SPEAKER_00 0:00

Hello, everyone, and welcome to a special edition of the AI Investor Podcast. Today we're diving deep into a topic we've talked a lot about. I think a lot of people listening will be excited about today, and that is robotics. I'm giving a complete presentation on the background of the field, what you need to know to understand the development of robotics, recent advancements in robotics, and I'm gonna end it with a review of five robotics stocks. Three of them we've previously recommended before in the AI Investor Podcast Portfolio. But we also have two new brand, new recommendations today. And I think that's gonna be really exciting for everyone listening. The presentation is it's about an hour and 15 minutes long, and and we'll cut to it directly after I finish this very brief intro today. And uh, you know, it's one I put a lot of work into. So I really hope everyone in out there enjoys it. I I think there's so much background in robotics. There's so much that needs to be discussed before we can really understand where it stands, some of the challenges ahead. And once you really grasp that, the right way to build an investing strategy for this space today, because I I do think it is a great time to begin having a plan for the robotics market, but it is also a market where, you know, as I discuss throughout the presentation, you're about to hear a market like self-driving cars in 2015, it wasn't just Elon Musk promising it would happen very soon. It was the CEOs of car companies like Ford and General Motors. And there's a fundamental misread about the rate of progress. So while we want to get excited about the future that robotics, especially humanoid robotics, could lead to, we also want to temper that, understanding, you know, some of the reasons that the market may not develop as fast as some people would hope for, but that doesn't preclude, once again, having an investing strategy, because I do believe that this is a market that will continue developing well for years to come and will be aided by many advancements in AI. So just a quick check-in on what's been going on in the market the past couple weeks. One of the more notable events was OpenAI doing a massive dump on some of the most challenging math problems uh facing facing humanity. I believe they solved something like 90 of the 500 most challenging math problems. It was definitely a um what you want to call it. I mean, it's it's a very niche moment, something that isn't hitting the radar. It's not going to be something that's picked up across, you know, evening news shows, but it's something that is a very momentous time that, you know, we are seeing many of these mathematical fields starting to see novel discoveries at a rate that AI simply wasn't capable of a year ago. And um, you know, it is truly accelerating today, and we will feel this across the world. And while it feels very niche and um uh esoteric, the reality is a lot of these math proofs that OpenAI released uh could be the foundation for cracking really massive problems that will benefit humanity in ways that we can barely understand today, such as in fields like fusion energy. So I think that's something not to be ignored. And once again, it there's there's so many reasons to be concerned about AI that we see in the media. This is something that is at its basis deeply positive. Now, are there some potential negatives as far as the progress of AI in areas like mathematics? Well, yes, uh for one, computer security and uh the implications for it in the future. But as I've said in the past, you know, there's it's not gonna be a straight and up to the right path uh with AI where we're going to see some peaks and valleys, where we're probably gonna see some very disturbing situations in areas like cybersecurity along the way. And we we just need to be ready for that and and believe that the positive outcomes from something like being able to solve so much of the universe's unknowns in areas like mathematics will outweigh the negatives. Um moving on to the market, it's been a weirdly few good weeks recently. I I would like to kind of toot my own horn a little bit to say if you go back to the July 30th episode, uh, I believe that was almost to the day of now, we have seen a localized bottom in the market. Um, it hasn't been clean the entire time, but since September 15th, we have seen a relatively strong rebound. Uh, the past couple days before this podcast comes out were a little rough. But again, the past three or four weeks, I believe since September 15th or so, we've seen a rally, especially across many stocks in the AI trade. So this has been a good time and it it reminds you why our recommendation is to always stay invested. And that doesn't mean that there aren't times that you can't be more bullish or more bearish, but timing the macro environment is very difficult. And what I see with a lot of investors is they get out of the market at times that look the bleakest, such as last April when when the market had quickly entered a bear market on uh new tariffs uh being announced by the United States. And and what happens when the market rebounds, uh, they don't get back in, and then they don't want to buy in because the market has surpassed the point in which they sold and they end up missing out on rallies. So that's that's not to say that you can't adjust your excitement for certain zones. You know, we have pounded the table, and I've been very proud of the moments we've pounded the table. I think we have called nearly every localized bottom since we began this show. And those are times that you can add more conviction, but it also shows as much as the macro environment is very scary right now, um, you know, the market often sees uh rallies at times that uh look the darkest, and and that is akin to what we're seeing today. So we have today's show, like I said, the focus is on robotics. This is the big robotics episode. I I hope everyone out there loves it. Um, but we do have some exciting things coming up in the weeks to come. After this week's episode, we have an expert in the field of robotics that is uh tentatively slated to appear next week. And then after that, we have an expert in the field of natural gas. And we're gonna kind of dive into AI's impacts on energy and discuss some ways that you're able to invest kind of in the growth of AI, but in fields outside of just technology stocks. And I think we'll follow up that expert show with another show behind it where we'll probably discuss some recommendations across the energy space. So we've got two brand new recommendations today, as I noted. We've got new recommendations coming out towards the end of the month. We have a full month of content plan, including today. So I hope you really enjoy this uh presentation. I should note, if I haven't already, that this presentation it is highly visual. We put a lot of effort into making this as visual as possible. And and I do think the visual elements really make it easier to follow kind of the storyline that we've laid out about robotics leading into those brand new recommendations. So we will do our best to get a video version uploaded to the podcast services. But if if we're not able to do that or it's not working on your end, I would recommend this would be a great episode to check out on our YouTube channel because this was predominantly built to be something enjoyed on YouTube. If you have the opportunity to do that, I highly recommend it. And with that, let's kick off the presentation. Robots. They're folding your laundry, loading your dishwasher, and making your bed. They're working shifts and warehouses and building cars, they're running half marathons, they're boxing, and they're doing backflips. Well, sort of. So while AI became the market's biggest story the past five years, the question is, will robotics be the story of the next five? Today, we're diving deep into robotics. What are the storylines you need to follow? How big could the industry get? And is it poised for takeoff? Or is a future with robots cleaning your house set to remain, well, just a pipe dream? And we'll analyze five key robotic stocks I'm recommending today. We'll cover everything from a stock with a call option on the future of robotics to one that could see 300x growth from a tiny steel ball that may feel unremarkable today, but could turn into one of the largest growth industries in world history. And we'll also look at one 10x moonshot you can't buy yet, but it needs to be on your radar. All that and more next. Hello, my name is Eric Bleaker. I'm the owner of 24-7 Wall Street, and I host the AI Investor Podcast, where we recommended 58 different stock ideas across the past two years, and our average recommendation, well, that's up 136%. But one area I'm always on the lookout for is new ideas in robotics. And the reason's pretty simple. If robotics takes off, the upside could be staggering. Just consider that NVIDIA's CEO Jensen Wong has said physical AI represents a $50 trillion opportunity. Brett Adcock, the CEO of Robotics Unicorn Figure, which we'll discuss a lot in this video, has said if you can solve the humanoid robot, it'll build the biggest business in the world by a large factor. Alone under half the world's GDP is human labor, and he continued it will add tens of trillions of revenue. And Elon Musk, we're gonna hear a lot from Elon Musk today. He has said robotic production will reach 1 billion humanoids per year with a total addressable market of more than 20 billion humanoid robots. He said Tesla's humanoid robot Optimus will be worth about 80% of Tesla's value. And then there's Wall Street weighing in. Morgan Stanley estimates a $5 trillion humanoid market by 2050. The bank said the more interesting way to think about TEM is the substitution of labor. Cities Robux team estimates the market reaching 7 trillion by 2050, and Bank of America projects as many as 3 billion humanoid robots by 2060. For perspective, that's twice as many humanoid robots as cars on Earth today. So some very, some very heady figures there. Now, there is a problem that you're probably recognizing. Those predictions are a long way off. 2050. That's as far away from today, well, as 2002 was into the past. So just think back to how different the world was in 2002. And more of the point, if you had tried investing in AI at the time, you'd be way too early. You'd need trends like smartphones, social media, and cloud computing to take off well before AI had its significant impact on the market. So the question I want to answer today: how close is robotics to becoming a mainstream story in 2027? Is it too early to invest? Or on the other side, are you risking out on missing investing's next big thing by skipping out on the space? These are big questions. So here's how we're going to attack them today. First, we'll explore the history of self-driving cars. Now, at first blush, this might feel like a detour. This is a robotics video, not a self-driving car video, but it's actually really essential to understand the future of robotics. Because you might not immediately think of them this way, but the reality is self-driving cars are the largest robotics project the world has ever seen. I mean, think about we taught these giant machines on wheels to perceive the world and make decisions and do it at a level better than humans. And the advancements made in self-driving cars are rapidly being applied to humanoid robots and beyond. And yet, there is another side to the self-driving car story, and that's the time that's taken for them to become a common sight on streets across American cities, it's taken a little longer than expected. After all, networks of self-driving robo taxis were promised by industry leaders to arrive nearly a decade ago and are just now undergoing rapid expansion for the first time in 2026. The key point here, if you understand why self-driving took so much longer than predicted, well, you're going to also understand the challenges facing humanoid robotics under development today. And you'll also understand why many of the breakthroughs that are making self-driving incredibly more capable today in a short window of time could be the key to a rapid robotics takeoff in the years ahead. Second, we'll dive into the state of robotics today. There have been some incredible developments so far in 2026. Humanoid robots working in factories, robots fulling laundry and cleaning houses, and even, yes, yes, there was a robot Olympics. And we'll dive into what the most important storylines have been across 2026 and what breakthroughs are needed for a robotics to become technology's next massive megatrend. And finally, what you're probably here for, I'll give you my robotics investment plan for 2027. Five stocks, and the robotics company I'm confident has the highest chance of becoming a 10x moonshot. You can't buy it yet, but it's one investors should keep on their radar. So if that sounds exciting, let's go ahead. Let's dive in. First up, let's look back at that piece of history we needed to discuss first. Why self-driving cars were first promised to arrive nearly a decade ago, but fell chronically behind schedule. A look at predictions in the fate uh space, I should say, tells the story. In 2015, Elon Musk said the world would have complete autonomy in, and I quote, approximately two years. In 2016, he called autonomy uh basically a solved problem and doubled down on the two-year time frame.

SPEAKER_01 15:09

Well, I I mean I think I mean I really would consider autonomous driving to be basically a solved problem.

SPEAKER_00 15:16

He predicted a Tesla would go from a parking lot in California all the way to New York with no controls touched at any point by the end of 2017. Did not happen. In 2018, Musk predicted we'll get full self-driving next year. In 2019, he predicted one million robotaxis on the road by 2020.

SPEAKER_01 15:36

But next year for sure, we'll have over a million robotaxis on the road.

SPEAKER_00 15:42

And, you know, we're not picking on Musk here. He wasn't the only one who proved wildly optimistic mid-last decade. In 2016, Ford cited a plan to have thousands of fully autonomous vehicles operating in ride-sharing fleets by 2021. And General Motors Cruise Division targeted robo taxis at scale in San Francisco by 2019. Now, it's important to note these predictions weren't wrong per se. They were mostly just very early. That's because today, self-driving is finally hitting a tipping point in 2026. Let's dive into the numbers. Waymo has seen its ridership grow 10x in the past two years and now serves more than 500,000 rides every week across 10 U.S. cities. The company has announced expansion to 34 cities across the world. And we've got development in China too with Baidu's driverless Apollo Go service, providing 350,000 driverless rides per week, and it has surpassed 22 million rides in total. And let's not forget about Tesla, their gold cyber cabs. When you see One in the Wild, trust me, you can't miss it. They're in production by the thousands and are being deployed to Austin and tested in other cities. The CyberCab is especially notable because it's purpose-built. It doesn't have a steering wheel or pedals, it's built to be an autonomous taxi, and it is the true fulfillment of the vision Elon Musk had laid out back in 2015. Now, the question becomes I just mentioned that 2015 number. If self-driving was promised to be completed first in 2017, but hasn't really taken off until now, 2026, and also still has a long way to go before it's truly adopted at scale. It's still not in most cities. Well, will robotics follow a similar path of hype and missed targets? Or could a new trillion dollar market emerge sooner than just about anyone, including Wall Street, expects? To answer this, I need to discuss what happened with self-driving cars over the last decade. First, the most important point is that the first 90 to 99% of driving was just much easier than the last fraction of a percent needed to remove humans from driving entirely. The early rate of progress is what Musk and those other industry leaders I quoted earlier saw in 2015. Imagine if you saw a radically new technology going from 0% capability to 90% in just a couple years' time. It would feel borderline magical. And logically, you'd assume the final 10% would soon be solved as well, right? Well, that ended up not being the case. And it led to CEOs of companies pursuing self-driving, underestimating the remaining challenge by what some in the industry have called, and I quote, a couple hundred orders of magnitude. Which brings us to the second takeaway. Major breakthroughs were required beyond 2015's technology. Here's what Musk himself would later say. And I quote, generalized self-driving is a hard problem, as it requires solving a large part of real-world AI. I didn't expect it to be so hard. Nothing has more degrees of freedom than reality. And Waymo's co-CEO said, and I like this quote, the demo took 18 months, the product took about 15 years. That summarizes it nicely. Think about this way. This is a very revealing stat. Tesla's Robo Taxi Network recently hit about 1 million miles driven. That sounds impressive. Well, until we do a little math, because Americans, they're not driving in the millions of miles. They're driving in the billions of miles every single day. 9 billion miles driven across American roads every single day. So across more than an entire year of operation, Tesla's Robo Taxi Network has driven just 0.01% of how much Americans drive every single day. So in the early period of self-driving progress, engineers quickly solved challenges like how to make a car drive on a highway. But that's the easy part. With how much people drive, an incident you'll only see once a million miles, which is about once in your entire lifetime, well, that happens 9,000 times per day across American roadways. The bottom line here, when it comes to solving self-driving, these edge cases, they're not edge cases. They become the entire problem at scale. So back in 2015, I first got to witness self-driving cars firsthand on a trip to Silicon Valley. In a visit to Google's Secretive X Lab, the company shared details of how they were building what looked like it almost looked like a movie set on the desert, but it was simulating real-world edge cases. And then what they would do is they would write code to capture exactly how cars should react in these cases. Now, the problem, the amount of weird stuff cars experience, you've got a uh Jaywalker acting erotic erratically, a kid running around in dinosaur suits. Mine do that all the time, so don't drive by my house with a self-driving car. And weird weather events, they quickly become infinite. Trying to hand code each of these situations, well, I think the easiest way you could describe it is a never ending game of whack-a-mole. So the solution that felt so close in 2015 that got self driving up to 90, maybe even 99% last decade, it was close, but still fundamentally broken. It could make self driving cars. That were very good, but not good enough to operate the large robofleets that had been promised. So now the question is: what changed recently and led to Waymo and Tesla launching robo taxi fleets across the country? Well, several breakthroughs have emerged and we'll explore them each. First, NeuralNets ate the traditional hardware stack. And we can we can summarize this in a little bit more plain English. The brains of self-driving cars, well, they got a lot better. On a more technical level, NeuralNets trained on billions of hours of driving videos and scenarios began replacing handwriting code. So instead of a self-driving car relying on an engineer writing thousands of rules for specific scenarios and the self-driving car executing if it identified one of those scenarios, well, they moved to providing huge data sets to cars and allowing neural nets to learn and make more human behavior in these scenarios. In short, we try we stopped trying to teach and code every scenario, and cars with the additional data got smarter about how to make decisions, especially in niche events. How big of a shift was this? Well, Tesla's announcement of version 12 of its self-driving software provides some clues. Here's their announcement. FSD Beta V12 upgrades the City Street's driving stack to a single end-to-end neural network trained on millions of video clips, replacing, and here's the emphasis, over 300,000 lines of explicit C code. Now, you can see how many explicit situations were removed as we moved to something more generalized. But this leads me to another key breakthrough. The amount of data created to train self-driving cars and allow them to operate by generalizing these scenarios exploded. The genius of Tesla's strategy is that the entire company operates a data flywheel at scale. Their cars have captured more than 15 billion miles of data from customers driving with their self-driving software. This is then used to train and improve self-driving models. Now, keep this number in mind. 15 billion miles of real-world driving. Because as I detail the amount of data being collected for Robotics in just a moment, you're gonna wanna remember it. And second, I should also add that as time passes, technology gets better, we have breakthroughs in other areas of AI, and software that can simulate real worlds has also gone very good as well. Google's world models can now create realistic models of everything from tornadoes to ice on the Golden Gate Bridge to yes, my kids once again in dinosaur outfits running outside my house. These simulated worlds are then used to further train self-driving cars on how to react to millions of unlikely scenarios. So we're getting data from multiple places. And finally, I need to talk about the more time goes by, the more time that's allowed to support the ecosystem better. Tesla, they pursued a strategy mostly using cameras for self-driving perception. Waymo, they use a broader array of sensors like LiDAR. Over time, the computers inside Tesla became far more powerful, enabling larger AI models that were more capable. If you have an early Model 3, you're not gonna be able to run the newest self-driving software. In the case of Waymo, the cost of mini sensors is plummeting as technology progresses and they're produced in larger volumes. This is especially important to remember with robotics, as the cost of humanoid robots reaches, in some cases, into the hundreds of thousands of dollars per unit. Um, and and we will need to get them into the real world, which will mean costs will need to come down. But more on that in a moment. So that's enough talk about self-driving cars. It's time to ring the bell on the main event. Ding ding ding. Let's talk about what's happening in robotics today, which is always exciting. The first thing to know is that robotics is already a fairly large industry. Um, it's it's split up between a lot of places, medical care, our homes, and factories that produce, also logistics factories from companies like Amazon. And here are the figures. There were about $62 billion in Robox sales in 2025. That number, it's it's growing at a pretty fast pace as well. Sales have nearly doubled from 33 billion in 2021. And Morgan Stanley estimates there are 128 million robots installed across the world today. That number is expected to soar to 343 million by 2030. So I'll do the math for you. That's a unit growth rate of 168% in four years. And that's before the stuff we're going to talk about below really starts to take off. Because where robotics revenue comes from today and where the future is likely headed, well, they they couldn't be any more different. Um, the largest segments of robotics sales today, they're all about pre-programmed, repetitive movements. The largest robotics uh market today is industrial robots. They generate about 22 billion per year in sales and see about 603,000 robots per year installed to handle jobs that, once again, highly repetitive. Welding, assembling, palletizing, most of them look like just giant robotic arms. You you've probably seen them before. And unlike humanoid robots that may one day work alongside people, these robots they're generally caged off or sectioned away from people. Now, I should note there is a type of robot that currently works alongside people in factories named collaborative robots, but these robots are still a tiny market at only about 1.2 billion per year. We get other markets, logistics robots that operate in areas like Amazon warehouses are about a $5 billion industry. And medical robots might surprise you the size. We've got companies like Intuitive Surgical that have grown this into a $17 to $19 billion industry. Consumer robots are where we see a lot of the volume of robots, things like iRobot Vacuum Cleaners. They're a large market. There's 34 million of them that have been sold. But as of today, most are built for those very specialized use cases: mowing your lawn, um, vacuuming your house, etc. And you also have markets like drones and military robotics that are growing extremely rapidly. But where we are most interested in today, where we are arriving at at the end, is humanoid robotics, the kind we opened the video with, where you saw videos of them folding laundry, working in factories, and once again, doing backflips. Kind of. It's the anticipated growth in this type of robot that's fueling the predictions I led this video with. Like NVIDIA CEO Jensen Wong saying 50 trillion in physical AI sales could happen in the future. Musk predicting a billion humanoid robots sold per year, and Wall Street firms predicting robotics reaching 7 trillion in sales by mid-century. That would be a growth rate of more than a hundredfold from today's levels. But here's something critical to know. Only about 7,000 humanoid robots were sold over the past year. So while many forecasts have them growing to be a hundred times the size of other robotic segments, today their sales, well, they're a rounding air for the $62 billion Robux industry. In fact, humanoid robot sales are just 1% the size of the legacy industrial robot arms that uh I talked about earlier, which make little use of AI, don't work around people, and perform repetitive tasks over and over. But but here's the thing: even at this tiny level, humanoid robotics in 2026, well, the best comparison I can give is they look a lot like self-driving cars did in 2015. Remember that quote I showed earlier from the CEO of Waymo where he said the demo took 18 months, the product took about 15 years. Well, the year self-driving demos truly went mainstream. That was 2015. That was the year early progress in self-driving cars began blowing people's minds. It was the year when Tesla first released autopilot to customers, and you suddenly had the videos all over the internet of people driving with no hands. Um we've moved past it today, but man, at the time it felt like what a time to be alive. And Waymo conducted its first fully driverless public ride in 2015, driving a blind man across Austin in what the company had called its kitty hawk moment. Of course, that referring to when we first achieved flight. The announcements around humanoid robotics in 2016 feel um just as mind-bending as what we saw from self-driving cars back in 2015. Here are a few of my favorites. We we talk about these on the show quite often, but they are cool. Figure, a lean robots company, now valued at 39 billion, has had a busy year. In May, Figure ran a live stream of one of its humanoid robots sorting packages. The company originally planned to do just an eight-hour live stream where it actually was competing against a person, but due to massive public interest, the live stream kept going for more than a week. It ultimately ran for 200 hours, and Figure's robot processed nearly 250,000 packages and was roughly in line with the person it was competing against, who obviously they didn't go for 200 hours. Uh they they quit after the first day. In June, the company's newest robot began operating under dynamic conditions in BMW's Sparnberg factory. So rather than performing a repetitive task, like robots in today's factories, figures robots are performing sequencing. So that requires manipulating heavy carts, sorting through bins where objects can move and stack on top each other, and also significant body control to be able to take some of the objects and be able to move them to the next part of their line. So this is a significant breakthrough in capability. And then in September, Figger sent robots to 30 unfamiliar homes, and with no home specific training, they managed to tidy the homes, fold laundry, make beds, and well, perform the cleaning tests you would expect from a home cleaning robot. Now, beyond Figger's announcements, Boston Dynamics Atlas robots are training in auto plants in Georgia. Uh Hyundai, which owns Boston Dynamics, is targeting manufacturing capacity of 30,000 Atlas robots annually. That line alone would produce more than four times last year's humanoid robots market. Agility Digit Humanoid is working at customers like GXO, Toyota, Amazon, and Scheffler. At Toyota, the robot has moved from a pilot program to a commercial program. And then OneX, you probably saw this. We we had talked about earlier. They announced a robot for the home. It's going to involve some teleoperation, but they sold out of their entire allotment a year's worth of humanoid robots, 10,000 of them, in just five days. And if you live where else for this story, but in in San Francisco, yes, you can now have robots clean your home too. Tau Robotics has launched a home cleaning service for $30 an hour. And I want to key in here. Well, roboticleing houses and working in factories, they steal the headlines. They're what you see on the news, they're they're what gets absorbed in social media. In the background, major breakthroughs on the software side, or you know, the brains of robots are happening. Google announced Gemini Robotics 2 with whole body control and cross-robot transfer. NVIDIA released Groot N17, which is a full-stack robotics control platform, and it adds some things that are important for the future, like generalized robotic skills and advanced dexterous controls for things like hands. And companies are beginning to announce one-shot or zero-shot models that allow robotics to generalize and perform tasks with previously unseen objects or in entirely new environments. So I do want to pause here. And I want to take just a minute, just a minute to dive into that last breakthrough. Because it's easy to just go buy it, but but it's really important. Um trust me when I say you're gonna want to understand this if you want to fully understand the implication of where robotics could be going. So, as an example, when Figure announced the robots had visited and cleaned 30 homes, they made sure to specify it wasn't just they visited 30 homes, it was 30 unseen homes. Because what Figure is doing here is they're trying to demonstrate that robots no longer need to be trained for specific scenarios. There's no fine-tuning, no data collection, or adaptations to any of the homes or objects. Figure's uh Helix 2.5 robot had to manipulate to clean the 30 homes. To put it in as plain English as possible, Figure's robots now understand which objects need to be picked up. It's not from a list, but from generalized, well, we could call it judgment, similar to how our brains work. It can perform tasks that now require incredible dexterity, like folding laundry, but are doing so in new environments rather than a lab where conditions are always the same. And they're performing tasks like making beds in rooms, well, rooms they've never seen before. So, what's different about 2026 is for years robotics demonstrations were always showing in controlled lab settings. But for the first time, we're starting to see them in the real world. And this is critical because whether we're discussing self-driving cars or humanoids, generalization is the holy grail of robotics. If we look back to self-driving cars, they quickly progressed from 90% to 99% of capabilities, but they hit that dead end we talked about because they couldn't sufficiently generalize for real-world settings. Hand-coating edge cases in self-driving cars or training to very specific situation humanoid robotics are both dead ends that don't work at scale. Because once a billion cars are on the road or a billion robots are in factories and in homes, as I've said earlier, the edge cases are limitless. So the question becomes if robotics gets very good at generalizing very fast, how quickly can the future arrive? Because self-driving was first predicted by Musk to arrive in 2017, but it's taking off today. Once again, this is this is thanks to breakthroughs and things like N10 neural nets, smarter brains, data collection, that massive amount of real-world data to trade on, and the supporting ecosystem and the components needed for self-driving cars to get both more capable and cheaper. And humanoid robots today, they're already benefiting from each of these breakthroughs, from each of these advancements. Robotics models keep getting better and are becoming massively more capable as large data collection projects are launched. Which brings us to that all-important question we asked at the start of the video. Could humanoid robots avoid the dead end that stalled self-driving cars for so long? Do they have a path that with enough real-world data and enough investment, seeing them deployed at scale could happen much faster than the market expects? Let's put it this way: could that quote that self-driving cars, the demo took 18 months, the product took 15 years, could it become for humanoid robotics, the demo took 18 months, the product took five years? That's that's the big question. One that could swing trillions of dollars of stock market value in the coming years. So I want to answer it as fairly as possible and look at the data coming out of the robotics space. First, let's look at some of the recent data that makes me really optimistic, right? But this is an investing show. We're not cheerleaders. We are trying to look at things as clear-eyed as possible. So, what I'll do second, I'm gonna balance some of this really optimistic data with some of my own concerns and where I can see this market still running into problems. So let's start with the bold case for humanoid robotics. One of the biggest stories in AI over the past half decade has been the continuation of something called scaling laws. As companies like Anthropic and OpenAI have released new models, they've been continuing on a trend line where performance improves predictably as the amount of compute and data along with model sizes have increased. It's exponential growth, but exponential growth along a trend line. And here's what's exciting there are early signs that this dynamic is playing out in the models that are now being constructed for robotics. So a few quick examples on this. Dyna, they released data on their world action model trained on videos of humans working. So they trained the model on a thousand hours, 10,000, 100,000, and then a million hours of footage. And then they scored the robot's ability to perform 15 real-world tasks in terms of a percentage of how much of the task could be completed. When trained on a thousand hours of footage, the robots scored 20% completion on tasks. On 10,000 hours, they scored 28%. On 100,000 hours, robots scored 45%. And at a million hours, the score jumped to 53%. This is early evidence that scaling laws of more human video leading to predictable robot performance gains is holding. Generalist AI is seeing another version of the same phenomenon. Its Gen 1.5 model can watch just uh three to 12 second demonstrations of new tasks and attempt immediately without retraining. Across 10 tasks, it succeeded 59% of the time. But if you give it roughly five minutes of demonstrations and just 10 training updates, the success rises to 83%. So generalist says these abilities emerged from this broad pre-training rather than being explicitly engineered into the model. And this is what we're talking about: this ability for the brains to take data and be able to create general uh ability to execute weather instead of us having to explicitly engineer, because again, explicitly engineering leads to dead ends at scale. And Figure is attempting to create the same data flywheel Tesla built in self-driving cars, but in human robotics. The company launched its index data collection platform. Humans wear a uh they they wear a recording device. There's a little something dystopian about it, but they're wearing a recording device that captures them working all day. And Figure is now receiving 43,200 hours of video every single day that's being used to train its robots across general tasks. And here's what's impressive Figure said applying this data has led to a nearly six-fold increase in their humanoid robots capabilities, which is to say, the game plan that's led to massive gains in both AI models and companies like OpenAI and Anthropic, and we all know how successful they are, and breakthroughs in self-driving cars as they finally reach their tipping point in 2026, they do appear to be working in robotics. New, more capable models. Once again, the smarter brains combined with more massive amounts of data from the real world are lean to capability gains that few thought possible a few years ago. Now, that's great. Here's where we need to just throw a little cold water on the excitement and provide some caveats to what's going on today. Because remember how I told you to remember a specific figure earlier in this video? It was a number. Do you remember? I said it would be coming back later. I know we've talked about a lot, but it is 15 billion. And that's the amount of data in Miles. Collected to train its newest self-driving systems on real-world events. Well, compare this to some of the figures we just discussed. 15 billion miles works out to something in the range of about 500 million hours and 57,000 years of driving. Dinah's model, I discussed earlier, that saw the early scaling laws evidence, was trained on just a million hours of training data. Generalist AI's was trained on 500,000. And while Figgers index project is ambitious and in its early innings, it's on pace to collect about 16 million hours of video per year. That's about 3% the data Tesla collected to crack self-driving cars. And it should be noted, not all data is created equal. Tesla's data is four cars, and it's being applied to the same cars that will use self-driving software. They all have four wheels and pretty similar bodies. If you collect data of humans working, well, it must be translated to robots that have different form factors. This increasingly brings up a risk that you hit a limit somewhere in a strategy like this. Also, data might not be especially useful for very valuable tasks like robots and factories, where you probably do need to get to a level of more specialized training. Well, Figure it has a project of humanoid robots in BMW's car factory. To date, these robots have logged just 1250 real-world hours conducting very narrow tasks. So you see, this is something that's very much still in the earliest stages of a demo versus something that is executing at scale of the things we know have led to significant progress, which is just to say, it finally became clear to Elon Musk how challenging self-driving cars would be when Tesla reached 90 to 99% of capabilities. And at that point they realized the most valuable capabilities were the remaining fraction that can't be solved. For this particular task, getting to 90 to 99% didn't mean a lot. All the value was in the remaining last few percent. So with humanoid robotics, the early results, they are very promising. But we need to caveat them as that. Figure their O3 robot reportedly costs around $25,000. That's expensive, but it's also impressive. That's because their last robot, Figure O2, reportedly costs $100,000 to $150,000 per robot. The reduction in cost is gonna increasingly come from growing scale. Their Figure O2 was built in small batches. Well, they're now building a bot factory that is initially designed for 12,000 humanoids uh created per year. And over four years, with expansion plans, figure plans to produce 100,000 humanoid robots. And looking at other companies, prices have dropped dramatically in recent years. One X's Neo Robot uh that sold the 10,000 units in five days I talked about earlier, that has a $20,000 price tag. And competing humanoid robots from China are, well, not surprisingly, even cheaper. Unitree's G1 costs about $13,500. Now, challenges definitely remain. The price tag for robots can rise dramatically with factory adaptations or things like more advanced hands. Which, you know, it it shouldn't be shocking, but it is kind of shocking that some of the most advanced robotics hands cost $150,000 each. But when you think about the precision that needs to go into building a hand for a robot, it is one of the most complex engineering challenges where there's still a lot to be settled and what the optimal way to build it is. But I do need to highlight, this does demonstrate a second flywheel beyond data that begins to crank once robots are built at scale for the real world. And that is the cost to build them will continue to plummet. How I talked about in self-driving cars, the more time that went, the more units you started building, the more the cost of components came down, the more capable they became. We are at the very beginning of seeing this flywheel work in humanoid robotics. Okay, now that you've learned probably more than you ever wanted to discover on the background of robotics, let's talk about how this could impact you. And I'm not talking about having a robot butler in your house, although once again, you can pay that robotics company OneX $20,000 for one of those today. They're sold out, but maybe you can get the next batch. I'm talking about investing in the space because I hope this video, I hope it's been a balanced look at Robux. A lot of people have been searching for this Chat GPT moment. They have this expectation that one day an announcement's gonna be made, and it's gonna be clear that there is this takeoff in robotics where their capabilities are about to suddenly explode and they're waiting for that moment to invest in robotics. But as I've shown today, evidence is building that robots at the beginning of a curve to get dramatically more capable in the years ahead, but it's driven by many of the breakthroughs that led to self-driving cars reaching their own tipping point, and it's gonna take years as improvement to the foundational models for robots. Once again, what we call their brains, continue improving. And it's gonna take a long time as we continue to collect massive amounts of real-world data to be able to continue improving the capabilities of Robux. So I like to think of this as rather than one specific breakthrough, I would expect continuing progress and more capabilities in the years ahead. And that's why my advice for anyone watching or listening today is to start building a strategy for robotics today. If you're looking for the next era of Robux to take off in 2027, as I just mentioned, that's probably not going to be the right way to look at this market. Humanoid robots won't hit the projections of Lawn Musk of a billion units annually by 2030. But I do believe we'll be looking at hundreds of thousands produced, perhaps if the right situations happen exceeding a million by then. And you combine that with the growth of other robotic spaces, again, you have massive acceleration in areas like drones, um, then you have a situation where the most likely scenario is that robotics will surpass the targets that Wall Street has set for the near-term future in years like 2030. So here's a few other areas that excite me and why I think investing in robots is the right move today with the right strategy, which I'll discuss some of the stocks I think look most opportunistic today. So, you know, if we look at AI, we just have to be very clear. A massive reason for AI's progress the past five years is how much capital flooded into the industry. Anthropic and open AI combined have raised about $300 billion. But, you know, that pills in comparison to what's coming from companies like Google, Microsoft, and Meta. They were all flush with cash and they've all poured it into AI as a priority. And, you know, this investment is going to be needed to solve the chicken and egg problem that's present in robotics currently. To make robotics advanced enough to work in factories, they need more real world data. But to get that real world data, billions need to be spent on pilot projects that often have minimal initial ROI. And robotics, what I hope I demonstrate well in this video today is I think they've now shown enough progress that we are getting past this chicken and egg problem, that this is now more of a snowball rolling down a hill, and significant capital is going to flow into the industry. Bigger AI, it's valued at $39 billion and uh has raised a little less than $2 billion. So that valuation is going to allow the company to do things like commit a billion dollars to that real-world data collection project I had discussed earlier. It's gonna allow them to begin construction on Bot Q Factory that aims to produce 100,000 humanoid robots across the next four years, and it's going to allow them to invest $3.5 billion in data centers specialized for robotics training. And figure it isn't alone. Venture investments into robotics were only $6.9 billion in 2023, and they didn't really move in seven uh in 2024 as investments started really flowing into other AI adjacent spaces, reaching just 7.2 billion. But in 2025, they saw a massive jump to 15 billion. And through June of this year, venture investment in robotics is at a $37.6 billion run rate, which is all to say the capital to make these early advancements to scale, it is arriving today. Second, as anyone who's subscribed to the AI Investor Podcast knows, I'm a believer in the progress of AI in general. Self-driving cars, they've benefited tremendously from broad advancements in LLMs that have contributed to more specialized self-driving models. Likewise, the largest models from Meta, Google, Anthropic, and OpenAI are all showing the beginning of signs of recursive self-improvement. That's a phrase uh where it demonstrates that models have become basically good enough to begin improving themselves, often with minimal humans or sometimes even none in the loop. And recursive self-improvement has um already been seen, some versions of it in other industries. OpenAI used elements of it to design its jalapeno, it's very spicy computing chip, which shocked the world and had publishers trying to break down how it was as capable as leaving chips from NVIDIA in an extremely short amount of time. And the reason is once again, because they're using elements of recursive self-improvement in that chip design. And I have every reason to believe gains from recursive self-improvement will soon be targeted at robotics themselves, robotics models, because the size of robotics, I I showed at the beginning, it could become the largest industry on earth. Again, the TAM of robotics is going to be labor replacement of areas. Um, so when you look at something that large, well, it means these companies that they are focusing on AI models today because of the potential for returns. Wouldn't you expect that every large technology company is going to have some kind of incentive to create their own robotics projects at significant scale in the not so distant future? Simply put, there's too much money to be made on robotics to not put a focus on it uh from RSI very soon. And speaking of AI, its build-out is also the next use case for robotics. So leading hyperscalers like Google, Microsoft, Amazon, SpaceX, Oracle, and Meta, they plan to spend a combined $1.4 trillion on capital expenditures, mostly for AI data centers in 2027. And robotics will have massive applications, not only for use inside data centers, but also in the factories to build the chips that then fill data centers. For example, Elon Musk plans to build Terrafab. Uh, that's a chip factory that very modestly would be uh about 19 times the size of the Pentagon if he reaches his goals. And one of the goals is to use Optimus robots inside Terrafab to build the chips that are being built to power still more robots. I'm sorry if that broke your brain a little bit. It is a bit circular, but we know that Musk has a history of predictions on overly aggressive timelines. Um, let's look at some real-world data because we are starting to see this actually playing out from companies today. So Pterodyne, and that's ticker symbol T-E-R, has been one of the greatest robotics success stories in recent years. Shares are up more than 200% since the beginning of 2025, and in large part is thanks to a lucrative Robux contract with Amazon. Now, I talked about logistics a little bit earlier. It's a massive market that robotics will continue expanding to in significant ways in the coming years. But recently, Teradyne, they've started talking about a new opportunity beyond logistics. At a recent uh investor event, Teradyne said electronics manufacturing into the semiconductor uh supply chain that includes robots and semiconductor test, assembly, and data center operations is their fastest growing vertical in robotics. And they've called this the most important trend that's going to drive our business over the next few years. So, again, it's happening in the real world already. And let's talk about specific robotics stocks that are part of the game plan. I'm building for the space. I know that's the reason a lot of people are here. You want to understand what's going on with robotics, but you also want to understand how to invest in it. The good news, today is a much better time to invest in robotics than a few months ago. Throughout April and May, robotics hype soared. As many of the demos I detailed earlier, they're, you know, they're all over social media, they're getting picked up. The hype is growing. But when we hit June, it was almost like a pinprick, you know, poop, letting the air out of a balloon on a lot of these stocks. And, you know, AI stocks that had seen the most momentum during the rally in April or May, they generally retreated the most. And many robotic stocks are now down from recent highs about 30 to 50 percent. Now, if you're a long-term investor in the trend, this is great news because many stocks that started having large amounts of robotic sales on uncertain timelines built into their valuation, they're now trading at more reasonable rates. Often, you can get the call option, you can get the optionality of robotics growth for very cheap relative to just their core business. So let's start by detailing three stocks I've already recommended in the AI investor podcast, but are likely new ideas to most investors watching this presentation if you haven't previously subscribed. First up, let's start with the Regal Rex Nord, ticker symbol RRX. The company is worth about 11 billion in trades for about 13 times next year's earnings. You know, the story with Regal Rexnord from a very high level is not gonna be exciting. It's a diversified industrial company. They make a wide array of power transmission, automation, and power efficiency motors. I know. Still your beating heart. It's pretty boring stuff. But here is what's important the company has a few powerful callus that could accelerate growth in the years ahead and accelerate in a very significant way. First, the biggest trend in data centers is modular designs. I've talked about that on the podcast. I've had guests like John Rotanti who talked about how big of a deal this is in the industrial space. And as I mentioned earlier, hyperscalers are expected to spend about $1.4 trillion mostly on data centers in 2027. And they're looking for suppliers that number one, haven't already sold all their capacity, especially in some in-demand areas like Switch Gear, and also have solutions that help data centers get built faster. It's it's very much a race right now to see who can get compute built and online the fastest. And Regal Rexnort has created an e-pod container that houses power management systems that are an extremely important component of data centers and often in shortages, and they're being built by the company. Regal entered this market only in early 2025, but they already have 735 million in orders that will ramp throughout 2027. What's important here is this company has something like $6 billion in annual revenue. So a product like this is a substantial ramp. And second, many of the company's boring product lines stand to take off if humanoid robotics sees substantial growth in the years ahead. Regal Rextern has an unusually broad portfolio that applies to humanoid robotics with precision motor control hardware like motors, gears, brakes, bearings, and actuators. Keep in mind, motion control components are about 50 to 60 percent of humanoid robots cost. And this is Regal's sweet spot. And here's a little tease just for some more recommendations we'll have later. I'll detail why one of Regal's product markets could grow more than 300 fold to become a $250 billion market by mid-century if humanoid robox take off. Um, we'll we'll get the details of that in an upcoming recommendation, but it does apply to Regal Rexnord as well. And Regal's stock had run up in early 2026, along with many stocks in the robotics space as investors raced. They were turning over stones, looking for any company with exposure. But again, once the market is fickle, once attention went elsewhere, we saw a lot of these companies return to the prices they trade at before then. And I wouldn't be scared by that. You know, this is an opportunity, um, again, because that hype for robotics has passed. And now what we have remaining is Regal is a fairly low-risk way to play the growth of robotics. If robotics takes off faster than expected, this is gonna once again be a name that becomes very popular. And if humanoid robotics stall in the years ahead, well, there's not a significant amount of downside, and the company could still see strong growth from its data center offers, right? So you're you've got you've got multiple tailwinds here. So I know I started out, this is probably more on the boring end of the spectrum, but I wanted to give, you know, a group of recommendations, including lower risk stocks. And I like this idea of something that has a call option on the growth of robotics, because that could be something I think a lot of investors out there are looking for. And next up, we'll look at a company, we'll go for a little more upside here, and that's ouster ticker symbol O U S T. The company is worth about three billion dollars and is a leader in physical AI. Um, you know, where they really have their kind of core appeal is robotics sensing what's happening in the world. Their biggest business is LiDAR, which, if you're not familiar with it, shoots millions of laser pulses, times the return, and you create a 3D map of the world. It's controversial technology. Elon Musk famously passed on it in Tesla's self-driving cars, but other companies like Waymo and other self-driving companies have generally embraced it as an essential technology. Uh, the key reason here is it just it adds a cost to the car. And some companies have made a decision that it's worth adding the cost, and and some like Tesla have really worked as hard as possible to only rely on uh a limited number of sensors like cameras for sensing. So while some companies are uh choosing to use LIDAR and others aren't, the bottom line is Auster sales in the space continue to grow at substantial pace. LIDAR unit volume grew 70% last quarter. In addition, ouster added camera technology that's specialized for robotics with its acquisition of stereo labs. And this acquisition adds to Ouster's overall pitch that the company is the leading unified sensing and perception platform. Robotics of all kind need to sense the physical world. And Auster's platform, it combines LIDAR, specialized cameras, compute, software, and AI models. So, you know, it has competition from predominantly from China, but a couple things on that. Number one, they maintain some technology advantages over top these companies. And also, there's just a growing concern about Chinese companies having access to such sensitive data, such as sense. Data of what's going on around every robot in the physical world. So I think increasingly this is a part of the supply chain that companies are going to look to non-Chinese suppliers and they're going to find the best option in Ouster. So to be sure, Ouster is on the riskier end of robotic stocks we'll discuss today. It trades for about 10 times 2027 sales, and profitability is expected to come until 2028. Yet it not only gives a pathway to humanoid robotics, but also self-driving cars, drones, logistics robots, and even other things that might not be an area you've thought about, but do have impressive growth, like smart cities. Its end markets are generally across the board very attractive, and the components it sells have a high, uh far higher chance, I should say, of being differentiated versus many of the other picks and shovels plays that investors will often select in the robotics space. And a final prior recommendation I have made in the AI portfolio, but I'll be discussing as part of my five Robox stocks today, ist Micro, which is Ticker Symbol STM. ST Micro and in a risk scale, it's probably somewhere between Ouster and Regal Rexnor, uh probably a little closer to Regal, though it's it's not especially one of the larger, or I should say it's one of the riskier stocks as it's large and diversified. It's it's a $52 billion company, and its shares bottomed in early 2025. And there's a simple reason. The company was heavily exposed to the auto industry, uh, which uh auto sales to semiconductors have been in an extremely long malaise. So in 2023, before this malaise began, its adjusted profits were 446. In 2024, they fell to 166. In 2025, they bottomed at 53 cents. But what's important is the rebound has now begun with per share adjusted profits expected to hit one 134 this year, 254 in 2027, and 387 in 2028. In the near term, a significant callus for ST Micro is NVIDIA. Thanks, NVIDIA. They are pursuing a massive shift in the way data centers are built. What they're going for is something called 800V DC power architecture. And what it's going to lead to is significantly more revenue for companies with a specialty in creating power management chips made out of compounds like silicon carbide, or you'll see it as SIC and gallium nitride, which you'll see as GAN. S T Micro, they specialize in both and they have a complete portfolio aligned with NVIDIA's new data center designs. Now, that data center tailwind should provide growth in the coming years, but after that, what I think could be so powerful and makes ST Micro an above average recommendation is that a robotics tailwind could take the baton from this data center surge. ST Micro makes more than 500 components that are applicable to humanoid robotics, ranging from processors to motor drivers to sensors and power devices. So, you know, at the end of the day, what happened earlier this year is shares of ST Micro skyrocketed across the spring as investors were racing to find winners in robotics and also who was going to benefit from this new architecture that Nvidia is going to be pushing across the coming years. But once again, when many of the stocks that had seen the largest gains, they kind of fell back as many investors rotated out of AI in July and August. Well, we got many of these stocks trading at significantly better prices. And that's true of ST Micro, it is down 30% from recent highs. That presents a much more attractive entry point for a company that's not only one of the highest likelihood winners from this massive data share center shift in Videos making today, but also if humanoid robotics takes off sooner than expected. So let's quickly summarize what we have so far. We have Regal Rexnord that has a safer play with a call option on the growth of robotics. Ouster, a leading sensing platform that adds growth in markets like self-driving cars, and ST Micro, a winner of NVIDIA's latest data center shift, and one of the most likely picks and shovels winners if humanoid robotics takes off. Now, once again, I've previously recommended all these stocks to listeners of the AI Investor Podcast, which I should mention, if you haven't subscribed yet and you have enjoyed the content of this video, the best way to get continuing stock recommendations uh in AI and robotics and get continuing news and analysis is subscribing to this channel and also setting up an alert. We publish new episodes of the AI Investor Podcast every single week. We break down the biggest developments across the AI space, but I also recommend a portfolio of stocks. As this video publishes, I'd recommend 58 different positions. I'm about to recommend two more across the past two years, and the average return has been 136%, which it should go without saying is absolutely smoking the broader market. Getting future stock recommendations and updates, it's totally free. All you have to do is subscribe. And if you're looking for more on-the-go financial news, well, you can also subscribe in other places. The AI Investor Podcast is available on Spotify, Apple, and other major podcast providers. Now, with you properly subscribed to the channel and ready for future updates on these brand new two recommendations I'm about to share. Let's get to the next stock, which is Timkin ticker TKR. I just mentioned past recommendations to the AI Investor Podcast. Well, in late 2024, I went big into optics recommendations. Some winners from that time include Credo, up 683%, Lumentum, up 1187, Marvell, up 194%, Coherent, up 192%, and Siena up 526%. Now I've continued making optics recommendations since and identified some other winners like Applied Optoelectronics up 420%, air test systems up 154%, and Semtech up 154%. Those last two happened just earlier this year. And you look at these returns, and it looks exciting. You know, not only that, but stocks like Lumentum and Marvell, they've become some of the story stocks, some of the most popular stocks among retail investors in the entire market. But the truth is, when I first recommended optics, it was a total backwater. No one was paying attention to. So the question is what boring backwaters today could you know quickly turn into strategic growth areas if robots rapidly accelerate and become the next big thing ahead of timelines Wall Street and most investors are expecting. One area to consider is bearings, which I know you're probably doing a double take. Bearings? Uh it sounds it sounds a little bit like uh from the graduate. The you know, the future is in plastics. Um, but the stats around bearings will probably blow your mind. Morgan Stanley predicts bearings demand for robotics will surge from less than a billion today to more than 250 billion by the middle of the century. Most of that demand is going to come from humanoid robotics. But markets like self-driving cars, drones, and home robots, they're also going to have their own contributions to this trend in Morgan Stanley's research. And the reason is every single motor in a robot requires at least one or more bearings to reduce friction and support rotating parts. And there's going to be a lot, a lot, a lot, a lot of rotating parts in the future of robotics. Now, there are plenty of names to look at in the broader bearing space that are publicly traded. As I had alluded to earlier, Regal Rexnord is another play in this area. There's RBC bearings and a German company named Scheffler, but Timkin is the next name I'm zeroing in on. It's an $8 billion company that trades for about 16 times forward earnings. A look at the company's recent history reveals uh look at that special. Sales hit 4.8 billion in 2023. They're expected to land at 4.8 billion this year. Um, not off to a good start, but when you look under the hood of this company, it reveals a deeper opportunity. Timkin has been extremely ambitious at building a portfolio all around being a leader in not just bearings, but robotics movements. So this includes acquisitions to build out that portfolio in key robotic segments like harmonic drives, high precision drives, linear actuators, and motion systems. I know these might not sound like household names today, but optics didn't sound like a household name two or three years ago. If I wrote an article about optics, no one would read it. And something like linear actuators is rapidly going to be something similar if robotics really exceeds expectations in the coming years. So while Regal Rexmark was more of a cautious play, it was aiming to reduce risk. Um, it has another tailwind from data centers. Timkin, this is an attempt to lean in more. This recommendation has more torque, it relies more heavily on robotics being seen as an extreme growth market for it to play out. But I like having some concentrated bets on robotics, in addition to some companies where their growth can be more of that call option. And finally, we're gonna arrive at the last recommendation I'll be uh discussing in this video today. This is another new stock, a new recommendation to the AI portfolio. I considered a couple different positions. There's aometry, um, ticker symbol XMTR. It's a marketplace for AI generated ideas with manufacturers. I like the business a lot. You know, it honestly feels really good for the early stages of robotics and generally just what's going to happen across the world the next few years from the broader advance of AI, but it is expensive, trading for about 30 times 20, 30 earnings, and it currently trades near 52-week highs. So for now, I'm gonna keep that one on the watch list. I also did a little bit more research into Vichy Preston Group that's ticker VPG. The company specializes in force sensing, which tells robots things like how tightly to grip an object. You don't want to grip an egg with uh extreme strength, uh, otherwise you're not very capable. Vichy is a stock I have had on my radar. I believe I previewed it um earlier this year in a podcast, but it had surged a lot in May, and shares have now collapsed like many other companies in July and August. So they're they are currently about 50% off their 52-week highs. But one problem with this company is it requires the highest degree of um patience as it relies on advanced robotics to see growth. Sales are expected to move from 344 million in 2026 to 395 million in 2028. That's growth to be sure, but it's also trailing the broader robotic space. So, what I'm gonna do is I'm gonna keep that stock on my watch list for now. And if my confidence in humanoid robotics accelerat faster than expected increases, it's probably a name I'm going to look to add. But instead, what I added for the fifth and final stocks is Delegro Microsystems that's ticker A-L-G-M. It's a stock I discussed earlier this summer, but ultimately thought the valuation was too rich and I added it to a watch list instead. Well, with the stock down 50% or thereabouts from recent highs, trading below level scene back in February. Well, I'm ready to add shares today because Allegro's specialty is magnetic sensing and power management chips. Historically, its products have sold mostly into the car industry, specifically for EV and some of the advanced safety systems. As I mentioned earlier with ST Micro, that has been a horrible market for companies that have been involved in. But the ray of light, the light at the end of the tunnel, is many of the products designed for EVs are now rapidly being installed across data centers. So this was a really attractive opportunity when you could add these stocks on the cheap and you were able to get the further out opportunities of EVs being down, but they're going to get some transition to data centers and potentially in the more distant future, robotics. We saw the stocks go ahead of themselves in summer. And now with more reasonable evaluations, they become far more attractive. In Allegro's case, this meant its data center revenue was almost nothing in fiscal 2025. Then it quadrupled in fiscal 2026, which was its uh last year, to reach 10% of sales. And the acceleration is continuing in recent quarters. In the company's most recent quarter, data center sales hit 17% of revenue. So from nothing fiscal 2025 to 17% of revenue and continuing to accelerate in the most recent quarter. The company sells isolated gate drivers for silicon carbide, which once again is SIC and gallium nitride GAN. I already talked about these, you know. Again, they're not household names, but if you want to understand this massive architectural shift that Nvidia is pushing, um, companies with expertise in these compounds, they're they're generally going to have some significant growth opportunities. And that's what Allegro is seeing today. And they produce a large host of sensors applicable to robotics. They expect a 30-fold increase in their revenue opportunity from today's consumer robots transitioning into humanoid robots. So the math here would be that at a million humanoids, they see their opportunity at about $150 million annually. And if we reached a point that the market size had increased to 5 million humanoids, their opportunity at that point grows to $750 million. That might not sound like much. Um, and and as I've cautioned many times in this video, humanoid robotics, it's going to take some time to scale. But it's also worth knowing that Allegro is a small company with just $890 million in revenue last year, and it has that existing data center catalyst, which we can allow to play out while we wait for this robotic situation to continue developing. So if excitement for robotics builds in the coming years, well, this is going to be a stock that many investors are going to flock to because they have a genuine niche and expertise that's going to be highly applicable to the future of robotics. And there we have it: a complete overview of the robotics market and a game plan for five leading robotic stocks you can buy today. Now, before I sign off, I want to mention two things. First, as I mentioned earlier, I host the AI Investor Podcast every week, but I also own 24-7 Wall Street. If you've been looking for a source for financial information, I strongly encourage you to search 24-7 Wall Street and Google or go directly to the site after this video and just check it out. The more you use the site, you'll be able to get on our email list where we send out our top articles with stock ideas and content like the video you have watched today. And I need to mention as well, our goal has always been to make 24-7 Wall Street and the AI Investor Podcast a community. We respond to listener comments and have recommended stocks that come from our community. So if you're watching this on YouTube uh or Spotify, uh thumbs up, a show rating is always appreciated. But we also love comments and would love to hear about the top robox stocks from the community as well. And last, I did promise a 10x moonshot today that's not a stock you can buy yet, but needs to be on your radar. And that's a company you might be able to guess from the contents of this presentation already. It's figure AI. The company has pursued its own data strategy. And as I discussed earlier, they're now collecting 43,000 hours of video data every day. They're building a robotics factory with the goal of producing 100,000 robots over the next four years, and they're making a large $3.5 billion investment and compute for their robotics model. So, simply put, there's gonna be a lot of humanoid robots companies probably racing to cash in on excitement uh very soon. We saw a recent one out of China that I think was oversubscribed by 8,000 fold. But my recommendation is if you want to buy a robotics company today, a broad robox company, wait for figure. With anthropic and open AI readying their IPOs, the company may target the IPO window that that is open today. And if it does, that's gonna be a stock that I'm deeply interested with. So, with that, I think I've covered everything. I I promise today it's it's been a marathon, it's it's been a longer video, but I hope you agree that this is the most thorough and accessible video for investing in robotics that you'll find on the internet today. And as I mentioned, if you're not subscribed to 24-7 Wall Street, if you don't have notifications on and a subscription to the AI Investor Podcast, you're going to miss out on future ideas and recommendations as the field of robotics keeps moving forward. So if you haven't yet, one last reminder to make that happen. And that's all from me today. I appreciate everyone for tuning in. I hope this has been enlightening, and I hope you join the investing community we're building at 24 7 Wall Street.

SPEAKER_01 1:25:12

The AI Investor Podcast is for educational purposes only and should not be considered investment advice.