Transcription
Right, let's get right to it. Um, again, you guys can read through it, but it was a big week for stuff, at least in things of uh I'm looking at uh I think you'll uh you get a lot from this one.
First of all, starts the week off. The next bubble in AI. So, we've got Sam Alman believes we're in an AI bubble. We've got Meta AI reality check. Howard Marx bubble. MIT failed to boost revenues. Palanteer tumbles after short selling and the AI winters upon us. All right, so I'll just keep saying this. This stuff is all stupid. Um, there's a truth involved in it in terms of the fact that there is speculation going on. There is a valuation issue, but I just want you guys to remember there's a difference between certain stocks being overvalued or building in way too much of the future, which I do believe is happening to some degree, um, and saying that AI is a bubble. This is the very early innings and as you get through this we're not even into the second stage which I think is the longest stage which I've highlighted before which is when the digital employees and the robotic employees start to come in and both of those will happen next year and so as you go through this that's the important point is use these AI bubble talks as opportunities for you guys to find things. I do believe this is a regime shift. You're going to hear this point again and again. and I've talked about it. I do believe we are at a very important macro regime shift. So for all the macro people out there, uh this thing is really more about the regime shift than it is about AI in general.
All right, so let's continue. Uh here's the positive side. So Goldman comes back from a trip. I think the important thing to look at is that the acceleration in Gen AI adoption is starting in 2026. We haven't even seen the adoption yet. That's the main point. Um, all we've had is the infrastructure buildout. That is stage one. Stage two is the adoption side. This is when we start getting people actually using it. So those stories you saw, there's a truth in it. It has not been adopted yet. And the reason it hasn't been adopted yet is because people are lazy. Because people have trouble with the culture of the firms to get their people to use it. The easiest way to get people to use AI is to have AI use it as opposed to humans use it.
The pressure on the Fed continued this week. Um, I won't go through all these stories. Let's just say whether it's uh mortgage fraud, whether it's he's late, whether I don't even care. Um it's non-stop. The pressure just continues. Zervos was on there saying the Fed's not independent and saying PAL is left-leaning. Uh and we get Pal flinching. So speaks of Jackson Hole. um you know you can go through and read whatever you want but the main thing is there's definitely a recognition that the labor market is weak and somewhere behind there whatever it is uh if the risks to employment are rising while inflation is believed to be temporary there's your justification for the rate cut in September which is only at 80 you know 80ish% coming out so it's not like it was a complete pivot um but basically this is paladopting the Chris Waller thesis, the leading candidate for the Fed chair. Uh I wrote a paper on this this week. Um regardless of what happens in September because we still have some more data for all the bears out there that want to believe that inflation is going higher and this and that, you are fading AI. And again, this whole paper, the academic Fed versus the inflation target of the future, a secular regime shift, however, is far more powerful. It usually reflects dramatic changes in monetary policy. That's what we're seeing, which is the the uh desire of this administration to run it hot. The reason they want to run it hot is partly because of the fiscal policy spend which is going towards AI. They've got all these artificial intelligence people behind there. So for everyone in X posting that they should be focusing on inflation, you are being academic. you're focusing on the today, the CPI today, the PPI today. The fact of the matter is we had five CPIs that came in uh below expectations. Then we had one that came in line and then we had a PPI that came above and everyone's freaking out because the tariff stuff is finally showing up. When you have a monetary policy shift, which we're seeing, which is they want to run it hot, you have a fiscal policy. They are not only spending money on the AI, they are focusing on so many different things on the fiscal side and not caring about the debt and deficit the way that people had thought uh not that long ago and you have a massive innovation shift. That's when you get regime shifts. I go through it in this I compare it to 2007 when we had the same type of situation. At that point it was the software boom. Besson argued that the US may be entering an environment reminiscent of under greenspan when rapid technological innovation allowed the Fed to run the economy very hot without sparking inflation as digital employees begin to replace human workers across industries that is beginning next year. Deflationary pressures from automation will be powerful and persistent. You have to be doing the Wayne Gretzky going to where the puck is to be an academic Fed and just be doing what you think based on today's data data dependent makes no sense in terms of a world where there's exponential change.
The labor market is weak. I've highlighted it. There's no way you can look at this which is basically the service side, the office job then no job creation over the course of the last two and a half years. I've highlighted all these different stats. We have negative X healthcare. You have the diffusion index below 50, which historically is a recession. And you have if the labor participation rate had not fallen the last three months, the unemployment rate would be 4.9. We wouldn't even be having this discussion. We'd be going 50 if the labor participation rate hadn't gone down. And again, here's the last three months. And here is just the health care side. So negative ex healthcare extremely easy to see there. They should be cutting rates uh for people who don't believe it. Have fun. Uh the revisions massively lower and again the BLS is set to revise down US job numbers by 550 to 950,000. This is what started the SOM rule uh added to the SOM rule panic last year. This is coming according to Goldman on September 9th. If we're going to get another revision, that'd be the biggest 12-month downward revision in 15 years. Again, they have the reason to cut. And oh by the way, whether it was in the teback from the most recent quarter, if rates don't come down, you have a problem globally. Central banks are undergoing pressure to keep interest rates artificially to offset the cost of record borrowing. You're missing the debasement side. They have no choice in this element. They are not fighting inflation as much as they are fighting so many other problems that are acrewing every day.
S&P gets back to new all-time highs after that little AI bubble fall. banks I mentioned globally they're going BKX right at the highs closed just a point below it uh a reflationary boom uh story uh sectors what you'll notice here energy best performing sector tech worst performing sector this is the rotation in the regime shift that I expect to happen as the time by the time you get through this entire presentation there are many many reasons for everyone saying that if they cut rates the big thing would be to short bonds. Yep. Continually the worst trade people can possibly do in an AI boom. So if your if your thought is they cut rates, they move 25, bonds are going to fall. They should have fallen on Friday. They didn't. Um I highlighted this last week. This is the or two weeks ago. This is the momentum in industrials. This just highlights this is the best performers in industrials. And then you can see the other half is horrible. These are related to the consumption economy, the auto economy, the real estate, whether it's residential or uh commercial. And then everything here is AI. If we are in a regime shift and rates are going to start to move lower, you should start to see these do better. And that's what we saw this week. This is the um industrial momentum. It's inverted. So we saw a sharp move the other direction after peaking. Uh this is overlaid with the PMI, the manufacturing PMI, which I will get into. Uh so again, another phrase that if you're going to have this move higher, that would mean that the things that are doing poorly start to do well as well the things that have been doing well. This is a tide that lifts all boats. The PMI is a diffusion index. You should start to see that go higher. We'll see what happens. momentum on the broader level. It was a a down week for momentum as well. Uh I just want to highlight these again. I've shown this chart multiple times. Regime shifts are critical. They happen for monetary policy for fiscal policy. This is the regime shift I I highlighted with software QE and austerity. Uh this was coming out of the great financial crisis. Then you had this one in in here related to uh the shift from oil when oil hits uh hit $25 and then we got the shift. Uh this one was going into COVID. This one was coming out. This was the shift in terms of the Fed pivot back in 22. This looked and to me was the beginning of the next shift. But then we had the tariffs which set us back higher. I still believe this is going to come. Momentum will shift. And if it does shift, you're talking about a regime shift that will have implications. It is always driven by the things that I mentioned. Uh in terms of the factors, size, I highlighted this last week. A 100% composite score, factor strength of 3.53. Here's momentum here. The two best factors. Uh I expect that size and momentum will be the ones that shift in a regime shift. And this week, the size factor, the pure size factor was down all five days this week. So regardless of whether it was the a the AI bubble or uh the fears over the Fed not cutting, then we got it built back in. So when you can have five days in a row of size under pressure, that says there's something going on. And this was the biggest weekly fall in size since right around the SAM rule back then when the rate cut started to be built in. So I do believe that size is going to be under pressure. small cap index. The S&P small cap index had its biggest day since the bottom in the stock market bacon based uh back in April and also the election back here which led to a rally. So when small caps have a big move like this, it's usually a turning point of some side and it comes at a time when they're finally having earnings. So small caps are beating by the widest margin since Q122. Uh we got all the press on how good the earnings season was, but the earnings season beats for a lot of these were not out of the ordinary. But for small caps, which should have been hurt the most by the tariffs, they still beat by I mean it's a huge be. Uh the positioning in Russell futures net short by the biggest amount, even lower than the amount in 2022 when rate hikes were going on. positions there. And John Rog, if you guys uh you can see John Rog's email here, I highly recommend reaching out to him. John and I have known each other a long time. I've shown him uh his work on here a couple times, but right now because of the regime shift side, he happens to be on the same side out of a lot of things. And John and I do not do anything together. I don't I use charts on my own. I'm an Elliot wave person. But what John is great at is these big bases, which usually when there's a turn happen at regime shifts. So, in terms of where the Russell 2000 is, he's bullish on it and he sees a potential big base. So, a breakout above here would have to happen in my opinion for a secular side. He's positive on it. I like it at that point. And it matches up because that big base coincides with the PMI sitting below 50. And again, for those of you who don't understand the PMI below 50, think of it as uh it's a survey. It has five components that make up the the actual number. And if we're below 50, think of it as there's not more than 50% of the economy doing well. That typically happens during recession. It's a diffusion index. It's very similar to the industries on the job side that I showed. We've been in this for a long time. And I actually believe that the entire time in here, there was no cyclical upswing uh that was durable. These were all the reversals of negative situations. This one's going to be different to me. I think it's going to last and it's going to look more like this period in here when we stayed above 50 for a long time. In this case, we stayed below here for a long time. Somewhat similar to what went on in here. So, just keep that in mind as we go through because as I've talked about, um, PMIs, the S&P Global released their flash numbers, best in more than three years in the US. Uh companies across manufacturing services are reporting stronger demand conditions but are struggling to meet sales growth causing backlogs of work to rise at a pace not seen since the p pandemic related capacity constraints recorded in 22. So last time we saw a supply demand situation like this was during during the the point where people weren't working around the globe. Uh, a lot of this is going to have to do with tariffs, but the bulk of this is going to have to happen because of AI. Combined with the upturn in business activity and hiring, the rise in prices signaled by the survey puts the PMI data more into rate hiking rather than cutting territory according to the historical relationship. I bring that up because again, they want to run it hot. You're going to have a time where PMIs are rising above 50 and the Fed is cutting rates. not going to happen too often and I think people need to pay attention and start to change their positioning based on that. Focus on small caps, focus on some of the lagards and some of the things that are rate sensitive because I think the Fed has or the administration has more surprises to help in the housing market and that would just be the final tailwind. So for people that have been bearish on PMIs have pushed back on me continuously on this, it's not just in the US. The Euro zone composite PMI rose 51, strongest private expansion since May of 24, highest in 38 months for the manufacturing component. Japan quickest pace in six months. India fastest activity in at least two decades. Prices rise. Australia good proxy for China solid 54.9. Morgan Stanley revised up their global capbook significantly. You're talking now well over a trillion dollars over the course of the next two years. Uh the data center power demand double now. They've moved everything up. This number here used to be about 95 or 92 according to Eric Schmidt. We're now up 15 20%. Uh meeting 100 gawatts which is less than that number would require a thousand natural gas turbines. Not going to go through this. You guys understand um everyone who's been bearish on China and I mean God when they came out with this stimulus here everyone was bearish on it consolidated in a beautiful chart formation. This is called a third wave folks. This is a parabolic move breaking out of a 10-year base. I'm not going to show the John Rog side but he's got a 10-year base on this too. This is not something to fade. And yet most people are still negative on China. China stock index rises to decade high as they're dumping bonds. Turnover on the mainland exchanges reached second highest level ever. So what's happening on big volume? 30-year bond. When the stock market was rising, I kept hearing that bond yields are not rising. This is something the fade activity is not happening. Wrong, wrong, wrong. China's stock rally has the makings of a durable bull run. Uh FOMO is setting in now. Finally, the scariest aspect of the ongoing share rally is that it has unfolded organically despite widespread skepticism in the wellrecoognized lack of meaningful policy support. So, everyone thought they needed more policy. And I've highlighted before that IP is well above the old traditional signals. This is a new world of AI. This is a new world of drones, of humanoids. It has nothing to do with what you've learned in the past. There will never be another property boom in China. Go read a new book. That being said, all the stuff out there in stimulus that's come out, this is all this week, loans, focus on property market, focus on property market, focus on property market, they are trying to get the stabilize the prop property market. It's never going to rise. And if you keep looking at it in a big way like what's coming, by the time that comes, they'll have runaway inflation. They'll have all kinds of problems. They don't want that. This is a new era. All right? And the new era we're talking about the regime shift. I wrote this in April from cloud so software to embodied AI. This is what we are now entering. This is the part that you guys this is a long uh paper I wrote back in there uh for 22V. We stand at the brink of a new era. One that may prove even more transformative. The rise in AI marks a turning point not only for how software is built and deployed but how it how in machines interact with the physical world. As Alain must put it, AI is the most profound technology humanity will ever develop. Bubble more profound than electricity or fire. If software ate the world, AI is poised to rebuild it. Rebuild the entire world because of sticking intelligence into machines. This has to be understood as I go through the rest of this because we are at that critical point. The convergence of AI and hardware advancements is unfolding against the backdrop of a profound macro regime shift driven by simultaneous changes in fiscal, monetary, and innovative forces. Again, I wrote this in April. We are now in the stage where every single one of them is in play. And here's what you need to start to focus on. Uh I've owned a Tesla or multiple Teslas since 2012. So I love the car. Could care less about the stock historically. This is the first time I've ever talked about it in any meaningful way. the robo taxis you have to understand I I'm this paper is not out uh I'm just showing you that I'm doing a long form paper specifically on robo taxis to highlight the importance the opening act of embodied AI if you guys haven't reached out to 22V yet and you're interested in Tesla reach out now this mustwatch this is from Adam Jonas who is probably my favorite stock analyst ever in the history of my involvement in the markets uh and auto analyst who took the same path I did about the same time to really understand how much technology was changing the old industrial world and how the economy was no longer representative of what it used to be and it was around the same time. So Adam put out this video. It's about 22 minutes. I highly recommend it. The robots are coming. You'll learn a lot. AI is about to get physical is where you'll find it in YouTube. The highlights are unbelievable. Embodied AI is a decat trillion TAM. So when people are fading AI and they're saying they're never going to get the revenues, they may not get on the software side, but his focus is how we're going to get it on the physical side. This is important for PMIs. This is important for macro people. This is important for everyone who needs to understand how the economy is going to change. You will have humanoids walking around you on the street in four years. You'll have flying cars. How can you not be paying attention to this? Go watch this for 22 minutes and at least get an understanding. There's a generational TAM expansion beyond software into robotics, logistics, mobility, and defense. The race for photons. This is critical in understanding the robo taxis. That's why I have this little blue line here. Companies with the best vision actuation data sets will dominate embodied AI. We are leaving software and we get it. We're getting into the point. This is about vision. It's about photons. This is what Adam does a great job of highlighting which I haven't seen in anything else. This is the next hyperscaler, capex, track robotics, supply chains, sensors, actuators, batteries, semis. This is all manufacturing. Tesla's advantage and their moat is manufacturing is the key. You have to think about that when you realize all the AI action plan and everything we're doing. Um, it's not an auto company. It's the first vertically integrated embodied AI company positioned for humanoids and autonomy. Multi- trillion upside. This is why you have to go pay attention to it. Robo taxi equals one of the biggest monetization opportunities in history. Think ad models, entertainment, productivity unlocks. This is what Adam Jonas highlights. These are all the things from the video. This is why watching it is important. China makes more drones in one day than the US does in one year. Hence the need for us to build up our manufacturing because it's a national security in for defense budgets. No way it's going to stop. Humanoids are for and $40 trillion labor market. This is the impact. This is why profit margins will sore as we get through digital employees and robotics. Um, manufacturing decline reversal. Manufacturing dropped from 30% to 10% over 80 years. Macro regime shift. He's calling for the exact same thing. Embodied AI position to reverse this trend. This is industrial policy plus productivity super cycle opportunities and semis. Robotics component supply all of this stuff. You can go read it all. This is not something that won't happen. It is here. and robo taxis are the first place where it's going to show up. Jensen Yuan talked about it and he said there was a chat GPT moment for robotics coming soon. That would be this moment in my opinion with um robotics. If you want to go listen to a podcast, this is uh from Gavin Baker. This is on invest like the best. This is a year ago where he was talking about the early stages of LLMs, Nvidia, reasons that they would have a little bit of trouble. Gavin Bacon is one of the the be Gavin Baker is one of the best out there in my opinion on what has been happening. He's been studying this for a while. It's very hard to find people who understand Tesla from both a technology perspective but also from the uh the embodied AI perspective. An auto analyst is not going to get that for you unless it's Adam Jonas. Here's one of the quotes from there and the thing you have to understand. So so you have to remember for Tesla Tesla is going to the same miles between disengagements. It's like if you ever built a new cities on Mars and it was populated by entirely different looking cars and streets and everything, you could drop a Tesla robo taxi in that city and it would have the same miles between disengagements in any other city. Where something like Whimo, which you may have been in, Whimo has more cars than Tesla, but their technology and approach is completely different. It's geoenced, which means it's using maps and sensors as opposed to what Tesla robo taxi is, which is sticking a brain in the car to learn as it goes. Tesla has launched its robo taxi. Now, what that was an arc research thing. You can go read it from this week. Tesla secures Texas permit for driver. Now, what I'm getting into is what it's important for investors and how quickly this can scale. He is seating cities cities across the country with little amounts of cars 10 to 20. That's why people will fade it once it starts to work. You will be able to extrapolate very quickly. So we are at a point now where all the signs are showing up that he's ready for this now. So he gets the permit in Texas. The Tesla robo taxi service area is already larger than any competitors in Austin and Bay Area. Again, different technology, different approach. an analyst went in it this week and there's been plenty of them that have gone through it because remember this was only released in June. So this is now you're looking for signs as to when this will start to take off. Uh we experienced a glimpse of the future and exciting the comparisons are meeting in Stark. I won't read the whole thing. You guys can go read it. It was on TV. William Blair analyst is impressed by the rides felt a lot more like person was driving it while rides still feel somewhat robotic. different technologies. Version 14 of Tesla self-driving feels sentient. Version 14 is not out yet. He's saying it's coming out soon. So that's another level. He's been testing this. Italy, Netherlands, Japan, Spain, UK, Germany. You can extrapolate again and get how big this can be. If you heard about this signing, a 16.5 billion dollar deal with Samsung this month, he posted, "The strategic importance of this is hard to overstate. This is the company's bet on an all-in one chip design that can scale from powering its driver assistance system known as FSD to Tesla's Optimus hum." This deal is because he expects this to accelerate, which should not be a surprise, but here's going uh he's been posting all week. This was from this week. This is him saying that last year. This will change the look of the roads. Robo trucking is also part of this. You should go watch this video from person who does next big future to go see that trucks, semi-truckss will be on here. Just think about what this does in terms of cost deflation coming both from not needing people and being able to drive down prices, but from competition. And Musk put out his warning. If they don't exit their short position before Tesla reaches autonomy at scale, they will be obliterated. There's no doubt about that. Uh John Rog, here's his second one. Again, big base. What's happened on the Tesla big base? This is what really started the timing of this. I use a fundamental side, which I already had all of the things. I had already heard the Gavin Bacon interview, all of this stuff. And then John says out of nowhere, I really think Tesla's going to start to break out. Okay. I believe charts give you an insight into matching up. And then when you go do the work, you put it in the bigger picture. When you put in rate cuts, when you put in a a a regime shift and you put all these things and you start searching and Tesla has underperformed, it's down 20%. Elon Musk is hated by Republicans. He's hated by Democrats at this point. He's hated around the globe. And yet somehow the chart looks good. another technician who I I have on my ex post who does different work than John, more short-term oriented, more trading oriented. This is the start of a coming technical breakout in my view. Today's move to six day highs exceeded the minor downtrend, blah blah blah. You get the point. I've got technicians all bringing up the same thing. At the same time, if you guys aren't familiar that humanoids are sitting here, then you should go watch the world humanoid robot highlights. Yes, the Olympia the Olympiad of robots was in Beijing. They wanted to show off how this is going. And of course, you know, the Chinese won the top medal in most of the world's humanoid robot games. But you can go see that this happened and we're at this point now. Okay.
So, I'm positive on AI. That being said, uh this is something I think people should read and this gets into some of the issues and the major issues I have with the Mag 7. So again, I'm talking about a major regime shift away from the software dominated world that has been in place for 18 years and into a new world completely dominated by the hardware side. If you are long software companies, as I highlighted last week, you are exposed to coding being free and being able to replicate everything on your hand. So this is from uh Harris Cuppupperman. He goes by Cuppy on X. Uh I saw this uh and I read it and I actually liked it and I think you guys should read it too. Um sometimes a hyperrowth company amazed me when it actually grows into its valuation. That's though that's rarer than you think. Usually cash flow is king. ROIC is the queen and everything else is simply stock promotion. Hence hence the skeptism I heard towards anything new. Okay, I want to bring this up because normally I would say okay this is just a perma bear. He uses AI a lot and gets huge value from it. I'm not here to belittle AI. It's the future and I recognize we're just scratching the surface in terms of what it can do. Okay, this brings back credibility immediately and I want to read the rest of the report. I recognize all this. I also recognize massive capital misallocation when I see it. I recognize an insanity bubble and I recognize hubris. This is the bubble that I agree with. Massive capital misallocation. It's not that they're misallocating. They're chasing something which I don't think they're going to get the benefits from. I recognize an insanity bubble. Okay. Whether it's the spending or whether it's people that are are buying these types of names at regardless of the valuation. And I get hubris. I think we're I think those are all fair comments. Now, this is one of the things that got people all freaked out. A hiring freeze at Meta AI operations in effect after investors voiced concerns about the scale of Meta stockbased compensation costs. So, it sounds like investors, board members, whatever. basically forced Meta to come out with a statement saying they're not going to be hiring people for $100 million anymore. So, they made their big bet. Okay, it's not a bearish thing for Meta overall, but it does show that the capex side, the spending side is becoming an issue and if they don't start seeing the returns, you're going to have problems with these stocks. So, the rest of what I want to show here, so Jim Chainos came out with this. I kind of highlighted it. this is getting into on its capital base. The depreciation of this um the accounting that's going on on this is just going to put more pressure. This is the critical detail that nobody can unravel in the AI trade. Free cash flow has turned negative for Amazon and Oracle and declined for Alphabet Meta amid heavy AI spending. Again, they're spending money to get something in return. This worked with the cloud. The question is, is it going to work now? I think this is a much bigger problem for them. A because of how much money they're spending and secondly because of this thing that software is going to already be eating away at some of their businesses and I don't know how they're going to get this. So it seems like it's all related to AGI which is some mythical thing that we're not sure how you can monetize. The Mag 7 tech giants outspend rest of corporate America 2 to1 on capex. Again this is a massive bet that becomes very important for the market. uh mag seven value to free cash flow. Now again, it worked in here. The question is, will it work expanding this? Now, I'm gonna say no. And here's the problem. There's no more room for for error. The Magnificent 7s EUR has outpested the rest of the S&P by more than 20 points for five consecutive quarters. So, here's the problem. It's outperformed in terms of earnings significantly, but it's underperformed the market. So, you have a divergence. I'd be worried if I were people overweight these things. John Rogue back again highlighting this little point here. Here's here's the Mag 7. Uh again, on a relative basis, it's down for the year. So, it hasn't worked. I believe this is a problem you need to start paying attention to. And the reason is here's another Alan Mus pose from this week. It's an easy prediction of where things are headed. No more traditional operating system or apps, but just AI rendering everything directly. This is another critical thing for people to spend time on on learning. You can go into chat GPT take that that exact quote and then bring it in. But here's the details. They envision a future where smartphones function as AIdriven engine instead of routing every task through a cloud service. The phone itself would act as an intelligent rendering engine. So I just want you to think about this. If you've been wondering why we haven't had AI on a phone yet, but we have it on our computer. This is really getting into it. So I'm using AI on my phone. I'm using chat GPT and I'm talking to it. But that's not actually using it to to do what I want it to do. Hey, go order me an Uber. Hey, go do this. Hey, go do that. It doesn't work well. And the reason is the bandwidth constraints make pure cloud AI impractical. So what he's talking about if realized the architect the architecture of having a phone that has AI on it. If you want an app, you just build it yourself. That's what Alon Mus is talking about. And without going through all the details, Mus AI is betting that the future of software isn't apps at all. It's AI native devices powered by a blend of massive centralized training and decentralized edge inference. This is very similar to what he is doing in terms of the car versus Whimo. AI AI just becomes part of the machine rather than having to use the apps on the machine. When you use an AI system like chat GBT or GR today, most of the heavy lifting happens in giant data centers and your device just streams the results back. So you're doing the work off the machine. That's fine for text, but when AI start generating full experiences in real time, the amount of data that's needed to flow between your phone and the cloud becomes enormous. Now, before I read this, just remember massive regime shift from software to hardware emphasized again here. The economics and physics of networks can't keep up with billions of people trying to stream full AI generated environments simultaneously. Again, you're going to need batteries. You're going to need semiconductors. Instead of doing it on the cloud and in the data centers, it's going to have to happen in each device, whether it's a humanoid, whether it's a car, whether it's a computer, whether it's a phone. The buildout of the hardware is massive. Do not fade it. The next phase of AI. So, I went through so this is about hardware moving forward. Next phase of AI isn't just about clever. It's about hardware and infrastructure. The training was in the cloud. That's phase one. But inference running the AI to generate answers, visuals or actions needs to happen locally on the devices, phones, cars, robots because of the bandwidth in them. If AI agents replace apps, then the real mode isn't the app store anymore. It's who controls the hardware layer and who can manufacture them at scale. Moving forward, the bottleneck and battleground is hardware, compute capacity, chip design, and adapt people. And again, this means I'm negative on the Mag 7. Not for them to go down, but for them to have trouble going up and to underperform. They will still be having earnings, but they are valued at a level that has built it that has not built in the risk of them not getting enough return rather than the capex. Think of them as bleeding rather than gushing.
Now, this leads us finally to Bitcoin, which is still the chart is fantastic. Again, it's consolidating up here. came to the bottom end of the range and then when Palism Smoke it got back up to 116 which is where it is this morning on Saturday. Um Ethereum made new all-time highs. Another big base. John Rog likes these as well. I'm just not including the chart. My favorite chart. If the Mag 7 are going to underperform, guess what? Bitcoin is the large cap that will start to win. I believe this happens into the end of the year and you start to see Bitcoin as I said it would be the second largest uh asset in the world. It's going to be hard to do. I'll stick with it because I did think we were going to get there. But regardless, this is it relative to the Mag 7. This is the trade that I think is the most important. And just two pieces of news. The Fed talked a lot about payments. Uh cryp changes coming to how the Fed views AI and crypto. Very important especially since again we have a regime shift. So the Fed's going to be more involved in this. And finally, uh Brian Armstrong said we'll see a million Bitcoin by 2030.
One more week of me in Maine. Hope you guys have had a great summer as we head into the final week. Uh, lot of stuff on here that I think is important. Reach out if you have anything and remember 22V research. I'll be doing the Tesla report there. Have a great final week of the.