Transcription
Back for, uh, another year. Quiet ending of the year, as always, but, uh, there was a lot of stuff that actually happened in the last couple weeks in the AI space, uh, both on the news front, in terms of posts and X, and also, uh, some deals that were done. So, I'll go through all of that, as well as just kind of go through a wrap-up, go through some of the themes for this year, and I'll finish up with, uh, basically, uh, a recognition from my front in terms of where we are with AI. I think the progress that has been made in the last three months in terms of making it something that you can no longer wait for, uh, based on how much you can do with it. Uh, and more importantly, I think for, you know, people my age with your kids, I think it's a necessity to, uh, get them more involved, especially since the school system doesn't seem to know, know how to, uh, deal with it yet. So, I'll go through all of that and, uh, hopefully leave you guys with a lot to think about on multiple fronts.
First of all, S&P finished the year, or so, with all the bubble talk. Basically, we ended up about the same as we did in 2020, in the COVID year, but you can see this was not some kind of freakish year. We finished up, uh, 16% in terms of price returns before dividends for the NDX, the, you know, the beta part of the market. It was, uh, the worst positive year we've seen, uh, over the last seven years. So, uh, definitely not showing, uh, bubble land. Uh, Russell, minor positive, you know, 11%, nothing big. I expect bigger performance next year. The headline story was gold. Gold had its biggest performance since 1979, uh, and not by a little bit. It was a, a very, very, very good year for gold, especially vol-adjusted. Bitcoin, uh, sent the way sentiment ended, you'd think it was a crash, but after back-to-back 100 plus, or 125% plus years, uh, it finished down 6%. Everyone hates it now. Sentiment is low. Technicians hate it. I love it. Um, I loved it last year, so take it for what it's worth. Uh, but I'll go through a lot of the, the reasons why, at least from a, uh, a tactical basis. Forget the long-term story where I believe people are just not putting into context what's happening with the financial guard rails and the disruption from AI, but let's just focus on the tactical side. Ended up at the highest level of the year in terms of the AI net. So, it's no longer a situ- situation where the market is negative, where we had been the entire year. Uh, basically, if you look at the zero line, we spent most of the year below. If you take an average, it was one of the worst years, and it was definitely the worst year adjusted for the fact that we were up. Uh, I'm treating this as coming out, believe it or not, of a recession year, where the cyclical upswing is coming. Uh, we'll, we'll get more into that as we go on.
I think the big news for the end, and something that, uh, you know, you may not have seen, but GDP came out, uh, for Q3, and the result was 4.3%. Here it is out here. Only one economist, uh, basically was even close. Uh, so you had a very, very big move. And I think more importantly, this 4.3 comes after a 3.8. So, what I wanted to do in seeing that real GDP coming in that high, basically 4% over a two-quarter average, nominal GDP went back to the highest level since '23. And again, this is when inflation was on the higher end. So, we had nominal GDP go up, but inflation in Q3 was not, uh, accelerating. And so, basically, what I did is I went into Gemini and asked it to go find out how many times in history we've had back-to-back, uh, quarters of at least 3.8% since, uh, I guess this century. So, one of them, obviously, the COVID rebound, where we had a bunch of big numbers. Then you had the mid-cycle boost in 2014. And before that, you got to go all the way back to the beginning of '03, '04, coming out of the dot bubble, and then you had the pre-.com bubble. So, the most likely period, or the one that ChatGPT, or Gemini gave us as the most likely period, uh, it's most structurally similar to 20, uh, 2003 to 2004. I'll go through the reasons why as we go through this, but one of the things that's unique, uh, about this is the fact that we just did this two-quarter run with no job creation. So, I want you to put that into context. We grew 4% a quarter, which has really only happened, uh, three times this century, and we had zero job creation during the two quarters. So, if you're looking for something in there that tells you this time is different, and as investors and as macro people, you have to spend the time understanding how to put this into context and what it means. And there are not many times in history where you can find no job creation at the same time as nominal GDP coming in at an annualized 8% in the quarter, real GDP over two quarters coming in at 4%. So, uh, you got that. And then you also have the fact that we've got, for next year, still two rate cuts built in. So, the Fed is still going to be cutting rates because of the job picture, but also because of the inflation picture. And gas at the pump is now down to $2.83, the lowest level since 2020. And basically, we're right in where we were from 2018 to 2020 in terms of the midpoint of the range. Without gas at the pump going higher, it's going to be really hard for the Fed to be raising rates. And anyone who's worried about inflation, I think should just sit there and say that we are probably going to have a reflationary boom where inflation will be sticky. Uh, but if we're not going to see a change in the payroll numbers, and I don't see how they're going to get better with AI, uh, and AI agents coming in front of us. So, we'll talk more about that.
Productivity. Uh, Dennis and the team at 22V basically highlighted it well. We are starting to get more and more of the productivity conversation, and I think this is going to accelerate now. Mark Andreessen started to, uh, talk about it, and the fact that we got 4.3, Elon Musk, double-digit growth is coming within 12 to 18 months. If applied intelligence is a proxy for economic growth, which it should be, triple-digit is possible in five years. I just want you to take into context this whole double-digit growth and what's going on. Even though I don't think that is coming, I do think the period you're looking at here with the ability of replacing labor or not hiring people as topline revenue, in this case, nominal GDP grew 8% in the quarter, and yet we have no hiring going on. So, think about that in terms of the productivity boom. If this compounds, then you can theoretically get to numbers that are insane. So, just keep it in the back of your mind because it's not a joke. Uh, copper moving higher. So, for anyone questioning whether where we are in a reflationary boom, copper accelerating into the end of the year. MSCI World X the US, which is my proxy for global liquidity and kind of a global boom, closing on the highs of the year. The dollar, after, you know, a bounce from what were, you know, a very weak beginning part of the year, is in a downtrend again, posting its worst year since 2017. With Fed cuts coming, the yield curve steepening, all of these on suggesting the same thing. The Cosby closing, Dr. Cosby, Dr. Copper, you've got everything kind of moving higher. In Korea, export growth posted back-to-back big quarter, or big months to finish out the, uh, uh, or sorry, quarters to finish out the year. And to show you the importance of South Korea from an export, this is year-over-year exports. The white line, this is the US PMI. We're going higher in terms of all of these fronts. You also had Taiwan exports, which again, another proxy for PMIs, blowing out to the upside. Capital goods, we got the durable, latest durable goods report. Capital goods new orders, non-defense X defense, X air, at levels historically equivalent with close to 60 PMIs. We just continue to march higher. The diffusion index, this is from Macro Micro. I thought this was a good chart just highlighting the global PMI diffusion. So, again, diffusion on top of a diffusion. So, think of this again as a second derivative, again implying higher levels. This is using it versus global monetary policy. I've shown charts like this before. We are doing rate cuts, and so not a surprise if you've watched my videos.
Next year, to me, the number one theme is connecting AI into the physical world. So, my outlook paper for this year, long one with a lot of the themes that I think are going to be important to support PMIs not only going higher this year, but the beginning of a mega, a mega trend on the back of commodity strength, on the back of enterprise using edge devices, edge device upgrades for cars, auto, or cars, phones, computers, uh, the NPUs that are going to be necessary for the embodied AI, humanoids coming behind it, the data center buildout, the bring your own generation, BYOG power trend that is in place. All of this stuff fits into the physical world. I'll be putting this out 22V. If you guys haven't reached out to 22V, get signed up. There's a lot of things I'll show at the end in terms of what I'm going to be doing on top of just the written work now, but giving you specific names in the space, uh, the thematic ideas within there, some of the ones I mentioned, but you get the picture. So, for all of these, including the BYOG side, uh, which is becoming a bigger and bigger theme, solving most of the power problems, uh, I think it's an important thing to get up to speed on. Colossus 2 is going to be, uh, up and running completely. There's more work going on on building out Colossus to get it to its eventual ending of a million GPU cluster. All of this stuff is going to be important. Uh, P1 podcast to reference for this week. I did listen to Facts Versus Feelings. I do from time to time. The reason I cared about this one were twofold. One is just wasn't a lot of podcasts out the last couple weeks of the year, but, uh, Ryan Dietrich did have on two technicians. Uh, and I thought they did a good job. I like technicians at the end of the year because I like to look for charts that are pointing a direction and then try to figure out what it means together. That has gotten a lot easier with artificial intelligence. Uh, my number one thinking partner on this stuff, which again, I'll show you at the end. Uh, this is going to be a big year for teaching whoever wants to learn how to be able to do this stuff I do. But also, I don't think you have time to wait anymore because I think we've reached a point where the capabilities of artificial intelligence have compounded at such a fast pace now that when I show you what I've been able to build in a matter of hours, uh, on things that I had a team of data scientists that were never able to finish for me, not because they weren't capable, but because of how difficult it was for me to give what my brain wanted and get them to be able to do it. That merging together of thought collaboration is what has happened with AI, which allows you to do this stuff and kind of get your answer, go back, get your answer, go back. So, with the podcast, what I did is I just said, take all of the charts that these two technicians talked about and let's break them down in terms of what went on. So, they talked about financials, talked about long decade sideways assets. I've shown some stuff from John Rog. I'll go through a bunch of charts, same way. These long-term bases we are breaking out of. This is the PMI side. Uh, a lot of these long-term bases are even longer. They're commodity stuff going back. That's the hardware versus software. You have hardware versus software. All of those multi-decade sideways things, which really haven't worked since 2003 to 2004. Uh, I think you're in the beginning stages of it. They talk about gold, what it means, especially when it's happening with stocks going higher. And they highlight another fact, which is the market breadth. So, when you take all of those together and you go look at gold having risen at the same time as cyclical strength, with breadth improving, with junk bonds, strong equities making new highs. These are the only time periods that you found where you could have gold going up alongside of risk-on assets, but in particular with inflationary type or reflationary assets, reflationary, not recessionary, global growth boom. So, these are the periods that's bringing up. Same thing shows for the breadth. And again, usually another 12 to 24 months using history, even though they only have a couple of them.
So, what regime are we most likely in? Well, financials breaking out. Early cycle reflation theme. Multi-decade sideways poised to break out. Commodity super cycle beginnings. Global nominal GDP accelerates faster than expected. Real assets dramatically outperform financial. This is never a recessionary regime. It is a global reflation regime. So, then you go through what does it normally mean? And again, I just showed you copper's going higher. Nominal GDP is accelerating, equities are going higher, gold's going higher, financial conditions are easing, we've got the Fed cutting rates, the dollar is weakening. This is all part of what normally goes on during it. What we would expect in this inflation stabilizes but remains above the pre-I era. This is not late cycle stagflation. And again, I've talked about this the amount of times I've heard stagflation. All of these things have kept people out of the market. Bubbles kept people out of the market. A full-scale reflationary expansion where tech and commodities rise together. This has only happened twice in 60 years, and both were extremely profitable multi-year cycles. Climb the wall of worry.
So, Bitcoin had a horrible relative year and had a bad year overall. How does Bitcoin, or how should it perform in this regime? AI-driven global reflation, commodity breakouts, financial strengthening, breadth expansion, and a weaker dollar. Bitcoin should be one of the strongest performing major assets in the world. I will say it will be the strongest. Thought the same thing last year. So again, this is not advice. I'm putting my money where my mouth is, but I pretty much always do with Bitcoin because my long-term view. The regime aligns almost perfectly with Bitcoin's historical environments. It loves a weaker dollar, steeper yield curve. It thrives when real assets outperform. Here is it versus PMIs. I've talked about this. When PMIs are rising, you get the best time of Bitcoin. Doesn't matter if it's a minor cycle or a bigger cycle. Anything going on, it is typically very sensitive to PMIs and altcoins, or let's just use higher beta, uh, crypto, and let's just take Ethereum as I'll get into in a little bit, uh, as another part. So, this is the perfect conditions for S-curve adoption. This is something critical. I've talked about this as my network effects, AI adoption, and decentralization equal the Bitcoin narrative strengthening, tokenization, and real-world assets, a big theme for next year, more on-chain liquidity, more money inside the crypto world, the digital economy, good for Bitcoin. Institutional adoption is now structurally embedded, ETFs and custody rails. All of this stuff is the technology side of it. Should be one of the best performing assets. And again, reminder, this is after it had a quiet year, but again, if you take the three-year performance, nothing has beaten it. So, just use it for what it is. I still believe this is what went on last year. We continued into the end of the year to see OG's selling. This to me was just a representation again of what makes sense ideologically, the people that were early and went into it. You've got the government more involved in sponsoring it. You've got ETFs. This is not the true decentralized ideological breakaway from the fiat system. So, you've got those people bailing out. At the same time, there's a new sheriff in town in terms of where you can find a five-bagger or 10-bagger. And Bitcoin's not in the same class clearly as what's been going on in AI. Micron Technology, a very big asset, went from a low of 60 some odd to over 300. You're getting moves like Palantir and things. So, it lost a little bit of its relative mojo as a beta asset, and I think that took a lot of people out, especially, uh, people that on the VC world went, we had some market maker issues towards the end of the year. I think all of this was a consolidation, and I love the fact that sentiment is dead, and I'm love the fact that I can't find one technician that likes it at this point. Now, I posted this on December 30th that we had closed below the 50-day moving average for 64 days. Uh, it finished the year with another three on top of it. We did close above it on the first day of the year. So, it had been 64 straight days, or 65 at that point. Since 2020, that's only happened four other times. So, we tied this one from early this year. That was obviously a major bottom before it went higher. Not all of these ended up with a major bottom that ended up going higher. But I will show you the four points. One, two, three, four. All of them were at least short-term entry points. This one got you a good rally. This one got you a good rally, but none of these were kind of the beginning of this. This one here, you had to deal with some more basing. But my main point in terms of going through this was to highlight that again, for me, MSCI World X the US, which takes into account the non-tech world in terms of the ex-US dominated by commodities and things like that. At the same time, when you have this going on, this has got a weaker dollar trend involved with it as well. MSCI World X US did outperform last year to show the correlation between it and Bitcoin. Here is the 100-day moving average of both of them, and over time, you can see Bitcoin has diverged since the market-making incident and since we went down. So, my belief, and the reason that I will not listen to technicians on this, and I will still be buying into any dip that occurs, if there are any, at the beginning of the year, is that liquidity is still positive. My outlook in general for this year is positive. There will be time in the future where I'm not as positive on the US side, and I think Bitcoin can have a correction, but right now, it is swimming against the liquidity stream, and I think that means the snapback will be violent, just like the snapback here, which also occurred during the same thing. Liquidity was going higher. I believe the same thing is going to happen this year. So, we closed above the 50-day moving average, first time. We'll see what happens. Uh, it wasn't the only one. Solana, only time Solana had been below the 50-day moving average again came in during the April period. We closed above it again on the first day of the year. We'll see what happens. What really has me positive from a beta basis, and what I like happening related to the PMIs, is what's been going on with Ethereum relative to Bitcoin. So, we look at a lot of relative, uh, positions in the stock market, but very seldom will you see people look at Ethereum relative to Bitcoin. Last year, during the upswing, or when Bitcoin was going higher, really beginning from 2022, Ethereum was underperforming. The fact that Ethereum is now going higher at the same time that we're getting the rise of acceptance for both stable coins and tokenization, to me means the ecosystem network effects are kicking in. I think you want to be positive on Bitcoin because Ethereum is outperforming. The network effects are in place. Uh, John Rog, who I show a lot, and I'll be doing a lot more work with, I'm just showing this because he's got a very different view. Um, and he's saying there's risk to $6,000. He's also still calling for lower levels, uh, in, uh, MicroStrategy. I'm going to go against John on this one, but on the rest of the ones I'm about to show you, we're on the same page. Uh, this is some of his big bases that again fit back in with the facts and feelings situation. Group a bank, as John likes to say, and as I've shown on here, you do not get bearish the market when banks are breaking out, but in particular, a 16-year base in Group. That takes us back to the great financial crisis. Banks are going higher. Deregulation is kicking in. Freeport-McMoRan, haven't taken out the highs yet since again, the great financial crisis. A big base, but you also have a PMI base going here. Major catalyst. I think copper is going much, much, much higher on the back of the needs that are associated with AI and the buildout. ExxonMobil, again, up at the top end of the range. PMI base here, long-term base. And again, this happened with oil trading down to $55 this year. DuPont, there's going to be a major, major bull run in chemicals in my opinion because of a lot of things that I'll cover in terms of the Nvidia Grok deal and just a lot of NPU packaging. We are moving away from GPUs and we are moving into advanced packaging, building out the next phase, which is edge devices. Chemicals will be a big part of this as well. You've got Hecla Mining. So, the gold-silver side breaking out, long base. Newmont Mining, you got to go all the way back to the 80s in terms of breaking out. Delta Airlines, I've highlighted transports, another big PMI theme. All of these are PMI themes. Eli Lilly, I've highlighted the advancement in terms of the shift. So, Lily fits into a different category for me. I think healthcare is going to be a surprising replacement for a lot of tech this year in terms, so biotech and pharma, as AI drug discovery starts to repric and rerate this group with higher multiples. Tesla, a big part of the robotics and PMI theme for me. John likes that one as well. He and I have been on the same page on on this back since it was about $330. Uh, both talking about it. And as I mentioned, Tesla, no safety monitor in the car, went on December 24th with Elon Musk in it. They're rolling it out for public rides. And again, AI still, I showed this last week or two weeks ago, but in case people didn't see it, this is the biggest risk to the market stability in 2026. We've never seen a single risk score so far ahead of the rest entering a new year, says Deutsche Bank. And again, AI bubble, you will be climbing the wall of worry the entire time. Uh, eye on the market, Michael Sembleis did a, you know, a great report. I like his work. I've highlighted before, but I think he brought up a lot of charts and tables that are worth looking at for you guys in terms of the AI bubble. Uh, and again, what we've got here is just total stocks. When you include all of the direct AI stocks, the AI utilities, the AI cap equipment, you're dealing with 42 stocks. Price return 78%. S&P X those names 22%. So, basically the bulk of everything, whether it's earnings growth, whether it's the capex, so all of this has been related to AI. Not a shock, but I just want to make it in terms of, uh, spending, how big this is. You got everyone kind of spending a wasteful amount of time trying to say how much of this economy is driven by AI. I keep saying it's almost all of it. We have no job creation. If anything, you can see there's no job creation. AI is driving all of these factors at this point. The Fed would not be cutting rates, uh, if it wasn't for the job situation. The job situation would be, AI is a deflationary force. It is a job-replacing force. It has societal impacts. That will be a major risk is going to be the backlash that comes starting with the midterm elections. Uh, all of the things of trying to restrict this buildout. You'll have probably a lot of issues that pop up and risk that scare the market this year. But the reality is the tech capex is massive right now. If you combine all of these, and again, these are in percent of GDP, so this is normalized. Manhattan Project, electricity, Apollo Project, you're dealing with this. You leave this one out, you're dealing about the same size as all of them combined. That's how big the capex buildout is. From a valuation basis, the reason I like showing this stuff is because I wouldn't say he's positive on the on AI, but I would say that he's pushing back against the valuation side. So, if you've got a negative view on valuation, you should go spend some time looking the report. If you're worried about the debt situation, he highlights that the debt is big relative to what it was. But then he honestly goes through and highlights the net debt to EBITDA of, so zero indicates an excess of cash and marketable securities. The majority of the space has is just loaded with cash. Free cash flow to revenue ratios of of, uh, of the AI companies. He goes to and compares everything to the rest of the market in terms of the Russell 3000. His argument is this, these none of these are are are issues for now. If they grow rapidly and we don't see the revenues, but that's the same argument that Jim Chanos makes, which I agree with. If we don't see the revenues in a year, you're going to have trouble for these names. They're still spending the money, and as long as they're spending the money, you're going to see it flow through the S&P 500. You're going to see it flow through GDP. The question is, are you going to get the benefits? I'm telling you as a fact, the improvements in AI over the course of the last three months are impossible to describe. It is moving so rapidly that if you don't use it every day, all day, the entire day for everything that you do, in your cars, everything. Go buy a new phone, go buy a new car, talk to it all day long. In both situations, I took a drive for three hours going back out to New Jersey over the course of the holidays. The entire time I was on with Grok, I didn't listen to a podcast. I literally had a conversation. It was beyond amazing for me to sit in the car and do it.
Michael Sembleis, his risks for 2026. Power generation constraints. I'm less worried about this, but I do think this can become an issue, even if it's just the states pushing back. China scaling the technology moat on its own. This one, I think, is overblown. China's approach to Taiwan, geopolitical risk. This one has the biggest tail. If there's any military fears on them invading Taiwan, obviously that is the biggest one. This one to me is the most likely, the metaverse moment for hyperscalar profits. Not that they're not going to get the money, but if it doesn't materialize as expected, it could become an issue next year, particularly with all the issuance that is going to be necessary to get more money. So, I took all of his risks. I put them into, uh, ChatGPT Pro and I said, "All right, go through these and give me your analysis. Give me the outcome probability, worry probability." So, again, outcome probability, what's the probability it actually happens? Worry probability, what's the probability that you, the investors, are worried about it. Look how high those numbers are. The nar- dominant narrative probability, market impact. And then so I just went through highest volatility spike, not surprising, is Taiwan. Most repeatable pain risk is the ROI capex, which I agree, you can go look at it on your on on your own. Uh, okay. So, for the final, uh, slides, if you haven't read this, bring it up in ChatGPT. I'm showing you the date. You go find it in X. But Andre Carpathy, who's, who's arguably the most respected person out there, and who's been very, very, let's say, uh, non-emotional on this whole situation, from mainly time at Tesla, but also at OpenAI, just a phenomenal, well-respected person within the space. You can tell, 15.6 million views. He did an interview with Darkh back in October. I received multiple hedge fund, uh, reachouts saying that this was a negative for AI. What was my opinion? Where he basically said AGI wasn't coming until for another decade and that AI agents were way behind. Well, this is what he put in. I've never felt this much behind as a programmer. So, I want you to remember, I've never felt this much behind as a programmer. I have a sense that I could be 10 times more powerful if I just properly string together what has become available over the last year. So, Andre Carpathy is number one doing podcasts talking about AI in a negative fashion. But then when he finally sits down and takes the time to use it a lot, I've never felt this much behind as a programmer. Clearly some powerful alien tool was handed around, except it comes with no manual, and everyone has to figure out how to hold it and operate it while the resulting magnitude 9 earthquake is rocking the profession. Roll up your sleeves to not fall behind. I'm going to emphasize this point to all of you because this is not just for programmers. This is for you if you're still going to be in the workforce, and especially your children who are about to enter the workforce. College is not teaching them this stuff. They need to be using it every single day. This post by Carpathy went massively viral in late December with over 15 million views. It captures a pivotal moment in the software engineering AI community, serving as a candid wake-up call about how rapidly AI tools are transforming programming. You have personal vulnerability. You are giving up on 10 times productivity boost. The Times of India, Business Insider, Final Express framed it as an open letter to software engineers, highlighting the urgency of upskilling and AI's rise. In essence, it's a landmark acknowledging that 2025 marked an inflection point where AI began fundamentally changing what it means to be a programmer, forcing a rapid evolution in skills. The first-order effects, job refactoring, productivity explosion. We get back to the GDP side. We're going to get into the AI agents, tools like multi-agent systems proliferate. Enterprise adoption surges. This is critical. When you get to this point now, the enterprise adoption is going to surge, which means the revenues for the hyperscalers are going to surge. The model producers are going to see it surge because of agents and enterprise adoption will go off. So, you're going to see a big movement in the market in this. The third-order effects. And again, if you guys aren't in it, second-order effects, third-order effects. I have a Santa Fe model, um, system thinkers prompt that I use to basically take any news item or any podcast or anything and get it taken out to the second and third order effects. This is a critical part that you should be using on anything you have in your life, regardless of what it is. I'm going to go through some of the healthcare stuff, some of the investing, but anything where you learn, if I do X, what happens to Y and Z on the other side? You have to start asking questions like that. This is what it's good at. Uh, is the polymath side, the ability of looking at things from a system, uh, systems thinker perspective. Traditional software modes erosion. This gets back into Salesforce. It gets back into everything in there. The ability of building apps on the fly. Most move to data pipelines, proprietary data sets like Tesla's driving data or agent ecosystems. We're in the mid-transition. It was hyped as a year of agents, but he noted that they fully don't work yet. So now we're getting into the point where the path to recursive self-improvement. He's still cautious, but he didn't mention that in this particular one. But potential for rapid pro, uh, progress 2027. Jensen Huang, the most important thing is you're not likely to lose a job to AI. And I agree with that. You're going to lose a job to somebody who uses AI. So, I would recommend everybody do what do what we do, which is use artificial intelligence with our work and do our jobs better.
So, that brings me into this. So, this year I will be launching this. It'll be coming out in January. We are just getting rid of all of the bugs. You can see my name is spelled wrong here. There's lots of little bugs that need to be, uh, need to be finalized, but the goal, and part of it's going to be not just this, the research, you can get the stocks, the sectors, the themes, but a big part of it is going to be the AI side and just showing you how I do all the things that I do and how you can incorporate AI into everything that you want to do. Whether it's cooking, whether it's fantasy football, whether it's health, whether it's whatever, anything that you find fun that you do on a regular basis where you've historically used Google or books or whatever. Uh, I'm going to basically go in and show you how to do this prompt engineering, everything along those lines. During the two weeks, Nvidia, and I just want to bring this up because this is kind of the way that everything goes for me now and what I've set up. So, Gavin Baker goes through the Nvidia versus Grok thing. I go out and write a paper on it in terms of the importance of it. The Nvidia Grok deal, why AI's next frontier requires architectural revolution over Moore's Law. This was a critical, critical relationship build between Nvidia and Grok. Um, it led to, and this is what I'll be doing for the people who sign up and the institutional clients of 22V. Basically, I have about 20 names that are associated with this change. This is a change. Nvidia is not on the list. So, this is a broadening out related to the PMIs of moving into a world where advanced packaging becomes the dominant theme. Now, this has huge implications for variety. So, the way that I'm going through this, and I'm just giving you how do I go from a thematic idea to an equity framework, give you the companies. John Rog put together the score. If you don't know John's work, his his, uh, work for is, uh, is what you want to be involved in in his work. It means the trend is your friend. Um, so these all have a Ford technical score. I go through the earnings commentary, and what I'm looking for there is I've got an algorithm which gives me the rankings in terms of is this company's progression in their earnings commentary showing the strength where the management is getting more optimistic as time is going on related to the AI demand theme. So, you can see these are in there. Lattice Semiconductor continues to grow on this, and Qualcomm does. And then revisions. So, revisions are just are the analysts on the street. So, here you have management, here you have the sell-side research, here you have technicals, and here you have thematic ideas. It's a broad picture. My goal is to be able to provide you guys this, uh, for asset managers, for separately managed accounts, anything along those lines. I will be putting all of this stuff together. There's a lot of themes, there's a lot of names, there's a lot of way to screen. I'm going to be doing this for a variety of people, uh, based on the ask, but also based on the fact that I've set all these up. So, Nvidia just admitted the general-purpose GPU era is ending. I can't think of a more critical period than kind of thinking about this and what it means. It means there's different names that are going to work this year. Now, at the same time, why did Meta at the end of the year buy Manis? What is the signal for your enterprise AI agent strategy? The Wall Street Journal marks said marks, one of the clearest signals yet that large tech platforms are no longer just competing on model quality, but on who controls the execution layer of AI-powered work. We're getting into the return on invested capital. AI agents is the critical part. Again, I talk about this in here in terms of this with the names that are going to benefit from the enterprises needing to upgrade their systems. This is a major, major change. We are going from the cloud. We are going to need to have these on-premise, onsite. Everything is on-site. Bring your own generation. It is very physical. It is very hardware-driven. You're going to have a lot of names that are going to benefit from this. I'm now created another thing which will be on the paywall for my X trending stuff. So, each weekend I will go in and basically look for the analysis of the most trending things out there. The reason I wanted to show this one. So, this is from the final week. The agentic and autonomous agents. So, AI agents is the theme you want to be investing in. If you don't have 20, 30, 40 names related to the agentic side, you got to start focusing on it. This is not about data centers anymore. This is not, that's what that's what was up there last year. It's not about the Palantir stuff in terms of the data coming through. You're getting a change in what's there, and the charts are showing that the trend is changing. And like I said, reflation is the theme, and reflation, you want to be less software, you want to be more hardware.
I created a market immune system. This is something that took me, as I mentioned at the beginning, uh, less than, less than two hours to build. I sat on my couch early in the morning and I basically said, you know what, I'm finally going to build my turbulence model that I had asked my data scientists to build for years. They were more than capable of doing it. But here's what was not easy. The way that I did this, which you couldn't do back then, is I sat there and I said, "I want you to go to GitHub and I want you to go find me the most highly rated code out there for a turbulence model." It comes back, it gives me this one. I go through and now I have to choose the assets. And the way that I wanted to choose the assets, and this is always what I believe when you're dealing with something like a covariance matrix as a warning system for the market. And that's what I wanted. A market immune system for me was basically something that would tell me when correlations and volatility were changing, especially when the S&P was not yet changing, because I believe the market will always show you the risk is moving before the actual index was. I, I use models like this back in in Brazil. They were not as sophisticated as this. So, I gave it 100 assets, 99 assets. A good percentage of them were related to AI because that's what's been driving the market. But then it also included things across stocks, bonds, commodities, currencies, crypto. When do they all start shaking at the same time? And what I end up getting is a model that basically is measured this way. The red line is where the turbulence is. When the red line is normally up high, these are ends up being bottoms. So, this is after something has already trended. But at the same time, when you get above this line here, this is the warning system, especially when the S&P is trending. So, the last one where we got this where it broke above it while the S&P was trending higher and above its 50-day moving average was here, and we went through a correction here. This has typically, and going back where the back test is basically gives you a warning signal of about seven to 10 days. So, this will be up on the paywall as well. Um, and I'll send it out occasionally, uh, on the institutional side. And I just want to remind you guys that for AI, you guys have to spend the time on it. I think the two best ways, uh, and this is why I've done it this way, are on the investing side, which is what I'm showing you right now, but the other one is HRV. For those of you who have Aurora rings, for those of you who monitor your heart rate variability, I'm telling you, I've spent more time on this. This has been a long journey for me. This has been about 15 years of focusing on this concept, using meditation, focusing on the microbiome, all of these different things. But if you don't know why it's important, if you want to be better as a parent, if you want to be better at your job, there is an element involved in this where the balance between your evolutionary side and your brain, so your body and your brain, that is what it measures. And so I've spent an enormous amount of time on this. I'll be rolling this out in Substack. My whole goal for this AI training side in the analyzing things, I approach things in this manner. It's a complex system. So investing in health. If you move one thing, if you change your protein, it has an impact on your fiber. If you do this, if you extract something, everyone plays around with these. They buy self-help books. They buy things for investing. With AI, it has to be a habit. This is the McCraven side. There's a bet thing. Every decision you're making under uncertainty. This is the Annie Duke side. Think with intelligence, dialogue, not answers. AI is the collaborator on all of this. You do not go for it like Google. Too many people are looking for solutions. It hallucinates. It does this. It is not meant for that. It is meant to help you make better decisions. But to do that, you have to be thinking with it, and you have to be asking it the right questions as you do it every single day. That's what I was saying. This has become so powerful for me, and this is what I'm going to focus on. Habit comes before outcomes. Every decision is a probabilistic bet. AI works as the best collaborator, not an answer, not an oracle. I'm going to basically get you to get it into a habit. The goal is not being right. The goal is making better bets repeatedly. That's the Annie Duke philosophy. The collaboration side, I can't, I really can't emphasize to you how much it goes through. On the prompt engineering side, what you're going to see is to give you an idea of my prompts. If you just type in and you go, "Hey, I want an answer for this Ultraink reasoning optimize operating contract." All I do when I ask it a question of something that I want is I'll put in what I'm trying to solve for, and then I'll type in I, I'll paste in this prompt which goes all the way here. It's two pages. If I want to do Socrates mode, your child might be someone who doesn't raise their hand in class, but if someone asks them a question, they're fine. This is the way you're not here. You, you literally put it in Socrates mode, and that way it asks you the questions. Think about what a doctor does when it says, what did you have for lunch? What did you have this? That's what Socrates mode is. The process in one sentence is show up daily, think in probabilities, collaborate with intelligence in systems that never stop changing. You have to do this. Investing in health are the easiest way in my mind. But if that's not what you're interested in, I guarantee you by reading the Substack and by going in and and watching the videos, whether it's cooking, whether it's fantasy football, travel planning, learning a new skill, anything, chess, painting, whatever, career development, any of those things, you will get the answer for. You'll be better in terms of thinking. If you're not using it now, this is where you're going to be at the end of anything involved with me. That's how I got to where I am with doing these videos and kind of showing you guys how to go from podcast thematic ideas down to stocks incorporating all the information. Everything that I do there, everything that I built, the whole immune system thing, it was all done basically verbally. And like I said, driving out now anywhere New Jersey with my kids going through this, I'm just on Grok verbally. If you guys haven't started to do it, you're missing out on something big. The goal for this year for all of you is to get you to the point where AI is something you're collaborating with. It'll bring more, uh, power to you. And if you are focused on the health side, I'm telling you, if you take a journey with me, there will be hundreds of things that you will learn. Even if you're not focused on HRV, but you're just focused on being happier and getting a better balance in your life between what your body is saying and what your mind is saying. Trust me, that's what the whole thing is about. Happy new year to you all again. Thanks for being here. Share the video. Do whatever you need. Uh, we're going to reset for the year. We're going to make some money together. We're going to be up on the news together. And for those of you who join me on an AI journey, I promise you I will make you much more powerful as an individual. And for the younger people out there, uh, our job is to find a way to make sure that you're prepared for what the world's going to look like. I'll see you guys again next week.