpot. Called how to get your first 100 customers. This isn't just some little pamphlet with generic ideas. No, it's a full framework. And that framework is filled with practical tips and real world case studies from amazing founders and experts, including my pal Sam Parr and Sean Puri. You know the hosts of my first million, which is a great podcast. I listen to every single episode. They're just awesome. So download the guide for free@clickhubspot.com twist1 or just find a link in the description of this show. And thanks to our friends at HubSpot for sponsoring today's episode.

Alex (11:05)

That's a lesson I first learned at a startup, actually, when I was at Mattermark before things got a little wonky at the business side. I had some headcount and my then boss, Daniel Morrill was like, hey, you know what you should do? You should go hire the most senior person you can find. And I was like, at the time, you know, 25, 26, I kind of viewed that as insulting in a way because, you know, you're younger, your ego is more fragile. But there's a lot to be said for hiring someone with a boatload of experience who can come also fill in your gaps.

Chris (11:31)

I love that it is an ego thing. You correctly pointed out why people don't do this. And what you have to learn in an organization is if you're the person who brings in an all star who makes the team better. If you were the person on the basketball team who recruited Kevin, if you're Draymond and you recruited Kevin Durant to join the warriors and you win a chip or two with him, that's like a very powerful position to have in the organization. Which is probably why the warriors organization coveted Draymond, even despite the fact that he's effervescent and unique in terms of his, you know, performances on the court.

Alex (12:06)

I mean, he's polite.

Chris (12:07)

Well, no, he's a basketball genius. His iq, and I'm friends with him, but his IQ is like way off the charts. But he also has that passion like Dennis Rodman or any number of people where, you know, the double edged sword of the passion can get you attacked. Technical foul. But you want to be that person in the organization. You bring in an all star, what the boss man or lady is going to think is, okay, this person cares about the business as much as I do and then you will always have a place in the organization because you can point to it and say, yeah, I've recruited that person anyway, there's an important founder Lesson in there. Congrats to all the people over at Discord. And another great example of, you know, not selling opens up a lot of opportunities. The thing that sometimes happens when you do an IPO like this is you force the hand of acquirers. So now you've actually put a price on the company. When Slack went out, I'm sure they had many offers to be purchased before they went for ipo, but people don't remember Slack was a publicly traded company. And then eventually that puts a price on it. The price discovery happens. The management team has to own the results every quarter. The discipline comes in, the processes come in. You can't wing it. Right? And Reddit can't wing it anymore. Steve Hoffman is one of the few founders who made that jump to being a public CEO and really guiding that company. Then acquirers can get really interested and say, hey, well, we know the price of this company. Let's start building a position in it. Let's start talking to management and then you can get taken out like we're seeing with Netflix and Warner Brothers. Okay, let's keep going really quickly.

Alex (13:37)

Strava is also filed confidentially to go public. Jason not as big of a name in this community as Discord is, but it's worth noting that the company was last valued at 2.2 billion in a round that was led by Sequoia. Strava, if you don't know, is an application for exercise people. I have used it to track my runs, for example. It's a super great application. They make their money off of consumers. And remember when Reddit, before it went public, had that big fight and beef with its communities because they shut down third party application access to Reddit data and everyone was pretty mad at them for a bit.

Chris (14:09)

Yeah, you know, people were creating a lot of value with their data, but that wasn't getting back to the mothership. And then the mothership said, hey, you know, we gotta go buy whatever it was. There was like blue something, which was the top Reddit app. And they were like, let's just buy it and we'll have a Reddit app that is actually up to speed. X or formally Twitter had a similar thing. When you're growing and you can't get to everything. Yeah, you provide API access when you become public. Sadly for developers and the community, you kind of got to capture that value. Which is why Zuckerberg shut down his APIs. X shut down their APIs or charged a lot of money for them. So at least if people were going to build something Competitive, that was on your roadmap. You could get paid for it. That is the, the natural tension there.

Alex (15:02)

I bring that up because in the case of Strava, they did make some API changes, some third party access changes, and that made people really, really mad. But if you want to ensure that you're capturing all that value because you're going to go public, you do it. So I'm not shocked to see Reddit do that and then List and then Strava go through a similar kind of internal fiasco with its user base and now List. The company was valued at 1 and a half billion in 2020, just to put a, put some historical context here. And it grew more than 50% last year and the information reports it's profitable. So my question is how the hell did Sequoia manage to get in for such a cheap price? But when this one does go out, it should be, you know, around 500 million revenue profits. So it should be a win for all of its prior backers. And I'm excited to see more mid cap public companies. Jason, I just want more.

Chris (15:45)

It's 80 bucks a year I believe is the subscription price and people are extremely loyal to it. And this is another important founder lesson. When people have their data in your product, you know it's very hard for them to unsubscribe. That's the best moat, the best lock in. Because I use as an example a product called Slopes. Slopes has my last four years or five years of skiing data for me paying 50 bucks or 40 bucks, whatever I'm paying for it, I don't even know. And if it was 150 bucks, I'd still pay it. I mean I might think about it at that level. But I like to compare my stats from this year to the previous four years. And that's one of the most important things you can do in a subscription business is have streaks, have friendships and community. All of those things equal the moat. And so while anybody in the Vibe coding era could pop out lovable or another tool and make a Strava competitor very quickly, the data lock in is what makes these very special. 90% of their revenue from what I've read, comes from all of these subscriptions. They could layer on ads and other things. We will see consolidation in this space as well. We have an incredible investment in a company called fitbod that just focuses on cross fitness. There's obviously whoop oura ring and then you have the function, health, the superpower.com and those blood testing. All of this eight sleep. They're all going to come together and at some point there'll be a roll up of five or six of these products and then you'll get a really good picture of your health.

Alex (17:19)

Where should that stack be built? Because I could see eight Sleep buying companies, I could see Strava buying companies, I could see the scanning companies buying other companies. But if you were to pick some place to build that vertical integration, where would you start?

Chris (17:33)

Well, Apple Health obviously really cares about this. Tim Cook cares about it. And they've tried to maintain an app store where, like the Google, the Apple Health data is not a really good app. It's just kind of like the raw data and then other people build on top of it. But people, I think, got frustrated, I certainly did with my data inside of Apple, because I was like, just tell me my Sleep score. And it's like, oh, you want to get something out of sleep? Purchase one of these seven different, you know, products. And I'm like, no, I just, I bought an Apple watch already. Just give me my damn sleep score. And they wouldn't do that. So I got Whoop. And I love Whoop. And now I'm addicted to Whoop. I can't imagine. And I think I paid 200 bucks a year for Whoop. It's like 20 or 30 bucks a month. It might be 250. So I think between and Whoop just launched blood testing. So they're going to put those two together. Everything's going to move to singularity. It's just a matter of who has the chutzpah to do it. Strava going public means they have a currency to buy. Let's say if they wanted to buy fitbod, if they couldn't buy Aura Function Health, those things are worth billions. But they could go down to the next tier of apps and services that are worth hundreds of millions or even a billion and start to do that. So we'll see who does it. And now that M and A is back on the menu and IPOs are back on the menu, this is what we want in our economy. We want to have a vibrant IPO market and we want a vibrant M and A market. And here we go. We finally have it. The best is yet to come. The next three years are going to be gangbusters. Gangbusters. This will be the year of the IPO and M&A. I would say last year, 2025 was the return of M and A. M and A will accelerate in 2026 and 7.

Alex (19:18)

Whoop is considering an IPO in 26 or 27, just to put some more context there. And Oura Ring I think is going to be like $2 billion in revenue soon, but they're not looking at an IPO soon. So that's kind of the state of to play.

Chris (19:28)

If you're an independent content creator or you're building any kind of community oriented business like I am just making great videos or awesome websites, it's not enough. You need to spend real time with your community and you need to do that with tools that allow you to see what's going on and communicate smoothly and crisply with your community members. That's why I'm so happy to partner with Circle, the complete community platform for creators and brands. With Circle, you're going to get an easy way to build branded websites that help you with email marketing strategies that's super clutch and AI agents that can help you design, manage data and even do the branding. We use Circle here at Launch to keep up with our founder university community. We've done 14 founder universities around the world and they've helped us grow that from this tiny little program out of our office in Austin to this bonafide international movement. And they can do the same for your startup. Twist listeners can get $1,000 off Circle's professional plan by going to Circle. So Twist, that's C I R, C L E dot so slash Twist.

Alex (20:36)

Now you're saying the IPO market here is nuts. You should see what's going on in Hong Kong. There's been so many interesting AI IPOs. You know, we talked earlier about how companies go public so late here in the United States. Well, two of China's AI Tigers went out this week. Zippu AI or Zeta AI as I think of them, and Minimax both went public and they both did quite well. Zippu AI or Zita AI was up 36% from its IPO price and Minimax doubled on its first day, all the way up to 345 Hong Kong dollars per share. They're both worth about 6, 7 billion dollars. And if you know the GLM models that do quite well, that's Z AI and minimax of course, put out the M 2.1 preview, which is doing quite well on LM Arena. These are companies, Jason, that we talked about a little bit on Monday, but they're small and I just love seeing them go out early. So if any of them do does well, well, you could make a lot.

Chris (21:26)

Of money and if you are willing to take the risk, just like SPACs or other markets like this vibrant Hong Kong market. There'll be a vibrant Middle east market, emerging Japan, Australia. They all will allow an IPO with 10 to 100 million in revenue. It's kind of their speciality. And then that means you get to play late stage vc. Late stage VC is not a guarantee. It's not like buying Netflix, folks. It's not like buying Nvidia. You're going to be taking a lot of risk. And I think these companies, if I'm reading correctly here, zipu Z H I P U A I, which is Z AI, according to producer Claude, had 27 million in revenue during the first half of 2025. So they're on a $60 million run rate or so. Minimax reported 53 million for the first nine months. So very similar, probably at 70 million or so. Maybe they'll hit 80 million in 2025, I think is the estimate. So these are tiny when compared to the IPOs we see in America, where I think a billion dollars is what people are looking for. Again, increases risk, increases opportunity. If one of these goes 10x year over year, 10x again year over year, like Claude did, or anthropic or OpenAI or lovable, you know, or micro one, a lot of these companies can have 10x year over. These AI companies can have 10x revenue year over year. You get to participate in that.

Alex (22:55)

I mean, Anthropic went from basically a run $1 billion run rate to a $9 billion run rate last year. That's 9x at scale. The thing, though, that I'm just a little worried about is these Chinese AI companies are punching way above their revenue weight in terms of model quality. But because they pursue this open approach to just kind of letting people take it, it does seem to be a little bit tougher on the revenue side of things, Jason. And I wonder if they're going to be able to keep funding their research at the pace that they are, because they're spending a multiple of their revenue today on R and D. And you can do that for a while, but eventually it doesn't quite math out.

Chris (23:27)

Hemant, Tunisia, from General Catalyst. He's the CEO of General Catalyst, which we talked about last week, raised 9 billion. He was also on my panel earlier in the week, along with Bob from McKinsey, and the three of us were chatting things up and I got to talk to him backstage and during the panel and after that. Anthropic investment for them has been tremendous. And he talked about, you know, we used to talk about triple, triple Triple, double, double double for SAS companies, you know, 10x 10x 10x or 10x 10x 2x 3x, this is a whole different ball game. This is why even though there's a bubble in areas, there's froth in the AI market, there's also just massive adoption of this technology, which we will see in the GDP prints around the world with people who are adopting it. And we'll see it in the headcount versus earnings versus top line revenue or headcount. As we talked about for three years now, static team size, you're just going to be able to do more with less.

Alex (24:24)

Now I want to talk about anthropic a little bit more because why not? So Jason, you mentioned them and their fundraise and how their venture capitalist partners have done so well. Well, the company is looking to raise $10 billion more at a $350 billion valuation. Now that number 350 is actually the reported price that both Microsoft and Nvidia late last year pledged to invest up to $15 billion as a pair. So we've kind of seen this number before. But it is a rough doubling of its last private round, if you will, in which it raised $13 billion last September at a $183 billion valuation. We seem to have scaled past what VCs can afford to put up themselves. So GIC, the Singaporean sovereign wealth fund and CO2 are set to lead this round. It's going to close the next couple of weeks. It's not done yet. The numbers could still change, but this is the reporting on the ground. And I just want to answer everyone's questions about this. Is this froth? No, because I went back and I pulled all the data. Jason, it's kind of crazy to me how this company's effective multiple on its run rate hasn't gotten out of hand. It's not crazy whatsoever. So if you think about the company's growth through time, it had a run rate multiple of about 31x March of last year, it was about 37x September of last year and then it was about 20x at the end of the year. And it's going to be about 35x again if this new round closes as expected. And I think frankly, for any company Growing this fast, paid 35x current run rate is just not crazy.

Chris (25:51)

And this is what happens between being a private company with a lot of promise and then a public company judged on performance. You go from a voting mechanism, hey, who do we think is going to win in the private markets? To and a weighing, a scale that just weighs the quality of these earnings and the quality of the company, the management, et cetera. And so you'll just see eventually this will be priced not on price to sales, but eventually earnings. That's going to take another two years because the earnings of these firms is going to be super challenged for a couple years. Just like Uber, Lyft, Doordash were losing money on every ride while they built to scale. And then at some point they flip the J curve which shows the investment. You're losing money, losing money, losing money investing, investing, investing, whether it's in factories or software or team or building a brand and a marketplace. And then all of a sudden you stop spending out of control. You stop losing money on every ride or delivery or delivered package for Amazon. And then you slowly come out of the J curve and then you become money printing. And a company like Tesla could be sitting on 30, 40 billion. A company like Met Apple could be sitting on hundreds of billions of dollars. And so this is, you know, anticipated and a healthy part of the process. And you'll see these companies come out of the JJ curve in 2029.

Alex (27:15)

How much money do you think General Catalyst is going to net on its anthropic investments?

Chris (27:19)

I don't know. But if they own between 5 and 10% of the company and this becomes a trillion dollar company to take down 50 billion is incredibly rare.

Alex (27:30)

Almost never.

Chris (27:31)

I mean Google, Apple, Facebook, Tesla, you would have to go into those ranges and typically people would be selling into that kind of strength. SpaceX would be another one. So I think SpaceX, if you look at Peter Thiel and Founders Fund, they might own 10% of it. And that is possible when you're the Series A Series B investor, which is why those rounds are quite competitive for elite management teams. But it's certainly not 20% anymore. It's between 5 and 10 is what the venture firms wind up having in ownership.

Alex (28:03)

I think you just really outlined the difference between today's venture capital market and the one in days past, because I just pulled it up Alphabet. Google when it went public was worth 23 billion total. And now today we have SpaceX, which is worth actually I don't know offhand.

Chris (28:18)

Jason, but 800 billion I think is what people are saying, whispering 600, 800 billion, something like that.

Alex (28:24)

30, 40x. As much private today as Google was at IPO. So you can make a lot more money if you can stay in these companies longer.

Chris (28:31)

The outcomes are much bigger because people stay private longer and because the Opportunity Space is not just niche software for enterprises, not just niche consumer products or even mass consumer products. These are industry changing and leading companies, eventually they wind up having two or three businesses, business lines that could be contributing to the bottom line. In the case of Amazon, you have aws, you have their traditional selling stuff, E commerce, and then you have their booming ad business. So you have three ways to win with that stock, right? Tesla, you have powerwalls, solar cars, self driving and Optimus. That's like four different ways to win if you put energy together.

Alex (29:13)

I just did an interview with Zipline, the drone delivery company for Twist, coming out to your podcast feed soon. And they started off, people kind of forget this doing medical deliveries in Africa. And now we think of them as the, oh, they're the Chipotle delivery company. Oh, they're the Walmart delivery company in Texas. But they have several large business lines as well. And I think it's just such a great way to have a durable company.

Chris (29:33)

And they went from doing fixed wing aircrafts being slingshotted into the air, they have these now, you know, backyard units. Forget about what they're attaching, like the drive through windows, they're attaching to, you know, the Starbucks and Chipotles of the world, but they have these sort of backyard, bizarre looking, netted out arrays I guess is what I would call them with their drones coming in, not fixed wing, but quadcopters or whatever. And man, they look crazy. And when you see them buzzing around, it's like, okay, that's the future.

Alex (30:06)

This is from their Chipotle launch video. It shows how they drop the package down, Jason, into the little catchment that you're describing. Yeah, this is, this is a company that has, you know, multiple approaches to the market and they're all crushing it. And I just can't tell you how excited I am.

Chris (30:18)

That is a. And for people who are watching, it's like a little box on a tether. The box comes down into the Chipotle, they throw your bag of burritos in it. It then in that little box gets pulled up the tether so the quadcopter doesn't have to land. This is critically important because you don't want to get injured with a quadcopter, right? There's rotors and they're big and they're noisy. But if they can stay up 100ft or 200ft and drop a tether down, put their little box, the box empties or gets filled without, you know, they get emptied without a human inter you know, interfacing with it. You just stand back and watch it happen. Hopefully humans don't do stupid things like grab the tether and yank it, but somebody's gone.

Alex (31:02)

Yes, yes. If there's a socket, people will fork it.

Chris (31:04)

You know, that's like somebody jumping in front of a delivery bike. If you're dumb enough to do that, you know, it's your call. But yeah, this is going to be an incredible new future. And I was talking to Sachs just the other day. You know, he's here in Austin. And in Austin we have kind of a unique experience because of the topography of Texas. In other words, there's a lot of land. Outside the city center of Austin are many Amazon factories or warehouses rather. So when you order something, it comes faster than I was when I was in LA or the Bay Area. They literally will tell you you order something at 10pm at night, they're like, do you want this from 4am to 8am or 8am to 11am and you're like, well, nobody's going to be at the gate until 9am So I don't care. But sure, leave it there. Between 4 and 8am you can get stuff same day as the default in the Bay Area, same day is the default. Logs are an essential part of just about any tech startup. You need to keep your eyes on how your product is being used and you definitely need to understand what happened when things go wrong. But logs are notoriously messy. It's especially hard to gather the insights you need when your logs, errors and performance data all live in different tools. But now there's a solution. Sentry. Sentry's logs are trace, connected and structured, allowing you to follow everything clearly and understand the context. Even if you're a non technical founder. Whether you're debugging your front end, back end, your mobile app, whatever it is, Sentry will give you the context you and your team need to get the problem fixed. And to get on with your day, you got other things you got to focus on. That's why more than 4.5 million developers are already using Sentry, including amazing high profile teams like Disney and Anthropic. Learn more by going to Sentry IO Twist and use the code TWIST to get $240 in century credits. That's S E N T R Y IO slash twist.

Alex (33:05)

Next up, China is mad about the Meta Manus deal. Now if everyone recalls, I think it was on Monday we talked about Meta buying Manus. They are the formerly Chinese, now Singaporean AI company. They've Built a very popular agent. In fact, they were probably the first viral AI agent service that we saw out in the market. Jason. And the deal was for, I think, $2.5 billion, including $500 million for employee retention. So, 2 billion to the company, a half billion to the staff. Everyone's very happy. Hunky dory Meta wins. No. China's very mad. They're looking into the deal. Now, from everything that I've read and understood, I don't think China actually has a mechanism to stop this transaction from happening. But the thinking is they're making a stink about it, essentially to signal to all internal investors inside of China and founders that no more of this, you're not going to be able to go move your HQ to Singapore and then sell to an American investor or an American company. So this might be kind of a one and done situation. A Manus only exception, which I think is too bad, because there's a lot of talent inside of China that American companies would love to purchase via a Singaporean domicile. But in this case, I think we've seen the end of it. So shout out to Meta for pulling this off and I guess the door is closed. I'm sad about this, but I'm curious if I'm being too pessimistic.

Chris (34:18)

Newsom. The relationship with China has broken down, obviously, from, hey, we can win together. You guys make things, you know, we design them. It's a great, you know, you got factories, you're trying to get people from farmland and no running water and living on, you know, a dollar a day. And hey, we can maybe create a. You can create a middle class and we can get iPhones every 6 months or 12 months that are incredible. You know, now we're in the standoff. TikTok being spun out, Meta buying Manus. This is all part of a theme around can the two empires coexist and can they do business together? I actually am hopeful. I think there's so much opportunity for these two great nations to work together that we may see some type of deal from our chief deal maker that results in a little more fluidity in investment, because China needs investment and maybe just a little less saber rattling. We'll see. But I am actually hopeful for it. This is a very dangerous thing for China to do because China has been trying to get their entrepreneurial chops back. Going after Jack Ma was relieved of duty. Re educated. Whatever happened into Jack Ma, Exactly. With Ant and the financial arm of Alibaba. Alibaba. Thank you. And now what we'll probably see is great entrepreneurs there who have the ability to move around the globe, are going to just move, then start something. So instead of redomiciling and moving the company, which, hey, listen, if you're in America and you want to move from California to the great state of Texas or Florida, as we've seen Larry Sergey and others, Peter Thiel, David Sachs, you have that ability here in the United States. What they're going to have to do is do it ahead of time. And Singapore is a great landing pad. The uae, Saudi, Hong Kong used to be the landing pad, but now that's obviously controlled by mainland. So I think what we'll probably see is Chinese entrepreneurs leaving the country. Not good for Xi. So if you take out winning a big prize through an exit or an ipo, that will mute entrepreneurship and they need to actually get entrepreneurship going again. So it's a short term mistake for them to try to block these kind of deals. And it's a big win for Meta and for Benchmark. Remember, Benchmark got really hammered for being anti American, you know, not patriotic for investing in the company. I think it was, it was kind of patriotic for us to engage other nations. Somehow that became not patriotic, which I understand, especially if it's like military applications or AI.

Alex (37:10)

Do you not buy into the idea then that the China versus AI battle is essentially to the death? Because I read a lot of people who are critical.

Chris (37:18)

I think it's important that we be the leading economy in the world as a democracy. That's important. But I think open source is going to win the day anyway. Or take 60%, 70% of the market and then maybe 60, 70% of the revenue will go to closed source like we saw with Android and iOS. That's the likely scenario. And we saw that this week at ces and we're going to get to our CES segment in just a moment. The most interesting thing in 2026 at CES in my mind was Nvidia announcing that they had an open source end to end AI like FSD from Tesla that was available to any automaker and they have the hardware stack available to any automaker. So I think we'll have an Android which would be the Nvidia stack versus the proprietary stacks which would be Waymo and Tesla. And so we're going. All that does is accelerate adoption. The fact that Android came out after iOS just drove massive adoption of smartphones and the mobile ecosystem. So that's what's upon us.

Alex (38:30)

Before we jump over to all of our. Jason, this is The Nvidia announcement that I recall seeing this drive Hyperion ecosystem to accelerate the road to full autonomy. Just to make sure I understand this, this is essentially Nvidia putting together the software and hardware necessary to turn any car into a self driving car car. But they're not going to be building themselves. This is just to enable OEMs and ODMs to do it on their own. So Stellantis and everyone else.

Chris (38:51)

Yeah, so Mercedes showed off a car that's, you know, in the 40,000 range with the complete sensor array and that's level two, which is supervised, which will very quickly get to level four. Very quickly being, I don't know, a year to three years essentially, they're probably halfway to where Tesla is now. In other words, they got to 98%, 99%. The last 1%, incredibly difficult. The last 1% is probably the equivalent of the first 98. So you know, they got, they got a ways to go. But what this will do is it'll give every single car manufacturer the ability to buy sensor arrays, chips and then do world building inside of their tool. So it's the end to end AI open sourced. It's the world building data, that tool. So you can build worlds and synthetic data and they'll have some basic amount of world data, whether you're doing robotics or self driving, both operating in the human world and then you can buy the sensor race. What is Nvidia's goal here? They just want to sell more onboard computers. So they're looking at. Yeah, they just want to sell. It's like selling H100. So I think in an arms race they're like, we've got ammunition here. Who needs guns and ammunition? You figure out what war you want to fight. Are you fighting a war to sell more Mercedes Benz or are you selling to Lyft, Uber and Doordash to enable their fleets to deliver stuff? Doesn't matter to them who's buying it, it matters that they're selling it. And how many can they sell? If you want to sell a lot of chips, you know, Android being open source and Samsung and HTC and Sony and countless others. Google being able to make pixels that just results in a lot of chips getting sold and a lot of batteries and a lot of camera sets. So this is going to accelerate the road to autonomy. The road to autonomy is going to accelerate and there'll be many more players.

Alex (40:59)

Yes, thank God. I love competition here because as I've said 10,000 times on the show, I hate driving. So really excited about Having this on all my cars in the future.

Chris (41:07)

I have to say, FSD is so good now. I just upgraded to 14 that I have to keep reminding myself, my spouse, my friends, please keep your eyes on the road. It's so good that I think people are now starting to trust it so implicitly that they are assuming that, you know, because it's better than a human. They can take their eyes off the road. They can sleep, they can watch a movie, they can text. You might be able to do a quick text. You might be able to change your podcast. You know, flip back and forth between this week in startups, this week in AI and all in. You can flip between those three. I think that's allowed. But not other ones. You can just. You have enough time to flip between my three podcasts, but no other podcasts. I think that's enough. That's enough for now. But don't go to sleep, because if it hits a construction zone or one of the edge cases, you could get in a bad situation. And that's why I'm very concerned about the pressure they're putting onto Waymo, Tesla, Uber, Noro, we ride Pony. They're putting so much pressure on these companies, Zoox, that they have to get this stuff deployed immediately. No shame in the stuff. No shame in the safety driver. No shame in a contained area. Please stop putting. And it's all these stock promoters. There are all these stock promoters who have built a position in Uber or Tesla or Waymo or whatever, and they're just doing 20 videos a day. I watch these stock promoters. It's a really weird phenomenon. People have placed bets on the stocks, then they've started YouTube channels, and every day they'll release two or three videos about. You won't believe that Tesla solved FSD. The stock's going to the moon. You won't believe that Uber has doing this. And they're just in this social media war and they're making money two ways. They're making money off of advertising and their revenue share. They're also making money through their bets on the companies. It's a very weird moment we're in. So I really hate this phenomenon. It's like GameStop, but in the real world, which makes it dangerous. Like, hey, this is the world we live in. People can place bets.

Alex (43:36)

I'm less worried than you, though, about the safety driver game, Jason, because every time you talk about fsd, you get slightly more rapturous. And every time I look at Waymo safety data, it still looks rock solid.

Chris (43:45)

But you don't own a Tesla, right? Or do you?

Alex (43:47)

No, I have a. I have a Hyundai Kona ev.

Chris (43:49)

I mean, it's time to buy them. You should buy The Model Y Hardware 4 for your next car. You can. I mean, the leases on these things are three or four or five hundred a month. It's basically free. I mean, they're getting close to free.

Alex (44:03)

$10 a day. It's like the same cost we spend on coffee.

Chris (44:06)

Well, and then if you can drive or if you can be driven by it. And you were taking ubers now for 10, 20 bucks a day for your car, it's essentially what you would have paid 30 or 40 bucks for ride sharing. It's pretty compelling. So you got to get on the FSD trend.

Alex (44:25)

Well, we do love to look to the future here at Twist. And we love to look to the future also at funny events like ces, something that I've been to many times, Jason. So we're going to bring up Oliver from this Week in AI. He's going to throw some big wins from CES for you and I to have some fun with. So, Oliver, come say hi.

Chris (44:39)

Oliver, how are you, sir? I'm good.

Jason (44:40)

Thanks for having me. Did you enjoy your time at CES? I saw you made front cover of a TechCrunch article.

Chris (44:46)

I did.

Alex (44:47)

You did, yeah. You do?

Chris (44:48)

Show me. What is this article? I didn't know that. What was the article? I didn't see this article. Was it because of my talk?

Jason (44:53)

I think it was related to the talk. It mentioned you, but it wasn't.

Chris (44:57)

Well, whatever it is, don't worry about it. We don't. Let's not make it about me. Let's get right to what the audience wants, which is what people thought the top 10 products were. We had producer Oliver go through and ask producer Claude what the top 10 consensus product was across all these top 10 lists from Engadget, the blog I started with Peter Rojas and Brian Alvey and Sean Gold and Ryan Block to Vox, the Verge, which was a bunch of refugees from Engadget. And here's what they. They said were the. So this is like a meta list. Yeah, we created.

Jason (45:28)

Yep, exactly. With the help of Claude. Obviously, there's a lot of reporting at ces, but there's also a lot of kind of what I would call CES slop that I'm less concerned with. Maybe like some smart shades or ultrasonic knives that don't concern us here at this Week in AI as much as maybe some other people.

Chris (45:45)

The weird stuff you're saying the weird stuff.

Jason (45:48)

There was a lot of really cool stuff as well. The first one, I thought that was mind blowing was one of the first things that was dropped at CES that I saw on my feed, which was Boston Dynamics new robot humanoid. And this was really interesting because we've all heard and seen videos of Boston Dynamics over the years. They were kind of the first, first company that was doing humanoids. They were doing flips. And it was a lot of cool stuff that kind of captured our imagination, you know, in the 20s, but there.

Chris (46:14)

Was no practical application of those. So let's play here and see what this one does. And I'm assuming the reason this is important is this one has an LLM attached to it and will eventually be available to consumers or businesses.

Jason (46:29)

It seems like they're targeting industrial automation. A lot of their kind of, you know, early clips and kind of advertising was geared towards industrial. But something that really kind of stood out in these demos. As you can see, it has a super wide range of motion, just walks forward and backwards, arms spin all around, which we haven't seen from any other humanoids up to this point. It also can handle, I think, around a hundred pounds of load, which is pretty heavy. This is kind of my first 10 products from CES.

Chris (47:00)

Okay. And everybody remembers there was a massive M and A movement at Google in 2013 or so, and they actually bought Boston Dynamics as well as Mecca Robotics and bought Dolly, which did computer vision, industrial perception, that did robotic arms and computer vision. You probably remember this really well, Alex, because you were a TechCrunch or Crunchbase at the time. They went on a massive buying spree. They were too early. LLMs didn't exist, which meant these robots couldn't interpret the real world. Even though they were getting there, they just weren't there. So once again, too early. They wound up selling all these companies, I believe, out of Google when they went on their cost cutting. So probably a long term mistake. Yeah.

Alex (47:51)

The reason why I'm excited about this one is that Hyundai owns Boston Dynamics Day. Well, 80% of it in SoftBank owns 20%. But Jason, what is Hyundai good at that mass manufacturing? They're good at getting things right at scale in a real world environment. So if this is what they're showing off on stage, I presume it's probably 80% as good as we think it is. But that's not too far away from actually being functional in a factory or a warehouse. So what I see here is Optimus and also the Boston Dynamics bot really going Head to head, which is great because it means the steel will sharpen steel and figure.

Chris (48:21)

Yeah, I mean, we're going to have a dozen of these. Fascinating that the car companies are particularly good at building these. I think it's two reasons. They, they know how to manufacture stuff, right, with chips in IT and batteries. And they have production lines and they also use robotics to build cars. So that's kind of interesting. The people who are applying robotics in their factories are going to be building the robots to go into other people's factories and other use cases. All right, really interesting. Anything else you want to add there, Oliver? Let's get to the next one.

Jason (48:51)

Just another example. Apptronic has a huge investment from Mercedes. So we see a lot of these humanoid companies and these partnerships with auto companies. It's going to become a common trend that we'll continue to watch. The second I thought was definitely another humanoid kind of, but definitely on a different level. So this particular robot, the LG's robot called Cloyd, and it really kind of rivals the Sunday Robotics robot, Mimo. I remember, if you remember, we talked about it a couple of weeks ago with the hat. So the main difference between, you know, this version of robot, which is really meant for to be in the home folding laundry, doing your dishes, maybe even doing some baking, is it doesn't have any legs. And we have seen that, you know, that can be beneficial. You can get more leverage if you're.

Chris (49:34)

Lifting items and you won't tip over and you can move around quickly and it can have batteries in it. So the not having legs means you can't go upstairs. But it does mean you don't tip over, does mean you can put a vacuum in it, does mean you can put batteries in it. So it does come with some advantages.

Alex (49:52)

That's kind of an ugly bot. I'm not going to lie. It doesn't do much for me.

Chris (49:55)

Lg, well, it's coming from lg. And if you don't remember, lg, that's the South Korean conglomerate. LG used to stand for Lucky because they were a chemical brand back in the day. And Gold Star, if you remember that electronics, they rebranded the Lucky and the gold goldstar as LG. And then in 1995, they started calling it Life's good for their international. But the Koreans, no manufacturing LG, Samsung make great TVs, great appliances, dishwashers, they're going to be able to scale this as well. So now you've got LG on top of Softbank and Hyundai. This is going to ramp up quite nicely. These companies will have no problem making hundreds of thousands of these and eventually millions. So if people want them, they'll get these down to that looks like 5 to $10,000 eventually. Optimus and the Hyundai one I would put at 20,000 eventual price point.

Alex (50:55)

All right, Oliver, what's next?

Jason (50:56)

I will add that the demo that they showed, it had the LG robot putting some laundry in in the machine and it was super slow and all the other demos looked really slow. So this is. They're obviously behind some of the other manipulation manufacturers but it'll be interesting to continue to watch. But. And then next I thought was I was super excited about this one, I think Jason, you commented about it. Neuro.

Chris (51:17)

Neuro, friend of the pod, the founder came on the pod.

Jason (51:20)

Yeah, Neuro, yes. And they have unveiled a kind of new production ready robotaxi with Uber Lucid and Nvidia is kind of the power behind the unit. And what we can see here is the, is the Neuro unit on top of Lucid car.

Chris (51:37)

The sensor pack is on the roof and that sensor pack is now down to under 10k. So they're going to be able to add these to the Lucid. The Lucid is made, this is the Lucid Gravity, the electronic maker and Uber has ordered I think 20,000 of these already. And this is in the Gravity, which is their best selling, highly regarded SUV slash minivan hybrid.

Jason (52:01)

What I'm interested about here is there's four different companies that are coming together to build this one vehicle and obviously this car looks much nicer than the Waymo, but obviously still pretty similar. So it'll be interesting to see how they compete.

Chris (52:13)

This is Uber Black essentially. So this follows the Uber playbook, Alex, of starting with the high end and then going down the Gravity. So we're showing the Zoox here as well. But the one we saw originally is the Lucid Gravity. That is a six seater, that's a big luxurious Model X, you know, Escalade competitor and it's a really nice car for an airport ride or a CEO, you know, somebody who doesn't mind paying for Uber Black Waymo and FSD Robo taxis Model Y, that's more of an Uber X experience. And eventually Zoox is more like the Uber pool lift line. Multiple people will be in those like a, on a route and all of those are gonna have different price points. And what's nice about it is we're moving into the mass production phase.

Alex (53:07)

So yeah, I want to get my own Zoox and put My peloton in it and that way I can just peloton around town, you know, almost like a real bicycle ride.

Chris (53:14)

You, you laugh, but there's going to be two different versions of the one world. There is a Palo Alto company, I forgot the name of it but they were also at CES producing a robotaxi for you to buy as a consumer and sit in the backseat of and the steering wheel will retract into the dashboard and you have a robo taxi like experience if you're in the geofence zone where they're allowed to operate. Then if you go outside the zone, Alex, that company then. And has the steering wheel come out if you want to drive it to Lake Tahoe. So let's say you're in the Bay Area and hey, you can go up and down the peninsula, you can go to Berkeley, Oakland, but you can't go to Napa on it because Napa has an approved or Tahoe and Sacramento have an approved full self driving or they haven't achieved it in that geofenced area. So you could, you know, hey, go to work every day in San Francisco if you lived in Palo Alto or vice versa and be driven in a robot taxi without paying attention.

Alex (54:15)

Is this the car you're discussing? The robo car from Tensor?

Chris (54:18)

Tensor, yes. This company came out of seemingly nowhere, but they're doing a ground up, I believe Hyundai might be making, it's a contract manufacturer making the car for them.

Alex (54:28)

Here we go. Here's the steering wheel folding up and going in. So you can see a retreat. If you're on the audio version, we're basically watching a car that has two large iPads on the, on this dash and the steering wheel literally retracts in like a turtle's head and then the screen goes over and covers it. So you're just looking at your own iPad Pro. Yeah, fantastic. Love it.

Chris (54:48)

So this will be, I think that's the, the bit my big takeaway of CES was it was robotics and self driving. Okay, let's keep going.

Jason (54:55)

So next we got the longevity mirror from Neurologic. And I thought this was really interesting. I think what, what their kind of, you know, thesis is that you could get a lot of, of you know, longevity, health, aging data just from simply, you know, camera that you sit in front of for 30 seconds. And I think that you know, there's going to be a lot more health tech and we've obviously seen a ton with Aura and, and whoop. And you mentioned Strava. They'll all be coming together. And I just think that we're going to be living for longer and there's going to be a lot more tech just like this.

Chris (55:25)

Okay, keep going.

Jason (55:26)

We have two pet health projects. One is a collar. So this collar will basically go on your pet. And it does. It's obviously works like a normal collar. You can attach a leash to it, but it also goes beyond. And it's kind of focused on the health side of things where it'll actually track the gps, the motion, the sleep and the temperature data.

Chris (55:43)

Fitbit for your dog. Brilliant. Everybody wants this, especially if you have a ranch or you have a dog that escapes. Next up on the top 10 list is.

Jason (55:52)

Yeah, so this is another pet health tech company that focused on cats, how your cat is eating. And this is just another example of more health tech.

Chris (55:59)

So it weighs the food you're giving to it, or it some way will tell you how much your cat is eating for health reasons. That's the goal.

Jason (56:08)

That is the goal. Because I know that a lot of animals will like eat less or eat a different amount when they're sick. So I think this is kind of will raise some alarms when that happens.

Chris (56:18)

People have fat cats. These cat ladies, as you know, Alex, these cat ladies overfeed their cats and then they get three or four of them. And this is more about the cat owners being too permissive with food. That's what this is for.

Alex (56:32)

You know, as a dog owner, I have no idea what you're talking about.

Chris (56:34)

Exactly. You need to meet only one cat lady. Okay, Next up on the top 10.

Jason (56:39)

Consensus list, the plod note pin s. So this is a another wearable AI notetaker sits on your maybe collar or you can just probably just throw it in your pocket. And this was announced at CES and what I thought was really interesting about this, this and I've played with the pocket. I made a demo about it, which will be coming out soon on this week in AI I realized that all these kind of note takers and, you know, AI wearables are, I think, for the most part all the same, where basically they have microphones, they have a certain amount of storage on device, you know, they have a certain amount of battery life. But what's interesting is they basically all will like go to your phone and then you'll be able to use AI features and just use the transcription data from there. So it doesn't really seem like any of these devices. Devices are that unique. And for someone who's on zoom all day, I can just use A granola or a or notion AI to do this exact same thing. I would say that the use case that makes the most sense for tools like this would be someone who's out in the field, you know, maybe door to door salesman, someone who's in real estate. They want to remember what their client was talking about. So I think like a lot of us in tech, we're on a computer all day maybe, yes, less good space is for us.

Chris (57:51)

Yeah, that's who this is for. And if you use plod or Pocket or any of these tools, what you do is you press a button on them, it starts recording. Then you press the button, you turn it off, it puts it into a large language model. And when you process it, Alex, you can process it as a doctor, you can process it as a meeting, you can process it as a therapy session, if you were a psychiatrist, et cetera. And then it will make notes for you. They are all extremely privacy based. They're giving massive disclaimers about, hey, please don't use this to covertly record people. But that's what people are going to do with these. You're being recorded right now on every call you do. All of these have a functionality, Alex, on the products where you can, if you hold the phone, it will use the vibrations and the device attached to your phone in phone call mode to record the phone physically without you having to add something to the zoom. So you can just take a regular phone call and we'll do it. Now this is illegal in some states, so they come with tons of disclaimers. And there is a rumor that The Jony I've OpenAI product is not a puck or a pendant. It's a pen. And I had a pen called livescribe. And that pen, when you wrote into a book, would take whatever you wrote in the book and make a visualization of it and then use ocr. And I was addicted to this. I gave it to everybody 10 years ago in the company as like a Christmas present. It's kind of neat to think of never forgetting things. And it's also super, super creepy.

Alex (59:33)

Yeah, I was gonna say bummer on the privacy front. I can see people using these, but I just don't want them in my house. Thank you very much.

Chris (59:38)

Much. Anything else that was particularly interesting? Before we drop you off, two more.

Jason (59:43)

One would be the Samsung Galaxy Trifold. And I think that this is interesting because I have a question which is how many times could you fold a phone? But I do see the use case. It basically becomes a full Tablet.

Chris (59:55)

I have The Pixel Fold 9 I think and you know, I used it a lot when I first got it. It was great for watching YouTube videos. My personal belief is it doesn't provide enough value for me to switch to it, but if Apple makes one, I guess I might buy it. I don't know. Do you have any interest in these, Alex? The double fold, the triple fold?

Alex (60:20)

I like to have my computers be computers and my phones be phones. I like to have my iPhone do iPhone things. But I will say my father in law has a folding Samsung that he carries everywhere and he never ever unfolds it.

Chris (60:30)

He never unfolds it. Ah, interesting. I was unfolding it like playing chess or watching YouTube videos and I think the software hasn't caught up. What really needs to happen is you need to be doing your Gmail on one side and on the left side have it summarizing it or doing AI. I don't think the application layer has been finished on this. So if you were playing a video game on these, like let's say you're playing chess and on one of the folds it's giving you coaching. On the other folds it's giving you the hit history of the match and then the main fold it's showing the chessboard. That could be interesting but nobody's built interesting apps yet. And that's the missing piece for these. Well done producer Oliver for your first hit here. Not bad, not bad.

Jason (61:16)

Thanks. See you guys next time.

Alex (61:17)

Thank you. Oliver. On the using your phone to do stuff and have like, you know, help on one side and the actual thing on the other. My first thought was oh, we could have our, you know, our pre flop charts next to our poker game.

Chris (61:25)

Perfect example. When you're playing poker, if you were doing a training app, you're playing and people do play with a heads up display on their desktops. In notion you've got your sidebar in your main document window. In Microsoft Word you'll have something similar. So this idea of the AI assistant copilot on a third of the screen and two thirds being something it hasn't been finished. And what really needs to happen is somebody like Google has to take Gmail, which they just released some AI features for today I saw trying to catch up with Superhuman. They really need to figure out what each window is for and then you could become addicted and then say, oh you know, I prefer playing chess in this modality. I prefer doing Gmail in this one because I'm faster, better, it's a greater, it's a More entertaining experience. So this really is when you. This happens all the time. Somebody comes up with this incredible hardware, but you need the application, the killer app, and the killer app for foldables, if you know of one, put it in the comments here. And what you can do as a startup is you could go to the people making these and say, can I get 250k to make something the next time you demo it? So you can get developer funding for. And I've seen it over and over again for these. Oliver says he saw somebody playing flappy birds and it goes up when you fold the phone interest thing. That's bizarre. Yeah. You're going to break your phone pretty quick if you're flapping it open and close.

Alex (62:54)

It's a terrible idea.

Chris (62:55)

Terrible idea.

Jason (62:56)

Yeah.

Alex (62:57)

I will say though, I just think that we're going to be building our own stuff. I've become fully.

Chris (63:00)

Oh, the phone did break. He said the phone broke.

Alex (63:04)

Well, it's ces, you know. No, I've become really a big fan of Claude Code this week. I've been building my own stuff and it's like whenever we talk about now, waiting for permission to do something because someone else needs to build the software, I just immediately now think, can I build it myself? Like, I built myself a financial news dashboard and then I made my own GTO trainer app for both 6 max and 8 max and also pre and post flop. It took me like 20 minutes.

Chris (63:31)

It's getting there, it's getting there. I mean, production ready versus vibe coding, obviously, you know, pushing stuff to production. You're going to need a developer to lock it down, but if you're just doing it on your computer, I think in a lot of cases we might be sitting here a year from now and you can just make an app instead of paying for it, you just make it right. You're like, I don't want to pay for this app. I'll just make my own CRM and you just tell it. Make me something like Salesforce for my 3 person company and I don't care if the data leaks or something inside my company. It's not a big deal. And there is a theory that soon we will have a phone that you open and there's no apps. You just open it and start talking to the LLM and then you say, I need a word processor and it pops up a word processor and then it pops it down and you don't need to open specific apps. It's a very interesting future.

Alex (64:29)

That's an interesting way to frame that, because what you just said is we are going to have a personal AI that we take with us and it will have a phone based distribution point. But at that point the phone is just a piece of hardware that's bringing you the software. So it's not really a phone per se, it's more like a small screen into.

Chris (64:47)

It's almost like when they would. There are people who sell the. For a car, the basic chassis, right? And you just have the chassis. You can build whatever you want on top of it. You could build it into a van for an executive, you can build it into a van for, you know, camping. You could build it into a, a food truck. Right. I think that's how people might start to look at these slabs of compute. And maybe you don't even need to have too much local compute. Maybe it goes to the cloud, maybe it goes to a satellite. I think if SpaceX does a phone or Tesla and SpaceX Chamath had a really interesting prediction that Tesla and SpaceX would become one company. And then you could think, okay, the Teslas have satellite dishes in them and they're creating their own network of Internet access across the globe. And you have satellite access and there's a SpaceX or Tesla phone that comes with your car or with your satellite hookup and it just connects to Xai and you put Xai into it and now you have those three companies building products and services together.

Alex (65:57)

Sounds more like Starlink plus Tesla instead of SpaceX plus Tesla.

Chris (66:01)

Yeah, I mean pick your poison. And then Xai building the back end for all of these things. So now when I get in my Tesla, it's got Grok so you can tell it like, hey, I want to go pick up bagels and then I want to pick up some donuts from the best bagel place and donut place along the route to my office. And you tell FSD to do that.

Alex (66:21)

How is the integration? How smooth is that? Because that to me sounds awesome. But I've also been in a lot of cars with kludgy interfaces. I'm curious, is it good?

Chris (66:29)

You know, you can see some people doing demos online and I'd say it's 2/3 of the way to perfection. Which is a way of saying it'll be perfect in 18 months. Like I think it's like it's right at the halfway mark to perfection. You know, you could even. You'll probably be able to walk in and I think Waymo is doing this as well. You'll be able to walk in and say hey, play me some deep focus music without lyrics. On my way to work and I need to stop for a cup of coffee. I want a pour over and I really want a great coffee. I don't mind taking 15 minutes out of my route to get a cup of coffee and just get me to my first meeting 10 minutes before time and park the car, drop me off outside and find me some decent parking. I'm going to be in the meeting for 45 minutes, so you could even circle and then pick me up in 45 minutes. Minutes. And it's going to do all that. So you're going to be able to give your car a series of instructions from everything of the comfort to the pickup, to the drop off and to the parking. Really interesting future.

Alex (67:30)

I just had a great idea. So remember we were talking about zipline earlier and it drops the droid down from the main drone and then you kind of pick up your stuff from the ground. Okay, flip that around and add self driving cars. You want coffee, you want to go to get specific coffee. Why can't you just, just drive over something, right? And the coffee shop is in the ground and then it opens up a little slot and then puts the coffee up through the bottom of your car. And then you could literally not even get out or stop really. You could just like pull through like a little lane.

Chris (67:56)

It's interesting you bring that up. You know, the company auto lane, which we incubated is doing like the navigation essentially air traffic control for the parking lot at your Target or at your McDonald's, etc. And I saw people are now saying FSD, Tesla's FSD specifically understands. Yeah, and there's auto Lane. They basically do the orchestration level for when you go to, you know, when you send your car to pick up packages for you. Like the person has to put it in the right car. They want to put your package in my car. So somebody has to be responsible for that coordination. The coordinator, the air traffic controller. And that's what Auto Lane is doing. All right, Another amazing episode. Have a great weekend. I'm off to Japan tomorrow. We're going to do four episodes from Japan. Incredible investors and technologists. We'll be interviewing at Founder University. Founder University. If you want to apply to the next one in one of three cities now. Tokyo, Saudi or in the US of A. And good luck with the baby next week. Alex, we'll see you in a couple of weeks after you get back from paternity. Good luck.

Alex (69:05)

Yes, sir. All right, I'll see you all then fly safe.

Chris (69:07)

Bye bye.