The AI Boom Is Splitting Into 5 Bull Markets At Once

Is the AI bottleneck trade completely over? On weeks like this one, it kind of feels like it. Mainly because we've been used to months of new all-time highs, but beneath all this chop, there's actually a lot of proof going on that this AI buildout was never really just a bubble. The frontier models are showing actual revenues now, and the application layer is actually working. There is a cure for cancer out there that relies on AI to work. So, the real question is not is it over, it's how to play the game going into the next inning. What's up everybody? It's LG Ducat here and welcome to Milk Road Stocks, a daily market show that had this date circled for months and still doesn't know what's about to happen. Today is August 26, 2026, recording on August 25th. Like I said, there's a lot more happening in AI than most people realize. Today we sit down for our weekly rollup with our top three AI analysts to look at the potential for biotech. How there's finally evidence of ROI in the market and how software as labor could be the next big thesis. We'll even find time at the end to touch on Nvidia before their earnings report later today. These guys are making trade calls every single day and the price to see those is about to go up. Tonight at midnight, Milkro Pro goes up 60% from $25 a month to 39. We were going to end that offer last night, but we had a few people beg us today to open it back up. So, you've got just a few more hours to get that deal before it's gone forever. And a reminder that our podcast today is free and that it wouldn't be possible without our partners at Saber. The stablecoin payments platform built for Asia. Keep an ear out for more information about them later in the show. All right, boys. Last show of the month of August. We've been doing this for a while and it's only getting better. Uh let's go around the horn and hear how you guys are feeling in this slightly slower week uh in the AI market. Melvin, how are you feeling? >> Fantastic, man. Uh we had a bit of a sell-off in the last yesterday, but that has largely recovered. I feel like um I think Nebius um went as low as one or like 200s or 210. Um they're up today. be he had a huge huge run up in the last uh few days because Nancy Pelosi bought some uh call options on uh Bloom Energy which is great uh for our all our trades. Um overall I think we're in a really good position. I feel good. I'm still continue to hold all my stocks. I I think I bought some new ones last week uh but nothing so far yet. But so far uh I feel really confident in all my trades. So I think we're going higher. >> Beauty. How about you Vincent? >> Interesting Melvin. I think this is the first time we where we kind of have different feelings on the market. Uh to me the the AI infra feels like there was a sentiment uh reset especially if you look on X. I think that's that has a lot to do with with with crypto, Bitcoin, ETH uh up a lot and kind of capital rotating at least AI retail investor capital rotating into that market. But I think the more important kind of take from as an investor here is that July and August looked much more mechanical than fundamental. I think there is a lot of news on the Leopold unwind and then Citadel bought the whole part but actually already unloaded 80% of the stack which is why I think the broader AI info trade is under pressure at the moment. There's nothing really fundamental that is um yeah challenging the broader trade around memory power info etc. So I think there's some buying opportunities at the moment. I think on Milk Road you can see that we are buying stuff and uh yeah so also remain bullish but kind of different take on how the sentiment is at the moment. >> Oh wow. Kyle's going to come in as the tiebreaker. >> Uh also bullish uh pretty well across the board. I don't think we're getting rotation into crypto just yet. I know we had a big move there, but I think just a lot of that was like short squeezes um and a little bit of life back in the debasement trade which is which is great. um kind of supports the everything bull market thing I've been talking about uh now that we have debasment trade on top of it which is even better for for risk assets. So crypto and even things you know kind of down the risk curve like biotech and stuff which we're going to talk about they're they're a very high risk asset um and momentum asset and so you want things like the basement trade to to power that. I think on the infrastructure side as I've been saying for a while we're just in a chop for the summer. Um, we had a huge runup in the first half of the year and I think all we're doing is kind of resetting. From a technical standpoint, it actually looks like very very crisp. I bought yesterday a few of the um them in Milk Road Pro. You guys can check that out. Literally just because of a technical side of things. Let me just share screen. I'll show you what I'm talking about here. But if you look at Melbourne's favorite uh tokens, tokens assets uh stocks, you got Nibius uh >> your mind is only at the moment. It's so bad. >> Can't be a better Freudian slip than that, man. That's so good. All right, go on. Yeah, tell us about NBS. >> Yeah, so Nibius bounced right pretty close to this line anyway. Um it it it uh off its kind of bull market trend line. Um, and so that was a good signal. And then Cororeweave did the exact same thing, which is in this kind of like wedge. It's been in for a long time. Um, after it hit here and came down, I was like, "Okay, this thing's going to hit this." And it hit exactly it and then bounced off. Bloom Energy did the exact same thing. So, honestly, I just this feels like um, you know, obviously this selling here on on Bloom and on Nibius and on Cororeweave was kind of the Leopold thing, the Korea Unwind, all that that we talked about and covered a bunch. Then I think we bounce. I think this selling here was the um Citadel uh selling afterwards and again we went right back to the trend line and all these assets don't forget Nibbius Corweave Bloom Energy the one reason I'm saying those three is those are three of the bigger assets that situational awareness held um in terms of like being kind of smallalish assets you know they're all around like 50 billion or whatever uh and so I think this is just textbook uh bounce here and I feel like they're going to they're going to bounce from this again I I don't know if we go off to new highs anytime soon. We're still in summer. We got to get through like Labor Day when everyone the big boys come back who who actually buy Um but I think fundamentally, as Vincent said. Um you know, I think it's still very very strong and there's still going to be plenty of demand. So I think good time to be buying infrastructure stocks at the moment. Even though their multiples are much higher than things like the application layer. The application layer, we've seen the rotation over the last two months. I think that continues because the earnings are going through the roof and we're actually seeing earnings growing faster than stock prices right now in the S&P. So I think the application layer remains kind of undervalued. These are certain SAS companies you could look at and I don't have these but some of our analysts do things like Uber and Apploving and like all these kind of traditional companies that are integrating AI and kind of reinventing their business. I think there's lots of opportunities within those kind of companies. Um and then now you're even seeing the higher risk companies that are have potential because of the debasement trade. Whether that's in crypto, biotech um even I mean biotech has both the debasement trade and what happened with Madna and um the AI assisted mRNA cancer treatment. Um so again using AI to to you know improve their product. So I just think the everything bull market thesis stands and I think we're going to go into a really good end of the year. Um and so I'm well positioned for it. I'm just going to interrupt here for a second to say that we are extending the Milk Road Pro offer yesterday. It was supposed to go up by 60%. The price was going from 25 to $39 a month, but we're actually extending it by a day by request of many members who were late to get in. So, with that, you get access to all five of our analysts, realtime trade calls from them. You can get that by email everywhere else. our pro platform where they write updates every day and of course their full portfolios where they show you how they are preparing for these amazing bull markets that are coming. So get into Milkroad Pro before it goes up by 60% at the link below. Tonight midnight Eastern is the actual deadline. So do not miss that. Did they specifically credit AI with that M mRNA cancer vaccine? >> Yes or it's part of it. >> It's part of it. Okay. So but it was in there >> that they that it was assisted them, right? >> It's it's part of the solution. It's it it's not how it's not AI help developing the drug. Maybe there was part as well. I don't know. But it's part of this of the treatment. The AI or the model is basically picking what kind of treatment or specific treatment you your individual cancer needs. So um it ups the probability of of success in treatment. >> But I don't even think it was an LLM. I think it was more machine learning that did this as far as I understood. But I did not do a deep dive. Honestly, this was just some people I saw writing this. I didn't actually double check. >> Yeah, I I don't know if you want to cover that now or not, but I have some pretty strong views on that. Uh >> let's do it. Talk. Let's hear for those that don't know just for context. Madna was up like almost 200% uh in one day or like one and a half days because um it came out and announced that it is um able to not fully treat but definitely having some success in its treatment through like phase three of a specific type of cancer patient. Um and so you know kind of one of the first times we've seen a phase three success for cancer. Um and then also one of the first times we've seen kind of technology like AI be involved in that. So it's um it was a huge catalyst for the biotech sector. If you've looked at, you know, a few of the assets that I have in my portfolio and Vince and I have talked about a few times, Eli Liy and and a few other biotech names, they're just going through the absolute roof right now. Over the last week, they've gone up, you know, a ton. Um and so it's kind of driving some interest into the biotech market. uh because it looks like AI is going to start to play a role inside either treatment or um you know diagnosis etc. So Vincent would love to hear your thoughts on what happened here. >> Yeah. So so two data points before I jump into the kind of investor implications from this. So there's 49% lower risk of recurrence or death at a $5,000 uh kind of price tag per patient from this. So for $5,000,000 you for $5,000 you can basically save your life if you have cancer, increase the the probability of surviving by 50%. That's probably explaining why the stock is up so much. Um and what the investor implications are from that is AI or software in a broader sense will not only be part of R&D or drug uh kind of discovery, right? it it becomes part of the of the pharmaceutical value chain um itself because again here the AI picks kind of the the treatments specific to the tumor or or cancer mutations right and what this enables is a much cheaper and personalized um um treatment of human being right personalized medicine basically and if you think that through I think the really valuable part in the entire uh value chain is not sitting at the drug itself. So who's producing the drug um but actually the the uh the best system or the the the platform that enables this drug creation, right? And I think if you listen to Kyle and me talking about Eli Liy, they're building exactly this broader system or platform that enables exactly an a AI becoming part of of this pharmaceutical value chain. I think this is kind of the deeper thinking where the value isn't. So just for for those that um don't fully grasp what's come going on here, the world we're going into at least I think that Vincent and I think is going to happen and I think will lead it but we're seeing kind of the first initial start of it here from Madna is the ability to when you say kind of personalized like right now you have cancer or you have whatever you get like this treatment or this treatment or this drug or this drug, right? Um but it's the same drug for everyone else. And what we're going to do here with AI is we're going to be able to see what's happening inside of their body. And then AI can look at it, all the different, you know, there's hundreds of different um variables here. And then AI can can basically look at that, you know, machine learning, whatever, go through it real quick, and then basically create a cocktail, a different type of drug that's personalized for the exact inputs that we got from that patient. And so you can have way better um results uh on different diseases. and like we're not there yet, but we're getting there. And this was kind of the first um first crack at it, but also way better results with way less uh side effects and downsides. And so, and you can do it way faster, right? And so, if this happens, we don't have disease anymore. And that's where a lot of the smartest people in this sort of biotech healthcare space that are kind of on the the the connecting the the dots here they think in 5 10 whatever it is years we actually you know cancer ends up becoming a as similar as like the cold right it's like you get the cold you take some Tylenol or you take whatever an antibiotic because you have a stomach ache boom it's gone like that's how the level that we can get this to now we're still many years from that but at any point whenever ever we get close to that, the companies that are going to provide that product are going to be worth multiple multiple trillions. And so part of that is why Eli Liy is valued at a trillion dollars today, but barely any of it. It's actually valued at that because they have the best weight loss pill in the world. Uh and you know that they're making just tons and tons and tons of money from that right now. Um you know, billions and billions of dollars and that's only going to continue to skyrocket because it's still a small market share. But once we get into other things outside of just weight loss, it's a whole different world. And I think, you know, I don't know how far we are from that, but I do think Eli will be the winner there. But that's kind of the speculation of the biotech movement that's happening. >> Right. So basically that this is the first of many similar developments and treatments and for other cancers and other diseases and stuff like that. So it's not just Madna, it's many other companies um >> that will be able to do this. Go ahead, Vincent. >> Yeah. And I think this is important to to dive deeper on Lily because what Lily is doing is they spend massive amounts of money on M&A, right? They're buying a lot of startups and and bigger established companies in the pharma space beyond GLP ones. So, uh beyond weight loss drugs in the end, right? And they're at the same time they're building their um so they have 10,00 GPUs and they're building their own models uh also in collaboration with other AI companies. Um, and then they're using that AI infrastructure and plug it into all the companies they're buying, right? And they're just increasing the chances that we're seeing something similar to what was what Mona came out this week, right? Because they have so different that many different angles into into spaces of of of of the pharma value chain which is just massively valuable. And I think people are not understanding it understanding that it's so much more than than weight loss or applying AI to just weight loss. It's applying AI to all the different spaces in pharma. And I I don't know if you guys remember this almost feels like an alpha fold moment um when Demis Sabas like you know decided to release that uh into the world for for free. Um this almost feels like a big breakthrough since that and we talk about if you look at like the top AI researchers like Demis and you know even Sab Alman talking about or like if you listen to Moonshot people podcast Moonshot they they predict that you know by 2040 2050 all these cures all the AI will essentially solve all cure uh will create all cures which is likely what we're headed to you know especially after a big breakthrough like this >> and I I I think This it's really important to connect this back to the AI infiltrate because why is the infiltrate lagging? It's because people are struggling to see the ROI on those investments and we need more of those cases proving that with yes I understand it's probably not full fullyfledged an AI case yet that modern is but it's a step into that direction and if we get more of this AI use cases across industries in finance in pharma etc then also this this overhang of is AI really that valuable and do we need this entire buildout is going way which will help the the the infrastructure trade around compute power etc itself. It's big news all around and and I think we're starting to see more use cases on the application layer as well. I I posted a chart the other week um that showed that those companies this is like um the top I forget what size of company it was but I think they make 100 million plus or 500 million plus and um those that are using AI are generating three times more new revenue than those that are not right um you're seeing it with the Palunteer Service Now and all these companies and their earnings um in terms of both their earnings are good and then the clients that they have are good are really good and so I you know, we're starting to see it kind of all over the place and it's only going to get better as token costs go down as well. But the interesting thing from this, and I'm going to just find the chart here and bring it up, is did you guys see the revenue numbers from OpenAI and Enthropic uh that came out late, I think last week. Um yeah, so OpenAI is accelerating. Everyone thought it was not accelerating because of its Q2 was only up 18% quarter-over-arter. But then if you look at the month of uh July um into August, it grew 20% just in that one month. It was absolutely crazy crazy month. So it's actually growing faster than Enthropic is right now. And Enthropic is growing incredibly fast. So we're starting to see, you know, it's funny because the market thought that those revenues were going to get cooked because of the um the open source models, but they're actually growing extremely extremely fast. Now we're going to get Enthropics IPO. Um, some of the preIPO markets have opened up on like Hyperlquid in that and it's getting closer to two trillion. So curious what you guys thought when you saw those numbers. I I'll find the chart and I'll pull it up in a second. The 18% quarter growth number from from OpenAI is raising one question right if Frontier models cannot increase their revenue substantially which OpenAI did by the way. they they post a number of uh around 100% quarter over quarter over quarter growth. But if that were not to happen, who's financing the infrastructure buildout, right? I think that was the risk of the market. It ties also back to this ROI question. But then I think you need to look at it from a a broader lens, right? You need to see the hyperscalers who fund the capacity build out with the existing cash flows. Then you have Nvidia and I think this is once again really important for the broader build up what Jensen did there with the 500 billion um of capital that's that's flowing into the market. You have Bank of America launching at 250. So and then you have Goldman Sachs coming out saying that the real estate funds and infrastructure funds have around a trillion left of dry powder to spend in. So the worry of investors that capital will stop flowing into the AI market because frontier models do not grow as much because they're one of the biggest buyers of compute, right, is complete nons nonsense to me. And I think when you saw those open AI news coming out, the the infotrade sold off and I think it was a great buying opportunities and I also added some some bits and pieces there as well. >> Yeah, I completely agree with this. Actually, there was uh a chart that was going around in Twitter. It was from Versil, I believe. Uh it was basically saying that open- source has gone from 28% token share to 68% 62% token share over the last two months. And uh that's you know that's primarily taken from OpenAI and Anthropic. Um but despite that they still grew their revenue by you know 18% year I forgot if it was quarter or quarter. um or and Anthropic just hit $65 billion in revenue as well and they're projected to hit most likely hundred billion by end of the year and I don't really see a world I mean I can see where they could slow down in revenue but their revenues comparable or their revenue is directly tied to how much gigawatt that they have in capacity. So >> yeah, >> if they if they could get more uh gigawatt, they're going to continue to grow at a at a great pace. And I I still believe that and I still I I still think they're in a really good position. The good development here though is so open AAI entropic because I know Vincent you're talking about well we got to look at the hyperscalers you got to look at Nvidia and you know their financing and how are they doing but ultimately you got to look just the very top and that has for a while just been enthropic open AI and Microsoft and if they stop spending doesn't matter what you saw down below all that's going to go down right cloud revenue is going to go down cloud revenue is going up because open anthropic are are spending on that right um and so it matters was those two, but we're starting to see it diversify out. I think Cororeweave announced they had a new client that was like a financial firm, like a a trading firm. You've got uh you know, Eli Liy, which is buying GPUs. We're starting to see a bunch of companies that are not Frontier models buying their own GPUs and >> open models. >> Exactly. So, that is really really good. Um and then the second thing I would say is like while that's happening, these revenues are also going absolutely through the roof. Like these are big numbers here on this chart. And the reason why I just want people to understand because I think a lot of people are getting mixed up with OpenAI. OpenAI wasn't growing as fast because it wasn't going after enterprises like Enthropic was. So Enthropic grew a ton because they brought in enterprises which means they can just sell a bunch of seats. Um companies uh and enterprises when they're using AI, they're happy to spend more and more and more and go past their subscription amounts because they get an ROI, right? The more they tokens they spend, the more money they make. Okay. Consumers, that's not the case. So your average consumer who's spending 20 bucks a month or whatever on chat GBT, they don't really get like a financial ROI from using it. Sure, they save some time and like they might get better responses than in Google, but like they don't have max ability to spend. They're not going to spend 20 grand a month on API credits, right? Um and so but OpenAI, they have a billion users. It's mainly was mainly consumers. So what Open AI has done a really good job of over the last quarter and a bit here is going after enterprises and getting them in. now because those are the ones that have unlimited spend. Then what they can do is with those billion users they have a consumers because they still do, you know, they're smashing it on the consumer side. They're starting to sell ads to consumers. So I think what we're about to see here is another revenue stream. Everyone's only looking at subscription models for these. I think at least for OpenAI, I don't know if Enthropic will do this. I know they put out the Super Bowl commercial last year that would never do ads, but OpenAI for sure and they've already launched it and they just opened into 32 countries in Europe as well last week is selling ads to consumers and then selling, you know, subscriptions and API credits to enterprises. And I think both of those are going to grow together. They're expecting to hit 2.5 billion in ad revenue by the end of this year. So, I think that's going to grow really fast. I think that's going to be a good development to be honest. You know, one thing we've talked about a lot on this show is that crypto is quickly becoming a huge part of the global payments infrastructure. And nowhere is that more obvious than in Asia. But if you're actually running a remittance company or a payment business, you know that the hard part isn't moving the stable coins. It's dealing with local banking partners, compliance, liquidity, and all of the operational headaches that come with sending money into places like India and Southeast Asia. That's why today's partner is Saber. They give payment companies stable coinpowered infrastructure to collect and make payouts across Asia without having to build all of that complexity themselves. They've already processed more than $3 billion in transactions across 40 different countries. So, this isn't just a concept. These guys are actually doing this for real. If you're building payment infrastructure or expanding into Asia, make sure you check out saber.money. >> Yeah, I just want to quickly comment on that um enterprise value that OpenAI is going after. Um keep in mind that Codax, you know, which was largely lagging for a while. Um since 5.6 release, uh 5.6 6 release, they've grown to essentially 20 million uh users. You know, that that is an insane amount of growth that's happening with OpenAI right now. And on top of that, if you look at the data on who's actually using that, um the fastest coding um growing codeex adopters since July, since February are legal, which is up 108x, sales, which is up 41x, recruiting up 41x, marketing 26x, and healthcare 24x. These are people with lots and lots of capital that that they're willing to spend on. >> Can you say that again? Healthcare 24x up in terms of what took us. Yeah. Uh Codeex adopters. Those are the Codeex adopters since February. >> So do you have more details on it? Are that doctors sitting there and using Chat GBT when when patients are in front of them or is that deeper? That's really interesting. Yeah, this was a chart that was posted from A16Z, I believe this week. Um, I don't have too much details on it, but I would assume a, you know, mix of both. My doctor is using AI now. I don't know if it's codecs or chatbt, but, you know, I asked questions and he's like, "Hold on, let me just check." And then puts in a question into into AI and I was like, "Why didn't I just do that?" Uh, but obviously his is trained specific to his like to to their thing, so like it's it's better. Um but I don't know if it's FTP or not but yeah it's it's happening for sure >> to the point and maybe Kyle you can bring the other slide on again that that you showed earlier was entropic and open AI >> I think the really interesting part here is that it's pretty well known that OpenAI has more compute than entropic right was this lagard in building out compute right and it was always the case that the more compute you have the more revenue you make um I can see cannot see that from that chart. Right? So I think and and and with the entropic uh um IPO coming up in Septemberish. I saw a post today on X. I'm not sure if that's true. It's from Amid, so it's pretty reliable source. Um that they're saying we are addressing a $30 trillion TAM. So they're seeing a market size sized at $30 trillion US. >> Yeah. And a bit out to lunch though. They also think they're going to be the only company that exists in the world and like you know >> that's absolutely fair. That's absolutely fair. It's just questions to me why is opening I mean yes I understand that that the focus was wrong and now they're shifting towards >> but still the difference is like crazy and you see that AI or compute shifted towards the right value cases will bring that ROI uh and it's just important to understand >> I mean keep in mind that Anthropic's whole strategy this whole time around since they released or since they you know made a company was going after enterprise value so they already got a huge head start on Open AI whereas OpenAI only like couple months ago started you know reallocating their compute towards the right thing. We saw the shutdown of Sora we saw like them like shutting down a bunch of things and that's when I think those revenue will start to come in in the later quarters as they you know get more focused. So Martin had a good way to explain this. So he's one of our other analysts and pro. We were talking yesterday and he's calling it um software as labor. And so it used to be that you would have software as a service where you would just like plug into a CRM or whatever, but you had labor had to do actually work with that software. Um and now the software is doing the labor for you. And so his kind of bet is software is still super valuable, but it's a different type of software. It's software that's now replacing labor, right? Digital labor at this point or intelligence. And so um the harnesses the the the companies that you use to interact and create and then interact and govern and manage your agents. This is where the value is acrewing. This is service now and Palanteer are are literally this they are harnesses for companies. You go in, you use this this platform and you create agents. They have like ones that are already made for companies or they can build them specific for companies and then implement them and then make sure that they're very cheap and keep improving the models etc. run the the the infrastructure for them. um these are becoming extremely valuable and so that's why I remain bullish on those companies and I think there's more and more companies doing it and that's even enthropic and open that's what they're doing right like Enthropic working with enterprises is literally to go and help them they have like a team that works with the enterprise to use agents right because the more they use agents the more tokens they spend which means the more API credits they use which means the revenues go up and so all we need right now is you just got to look at who are the harnesses that people are tapping into to use AI they're making a ton of money Open router just sold for $8 billion. There was a reason for that. Yeah, I guess that's a good segue. Uh, so Stripe bought Open Router for 8 billion. For those that don't know, Stripe is a is a payment routting company targeting small companies, right? Basically startups and smaller midsize companies. Um and the obvious question is why are they buying open router a company that you use for rooting AI demand across kind of 500 uh plus AI models. So so you have one subscription with them and every kind of question you add uh on open router will be and they they basically decide which model is used to answer your your prompt. That's that's basically open router, right? So why is Stripe buying that company? I think it's a bet of Stripe on the AIdriven entrepreneurship boom that's about to hit. Um because what you need to understand is I work at a really big company next to Milk Road. Um we have 10,000 plus employees. AI adoption at our company takes a lot of time. And at Milk Road, Kyle comes out say, "You guys use AI agents now or you you're gone." Boom. Everybody uses Everybody uses AI. I'm still here, man. I'm still here. >> But that's your portfolio needs to be up 50% year to date or you're gone if you don't use. >> But then you need to Okay, we're not going down that road. Uh but that's basically what happens right and what this will lead to is much smaller teams and companies will be able to scale much faster and Stripe is going to benefit from that because in the future they will not only be able to to root payments but also to root intelligence AI that is being adopted much faster in those entrepreneurship companies right and yeah th this is essentially the merge between AI agents and tokenizations digital assets because stablecoin has a uh stripe has a big bet on stable coins as well right they're moving their entire uh payment routting platforms towards stablecoin right and this is just the first or this is another proof point that this intersection between AI agents and tokenization and digital assets is real there's sort of like it's almost like there's a new currency and it's it's AI tokens, right? At least that's the way Stripe is kind of seeing it. So, it's like they control the currency in terms of stable coins to pay for all your digital payments and now they're trying to control the way that you pay for your tokens for all of intelligence, right? And the combination of those two, I don't exactly know what they're going to do with it yet. Um, but I think it's a it's a huge uh huge thing. They're selling two things to startups. They're selling not only the the payment routting part but also the intelligent routting part because the the USB >> which are the two things you need in a company that's it right you need a way to get paid and then you need labor and it used to be you needed to build a team of people and now you don't need that you just build a team of intelligence interesting >> if if you connect it back to Martin's point yesterday software is labor this is the open router part of the story now stripe is it's a private company otherwise yeah looking forward to the the IPO >> I don't think they're going to IPO Oh, that's a problem. I wish they would. >> Yeah, that's probably true. And I think there is one step further ahead which I'm talking my book here. But uh yeah, I think this is covering the kind of payments part of things. But when we're talking about finance, it's not only about money and pay, it's also about capital, right? And markets. And this is the the teases for Robin Hood essentially, right? Because what they're building is that AI agents can not only spend money but buy stuff, buy assets, right? And they have already 100,000 users opened up aic trading accounts. But now you need to bring in the Robin Hood blockchain which essentially puts not only stocks but real estate, private equity, all other as every asset that is out there onto a software layer that AI agents can read and trade. So they will have AI agents on the platform that will trade many more assets on a much higher uh volume note. And this is just a future that I think people are not understanding yet that humans will be a side note in the financial markets of the future. And you need to in terms of investing, you need to think about the platforms that are building towards that future. Stripe and Robin Hood and probably also Coinbase and others are really good bets here. >> Is there no other way to get exposure to Stripe? Like there's no other companies that work with them or anything like that that are heavily partnered with them? Not that I'm aware of. >> So I Well, >> I don't know. The problem is who invests in Stripe? Anyone? >> Before SpaceX IPOed, you could buy like uh you could buy Google and you'd get you get exposure to SpaceX and Enthropic right? >> Um or there was like other companies that had investments in them that you could buy. I don't know of any for Stripe, but I'm sure there's some out there. >> Maybe the Robin Hood venture fund or there are some venture funds, right, that you can invest that that basically IPO that hold private equity stocks. Um, need to check that, but that's usually really expensive. >> We got Nvidia earnings uh today. Uh, well, tomorrow we're recording the day before. What? Um, I know Vincent, you just did a full pot on that, so that was great. >> Melvin, I know you were looking into some stuff on the, um, was it the chips, I guess, comparing them with Google and some others. >> Yeah, there's a lot of Nvidia news actually that came out um, in a conference yesterday. Um first of all there's reports that came out that Nvidia is uh raising it reportedly raising its price by 15%. Um because of all the memory cost that's happening that taking place right now and obviously this will greatly benefits you know benefit companies like Micron Skhinik Samsung. So we're not going to go in through uh too much into that. One of the reasons why Nvidia can also Nvidia can also like offset some of his um raise can afford to raise his prices uh is that customers still make a ridiculous amount of money using their hardware because there's estimates that data centers selling tokens from Blackwell systems could generate like net margins of roughly 58%. And that could actually climb to 78% with Reuben and 90 uh% with future you know um chips as well. So um if you're if you're like a cloud provider like if Reuben allows a crowd provider to produce and sell like more tokens um while using the same amount of electricity yeah it's a no-brainer for companies to like accept that price deal and you know um you know to offset that. Um and also what's interesting is uh at the at the um conference yesterday was called the hot chips conference. Um it released some early performance uh results showing that Ver Rubin NVL72 can deliver up to 30 more times 30 more times more agentic AI throughput per megawatt and up to 35 times lower cost per million tokens than the current GB300 NVL72 platform. I mean, these results are still pending like semi analysis review, but if you just look at the face car that usually like Nvidia puts out, they're pretty accurate on what the numbers they um numbers they project because if you think about it, this like matters greatly for like uh power because power is like the biggest, you know, bottleneck in the industry right now, right? like the Cororeweave, the Nebias, the Oracle and all those like companies cannot keep adding unlimited GPUs if they cannot like secure more electricity and that makes like tokens produced per megawatt one of the most important metrics in AI infrastructure. So Nvidia coming out and saying this and this is like a huge huge efficiency gains from their end. And on top of that they also made bunch of like Grock um I don't know if you guys remember um there was now they bought Grock GR OQ not not Elon's Grock couple about four or five months ago um they announced that their new chip is fully in production and this is this is actually a really good chip. I'm not going to go into too too much details about it, but this is a chip that's designed for inference era, which is currently what we're in in terms of like the engetic world. So, this is a major major catalyst for them. And also, guess what? Nebus will be the first AI cloud provider that'll uh you know that will uh they picked to put this out with. So, this is this is a really good news for bullish news for Nebius as well. U they made a bunch of other announcement. I'm not going to get into that. But the one of the one of the things that, you know, I did want to touch upon is there's the competition, you know, that is going that is happening between Nvidia and the hyperscalers because I don't know if you guys saw the news this morning, there was a open AAI jalapeno chip uh that they developed with Broadcom a couple months ago. Um they revealed some crazy performance number. Um, in testing it completed 1.5 to 1.9 more times using the same amount of electricity as Nvidia's Blackwell system while responding up to 3.6 times faster. Now, this is this is compared against Blackwell, not their newest like chip design, but this is still such a big deal because if competitors can develop these chips, you know, at a at a, you know, lower cost, this is going to break away. this may could break away some of Nvidia's moat that they had on their on their chip um on their entire chip um you know moat that they own. So this is actually a huge deal. And also there's like reports coming out from um Google that their TPU shipments could climb from 2.76 million units in 2024 to 8.84 million by 2027, which is insane if you if you think about it. But there's still bottlenecks. Like the bottleneck being TMCC still have to produce all these chips. But at at the end of the day, what I'm trying to say is there's a growing competition from Nvidia from all angles. So, how do you guys feel about Nvidia? How are you positioning? Are you guys still bullish? Um, I I would love to get your thoughts on this. Let me just real quick answer mine and then Vincent, I'll throw it to you because I'm sure you got more details on this. But for me, I think the real problem is they can't do this at scale, right? So, like, yes, OpenAI has this new chip, but can they scale that thing out to actually make it a dent on Nvidia? I don't think so. Google would be the one that eventually could, but they can't today, and they're probably still a ways out from it making any dents. And then the other thing is that Nvidia has um CUDA, right, which is the kind of the software side that connects to them. And so, it makes it more um what's the word I'm looking for? like flexible um where you can kind of change its purpose whereas the others don't really have that, right? And so they're very like specific uh and what a TPU can do or what the Jalapo chip can do. Um and so I just think Nvidia can has kind of a bigger surface area that it can sell its chips into and the demand of these is just continuing to accelerate and grow. So um and I also think that Nvidia stays ahead, right? Nvidia has the fastest sort of like R&D uh and innovation cycle of any company. That's what they've always done so well for like 30 years now or whatever it's been. And so like like you said, OpenAI, you know, comes out and they're, you know, a little bit better than Blackwells. Well, Blackwell is old, right? It's really old for for for them. They've already got Ver Rubin and then they have their next one already that they're working on that they're coming out with, right? So like Blackwell's old news and so they're comparing to legacy stuff. So it doesn't really matter. Um, so I just I think Nvidia continues to hold its margins as a result. What do you think, Vincent? >> Yes, I I did a whole deep dive on Nvidia covering all of those points. I think it's on on on YouTube already. The TLDDR is their innovation cycle is 12 months. A6 from the hyperscalers take between 18 to 36 months. And if you compare the Reuben chips to the or the the the Blackwell chips to the Reuben chips, the Reuben chips have between two to 10 times more compute efficiency. to produce more computes per watt. This is ultimately their core mode against peers. And then beyond that, beyond the GPUs, they're building products like NVLink, which is um a product that for instance Amazon uses where they can use their own A6 chip to plug into that um architecture of of Nvidia um so that it works more efficiently, it's connected to the rack, etc. So even though they're plugging in other chips in the data centers, Nvidia still makes revenue because they have that product. And then beyond that, they're moving into CPUs. They're moving into storage. So they're opening up new markets to get incremental revenue in. And I think this is kind of the the broader TLDDR of of of yesterday video. The biggest issue I have with Nvidia is if you look at the chart over the last two months, this was the first slide of yesterday, it performed round about 67%. 65% of that total return came out of three months. If you take those three months out, the stock is flat over two years. And this is basically the fear of investors that Nvidia cannot keep growing at the rate it does between 30 to 40% EPS growth, right? And as long as we cannot prove the AI ROI, investors will always think that we're overbuilding capacity and that we do not need as much compute. And if that's not going away, I think the Nvidia multiple will be compressed at the same time saying it's at the very lowest point in the last two to three years in terms of the multiple, which I think is a great uh entry opportunity here. Okay, so we got earnings tomorrow. As we wrap up, what do you guys think? Are we got to answer >> numbers good or bad? Price up or down? >> Price is gonna be flat like they always do or like flat to sell off. No, seriously, this is like no matter how much they numbers they put up. It always seems to be the case. Hopefully this time around it's different. I'm expecting some phenomenal numbers out of them for sure and great guidance. >> Yeah, I have pretty aligned views. The funny thing is every time we're approaching end of earnings quarter, we're saying the same thing. Nvidia needs to save the show. It's it's every quarter it's the same, right? This time again, the sentiment of the infrade is not good at the moment, right? Crypto is yay. Semis are nay at the moment. It feels like >> I bet you that'll flip by like next week. Okay. Like >> Yeah. Yeah. If Nvidia if Nvidia comes out and I think you're right on that Melvin guidance will really really important and not only the headline numbers I think the commentary leadership commentary of Jensen will be really important and he can save the show once again the good thing is he he's really aware of that he's the king of the eye trade he understands his broader responsibility if you look at his performances in the earnings report I think he did a really good job and yeah in terms of stock performance I also think it's flat to down Well, it doesn't need to be black or white. It's not crypto or or or AI infra. It's the everything bull market, guys. Come on. It's the everything bull. I've been telling you this for two months now. What are you not understanding? We're going up. We're going higher. Uh Nvidia is going to crush their earnings. I have no idea about their price in the short term, but I'm definitely a buyer here. Um I think it's way too cheap. Makes no sense. >> Excellent. My only prediction is that Melvin's going to top Last Bitcoin at some point in the next 6 months. I don't know when. I don't know what the price will be. I'm just telling you right now. He's or or he'll find some other obscure Melvin style thing. Like it's something that we're going to be like, "No, don't buy that Melvin." He'll be like, "I need to get this." And it's going to be the most random token. Anyways, that's that's not going to be on the show. Uh >> going to go for the he's going to go for the butthole meme meme coin. >> Don't you start with the DJ talk. All right. Uh anyways, guys, uh great show. Thank you for the thoughts. We'll see what happens. By the time people are listening to this, we'll know probably or at least we'll be just hours away from knowing what happens with Nvidia. Otherwise, um we'll see you guys again next week. >> Thank you, guys. >> Thank you. >> Thanks for listening to Milk Road. If you enjoyed the show, make sure you like and subscribe. And if you're struggling to find winners in the market, that's exactly what Milkroad Pro is built for. Our analysts have called some of the biggest winners early, and Pro lets you see what they're buying next, every trade they make, and the research behind every position. Check out Milkroad Pro at the link below. Everything you hear on Milkroad is forformational purposes only. These are our personal opinions, not financial advice, and we may own some of the investments we talk about. Always do your own research and make the decisions that are right for you. See you next time.