Transcription
To it. Um, I'm not going to read all these. Uh, we'll go through it, but I do want to make the point that uh, this week's video is mainly again uh, about AI and stable coins, but the most important thing is it's I'm I'm going to go through how this is the exponential point and how a lot of the other things that people are talking about and going through now—that the China-US situation, in my opinion, as I said last week, is officially over—and we are now in a, whether you want to call it a truce, uh, a separation, as I called it, with Pomp. Uh, it doesn't really matter. China and the US have come to some kind of agreement that they have to live with each other while they're figuring out how to get this divorce done. So, they're in mediation at this point.
Um, so I'm going to focus a lot on just the places that I think people should be focused on because the economy is going to become more confusing, not less. Uh, S&P, we're basically just below the highs. We obviously saw a little bit of a correction on the back of the Iran uh, Israel situation. Uh, best guess here's what we did last year during the Somu panic when we sold off. Uh, see that three-wave move in Elliot wave term. Same thing here. Impulsive move higher. We did a correction. I would not be surprised to see us kind of get back into this area. Maybe even fill the gap here. So, I'm expecting uh, some some bit of a correction here. Uh, we should be getting into the blackout period for buybacks and things like that as we get closer to the earnings report. So, it would not surprise me to see us spend time before we get through the all-time highs.
Uh, the S&P last week—a lot went on—China-US and then uh, you had the fears about Iran about Israel uh, in Iran, and then we had the reality of it, and the S&P basically was down 39 basis points for the week. So again, a lot of drama both on the positive side and the negative side, and nothing came of it. Small caps broke out this reverse head and shoulders formation, got to the 200-day, and then pulled back. Uh, I still think that's a place to watch just to see if we get any kind of expansion.
Uh, the big story to me continues to be that no matter what happens, sentiment is bearish, everyone's bearish, positioning is bearish, but the dollar does not bounce. There is clearly more and more hedging uh, going on. And even though we're going to have a bounce at some point and maybe we consolidate here, you have to start looking at—if you can't rally on anything, stock market falls, dollar falls. Stock market rallies, dollar falls. We have a geopolitical event with a a war and the dollar doesn't budge. Uh, 30-year auction. This is what's going to happen. And this is why I keep saying the most frustrating thing for macro people is continuing to play the bond market as the easy trade. It is not the easy trade. It will never be the easy trade.
Um, Morgan Stanley, who's had a good call this year. Mike Wilson's had a good call. You know, really did make the argument to be buying stocks down in the near the 5,000 level. Um, now saying there's a new bull case for stocks that's emerging. Completely agree with as I go through this. Uh, but I'm not going to call it a bull case. I'm just going to call it this messed up economy and messed up macro world that is all about artificial intelligence in crypto when very few people actually understand artificial intelligence in crypto, which is why I do this video every week for you guys.
Um, earnings revisions are heading higher now. So, you've got revisions breath going down. Last time we saw a bounce like this, not surprisingly, this was at the end of 22 um, when we came out of the rate hike cycle. So again, people scared about tariffs going through their usual doom and gloom stuff. And again, we have another doom and gloom thing that people can start to focus on. Uh, it's amazing that if you go to Bloomberg Saturday and you go look, the top five stories are all on this—um, macro is now going to be dominated by this. Just to go through AI. Historically, over the last 25 years, when Israel involved in a conflict of war, how long does it normally take for markets to stabilize—this is why you don't have to read the papers. You don't have to do anything. Just go to artificial intelligence and ask it the question on everything in your life. I can't make this any more clear than I do in these videos.
Uh, historically, markets tend to stabilize quickly after the onset of Israel-related conflicts, usually within days to a few weeks. Are there scenarios that it could be worse? Yes. But we're not in the game of focusing on the tails. We're in the game of focusing on the probability distribution. And you use history. The world typically isn't ending regardless of what goes on. We've been uh, Russia-Ukraine when it first came out. Uh, oil, the highest price for oil over the last 15 years occurred less than a week and a half after the invasion. And it's still going on, and oil prices went from a peak of 133ish all the way down to 55ish. Uh, so it does not pay to trade geopolitical events and sit there and waste money on puts and things like that.
Uh, I believe there will be consolidation here, but I do not believe that the China-Israel or the Iran-Israel situation is something to be bearish on. And by the time we get through the end of the week, my guess is there will be arguments as to why this could actually be bullish—uh, meaning maybe this is kind of another step forward in the Middle East getting through a lot of the issues. Uh, I just don't think anyone knows, but I do know this—only focusing on the left tail of the distribution is a big mistake.
Uh, the China-US situation—new step forward—basically agreed what they agreed in Geneva, sort of uh, China found the world's pain point on trade and we'll use it again. Rare earth—I talked about it for two weeks. Uh, the papers finally got a hold of it. I think everyone started to reach out to me and wonder where I came up with this stuff and went through it. This is the beauty again of focusing on artificial intelligence and crypto as the bulk of your reading. Uh, for those of you that have read seven articles on Iran-Israel today, I'm going to tell you, you're not going to get anything useful out of that from a trading perspective because you're missing the stuff that's happening on AI, which I will go through.
Um, this is the interesting part that wasn't announced. So China agreed to give us rare earth. The question is what did we give them? Uh, certainly not just the visas and things that were announced. So a lot of speculation that on the semiconductor side there will be uh, certain semis that are announced that are going to go there. We'll see if that's the case.
Um, one of the things Scott Besson said, which I think is important as part of the mediation—China leaders want their economy to focus on more consumption while US leaders want their economy to focus more on production. The good thing is for for the global economy on this, this means both focused on growth. I'm going to reiterate this. Both China and the US to get out of this problem are focused on growth. End of story. So when you're looking at the tariffs about what go on, yes, whatever number you come up with on the tariffs, great. Good calculation. 300 billion. I don't really care. It's a small number when you compare it to what's going on in AI. But these guys are going to spend money on growth. That's where their focus is going to be—is the same exact thing. We need growth. We need negative real rates. We need whatever we want. The debasement continues. Spending money continues.
Uh, China vows more efforts to stabilize property market. This came out the two days after uh, the whatever went on in London. Uh, China taps $1.5 trillion fund to boost home market support. If you didn't read this, basically go read it. I mean, they're doing stuff similar to what we had to do in the US post uh, the great financial crisis. So, they're printing money.
Uh, inflation, this is becoming a bigger story. At at the end of last year and into the beginning of this year, I talked openly that at some point this year, there would be an inflation fear. Um, we've now reached a point with the tariffs that I think people have to start reassessing, including myself, what's going on in inflation because there's something else happening here. And I think it's related to AI and this is not some deflationary story on the service side. I think this is a deflationary story on what AI is doing to the jobs market in terms of fear, what the tariffs are doing in terms of fear and the spending that is being impacted from it. Not in terms of again a recession, but in terms of this bifurcated situation where if 50% of the people in the country feel like they're living in a recession, inflation is going to reflect that. It's not going to reflect the wealthy people.
Um, and so if the stock market is going higher and it's creating more wealth, no chance of a recession when you add in the the AI side, but I think the inflation stuff is starting to get in there. So Warren Pies, he basically, you know, he put out the core goods negative month over month. Core goods. We're already starting to see some of the tariff stuff. So whether it's appliances, there are little parts in here that are showing it, but in the aggregate it's not super corp—second lowest reading since last summer. 06 month—if not for tariffs, Fed would be cutting. I completely agree uh, that they would be cutting at this point despite you know, everything happening. The the economy is is uh, is fine, but it's only fine for a very small portion of Americans.
So, here's the super core number. We're at 2%. We kind of stabilized here. But when you go into the guts of things and you look at the month-over-month numbers, there's a couple things that stand out. First of all, if you just look at the last four months, you know, you're basically dealing with a very very small number. If you look at the change from last year, so look at the super core month uh, monthly numbers. So June was low. That's where we're entering next month. But then the July reading, the August reading, the September reading, the October reading, the November reading, the December reading, January, which is an outlier. Again, these two January numbers are big. But regardless of that, you've had this softness in here. And the biggest change to me is this thing here, which is effectively airline tickets. Airline prices are down 7.9% year-over-year. And again, I think this is an economy thing. I think this is people worried about jobs. I think this is people worried about the tariffs and the inflation. So, it's acting as a suppressant right now. And so, no matter where you look in the super core numbers, uh, this is the 12-month change. This gets you that 2.8 number. But no matter where you go, this is six-month, which is coming in right now at about 2.4% annualized. Then you go here, this is the four-month, which means this is coming in right now at about, you know, the last four months at about uh, less than 1%. So no matter where you go right now, the inflation stuff is weak. And go another route. We had 73 estimates for this uh, inflation number. The core number was lower than any of their forecasts. When you start seeing those kinds of anomaly, you have to shift it lower. And Goldman is doing the same thing. They say in light of the more limited tariff effect this month, offsetting deflationary maintenance is coming from forces.
So when you start thinking about how wrong people were on inflation, how wrong they were on the recession, the question is, is the economy weak, and I'm going to keep saying it over and over again, that the jobs market is not recessionary, but it is softening, which means wages are going to come down, which again is a negative inflation um, impact, but in particular, you're dealing with high net worth people are doing fine, which is a very small portion of the economy, but inflation is about the average person. It's not about the wealthiest person. And that's the way a lot of these indices are going to show up. And so the data is going to be there. This is here. Uh, if you can't see it, make your screen bigger. This is the services PMI which came back down near the bottom end. The claims four-week claims inverted. It continues to just soften. This is continuing claims which broke out and people are having trouble getting a job once they get on unemployment.
Um, and that's that's just what it is. We're not at any levels, and again, I'm going to keep saying it. We're not going into recession, just like I said last week, but I do think jobs are heading towards very low levels of close to zero. And healthcare and leisure and and travel are the only places that still are creating jobs and there's less travel going on. So, would not be surprising from for in my opinion to see the the jobs numbers keep tracking towards zero.
Uh, Trump obviously wants rates cut. He was out there again saying we need to cut them. But this time everybody joined. So JD Vance goes in and calls it monetary malpractice. Howard Lutnik went on there and said it should happen now. And Trump called Pal a numskull. So the pressure is on there. And Paul Tudtor Jones was on Bloomberg this week and he basically said Trump is gonna—we only have a we have less than a year left of Powell—and he's probably gonna appoint an uber dovish thing. He talked about exactly what I brought up which is they're they've they've made the decision to grow their way out, which means they're going to run inflation hot. He specifically mentioned what Japan is doing where Japan is reluctant to raise rates higher because when they tried to raise rates higher it created uh, some sort of a tantrum. So what they have is inflation in Japan is running way above where where rates are. So in the US, he wants rates below the inflation rate because that's the way they're going to try to get out of this problem. So uh, he spoke on this. I'll get more into what he said later on. Luke Groman, same thing. Put it in there. Uh, he was highlighting the fact that De Besson's already his 333. Um, they're trying to get it under four. So now he's mentioning four. Uh, again the same thing—significantly negative real rates for a sustained period or a brief period of face-peelingly negative real rates, which mean really juicing inflation uh, which uh, again you're you're dealing with a point where uh, they're going to have to do something to debase—foreign investors continue to focus and talk about this is becoming again this capital issue—this taxes on foreign investors—a proposal in the house version of the bill levy a 20% tax on foreign investors—I've talked about it on here every week, but it continues to be a news item because it isn't out of it yet. But you have another issue, too, which is this was on the front page or yeah, in the Financial Times this week, central banks are beginning to fret about dollar swap lines. This is another place where in the article, ECB has asked its banks to report vulnerabilities around their dollar exposures and the influential think tank has now called for US central banks to create a mutual pack to prepare for worst-case things. So, the non-US central banks, they're trying to figure out how to deal with this stuff. Uh, they're raising their purchases of gold. They're making deal swap deals with China. And then it brings up the Kindleberger trap. Uh, warning that turbulence erupts when a dominant geopolitical power loses the ability or desire to support reserve currency. That's where we are. I fully believe the American people voted to not have the global reserve currency. I think AI is going to worsen the situation for the voters, meaning it's going to hurt the jobs market. It's going to suppress wages at a time when inflation is high uh, higher. Uh, and this is going to continue to be an issue because inflation is going to be both good for the stock market and profit margins, but it's not going to be it's going to be at the expense of labor, which means this pressure that has been here and been causing the debt to go higher will continue to be there. That's why we're doing these videos.
Um, tenure rates did not act well on the Iran-Israel situation. Uh, tenure nor will rise today on Friday. Further reinforces that it's they're acting unusual. Again, just another form of this not being the same type of environment that it was in the past. Again, I highlighted the dollar. Here it is down here. Didn't move. Tenure rates. You can barely tell that there was any kind of a major conflict that erupted uh, with Israel and Iran. This was the big macro news of the week. If you guys didn't find it in your macro writeups and your cross-asset market, I'll take you through it. Oracle hits record after seeing dramatically higher sales. So let's go through what specifically was said. Go to Perplexity. Here are the growth rates. Okay, this is a half a trillion dollar company. These are the kinds of numbers that are not a recession. So I'm going to keep saying it. The US economy is growing nominally maybe at 4% this year. Maybe we can get back. We can we can stay around five, but let's assume it's 4 and a .5%. These are the types of numbers during a boom. This is what you want to invest in. This is why the stock was up 13% despite the fact that it's a half a trillion dollar company. As you go through all of these different parts of their business, all of them growing 100% greater growth expected. You you're dealing with something right now that people faded the AI thing not that long ago. Here are the words that Larry Ellison described on the demand. Astronomical and insatiable—highlighted that Oracle received an unprecedented order from a customer, one customer willing to take all available capacity regardless of location, Europe, Asia, or elsewhere, underscoring the breath and urgency of demand across the markets. Never again should people fade the AI demand story. As I go through this, I don't know how to make it more obvious to people. This is the story. It's not about bombs happening in the Middle East. Aside from the human element, it has nothing to do with the market over this at this point. If it's going to end the way that historically has ended, they need to build data centers, deploy infrastructure. It takes time. Adding 47 new multicloud data centers. The astronomical demand is driven primarily by enterprises rapidly adopting Oracle cloud infrastructure for AI workloads. This is just the beginning. I want to repeat—enterprises rapidly adopting. No enterprises are using AI yet. This is just starting. This only happens with inference. This demand is so large that customers are willing to consume all available capacity, prompting Oracle to take massive capital investments to expand data centers and network abilities. The numbers that they're doing and they're adding in—they said that they will be doing over 25 billion in capital expenditures for 2026 as to deal with the bottlenecks that are there. So I this is Oracle basically coming out again—half a trillion dollar company—been around a long time—go look at the chart—up 13 14% on f uh, up uh, on the day of the earnings—OpenAI who has obviously a major uh uh partner in Microsoft signing a cloud deal with Google for more compute—Microsoft is completely supply constrained when it comes to GPUs—they're growing—Open is growing so fast. And again, I've shown this every single week. They're adding Oracle to those compute partners. This is just nonstop in terms of seeing it. And it's global. Nvidia, we had Jensen Yuang uh, speak this week and he said 10-fold increase in Europe's AI computing power on the come. So, everyone globally, we've seen the Middle East, we're seeing Europe, this is not just a US story, this is a global story. This is why regardless of how weak the economy looks from the traditional measures, it is because of the distribution of wealth problem.
Uh, I wanted to highlight just Anna Nicole Yfski because uh, I've been highlighting semis. Um, she's obviously a well-respected person in the space, but I cared because she was also a negative person in the space. So over the course of the last two weeks uh, she's been highlighting—Microchip pre-announces positively—business performing better than expected. I'll tell I'll I'll describe why this is important. O semiconductor confirm seeing signs of recovery in analog. The reason this is important—this is the
Embodiment side. This is the industrial side. This is the auto side. This is all of the things that I've talked about is showing up now.
Uh, Micron, which I've highlighted in here, has ripped higher from, you know, it's doubled in the last month or two months—no, month and a half. Uh, their DRAM price is jumping after memory makers announced they will cease module production of the DDR4. So I will show you DRAM prices in a second, but the reason she's important again, she was on TV because she finished fourth in the in the Robin Hood Foundation behind Mark Allen, Bill Aman, Dr. Miller. She finished fourth. Her short pick was tech. So, this is someone that was negative, not short tech anymore. Long DRAM and analog semis. It's a big shift. Um, I don't think people expected it. I just wanted to highlight again another thing.
So, I showed you last week the um S&P 1500 semis. The white line here is the equal weight socks. So, equal weight uh semis. The reason I wanted to show this is because any kind of cap-weighted semi-index is going to have a very high percentage of it being Broadcom and Nvidia. The benefit with the equal weight is now you're getting to see more of the analog and the things that will benefit from the old traditional hardware story. The thing that I'm saying is now about to spike violently. That also matches up with the pink line here. This is the Cosby. So the Cosby overlaid with the semis. They all peaked. This is also the Taiwan stock exchange; things Taiwan semi, all of them peaked together. Asia's leading the way higher, but Korea in particular is busted out. I think of these as a supply chain story—everything is happening within the semiconductor spot price update. So again, DDR5, four, and these these are the commodities you need to know—not oil; these are the commodities you need to be focusing on. This is where we have massive shortages. This is like the colonization of China, which was happening from 2002 to 2008. Uh, this will be a repeat with the data centers. What Oracle showed you, think of it as the colonization of China. What do we need to own? What, as a hedge fund, should we be focused on?
Here are the DRAM prices. This is for the uh the five uh lap, the new one, just going higher. Oh yeah, I didn't show on that. I just want to make sure you guys can see this is all the price changes. This is now for June. So, it's actually advanced from where I just showed that 513 was the chart I just showed you. Here it is, here; every single one of them higher in the subsequent months. So again, 513, 514, now it's up at 570. This is from the automotive uh uh side. This is an S&P global mobility march tracker. I just want to highlight out. Uh, the automotive semiconductor market is undergoing structural change driven by evolving vehicle architectures. The embodiment of AI, this is a completely new thing. Uh, rising demand for electrification, automation, and connectivity. Hybrid vehicles are gaining share, uh, shifting towards high-level autonomous driving. This is where we are. We're going to have shortages. Um, and they mention it in here. So the semiconductor side is it, and what's going on. This is the SMH. This is SMH over the S&P. This is semis versus tech. This is the XLK. And this is the percentage of names above the 200-day moving average. No matter where you look right now, semis are in a beginning thing. And again, they are down from where they were in July of last year. So you're getting an acceleration here.
Why are we having this now? I listened to this interview. Uh, it's about two hours with Sundar Pichai and Lex Friedman. Uh, and he made the comment that I'm I'm a little mad I didn't um see this, but he made the comment that the tokens that Google processed uh themselves in the month of uh I guess it was April was 50 times higher than a year earlier. So, here's the growth. There's two things I want you to highlight. Number one, 50 times is a big number. This little inflection point here, this is two weeks after DeepSeek. So what DeepSeek ended up doing is forcing all of the US AI companies to release their reasoning models faster. So at the time he talked about Jevons' paradox, that what DeepSeek would do was get more people using reasoning, using inference because it was free, it was open source, more people were going to use it. Well, it not only helped in that manner, but it forced all of these companies to release things faster than they would have. So, this is the growth rate that's happening. It's parabolic. This is what you want to invest in. This is what you want to talk about. This is what you want to be on top of. Not whether oil goes from 68 to 73 and then back to 69. You're not going to make money trading that. Whether 10-year yields go from 438 to 452, this is the story. This is the digital economy.
So, what does that mean? Well, it's also across industry, and again this is now not using Google's. This is using from AlphaSense, putting stuff together. This goes further back. And again, right here, this is DeepSeek. Nothing had changed in terms of token consumption. It was still sitting in the same level. The LLM usage was all there. This was the moment inference kicked off and we got this parabolic move. We're about to see it go into machinery. This is just on a computer. This is before agents keep going. This is before the enterprise adoption. Go back to what I showed you in Oracle. This is the beginning of this. This is not the end of it. So, everyone who is worried, everybody I've met over the last 3 months through 22V who's a power person, we are now going to go through what I've talked about. It is time—is now.
So, now I took this snapshot that I took from from uh from X's post, and then I put it into ChatGPT and I said here's a chart. Please analyze this and give me an estimated electricity needs today and extrapolate this growth rate out next year. So, two things I want to show you. If you're not using AI like this, call up 22V, pay them some money. I'll come in and I'll go through and I will do a demo of how you guys can do this stuff instantaneously. Same thing with the Iran-Israel side. We are multimodal. The need to use these tools is astronomical in terms of the way that Larry Ellison said it. You have to be using it. So, here's the analysis. I just put it in—convert again tokens into electricity. It does the math. I didn't give it any of the numbers. It looked at the snapshot. It goes through the growth rate—50 times in 20 months, 3900% year-over-year. Assumption energy cost. Now, again, this is saying GPT3. And I'll go through why that's important later. Um, but here's the approximate tokens. This is the approximate energy use; that's comparable to the annual electricity usage of the Netherlands. Extrapolate that to May 26. So it now is doing this. It's again using GPT3 inference. I will go back and talk about why that may not be exactly the the right way to go, but regardless, that's more than half of global annual electricity production today. So this assumes that it's 50 times and it assumes that there's been no efficiency gains, which is not going to happen. They will—the inference models as they get released will continue to get more efficient and they will use less. So 50 times that's being there. So what's a realistic view? Again, I didn't ask it to do this. It did it on its own. A second 50 times jump is unlikely due to the physical infrastructure limits of GPU availability by Nvidia and electricity supply by power companies. But 5 to 10 further growth is plausible. AI agents would be part of the reason. Autonomous AI pipelines, widespread LLM usage via smartphones and enterprise. These are all things that we're preparing for an embodiment. This is just confirming it. This is not me telling it what to do. This is it doing it. So, it's saying using these more reasonable. And again, I'm not so sure that we're not going to far, far, far surpass that. The reason that this is more reasonable is they're getting back to forecasts that Goldman and McKenzie. I'm gonna say this is going to be higher—that whatever efficiency gains we get will probably not be there. The reason that we wouldn't be able to get to 50 though is because we just don't have the physical infrastructure. So this is where the buildout and the frontloading of GDP has to happen because this is a race at two levels. And I'll say it again. You have a race for AGI for the hyperscalers. They believe if they lose this, there's obsolescence. Their business is gone. I'll get into that. At the same time, they're in a race with China. This is a powerful thing where it's coming from free cash flow from the hyperscalers. And we have to do this in the race against China in terms of the way that the Trump administration believes it and the way that it's being—whether that's true or not, AGI is there. That is the reality is what is believed in. And that is why this is the economy. This is what matters. This is where you want to focus.
AI token usage is becoming a proxy for energy demand. The hyperscale energy footprint could exceed some countries in less than 18 months. Many main major beneficiaries: power producers, utilities, gas and nuclear players, and AI chip makers. Constraints will appear in the grid capacity situation—bubble. So if you don't believe that and you go, I believe there's hallucinations in there, I'll show you what you do. I use Gemini 2 and a half pro, the highest-rated model at this point. I put in what its results were and I said, "Do you agree with this analysis?" Again, this is how you eliminate not only hallucinations, but you get a second analyst that is equally as smart as the first analyst to give you their views of the situation. And this time, in short, I agree with the general conclusion mathematically, but with significant caveats regarding the... This is why you do this stuff. I'm not going to go read all of this, but what I am going to show you as you you get in the conclusion. While there is a mathematical error in the state of Earth growth, the core calculation is sound. The analysis hinges entirely on the validity of two key numbers, the projected tokens. And again, this is where it is. If AI usage continues to grow exponentially, its energy consumption could soon be on par with that of entire developed nations. It effectively highlights a major challenge for the AI industry and global energy infrastructure. You have to have major positions in energy at a time when energy is not built up. There's no movement going on. Amazon this week joins the party. We've—I every week I show a different—last week it was Meta with Constellation. When you see these companies just hoarding for it, EIA projects record US data center power. AI crypto because of uh AI crypto boom, uh BlackRock growing power man as policy managers rethinking the trade-off between... This is just a major topic at this point, and honestly, the energy space is not moving that much.
Now, on top of the embodiment side, on April 24th, I released this long-form deep research—uh, Colin Fenton, phenomenal works with 22V; we went through it. He wrote what glut and talked about it uh about—again, we've got issues—electrify and amplify; this was all on the connection between them. If you haven't seen it, you can go read it, uh, but I I would just go through it. So one thing I'm doing this week, I will be doing a uh some stuff on Chevron just to show. So can you go through all of the company commentary and earnings highlights on Chevron? I wanted to go see its connections to AI and what's going on and some of the things that are in the public space on this that this came up with, which I'll include in the trade piece. It just shows how you start using AI to get a top-down side, and then you end up in Chevron's AI data center power initiatives, and you get to read all of the things that are already there. Like I said, some of the statements that I'll highlight, they're mind-blowing in terms of how prepared they are for this, the gap that exists before the power gets on from nuclear. There is a a massive need that is here and we don't have the power needs, which is why these guys are buying up as much nuclear that already exists, but that's not enough for the next 5 years. So, this is where we are now. We've gone from the AI demand side, which we saw from Oracle, the semiconductor needs, which is the short of exponential growth that's going to happen and people have not yet seen what embodiment means, at the same time the power needs—those are not built in. This is old-school hardware commodity-related investing at a time when there's been underinvestment for the last 17 years because this has all been about the cloud and software. SaaS companies are going to be obliterated by AI. Hardware companies are going to be—need for buildout—phenomenal uh uh YouTube podcast to read to watch this week. It's two hours and 20 minutes, but it's Pier Dearandis, three other people. They go through and debate the job situation. They also go through the AI warfare. There's a massive amount of buildout that needs to happen on the military side, which is part of this hardware as well. You're already seeing it in the differential between what Israel is able to do and what Iran is. You have to start paying attention to what's happening on that war more so than Ukraine and Russia. Mainly because Israel is using the most cutting-edge technology at this point, and a lot of the things that you're reading about are unbelievable in terms of their ability to do stuff. Everything is changing in the world depending on whether you have AI or don't have AI as a big buildout. You need the power. You need this. It's a huge differential between the haves and the have-nots. Spend the time on it.
Um, inside the video, one of the things that's interesting: startup to billion-dollar valuation speed. Again, this is going to become important as I go through this. In the past, it took 100,000 people to build a billion-dollar company. The number shrank to 50,000 in the 80s, 10,000, 2000s. A future of zero-person startups is possible. Agents founding and running companies. This is why all software companies are going to be obliterated over time. Uh, coding is now ubiquitous. I don't care who the software company is. I don't care if it's Google. I don't care who it is. Eventually, over time, whether it takes 10 years, 9 years, 8 years, 7 years, 6 years, but the future long-duration assets, I'm going to say that again, if you were invested in VC and private equity companies, my own opinion is how you can do that where rates stay up at higher levels uh than where they necessarily would on the long end. And at the same point, we have AI accelerating, making it very difficult to understand where things are going to be in the future. That combination is very, very dangerous for long. The competition will be accelerating at a pace that is beyond belief. Uh, competitive intensity is so high that a singularity sprint is underway. An anxious rush to launch before others make you obsolete. Listen to it. You'll get the smartest people talking about it. The jobs destruction. AI isn't just replacing jobs. It's doing so faster and more broadly. They talk about why this isn't the same as the steam engine and all that stuff, which if it's not clear to people, this is a completely different scenario because the machines are figuring out ways to solve the problems on their own. They're working 24 hours. They have brains. There's reinforcement learning. They're learning along the way. They're not being helped by the human beings. They're not button-pushed. The only thing that needs to happen is they need to be plugged in. That's it.
Um, Moaba Gadat uh predicts 10 to 40% unemployment in some sectors. Same thing as Dario Modai. That's where they go through. Governments are not prepared for the speed of change. This gets back into the unemployment rate and the fact that their job is to print money to keep this change from happening fast. They're going to have to print more money. Call for UBI trials, 4-day work week pilots. These are all things that they're saying are going to have to happen. I wrote a paper on this. Entrepreneurship is the last job standing. The only future-proof career is entrepreneurship, education. You have to get away from training to get a job. Don't pick a major. Just go and go through this. All of this stuff for your kids. Pay attention. Just go through it. If you had any doubt, this was Ilia, formerly of uh AI and one of the smarter people out there on AI saying the same thing. Uh, the day will come when AI will do all things we can do. The reason is the brain is biological computers. So why can't the digital computer do the same thing? It's funny that we are debating if AI can truly think.
Um, Zuckerberg is basically putting all his chips on the table. He literally invested 14 billion in Scale AI to hire Alexander Wang. So he's creating a secret super intelligence lab to pursue beyond-human AI. He's going all in to speed up this process. And one of the reasons why at this point is ChatGPT 5 rumored for July 25th launch. Um, this is the fir—if you haven't—if you're not prepared for this, I'm about to take you through. Um, this has been talked about as the most serious candidate yet for a pre-AGI foundation. So, basically, this would be the acceleration at this point of what's happening with AI; it would be the point that gets us to the next level.
Um, if you want to read something or bring something into ChatGPT as I did here in the paper—situational awareness by Leopold Aschenbrenner—what did he say about the importance of eventual releases, the eventual release of GPT 5? So, he's an OpenAI uh researcher. He wrote this—I think it's 165-page uh paper called Situational Awareness: The Decade Ahead—and although he didn't specifically mention GPT5, what he talked about was effectively sometime before 2027 we'll move from chatbot—so what most people know—to agent. In his words, by 2027, rather than a chatbot, you're going to have something that looks more like an agent, like a coworker, and strikingly plausible that by 2027 models will be able to do the work of an AI researcher, engineer here. So the fact that GPT5 is looking to be released next this next month um this would be the accelerant. GPT5 equals AGI catalyst, uh, and again this is from his paper—operate as an agentic colleague; trigger the recursive self-improvement path to super intelligence. So again, the recursive self-improvement where they're teaching themselves. Uh, this is now we're getting into another level. Now, did he mention the power needs associated with that world? He forecast that by 2026 training clusters will demand, and I'm not going to go through all the numbers, representing 20% of current electricity production; uh, it'll be limited due to data centers and warehouses comparable to building or integrating hundreds of power plants into the grid. He even expects national-scale mobilization akin to a Manhattan Project for AI, which would include a significant expansion in US electricity capacity by the last decade. Now again, the late in the decade, but again this is assuming when we get to AGI or at least get to GPT5. The main point is the same way that DeepSeek was an accelerant moment, think about what happened to the inference needs. That was in January. Those charts you saw are from January to April. This doesn't include V3 and the video side which I said is already melting GPUs. The demand side is growing faster than the supply side in an exponential way because the models are being released, allowing people to use them—there. I use them more than I ever have, and I've mentioned that on these videos. So, um, he warned that without massive new infrastructure projects, we could hit hard limits on scaling. Uh, and this is why if you're ever wondering, yeah, the massive power scaling challenge, there's no way that we're not going to continue this buildout. So, if you're thinking that they won't spend the money, that we're going to reach some point, you're reading people that are just feeding you garbage at this point. Read the companies and read what's happening. This is a race that has to happen.
So what does Marc Andreessen say? Reasoning. AI isn't the next cloud or the internet. It's the next microprocessor. I think this is a new kind of computer. All incumbents are going to get nuked. Everything gets rebuilt. This isn't an upgrade. This is a reset. So think depression, but not economically. That's a reset is a depression. It is—everyone gets washed out, and then we go through. So, when he says the incumbents are all...
Going to get nuked, let's start with one that so far is in that case. Shares of Apple sliding after this year's WWC event. I mean, I don't know how they're going to catch up. If someone else had a phone right now, they'd be dead already. Um, because they have no AI business. It's just staggering what's happening. But, uh, the fact that they're behind and you can't use Siri is just—it's unbelievable.
Back into the job situation. Google's offering buyouts to US employees throughout the company. Here are the another look at the numbers, longer-term number of employees. And I want you to see two things. If you add up all of these, basically you're growing across the board always except since 2022. That was when we got the rate situation, but that's also when Chat GPT came out. And now there's no more hiring going on. So, they're growing earnings rapidly, way above 20%, and they're not hiring. Super interesting.
Anecdotally, most VCs are seeing companies grow faster with fewer employees over the last two years, but this is the first concrete data I've seen. This is from Patrick Collison. Um, I've talked about this. I talked about the Stripe session and gave it as like one of the most important macro pieces to look at in terms of their presentation. The reason is you've got new businesses growing that are eventually going to compete with the behemoths. No employees are being hired by those people. On the other end, so think of it as a barbell. The other side of the distribution, the mega caps are not hiring anyone anymore. Uh, Disney's looking to take it down. They're all focused on revenue per employee. And what you have is the rest of the economy where people are stuck in jobs where the companies are eventually going to go out of business. And so this process is not happening simultaneously, which is a recession. It's a recycling that's happening, and people can go out and get a job because of the the flexible labor side, but it'll be for less money. And that's what you're going to see is the wage pressure to me is just going to continue.
Um, Microsoft working on a deal to add 1 million. This is again just to show the growth that we're about to see. The co-pilot side gets into the agent side. You're now reaching a point where more and more companies are going to do this. So how do you protect yourself and your kids? I wrote this paper, brainstorming with the future, uh, Substack. Go read it. I'm not going to go through the details of it. Bill Gurley, uh, another good, um, thing to listen to. This was on the state of VC and the entire private market. It's important as I make the segue now into the final part of crypto. It was on Invest Like the Best. Uh, definitely worth listening to.
The highlights from Bill: Zombie Unicorn Overhang. There were about a thousand companies that raised between 200 and 300 million during the zero in interest rate policy, but now lack the growth and can't justify their valuations. Again, they took in a ton of money. They got up to a billion dollars. They haven't taken the marks down yet. They're still sitting at high valuations. He goes through this. He goes through a litany of why the private markets are in trouble. I'm not going to go through all of them. The second thing was the post-M shock to AI. All of this stuff has created an overhang. And this is why the the IPO market at least for these companies is very difficult. The valuations are hard, and it's also why you've seen this Harvard—this is from their local paper—plans 1 billion sale private equity stakes. This is back in April. Um, they're all out there trying to do this. Yale consider selling part of its $6 billion uh or the transaction could approach $6 billion, near to sell $2.5 billion private equity stakes. Again, they're taking out debt.
So, as I've said, AI is messing with the education system. The administration is messing with the education system. Private equity and their investments where between 35 and 60% of their investments, according to things I've shown on here, are in the private market. The entire private market, in my opinion, because of less liquidity, is just—it's going to be in trouble in the environment that I see. Uh, you won't hear the private equity people say it. I've seen them speak out, but the reality is the pressure is growing. The pressure is growing. You need to produce way outsized returns, and eventually Bitcoin will be the measurement. Daily liquidity 24 hours a day producing well over returns that private equity funds can produce, well over with VC funds. Uh, and as we saw with Circle and as we've seen with any of Core Weave and things that have come out, I think this is going to be more and more around the AI and crypto world and people realizing they just don't know how to invest it, and all these long-term duration investments are getting hurt by what AI is doing to everything. It's causing fiscal dominance. It's causing rates to be higher. It's causing the need to run growth at a level, bring short rates down, but long rates are going to stay higher. It just creates an a very, very difficult environment. And I think Bill—go—Bill Gurley, highly respected, but more importantly he's a systems thinker, Santa Fe Institute person. Go listen to him on anything, but in particular he gave 10 reasons why this is going to be a problem. That is a systems thinker going from a top-down basis of the problem.
Uh, this is just highlighting the returns and how the public markets have already been outperforming. This is going to get more of a problem. uh, something Julian Patel and Raoul Pal have highlighted many, many times, but I just want to make sure you guys see this. So this is showing the the problem we're running into in terms of demographics. So you've heard—if you've read Elon Musk—the biggest crisis in the world is that the world's not having—in the developed world's not having any more babies. So this line here is the demographic problem that's happening, and this is leading to this labor participation rate which continues to move lower. We've got aging demographics, falling birth rates, and now the rise of automation. So, the rise of AI, I'll continue to say these two things are what forced the government. Right now, at this point, demographics has us with a fiscal deficit that is unfixable between Social Security, Medicare, and Medicaid. Humans are already being replaced by AM robotics at a staggering pace, and the shift is only beginning. This is deflationary. It also reinforces the need for ongoing stimulus to keep the system afloat. Everyone talks about the problem of the math. They don't talk about the reason and the fact that there's no way to fix it. AI will eventually fix it, but we're ways away from that. In the meantime, Bitcoin goes higher, and all of the crypto world continues to accelerate in terms of what's happening.
So, I wanted to highlight this—this level here, debt to GDP. So, this is two things. This is debt to GDP. The S&P 500. This is why we don't have recessions anymore, people. I can't make it any more obvious than this. There are no more recessions because we don't have this fiat system of debt in the equity market. It's been replaced by the government. So the government interceded, and what happened at 2009, this is when the Bitcoin white paper came out. So now we've got this gap completely reversed where it's now public sector debt over private sector debt. If I brought in households, you'd see the same exact thing. This is the situation here. The government has the debt. They have a printing press. They print money. And this problem here in terms of S&P 500 is down here because of the MAGA 7 which don't have any people. They're not hiring anyone. So the companies that are winning are not hiring one. The startup businesses that are competing are not hiring anyone. And we're about to have companies worth a billion dollars with no employees. So this is accelerated to a point where if you don't focus on AI and you don't believe it, you don't understand it, you don't use it, you have no chance of trading macro in this environment.
Uh, Paul Tudor Jones says the road map is clear: inflation will be allowed to run hot with three and a half percent inflation as the US tries to grow its way out of the debt. Bitcoin will be the fastest horse in the race. He said that uh, 5 years ago. Um, so here's the Bitcoin chart. We went up to the highs. We're consolidating here. Um, every time that something in Iran happens, everyone points to look, gold's going higher, but Bitcoin's going down. So everyone hears this. The only buyers of Bitcoin are either corporates that are buying underneath. They're—think of it as Bitcoin buybacks. Uh, they're not chasing the market higher. That's made for hedge funds. Uh, hedge funds jump into gold when there's a political event. And they short stocks. So of course, Bitcoin is going to go down because it's being driven by the knee-jerk reaction of people who think of Bitcoin as going to zero all the time. The people who believe it's going up, they don't chase the thing higher. It's not—that's not the way it trades.
Um, this is what's keeping Bitcoin from going up uh, a ton. So, the change year to date up 13% for Bitcoin. Look at all of these numbers here. Forget Tether, that's a uh, stable coin. Let's leave Ripple alone at this point. Down six, down 22. Look, Solana, Ethereum, both down 22. You can go through this stuff. We are in a bear market. SWE down uh, 26%. Avalanche down 44. These are the types. Tron down 45. These are what you would expect to be going higher in a bull market, but we're not in a bull market for the utility tokens at this point. And I just want to say this and highlight this. This was the NASDAQ bubble to me. This is what happened with FTX. FTX and the overhang happened here. So this was 2021 in my opinion. 2022 the regulations came in, the Democrats kill it, goes down. This is the NASDAQ post.com bubble in 2000. Every time you get a rally, then you get this. I think we're in here. And now I think we're kind of in here. And the reason I say we're in here is because this came out of the financial crisis. This to me is the tariff and the deep-sea moment. People do not understand what AI is going to do to the economy. They do not understand what AI is going to do to jobs. This situation is going to get worse. And we're at the network effect moment for Bitcoin and for the tokens.
So Paul Tudor Jones made one mistake. I love Paul. He knows that. I—I've known him a long time. But he said Bitcoin has five times the V of gold. I just want to start highlighting to people—this is a one-year—well, yeah, this is 365-day V or 360 V. This is 360 V. Gold's at 17. Bitcoin has declined now to 41. If you—I did shorter term, you're doing about two times the V now. Bitcoin V has come down. This has gone down. I have said before and I believe it—there will be a massive short squeeze in Bitcoin coming up at some point this year. Uh, I think it'll happen either over the summertime or sometime in in Q3 is my guess, but I'm not going to try to pinpoint it. I just think there's a short squeeze coming. Gold's V is now in that 17 area. Five times. It would obviously take us up more into where it was before. Paul needs to adjust his V-adjusted target. It's—I think he needs to have more Bitcoin than he's talking about. A little less gold, little more Bitcoin. V's been coming in.
Um, I want to go back to Patrick Collison. Uh, Stripe's macro figures, payment volume from customers that signed up for Stripe is tracking way ahead of prior years when the lockdowns triggered a high surge in last week was 116% ahead. Now remember this as we keep going through this because the reason of this is going to get into crypto. Stripe is playing 4D chess, circumventing the banking system via crypto rails. Stable coin payments are growing rapidly. So again, you want to invest in things that are growing exponentially. I've already shown you the fact that we're going to have that for power. We're going to have that for semis because we have it for tokens. So you want that for AI. Oracle highlighted. So here's the other thing: stable coins. If you haven't done your homework yet on stable coins and you don't understand why Bitcoin benefits from stable coins, call up 22V or watch me on on Pomp, and I go through it every single week. Stable coins now present what I believe is the first credible opportunity to onboard a billion people into crypto. This is the network effects. If you haven't checked in the latest stablecoin data recently, you might be surprised. Stable coins have done 33 trillion in transaction volume in the last 12 months, consistently hitting new all-time highs. That's close to 20 times the volume of PayPal, close to three times the volume of Visa, and approaching the volume of the entire AC. This is what's happening. This is what's going on. You want to invest in growth. Stable coin transaction volumes hit 752 billion in May, up from 409 a year ago. So again, token growth, stable coin volume growth. It's happening across Latin America, Africa, Asia, Pacific, Europe, Western Europe, North America. This is not just a US thing. This is not just one. This is everything. Stable coin supply is at an all-time high. Another great set. Over 1% of the total US dollar supply is now tokenized as stable coins.
Now on the infrastructure side, I want to get back to what I put. Ethereum, Tron, this down 22% so far year to date. This down 45%. They continue to dominate growth on chains like Solana, Arbitron B. At some point here, this will have an effect. You will start to see the benefits go through because unlike the .com bubble which took a long time, crypto—it goes faster, but you're looking for major inflection points, and this Wall Street Journal article on Friday—was it—Walmart and Amazon—basically the entire US economy—so if consumption is 70%, Walmart and Amazon are the bulk of consumption on staples and on regular things—exploring issuing their own stable coins, corporate coins could take payment activity away from the banks and the trady system, accelerating the volumes and accelerating the dollars that move into the digital economy, accelerating wallets being set up. And for the first time, Visa and Mastercard stocks actually fall after a stable coin report. And not fall a little, it was the largest fall for these companies in—I think for Visa, I think it was the largest daily fall since 2021, but more importantly on a stable coin report. So this is no longer a joke. Everyone's been waiting for when stable coins would have an impact. It's not just Walmart and Amazon. Shopify will enable stable coins in checkout via Shopify payments. Circle stock went gangbusters. There's more IPOs coming out. Shares soared 25% on Friday even though the broader market plunged on the military conflict between Israel and Iran. So, Bitcoin may have traded lower, but the stable coin company that came out last week did not.
Uh, Societe Generale will launch a dollar-backed stable coin. Yes, a dollar-backed stable coin. Uh, first major European lender to launch a dollar-backed cryptocurrency in the booming market for stable coins. Genius stablecoin bill passes key vote, advances in the US Senate. David Sax says stable coin bill will be a historic bipartisan win. He believes the Genius Act, the new state calling legislate, could become the first major bipartisan bill, the first major bipartisan bill to pass in years, marking a turning point for the US crypto regulation. I don't think it could be more important than that. Blockchain initiatives, so the blockchain, a completely separate side, but it'll start to get into tokenization and other things outside of it, uh, have been adopted by 60% of Fortune 500 companies at this point. I believe Robin Hood will launch tokenized stocks in 20 days. That means you'll be able to trade stocks round the clock. Uh, the announcement could shake the financial world and fuel its global expansion. If you haven't looked at Robin Hood stock, you missed last week's time and the last slide for this week. Uh, I think Bitcoiners will be extremely pleased—Trump will soon announce plans to accumulate more Bitcoin. Uh, I'm going to leave it there. Have a good weekend. I'll see you guys next.