Webinar Replay: Investing in Robotics & AI
The replay of our webinar, “Investing in Robotics & Artificial Intelligence” is now available. In this webinar, we are joined by Eugene Demaitre, Senior Web Editor of Robotics Business Review, to discuss current trends in the robotics and artificial intelligence industries and how investors can approach this space.
Jay Jacobs: Hello everyone and thank you for taking the time to join today’s webinar on the topic of exploring robotics and artificial intelligence. My name is Jay Jacobs and I’m the Director of Research at Global X Funds. We are an ETF provider based out of New York with over 57 funds, including 16 thematic funds including one focused on robotics and artificial intelligence, which is what we’re going to talk about today.
We’re also really fortunate to be joined by an expert in the field, Eugene Demaitre, Senior Web Editor at Robotics Business Review, which is a leading publication in research news and information on the global robotics industry. We’re very happy to have Eugene join us today.
In terms of today’s agenda, I’m going to spend a few minutes just going over some basics on robotics and artificial intelligence. Set up, what these disruptive technologies are, what’s their potential, what types of sectors they’ll impact. Then we’re going to switch to a little bit more of a radio-style Q&A with Eugene, to really pick his brain about current trends in robotics and artificial intelligence. We’ll do a hybrid, more traditional presentation with this radio-style interview.
At the end of our Q&A with Eugene, I’m going to close it off with a few more slides on thematic investing in general, our perspective at Global X on how to best approach the thematic investing space. We’re looking to leave about ten minutes or so at the end for questions. I believe there’s a Q&A box; please do feel free to send in your questions along the way. If we don’t get a chance to answer them during the body of the webinar, we will try to answer those towards the end.
I will pause for just one minute here to show some disclosures. I think we start off with a quick poll question here for the audience.
Moderator: Alright. This brings us to our first poll question. What aspect of robotics and artificial intelligence peaks your interest most? The options are: the high growth potential of the theme, concern about the future of employment, want to learn more about how the technology works, and other. I’ll give you a moment to answer that.
Here are the poll results.
Jay: Well good to see most people are focused on the high growth potential of the theme. I think that aligns a lot with what we’re going to discuss today, but we certainly will touch on these other topics as well.
Specifically, let’s just dive right into it. Let’s start with the basics on robotics and artificial intelligence. We see these as two separate technologies, the first with robotics, really revolving around the idea, creation, design, and application of programmable mechanical devices that can perform tasks and interact with the environments without human input. Robotics has really been around for a few decades. We’re obviously seeing a lot more interest in the space today. Eugene is going to touch a little bit on some of the reasons why that’s happening.
There’s been some incredible advancements in things like advanced materials, batteries, additional sensors on robotics that are really starting to accelerate the adoption rate. I really think of robotics as the body, artificial intelligence is the mind. That’s the second technology. It’s a division of computer science that emphasizes the conception of intelligent machines that can work, react, and learn like humans in order to recognize speech, plan, and solve problems.
When you combine these two technologies together is really when you start to see some very impressive technology that is starting to disrupt a variety of sectors. When you think in the past, robotics really has been around for a while. We’ve seen it in places like automobile manufacturing. It’s a really simple process, where the robot can pick up a piece sheet metal, punch it into a mold and bring it down the conveyor belt. When you include artificial intelligence with robotics, especially as those robots start to have more sensors, they start to be more dexterous.
What you see is no longer robots that are just taking a piece of metal and bringing it from Point A to Point B, but really robots that are beginning to interact with their environment. They’re starting to sense things around them and in real time make decisions about where to go and what to do. This is really where we start to see the J curve, if you will, in terms of technology that’s really starting to get wider and wider applications.
In terms of the macro trends, why is robotics such a popular theme today? Aside from the technology aspect, which Eugene will touch on, I think there are three important macro themes that are happening in the background here. The first is aging populations. This is really its own theme that’s happening around the world. What we’re starting to see in some societies is really peak labor force. Meaning in places like Japan, we’re seeing that by 2060, its labor force will actually be about half the size of its peak.
We’re seeing other places in the US and developed Europe and in China, where we’re seeing a leveling off of labor force size. That’s really starting to not only hit GDP growth going forward, but it’s actually going to start to make these economies very imbalanced. You’re going to see a lot more of the senior population, depending on a smaller group of workforce participants. That’s going to really challenge areas like healthcare or the services’ industry, where it really takes a lot of people to support the senior population.
What we’re seeing, especially in places like Japan, is robotics growing out of necessity. We’re also seeing it grow out of opportunity. I think in one area, specifically with labor costs, if the 1990s were really characterized as the heyday for firms searching for outsourcing and looking for cheaper labor costs abroad, what we’re seeing is that now robotics is getting cheap enough where it’s actually surpassing offshore labor in terms of its costs. We’ve seen some predictions that by offshoring you can save about 65% on labor costs, whereas robotics could save up to 90%. We’re seeing a lot of interest from robotics as a way of saving on labor costs.
Third is performance improvement – using machines because frankly they can do some things better than humans can, whether it’s in quality, speed, waste etc. This is not just in the manufacturing side; this is even transferring to places like medicine, where we’re seeing surgical robots that can do things even better than human surgeons. We’re seeing three really broad trends outside of just the technological capability space that are starting to really propel the robotics and artificial intelligence themes going forward.
The question that we get probably most often is which sectors are going to be disrupted by robotics. Frankly, Eugene and I were talking before this call, and we were almost joking what sector isn’t going to be disrupted by robotics and artificial intelligence. When most people think about the theme, the first thing they think about is manufacturing. Of course, robotics has been installed for the longest period of time but increasingly we’re seeing its involvement in military and defense with unmanned vehicles and drones. We’re seeing it enter the medical space, with both healthcare providing robots, as well as surgical robots that are able to perform tasks, in some cases better than doctors. We’re seeing artificial intelligence being used in diagnostics and trying to really try to figure out what’s happening in the medical space.
On top of that, we’re seeing it in transportation. We’ve heard a ton about autonomous cars over the last few years, as well as agriculture, where robotics are being used in place of tractors or to monitor things like crop yields across a very large acreage. We’re seeing this as a very powerful theme that is not just limited to one sector, really something that can be disruptive across the entire economy.
With that introduction, I will call this the end of the intro and we’re going to turn it over to Eugene here and start to enter into the Q&A. First off Eugene, thanks for being on here. We’re really excited to have you. Again, Eugene is coming from the Robotics Business Review, where he is looking at this stuff every day, publishing a lot reports and news and insights on the robotics space.
We have a few questions for you here; we’d love to get your insights. Again, we’ll follow up with a little bit about thematic investing towards the end and allow for a little bit of Q&A as well.
To kick it off, one of the questions we’re getting pretty frequently is why now? Why is robotics suddenly in the news all the time? Why are people asking questions? Why is it coming out in earnings calls, where companies are suddenly talking about how they have to respond to the rise of robotics and artificial intelligence? What’s happening now that wasn’t happening five years ago, ten years ago, fifteen years ago?
Eugene Demaitre: Well, Jay, it’s actually a culmination of factors. Obviously industrial automation and manufacturing has been using robotics for decades, as you said. But more recently, improvements in sensors, thanks to things like cellphones and video games, and cheaper processors and the addition of AI capabilities combined with big data IoT abilities, all that has converged to make robots more flexible, smarter, more nimble and capable than they were before. Before, you had huge industrial arms on an assembly line behind a safety cage. Now, we’re starting to see robots that are more mobile and are more capable and can do multiple tasks.
Jay: Great. That’s very helpful. I do want to move to our second question. We’ve been following this pretty closely, but in your opinion what do you think is the most cutting edge applications of robotics and AI? We’ve seen so much in terms of what people are planning on coming out with, some moonshot ideas, as well as real applications of robotics and AI that are really impacting everyday lives, whether it’s people’s personal assistants on their phones or vacuum cleaners that are cleaning their houses. What are you seeing as the latest in the cutting edge technology in this space?
Eugene: We hear a lot about drone deliveries. While that’s probably coming, I would say that drones are already being used to inspect bridges and to help out on construction sites in ways that previously people had not thought about. Also you mentioned agriculture earlier. There is a lot of use with drones for precision agriculture.
Another area, in addition to the ones you already mentioned is with artificial intelligence, we’re now using that not only for diagnosis but also for therapeutic uses. We’re being able to interact with patients and I think that’s definitely an area for likely growth in the near future. Then logistics/transportation, we talk about autonomous cars, but really it’s not just about passenger cars; it’s about delivery vehicles. It’s about fleets of vehicles being controlled centrally, whether it’s in warehouses, on a campus somewhere or eventually out in the world. Those are areas where I think there’s a lot of interest and a lot of development right now.
Jay: In your opinion, do you think the regulatory environment is going to have a big impact on the growth of this technology. In some instances, some people are saying this is a technology that just has so much momentum behind it there’s nothing that can really stop it. But on the other hand, many people worry about the impact of regulation and how it could, potentially, derail a technology as it grows. Where do you think that stands in terms of the robotics and AI space? Is there something people who are considering investing in this space should be worried about as a risk or is it just something that’s impossible to be tamed at this point?
Eugene: It’s impossible to be tamed in that it’s happening in so many areas and so many ways. I don’t think you’ll see one set of regulations governing all kinds of robots. On the other hand, there is definitely differences in how regulations that affect, say aerial drones, is being developed in Europe versus the US, or self-driving cars in US versus Asia. In addition to the investment climate, which I’m sure you know much more about, the regulatory environment does matter. It matters in terms of trade. It matters in terms of supply chain. It matters in terms of the labor pool that may be affected or that has to work alongside these robots. For the moment, I don’t think regulations are a threat but they definitely bear watching.
Jay: One of the things that we’ve been looking at is just how much of robotics is happening outside of the US. I think this is somewhat related to the regulatory question because, regulations could pop up in the US but that doesn’t mean they’re going to exist in Japan or in Europe, which means there really is an opportunity, through regulation or the lack of regulation, for the US to emerge as a leader in the robotics space or to completely fall behind the rest of the world. Where do you see the US right now compared to places like Japan, in the development of robotics technology, in the application of robotics in its manufacturing and other sectors? How does it stack up?
Eugene: I think the US is still the leader in terms of investment and innovation. We have many excellent universities. We have a climate that encourages a certain degree of risk, which you need for creativity. At the same time, we should not rest on our laurels. We definitely have to watch China, which is a huge end-user market. We have to watch Japan, which culturally is more friendly; there are certain types of robots like as servants and social robots.
Even Europe has made a really concerted effort to develop AI and abilities around that. If any one country can be a leader across the board, I would say it is the US. At the same time, we face differential competition, depending on what technology, and what part of the world, and what news case you’re talking about.
Jay: Now, we’ve seen robotics make a lot of headway in commercial capacities. Obviously the manufacturing applications that we’ve discussed but we’ve also seen them make headways in places like agriculture and healthcare. Is it already in the consumer space and we’re, kind of, missing it because it’s going on behind the scenes? Is it going to start to take off more, so it’s going to be more in our face? Why does it seem like robotics technology is taking off and yet I don’t have a personal robot assistant yet?
Eugene: That’s a question I get asked all the time. I will say that in the past few years a lot of companies that started out on the consumer side, whether it was for drones or for social robots, actually pivoted over to the commercial space. So you see more efforts right now for robots in hospitality or in big box stores behind the scenes. I do think eventually that is going to change back toward the consumer, as, again, the robots continue to become more sophisticated and more capable.
You’ll probably see it at first in areas like retail, where you may to go an airport or a shopping mall or a hospital and a robot will help you find your way. It may also help you do inventory at the same time. Eventually, those capabilities will become cheap enough and common enough that Rosy the Robot that you hear about so much will start to come into the home. I think at first, it’s going to be robots that come into the home will be for people with special needs, whether it’s the elderly or the handicapped. It’ll be therapeutic but it’s not a matter of if; it’s more a matter of when.
Jay: You actually mentioned in one of your other answers, when we were talking about Japan, that there are some cultural differences where the Japanese are just more comfortable with robots as a society. Maybe that’s one of the things that’s holding back the consumer adoption in the US. Is that a big hurdle for consumer adoption here? Are people just uncomfortable sitting in a doctor’s office explaining to a robot what’s wrong with them or is that something that you think is really going to become more every day and people are going to get more comfortable with over time?
Eugene: I think right now there is still a lot of discomfort in the US and in parts of Europe. We’ve seen all the articles saying robots are going to steal jobs or here’s the robot that’s going to cause the apocalypse or whatever. A lot of that hype will subside once people start seeing robots in their daily lives. In fact, the consumer robots that we have seen, as soon as people are exposed to them they usually start giving them names. They in fact may even give them clothing and really treat them, maybe not quite as humans but really treat them with an enthusiasm that you might not expect.
I think at first, yes, there is resistance in the consumer market, particularly in North America and in other parts of the world. But I also think that, as with cellphones, we’ll catch up pretty quickly, once they start becoming more common.
Jay: This reference is a little bit – the slide that we have up here and another slide that I’ll show – but what inning are we in for robotics? I think it’s a little bit unusual the theme in that in some instances it’s been around for so long but in other instances it’s such a newly growing theme because of some of the new technology that’s taking this to the next level. Are we in the early innings of robotics right now? Are we in the later innings and this is the last push? Where do we stand and how do places like China and its adoption of industrial robotics affect that change?
Eugene: Now, I would say robotics is maybe top of the fourth. It’s farther along than people may realize. It’s still not, again, ubiquitous and easy quite yet, but it’s getting there much quicker than people may realize. I’d say, again, because industrial automation has been around for so long and there’s a big push in logistics right now for automation, so I’d say maybe we’re at the top of the fourth. Artificial intelligence, on the other hand, it’s still relatively early days. I’d say we’re maybe at the bottom of the second (inning). We’ve had a couple waves now with development of AI technology and we’re trying to figure out okay, we can play chess, or we can play Go, but now, how do we connect that to our existing … How do we derive useful insights that we can act upon right away and AI is beginning to really move into that area.
Jay: Yeah, I did want to share this one slide because I thought it was really interesting because so many people think about robotics as really sort of this industrial technology, but this shows kind of expected future growth rates over the next eight years or so in terms of different segments. Of course, we’re seeing growth in the industrial segment but also other commercial uses, military use and then of course one of the fastest growing segments is that personal kind of consumer side of robotics. I was just wondering if you have any commentary on this slide and why you think kind of the non-industrial components are actually growing even faster than the industrial components?
Eugene: Well, again the industrial has been around for a while, although it is still expanding considerably. I have been in factories where they have one part in the assembly line that is automated and then the packaging or some other step remains to be automated, so there is still a lot of growth potential in industrial. Commercial, we talked before about the use of drones. That’s likely to continue growing. Military, we’ve written robotics business review extensively on how different militaries are pursuing different types of automation in the hopes of saving lives and defending their troops.
In personal, you’re right; it is growing and partially because it’s starting from the smallest number of the different proportions and also because, again, once people become comfortable it, with aging population, with shortages of skilled labor, there’s going to be a need for both home care robots but also assistive robots and right now the one household name is a robotic vacuum cleaner. There are lots of other things that robots can do around the house and around yard that once people figure out the best way to automate it, they will. There will be a lot of attention there, I think, in the near future.
Jay: You mentioned something there that kind of piqued my interest which is the fact that you visited some factories and you see parts of it is automated and parts of it isn’t. Is that a result of a lack of capital or just simple lack of skill of the robots, that there’s certain things that humans can do that they can’t? What’s the biggest hurdle for a lot of these factories to just go completely ‘lights-out’ and be fully automated?
Eugene: For the largest companies and the companies that are the most heavily automated like those in the automotive industry, for them, it’s more a matter of the edge cases but for small and mid-size enterprises, which we expect is an area of rapid growth of automation, there’s the idea of ‘oh, we automated the stuff years ago that we needed to do, but we’re going to have to re-engineer our processes. We’re really going to have to take a look from end to end; how do we tie in the robots in production with our AI systems that go back to the customer with our supply chain management.’ And so, there is a little bit of a reluctance but again, the companies that do that will be much more competitive. Once people begin to realize that, I think it will pick up. There is an install phase that people are going to say oh, I have all these interesting robots, but I think competitive pressure and the rise of more and more capable collaborative robots is going to change that.
Jay: Got it. I want to kind of discuss the actual technology a little bit itself more. I kind of eluded to the fact that there’s the two sides to the equation, right? There’s the artificial intelligence side, which is kind of the newest, maybe fastest growing aspect of it, which is making machines more intelligent. Within artificial intelligence, there’s a few different types of artificial intelligence, right? There’s machine learning and there’s deep learning. I was wondering, one, if you could touch a little bit on those two differences but then in addition to that also talking a little bit about the robotics and kind of more of the machine side of it. These are in some instances more basic machines to build, but we’re still seeing a lot of advancements because we’re using things like advanced materials, better batteries and things like that. I was wondering if you could talk a little bit about robotics as well as making these robots more capable outside of being smarter but being actually more physically gifted as well.
Eugene: Yeah, absolutely, I can talk about that. Let’s start with the difference between machine learning and deep learning. Again, I’m not a programmer, but I can tell you that artificial intelligence right now is being used in a very broad way. Machine learning, again, is thanks to a most recent wave in neural nets, and algorithms, and understanding that it’s easier in the long run to teach a robot to do something than to program a robot to do something. You have a lot of data, you train a robot on that data, and they can then adapt and learn to do things efficiently. Deep learning we’re to applying the next level of AI. We’re applying insight. Not only is the robot learning how to do a task, it’s learning how to do it better. It’s learning how to share that knowledge with other robots. It’s observing its environment and maybe reporting on that. There’s a whole set, another, if you will, cognitive level that’s arising with that. Again, most people don’t realize that we take a lot of things for granted as humans. Robots and AI are still catching up. There’s still a lot to be done.
In terms of the physical or the hardware side of robots, the improvements in material science mean that we can have soft robots that have flexible grippers that can grasp a variety of objects. For pick and place operations, that’s a big deal. You don’t have to program it to rigidly pick up just one object over and over. That’s useful, but it’s much more useful to pick up multiple objects without crushing them or to place them in various types of containers. Also, soft robotics means that they become able to interact with humans better, that you’re not worried about safety issues as much.
Another area you mentioned was battery. Honestly, that’s a big issue. If you want to have a drone that is persistent in the air for a long time or you want to have a robot that rolls around the warehouse or shopping mall for a long time, you need a significant energy efficiency or battery storage to do so. All those things are steadily improving, and as they improve there is a greater flexibility and a greater capability that they offer.
Jay: Got it, now that’s very insightful. Why don’t we take a little bit of a break in terms of discussing robotics and artificial intelligence? I want to tap a little bit more into thematic investing because the way that we see robotics and artificial intelligence really fitting into the broader investment spectrum is within the thematic investing sleeve, which is really a top-down bottom-up approach starting with identifying powerful macro level trends. That’s kind of the top-down look. What are some big trends happening around the world that we have a lot of conviction in? Then the underlying – the bottom-up analysis of looking at the underlying investment that stand to benefit from those trends.
If you think about this as a two-step process, the important thing is really on one hand looking for trends that we think are going to be very powerful and disruptive, that are highly likely to happen. This is what we call high conviction themes. You’re going to bet on a theme that you don’t think is going to play out, that’s probably not going to work out very well as an investment. The second aspect, that bottom-up approach, is really about investability. You could have a theme that sounds like a great idea, could be very high growth, but there simply aren’t that many companies out there to invest in. Through the ETF vehicle, of course, they have to be publicly traded as well. Some themes might just be not public in terms of these are private companies. The example I would like to use is space exploration could be a great theme, could be the final frontier, not very investible right now because there simply aren’t pure play companies that are listed on a stock exchange that could be invested in in an ETF.
The third thing we look at is time frames, so looking for themes that aren’t going to play out tomorrow or even over the next year but something that is really multi-decade in nature, perhaps even evergreen, if we apply this to robotics and artificial intelligence. Eugene, you were talking about how we are really in the fourth inning of robotics and AI, perhaps even earlier for AI. That’s exactly what we’re looking for within thematic investing. If you’re trying to bet on something that’s in the eighth inning, it puts a much bigger premium on trying to get the timing of the trade right rather than really doing your research on the trade and making it when you feel comfortable when there’s a time and a place in your portfolio to make that trade, and then really letting it grow over time. That’s why we prefer the long-term timeframes over shorter term. In summary, really those three points, high conviction in the theme; highly investable, meaning a broad group of publically traded companies they can invest in; and third, the medium to long-term timeframe for that investment.
Specifically when we’re looking at robotics, we launched an ETF last year just about ten months or so ago. It’s actually been one of our faster growing funds. I think it has about $180 million in assets now. The way we define robotics and artificial intelligence is really in four different categories, the first one being industrial robotics. This is really the traditional manufacturing companies, a lot of Japanese robotics manufacturers that are involved in everything from automobile manufacturing to the production of other industrial goods.
The second category is kind of a catchall for the non-industrial robotics. Not the automation aspect, but things closer to surgical robots, agricultural robots, etc., really everything outside of that industrial space. Third is unmanned vehicles and drones. This is where we’re seeing a lot of growth both in the consumer market as well as the commercial market. Obviously, a lot of people think drones are fun to fly around and a cooler version of a race car, but what we’re seeing increasingly is the usage of drones in terms of 3D imaging and mapping large spaces, whether it’s a building or doing quality control over a construction site, etc., really being able to use those imaging capabilities from a variety of perspectives.
Fourth is the artificial intelligence component. This is primarily a lot of imaging software. It also includes some hardware companies that are very crucial to artificial intelligence in terms of being able to compute lots of parallel computations at once. These are really the four sub themes that we’re looking for within the robotics and artificial intelligence space.
We can probably glean from this that we’re not limited to just one geography or one sector, which is another important point to us about thematic investing. I like to think of this as the new sectors in a way. If you look at something like the technology sector, a lot of people will invest in it because they see it as this high growth segment of the economy. But that isn’t a very precise way of targeting, certainly robotics, but also even just growth in general.
When you look at the technology sector, for example, about 20% of that sector is things like desktop computing, printers, credit cards, very basic technologies that are really growing at the pace of GDP growth around the world. It’s not really the higher flying industries that people are looking for when they’re trying to access growth. Rather than just playing a broad sector that’s going to include a mixture of these high-flying ideas with some of these slower-growing components, we think thematic investing is one way to isolate those higher-growth themes, whether you’re looking at robotics and artificial intelligence, which some of these companies fit in technology, some of them fit in industrial, some of them fit in consumer staples or even healthcare. Also, looking at things like social media and looking at how that’s a component of the tech space that’s really faster growing,
The last thing that we like to look at for thematic investing is how these themes really interact with each other. I think a lot of people like to look at these themes in isolation. Robotics and artificial intelligence is one theme. The internet of things, which is about connecting ordinary devices to the internet, is a separate theme. Lithium is a separate theme on batteries.
What we’re seeing and what a lot of the research in the thematic space is showing is that when themes converge, they start to reinforce each other, which makes the themes all the more powerful. For example, what we see as the intersection of lithium, the internet of things, and robotics and artificial intelligence is a theme of autonomous vehicles. Why is that the case? We’re seeing the car itself or the vehicle itself is the combination of robotics and artificial intelligence, being a machine that’s able to operate in this environment. Increasingly, it’s relying on the internet of things in terms of we’re attaching a lot of sensors not only to the car itself to be able to understand it’s environment, but also to interact with other cars on the road, to interact with infrastructure, and to figure out if there’s traffic somewhere or if there’s a change in traffic patterns and it needs to move. Thirdly, the lithium theme in terms of autonomous vehicles, it’s much easier to refuel a car with batteries rather than with gasoline, which means that we’re seeing a lot of extra tail winds behind lithium because we’re seeing this rise of autonomous vehicles.
What does it mean that there’s an intersection of these three themes? To us, if one theme is starting to lag behind, we believe the intersection of them together is going to pull it forward. For example, when we looked at autonomous cars a few years ago, what was the biggest hindrance to them being possible? It was actually a little bit of all of these.
On one hand, battery technology was too expensive and not dense enough for cars to travel large distances. Cars were not able to network with each other as effectively because of the lack of broadband, and also the artificial intelligence component wasn’t there. Simply powerful computing was not able to be transferred to individual vehicles very effectively. What we’re seeing is a rise across all three of these themes, but it’s not necessarily felt equally across all of them.
For example, the robotics technology we’ve talked about has been available for decades. The machinery behind the self-driving cars is the same as the machinery behind a normal car. However, the artificial intelligence component and the lithium components have really been some of the faster growing areas that are necessary to build the larger theme. When we see the convergence of these themes, we really think it helps build the power behind all the themes collectively so that one really isn’t going to lag behind the other. With that, we’re going to take a quick pause with another poll question here.
Moderator: This brings us to our second and final poll question. Which category of thematic opportunities is most appealing to you? Emerging technological themes, themes based on changing demographics and consumer habits, policy-related themes based on changes in regulations or fiscal spending, or other. As a reminder, you can click your response directly on the screen. Again, the question is which category of thematic opportunities is most appealing to you? The options are emerging technological themes, themes based on changing demographics and consumer habits, policy-related themes based on changes in regulations or fiscal spending, and other. Let’s take a look at the results, and I’ll hand it back over to you, Jay.
Jay: This certainly aligns with people’s interest in joining the robotics webinar today, looking at emerging technological themes. Also, probably somewhat of a relief to people that were taking a breather from policy-related themes, I know that’s been very popular in the news today. We’re obviously focused on technology, but also a little bit of the changing demographics and consumer habits. Some very interesting results…
Eugene, I think we’re going to loop you back in here and do part two of the Q and A. We have a few more questions queued up, but we also see a lot coming in to the Q and A box here. Let’s start getting you going again. First question for you, I’ll probably chime in a little bit with this answer as well, but how do you think robotics and artificial intelligence is going to impact employment in the future?
There’s obviously a lot of fears that this could be such a disruptive technology that it actually starts to flip the economy on its head in terms of its really widespread replacing jobs. What does that mean for the economy going forward? What are your thoughts? Is that possible? How soon would that even be possible?
Eugene: I think we have to be careful not to panic. There is definitely going to be changes, and there are already changes just as there have been through the previous investor revolutions. Manufacturing is changing and is continuing to change, truck driving and farming is continuing to change, and so too is surgery. At the same time, we have to remember that at the end of the day, robotics and AI are heavy tools that companies and people use to make more money, be more efficient, so there’s a choice in how those technologies are developed, and how they’re used. We don’t have elevator operators in most places. We don’t have people going out and harvesting wheat by hand anymore. That kind of change is going to continue with automation. I also think in terms of how soon… the lower functions, the repetitive, boring work is likely to be automated, and that poses some problems for low-skilled employees, but the idea is that people with higher skills or who are able to operate these machines will actually find more opportunities.
Jay: Yeah, those are certainly interesting points. I pulled a couple of stats before this webinar that I thought were interesting. In 1840, 70% of the workforce in the US was involved in agriculture, which is down to 2% today, so we’ve already seen an entire industry that was even more powerful than manufacturing has ever been in the United States be completely collapsed into next to nothing. Of course, we’ve, over that time period experienced tons of wealth creation and growth across the country. In terms of manufacturing, if we look at 1950, this was peak manufacturing for the United States. It made up about 35% of the workforce, and now it’s down to about 20%, so a little shy of half the workforce in terms of percentages has gone out of manufacturing.
A lot of those gains have been felt by the services sector, but this isn’t the first time this has happened. Technology has constantly been disrupting the economy and changing the nature of employment. I do think people’s worries are somewhat justified in the fact that technology is accelerating faster and faster, but I do think this also creates new opportunities for employment, whether it’s in programming, whether it’s at higher-value-add aspects of people’s jobs.
There’s actually a very interesting survey done by McKinsey, where they basically said 60% of jobs are going to be affected by robotics and artificial intelligence, and they found that about 30% of those jobs are going to be changed meaningfully. What they did is they specifically looked at the tasks that people do in their day-to-day job and figured out which ones can be automated and which ones can’t with fairly proven technology. I think this really resonates with a lot of people because if we look at our day-to-day tasks, how often are we dialing a phone, setting up an appointment, responding to basic emails, doing the day-to-day household items in our email or in our office that isn’t the best use of our time, or our intelligence, or our creativity?
I think those are going to be the first areas that really start to be automated with digital personal assistants or just smarter software. That’s not going to be a bad thing for the economy because that’s going to free up our time to do the higher value add things, whether that’s doing, from my perspective, researching and posting to our blog, whether that’s from our clients’ perspective, being more client-facing, being able to get out in front of their clients more. I do think reasons to be worried because of the pace of technology, but at the same time this could actually be a very good thing for a lot of people who are being bogged down by tasks that really aren’t providing a ton of value.
Alright, the second question I think Eugene, you might need to explain the context behind this one a little bit, but we did get a question how large is the peril of AI entering in a new winter in the next four years?
Eugene: I think that’s an interesting question because I’m guessing that the questioner is referring to the fact that technology development is not always at a steady rate, and it can sometimes hit a period of slow down. In fact, that did happen with artificial intelligence. There were assumptions that we would have robots in our homes already back in the 1950s. People were designing things or imagining things, but we hit the limit of what we could program at that time. Then, again, with improved sensors, and neural networks, and machine learning, AI has again made a quantum leap forward.
Could it slow down again? Yes, but I don’t think we’re at the limit yet of our current wave of innovation, so there’s a lot of areas to which AI can still be applied. Also, one of the enabling technologies behind AI and as processors, and as processors become more dense, more capable, faster thanks to Moore’s law, and new materials beyond silicon make processors even more powerful, I think that’s going to improve.
In the short term, we’ve also moved to cloud computing, so we’re not just trying to cram all of the intelligence of a robot into the robot’s body. If it has a proper 5G connection, in theory, it can share knowledge with other robots. it can tap into the wisdom of the crowd, so to speak, so there’s going to be a lot of area in the near future where AI and robotics can still grow, but we should not assume that all technologies progress evenly, at all times, everywhere. I think that’s a mistake.
Jay: Great. Thank you. We’ve had a lot of questions about specifically our BOTZ ETF, so I’m going to address a couple of those quickly before we turn it back over to Eugene for a few of his insights. First off, to bucket a few questions here together, so one, we do have an ETF with a ticker BOTZ, which is the robotics and artificial intelligence ETF. It invests in about 30 companies around the world. Specifically, we see about half of the exposure in Japan, which has been quite the hotbed for industrial automation companies, but there’s also quite a handful of companies coming out of Europe and the US as well.
When we specifically look at which companies are to be added to the ETF, the index provider is looking at these four categories. I think the important thing to note is I made the comparison to a sector ETF. This is really how sector ETFs are constructed as well. If you think about the financial sector, how is a company classified as a financial firm? Maybe that question seems obvious to some people. Of course, a bank would be a financials firm, but really the way that it’s built up is the index provider starts to lay out what are the criteria that makes up financials? It could be regional banks. It could be trading firms. It could be asset management, down the line.
That’s essentially what was done with this ETF as well in the index that it tracks, which is looking at the specific segments of the robotics and artificial intelligence market, whether it’s industrial robotics, non-industrial robotics, unmanned vehicles and drones, and artificial intelligence. Those are what I guess I would call the sub-industries that fold up into the robotics theme, so we’re taking from those four different categories. The fund reconstitutes on an annual basis. We actually just underwent a reconstitution two weeks ago and saw two new companies added to the fund, so that, in one way, continues to stay updated as the industry matures. Hopefully, that answers a few people’s questions that were coming up.
I’m going to pick through a couple more questions here for Eugene to answer. If you don’t mind, I’ll just pause one second. Alright, I think we’re getting a few questions on related topics as well such as 3-D printing and virtual reality. In your mind, Eugene, are these related to robotics and artificial intelligence? I guess a better question is how far and wide are these things going to impact all aspects of technology? Is it going to be involved in printing? Is it going to be in gaming? Is it going to be entering the energy space in oil and gas? How far can we really see this theme spread?
Eugene: That’s a great question, and in fact, we do see them as related technologies. Robotics Business Review, we cover robotics in a very broad sense because 3D printing is being used in automated manufacturing. Virtual reality and augmented reality are being used in the control of robots. Not all robots are fully autonomous, so if we’re talking about undersea drones for say, oil exploration, or maintenance of offshore oil rigs, that may be done with a human pilot but using cutting-edge technologies. I would say they’re all connected, and they’re all advancing, some faster than others.
Building manipulation are definitely key to the eventual development of household and consumer robotics, but at the same time, VR and AR have proven to be useful for inspections, whether it’s of facilities – they’re even talking of using drones inside warehouses to help with supply chain flow. AI may eventually become ubiquitous in that it’s no longer just something that is a special characteristic of a robot but that your household is connected to whatever intelligence, or your factory relies on the intelligence to maintain optimal throughput. They are all related, and they are all advancing.
Jay: What does that mean for the owners of that data? We see search engines have made data so easy to access and so ubiquitous, but it also sounds like proprietary data is going to be a real competitive advantage for certain companies as well. I mean, if you’re saying that the artificial intelligence is really the ability to tap into these databases, do you think people are going to start trying to hoard their own data because that’s really where the value is? As maybe we progress, the computing aspect becomes a little bit more commoditized?
Eugene: Yeah, I think you’re correct. I mean, we’re a long way from the hardware all being commoditized with, again, advances in materials handling and the materials science underlying it, but the differentiator for robots is, in fact, going to be their capabilities, their programming, their AI ability, and the amount of data that’s being generated, and then the ability to analyze that data in real time and derive an insight. Imagine if you had a drone inspecting a bridge. Yeah, the human pilot may have sent it out, but the drone spots corrosion and can immediately start to repair it, without human instruction on where to fly, and how to apply the sealant, and all that. The drone can do all those things on its own.
That’s going to be a big area of differentiation and an area where innovation will still matter. They’re not all going to be equally smart. They’re not going to be equally capable. You’re still going to have single-purpose robots, but what really makes the robot useful is can it understand and interact with its environment immediately and not necessarily wait for human interaction or wait for the data to be collected, analyzed, and then flown back out. That’s a technical challenge, but it’s also an opportunity.
Jay: This actually, I think, dovetails nicely with the next question, which is how do you really weight the opportunity between autonomous robots and robotic assistants? Or the difference between robotics being able to just be sent out and do their own thing and not need a lot of input versus machinery that really enhances what humans are capable of, whether that’s advanced tools, or exoskeletons, and things like that.
Eugene: I think a lot depends on the use case. In manufacturing, or in energy production, or in agriculture, you don’t necessarily need a human on sight the whole time, but ultimately, the robots’ that interest people are the ones that interact with people, so whether it’s an exoskeleton helping someone walk, or a robot reminding your grandmother to take her pills, or a robot that goes out and mows the lawn, there’s different amounts of autonomy that are desirable, and there are different amounts of interactivity that are desirable depending on the function. I think, again, that’s an area where there’s a lot of room for development.
I covered the DARPA Robotics Challenge a year or two ago, and we think oh, a robot walking up the stairs and opening a door by using a gripper to hold the hinge. We think that’s easy, and it proved that it’s not easy, let alone looking around and understanding oh, that’s a flower, or that’s a dog. There’s a lot of room there to apply all kinds of AI. Again, the interactivity with humans I think is going to be an area of a lot of interest as they become more capable.
Jay: Great. Here’s another question that I think is interesting. Who is funding the growth of robotics and AI? Is it coming from private companies, from private equity venture capital, governments, universities? Where do you see those in the pecking order?
Eugene: That’s also a good question. That’s one that actually varies a lot by region. In the United States, we do rely a lot on private investors, such as your audience. I think that’s one of the reasons why there is a focus on useful robotics. That’s an important trend both for AI and robotics is that they’re not just interesting projects. They’re projects that have a practical application. At the same time, a lot of the early stage research in robotics and in AI is done through universities, is done through government funding like DARPA, or the Horizon 2020 program in Europe.
What’s interesting is that a lot of other countries have national strategies around automation, and so you’ll have a country like China, or Japan, or France that will say okay, we’re going to make an effort to develop automation in this area, or in these areas. In the United States, we rely on private funding and on major institutions to do that.
Jay: Oh, that’s fascinating. They’ve actually come out and said we know that in this segment of the economy we specifically want to target it with automation, and essentially either protect the jobs in the other segments or make this segment as cheap as possible through automation. Is that essentially what’s happening in places like Japan and France?
Eugene: In China in particular – obviously, we know that China is the world’s factory, and there is a definitely policy to try to encourage domestic innovation, domestic production of robots. Right now, China imports most of its robots, but that’s something that they’re working to change, understandably. At the same time, the United States as a net exporter of technology has an interest in finding new markets and in places that you would not think. Oh, agricultural robotics is big in Australia, or healthcare robotics could be really big in Africa, or drone delivery could be big – there’s so many areas where these technologies are just beginning to be applied.
Jay: Alright, I think we have time for just one more question. I think this is an interesting one as well. We don’t focus on this in the BOTZ fund, but are there any specific materials that are really crucial to the advancement of artificial intelligence and robotics? Any unique metals or other types of materials that we don’t really think of as being a crucial input?
Eugene: The one I think that’s interesting that goes back to the 3D printing question is what materials do you want to build your robot out of? In general, the things that you’re looking for are strength, and weight, and reliability. One of the challenges has been historically the motors, the actuators that are involved, the components of the robot. You can make a case that you actually want the best possible components because the sum total is going to be more reliable, or more lightweight, or have greater endurance because of that.
We were talking before about lithium batteries. If a new technology displaces that, that’s going to be huge. In terms of soft robotics, we talked about compliant grippers, so we’re not talking about necessarily metal on metal contact. We’re talking about new, flexible materials, and 3D printing has already revolutionized a lot of that manufacturing, not just prototyping, but within two or three years, all hearing aids in the US went from being produced by plastic injection mold to being 3D printed. That is a tremendous change because of automation and related technologies. We’re going to see that kind of thing continue.
Jay: Excellent. Well, Eugene, thank you so much for joining us today. I really appreciate your insights. Please do check out Robotics Business Review where Eugene’s writing constantly about everything across the industry and across the globe. Really some of the most cutting-edge information out there on the robotics space. Also, thank you everyone on the line for joining as well. I really appreciate you dialing in today. We did use up all the time, but we will do our best to reach out and continue to provide the answers that you guys sent in. Thank you so much for your time. If you do have more interest in robotics, I would suggest that you visit our site at www.globalxetfs.com/research, where we’re putting out constantly different reports and insights on what’s happening in the robotics space, but thank you everyone for joining, and Eugene, thank you very much.
Eugene: Thank you.