Talk Your Book: Investing in the Rise of the Robots

0:00Ben Carlson:Today's Animal Spirits Talk, your book, is brought to you by CraneShares. Go to CraneShares.com to learn more about their Rise of Humanoid Robots ETF. That's ticker COID, K-O-I-D. It's CraneShares.com to learn more. Welcome to Animal Spirits, a show about markets, life, and investing. Join Michael Batnick and Ben Carlson as they talk about what they're reading, writing, and watching. All opinions expressed by Michael and Ben are solely their own opinion and do not reflect the opinion of Ritholtz Wealth Management. This podcast is for informational purposes only and should not be relied upon for any investment decisions.

0:36Clients of Ritholtz Wealth Management may maintain positions in the securities discussed in this podcast.

0:44Michael Batnick:Welcome to Animal Spirits with Michael and Ben. One of the dominant investing themes, or I should say, let me rephrase it differently. the dominant investment themes over the last couple of decades. It started with the internet, and then it went to mobile, and then it went to cloud. I don't know if cloud is quite as big as the others, but that was like the cloud took the baton and powered the next, the bull market. And then of course it was digital AI, NVIDIA, data centers, the build out. And it does seem that very naturally the next thing coming is humanoid robots. At least that's the pitch. The future is going to be weird, man.

1:22Michael Batnick:And I can see it now. 2032, the S &P at$14 ,000, making a run to$16 ,000. And the bear is screaming, well, if it wasn't for these damn robots, the S &P would be back down to$10 ,000.

1:34Ben Carlson:It is weird to think that people are worried about an AI bubble when the robot stuff hasn't even happened. And we're worried to say when this could be a real thing. So that's why we brought an expert on today. So Derek Yan is from CraneShares. He's a senior investment strategist there. And Craneshares has an ETF now that is investing in the rise of humanoid robots, which is kind of crazy to think.

1:57Michael Batnick:These are things that you would not pick out on your own.

1:59Ben Carlson:No, I would have thought this would have been like a MAG7 ETF, to be honest. Like a lot of these are. I've never heard of most of the names in this fund, which tells you that it is actually going for this stuff. So I met, quote unquote, met the robot down in Miami and interacted with it. I was telling Duncan, do you think the robot remembers me?

2:17Michael Batnick:You're very easy to forget. You're very tired.

2:19Ben Carlson:If I had to see the robot again, I wouldn't know who I am. But the idea that we're going to have these is both scary and exciting because we probably will need them at some point to take care of old people. I think it's way more exciting than scary. I guess the scary part would be people saying, oh, my gosh, they're going to take all the there's we're going to have robots do all the jobs. What are people going to be even for anymore? But the exciting side of things is like, what if in the future we have just too many old people that need health care? We need we need help at the hospitals. We need help taking care of old people.

2:47Ben Carlson:We need help around the house, whatever it is. I'm bullish on robots for humanity. I think that it's actually going to come at a good time. Now, is it the next five years, the next 10 years? Remember in iRobot, it becomes the biggest company in the world. What's the name of the company? Is it just called iRobot? I can't remember.

3:01Michael Batnick:I can't remember. But I do love the bad guy in there, the giant dude with big nose from LA Confidential. Yeah. He's not that guy. What's that guy's name?

3:08Ben Carlson:James Cromwell.

3:08Michael Batnick:There we go.

3:10Ben Carlson:But the interesting thing to think about is, is it going to be just one of the mag seven companies that takes, is it Tesla or is it Google or one of these companies? Or is it just this unknown company we've never heard of that comes from nowhere and goes, oh my gosh, this is the new NVIDIA. This is the company that figured this out.

3:25Michael Batnick:It's a good question. I feel like any, and I'm pretty much making this up because I don't spend a lot of time in this space. Anytime these companies reach some sort of promise, they just get bought.

3:36Ben Carlson:Yeah, you're right. You wonder if they will try to stay independent. But it's interesting. So this was a fascinating conversation for me. The presentation that Cranesha sent us was really interesting. The future is going to be wild, man. I don't even know what to say. If this stuff transpires, I'm going to let you get a robot first in your house to like fold your laundry. Okay. Let you have it for like six months before I'm okay. All right, fine. Didn't kill Michael in his sleep.

4:01Michael Batnick:Can you return your robot if you don't like it? What if you're like, me and my robot aren't getting along?

4:06Ben Carlson:That's true. What kind of personalities I'd have? I think it's weird that all the robots have to look like humans to make us feel comfortable. Like why does the robot need a head? It doesn't need one, but it makes us feel psychologically better to think that it looks like a human.

4:19Michael Batnick:I wonder if you could order a headless robot.

4:21Ben Carlson:Why can't it have like six arms instead of two? Why does it have to have two arms and two legs? I want it to be like Rocky on Project Hail Mary. Anyway, here's our conversation with Derek Yan from CraneShares.

4:32Michael Batnick:Derek, welcome to the show. Thank you. Thank you for having me. All right, so we're talking about the CraneShares Global Humanoid Robotics and Physical AI Index ETF. The ticker is KOID. We've had you guys on a lot And there's usually something, I usually make a joke about the ticker, K-LIP, K-Web, or CLIP or Q-Web. What's COID? Where'd that come from? What is this? Well, it's kind of like a humanoid, right? Like COID. Because Cranesha is humanoid. Yeah. Such a simpleton. Simple, simple. Embarrassed. I should have been able to crack that code. So the whole market is resting on the shoulders of artificial intelligence and semiconductors.

5:15Michael Batnick:And one of the supposed hopeful biggest beneficiaries of this AI build out is a completely new category of robotics, both in, I guess, factories where, you know, not like the human robots, but they've been obviously in place for years and decades now. But the humanoid part of it, am I going to have somebody in my house, a metal iRobot type thing, folding my laundry? And if so, when? Yeah, I think that's the ultimate format of physical AI, right? So today's physical AI, I think that just like doing general purpose robotics in a factory, you have like Amazon deploying those logistic robotics. That's potentially be like a natural round of physical AI.

6:04and you have quadruped, the dog shape. So in three-year, five-year, I think the breakthrough in the model is going to allow a humanoid in your home that can do actual work. Today, we already seen a company doing that. Like, it's not really, I think, one or two year. Down the rail, it may need several years to be that general purpose in your home to doing a lot of things.

6:27Ben Carlson:So I got to meet the robot, the crane chair's robot when we were in down Miami. Meet it, interact with it. I don't know how you call it. So you talk about this idea of physical AI. So I'm just curious, like, obviously robotics have existed, not in this like humanoid form or whatever you call it. But what does AI do that unlocks the ability of these robots to potentially like change the way we interact with them and use them and work with them? Yeah, I mean, like robotics is not new, right? It's been talked about for 20, 30 years. It's all over the factory. But like those robotics are single purpose, right?

7:02They're caged, kind of like industrial arms doing specific tasks. And if you want to do another task, you have to program it again, right? So those are hard programmed, like single-purpose robotics. But I think the next wave of robotics is AI-enabled. So a lot of scientists, a lot of big tech companies are investing into the physical AI model that allow kind of like the machine to think, to plan, to have visual, to have thinking ability so they can do multiple tasks. They can understand the environment, right? So that's kind of the next wave. That's where physical AI is going to play to do like manufacturing smartly, to do logistics, to really help us in the service sector, even at home.

7:54So that's where opportunity is totally different compared to single purpose. This is going forward is more general purpose.

8:02Michael Batnick:It just seems like such an impossible thing to replicate the amount of tasks a human being does. So for example, you walk into my house and you walk up the steps and then there's another flight and another flight to my bedroom. And I've got one closet. My wife and I share a closet. We've got whites and colors. Do people still separate whites and collards doing laundry? I have no idea. Not a laundry guy. And then there's like things that don't necessarily go in the laundry, things that maybe go to the dry cleaner. What if a robot like falls on the steps? I mean, for example, how in the world – so how far away are we?

8:44Michael Batnick:Because Elon keeps talking about this on the calls, and it seems like nobody cares about the cars anymore. That's like an afterthought. But every call, it seems like it's just getting longer and longer for delivery. And actually, I want to talk about this. Tesla is not even the, I thought it's the 37th largest holding in the ETF. We'll get into the details. But this general purpose stuff, is this really going to happen? Like you think like, yes, it's 100 % going to happen. It just might take another 10 years. Yeah. I mean, like Elon's always right on a direction. It's always late on delivery, right?

9:19So that's been the pattern over the last, I don't know, 20 years on EVs, on the even space ads. So I think like this is definitely the directional is correct, right? So we have the similar pattern that people use to train the digital AI, the large-action models. Now they train the physical models, right? So the top scientists today, they do the VLA model, the vision-language-to-action model or the world models. So you can actually use the same pattern recognition, same like large dataset training. The whole framework is already there. that's been successful, training the chat GPT or like cloud.

9:59So using that framework to apply our three-dimensional world, incorporating action data is well doable. We have seen a great breakthrough over the last two years that we made so much progress compared to last 20 years. So if that trend continue, we could see the GPT moment for our physical AI, right? So if the GPT moment come, you're going to see this, oh, shock. Oh, that's like previously it's GPT-2, now it's GPT-3. That makes sense. So when that moment came, everybody's not going to question, oh, is this going to be another 10 years? We're going to see accelerating progress on the model capability, just as we did in the digital AI models.

10:46Ben Carlson:It's interesting because you guys have a presentation on this, and you talk about the things that robotics could be doing at factories, warehouses, hotels, and hospitals, and even in our homes. Now, some people look at this and they go, oh my gosh, these robots are going to take away all the jobs. Other people might look at this and go, actually, we need the robots in the years ahead. We need them to fill a hole because fertility rates are declining. There's going to be more old people that are going to be dying, and we might actually need these robots. Do you think that there's a time frame where it's like, actually, the timing of this could end up being perfect for humanity, not like this end game where it's going to just destroy everyone's jobs?

11:24I think currently is the right timing. Because while first I said, like, AI model progresses so fast, and those hardware, just because investment into the EVs and smartphones over last 20 years the hardware's actuators sensors batteries become like so cheap and capable right so we have worry um good robotics um then then as you said um the aging population the the demand the shortage of labor uh is is a structural problem not only here in united states but like europe asia this is global challenge for many countries um that eventually you need a solution right um to keep up the productivity. And of course, as today, there's a lot of repetitive and dangerous jobs out there that our human is probably not suitable to do.

12:19So that's kind of like the easy replacement for a humanoid to be there, like firefighter, right? So we have like quadruped now going to those fires and save people's lives. So those like obvious use cases.

12:34Michael Batnick:Yeah, the use cases are incredibly obvious, too many to name. It feels like the cost is going to be prohibitive, at least at first, inside somebody's home. I think I've seen estimates of 50 ,000, 100 ,000, whatever it ends up being. Do you think that these are going to be borne by the corporations to start? Like Amazon is going to start building these humanoids, Tesla, and then eventually after a couple of years in training, whatever, like obviously, the mega, mega wealthiest side. do you see it playing out like that? Yeah, I would think this playing out where it's similar to smartphone in early days and EVs in say 2012 initially the manufacturing the cost of components is going to be high think about like EV, take example the battery cost of say CATL went down 90 % since early days to today So that drop of cost is enabling the EV to be affordable, right?

13:34So we could see a very similar trend happening. We're already seeing that in actuators, in the detector's hand, in those sensors, in those batteries. Today, those actuators, those bearings, those reducers, precision control, those take most of the cost for building a humanoid, building a quadruped. But I think as the economy of scale continue to happen, we're going to see more affordable humanoid players you can own at, say,$20 ,000 soon.

14:10Ben Carlson:So I'm curious. I want to get into how you guys end up taking this idea that's potentially far off into the future and then investing in it. But I'm curious where most of these breakthroughs are coming from, because it seems like everyone assumes that all the innovation comes to the United States. But it also seems like more people are coming around to the idea that, especially since CrainShares is such a China-focused investment firm, that a lot of, especially on the manufacturing side of things and technology, a lot of it is coming from places like China and South Korea. How much of this is a worldwide phenomenon?

14:41We believe this is indeed a global phenomenon. I think this is going to be a global demand because we have seen labor shortage across Asia and the U.S. but I think differently, right? US, a lot of companies focus on the intelligence side, the model side. You've talked about like private leaders, figure AI, right, like Uptronic. Many like companies are just like bragging their kind of like AI model capabilities, right? The ability to think, to feed, to do actions accordingly. But in China, it's very different. In South Korea, in Japan, And a lot of companies are really the supply chain company or the manufacturing company that can provide very efficient solutions for the hardware.

15:32We kind of see that trajectory for iPhone, for EV. But I think like today, we have seen early shipment coming out of China, right? So last year, there's 20 ,000 unit of humanoid actually shipped globally. and that's from AGI bot, from Unitree, from Ubitech, old Chinese companies. So early days, I think China's been achieving some scalable manufacturing on this. But the U.S. is working on, I think, more the model side, which likely capture more of the value. We're seeing a lot of startups in the U.S. actually training their AI models using Unitree's or a lot of those hardware. So potentially, you can have open-source, humanoid platform that leverages the low-cost China manufacturing, but adding a lot of value using the US intelligence part that can create a piece that's affordable and super intelligent.

16:34Michael Batnick:I want to talk about the actual companies, the stocks inside of the ETF. So I asked Claude to tell me about the combined market cap of the top 10 names, because I was just going through this list and I don't recognize any of them. The top 10 combined market cap as of March 12th, we're recording this on May 4th, but the computer says as of March 12th, according to Claude is$168 billion. And you've got a global portfolio. The top companies from Australia, it's a$12.5 billion market cap called Linus Rare Earths. It is the largest rare earth producer outside of China. The next one, you've got Rainbow Robotics out of Korea,$9 billion market cap.

17:14Michael Batnick:It says they do humanoid robots. Samsung is the top shareholder at 35%. If you've got Doosan Corp, also out of South Korea, the next one is from the US, a company called Regal Rexnerd. Next one is from Taiwan. You got one out of Europe. Teledyne. Okay, I've heard of that one. But these are all, the point is, if the top 10 combined market cap is$168 billion, and these companies have had a hell of a run. The ETF is up, I think, 50 % year over year. But that still sounds tiny. Like, it sounds like the opportunity, like, is this a$5 trillion opportunity here? Obviously, it's going to be volatile, you know, like startup-y type of investments.