First Trust ROI Podcast
On the ROI podcast, we discuss some of the most important questions facing investment professionals today, ranging from macroeconomic views, to perspectives on the equity and fixed income markets, to insights on practice management. We aim to cut through the noise, examine the data, and provide fresh insights to investment professionals as they help their clients find better ways to invest…seeking to generate attractive returns on their investments.
First Trust ROI Podcast
Brian Comiskey—Cybersecurity, Cloud Computing, AI/Robotics, and the New Era Of Digital Coexistence
Brian Comiskey, Senior Director at the Consumer Technology Association, joins the podcast to discuss how recent trends in technology are reshaping the spending priorities of businesses, consumers, and governments around the world.
----------------------------------------------------------------------------------------------
Subscribe Here to the ROI Podcast & other First Trust Market News
Website: First Trust Portfolios
Connect with us on LinkedIn: First Trust LinkedIn
Follow us on X: First Trust on X
Subscribe to the First Trust YouTube Channel
Subscribe to the ROI Podcast YouTube Channel
Hi, welcome to this episode of the First Trust ROI podcast. I'm Ryan Isakainen, etf Strategist at First Trust. For today's episode, I'm joined by Brian Komiski. Brian is Senior Director of Innovation and Trends at the Consumer Technology Association, or CTA. I've been very much looking forward to this episode. Brian is always very insightful when it comes to talking about what's coming next in technology, especially when we think about things like cybersecurity, cloud computing, artificial intelligence, robotics. These are some of the topics that we're going to talk about. Cta is also the provider of some of the intellectual capital for indexes that First Trust ETFs replicate, and so there's that linkage with the CTA and First Trust as well.
Ryan:Thanks for joining us on this episode of the podcast. I hope you enjoy. Well, brian, thank you for making the time to join us on the ROI podcast. Appreciate it. I think a good place to start would be the Consumer Technology Association. As I said in the intro, you're the Senior Director for Innovation and Trends at CTA, which, by the way, is a very cool title, I think. So it's a fun title For the layman who's not familiar with the CTA. Can you tell us a little bit about what the organization does?
Brian:Yeah, of course. So the Consumer Technology Association is a not-for-profit trade association, largest representing the technology industry in North America. So we have members from the United States as well as Canada, and our goal really is to help push forward innovation. And so that's representing the hyperscalers, those large multinational technology companies, so you think those Amazons and Googles but a lot of our membership is actually at the startup level as well as small business, so I think it's something pushing like 80%. So, when you think about who we're representing, we really are representing the entire technology ecosystem, from the garage all the way to the corporate boardroom, and so that's really CTA in a nutshell.
Brian:And so we have efforts in government affairs, where we're working to ensure that we welcome regulation, but welcoming regulation that doesn't inhibit innovation. We have a robust standards team, so if you think about airplane mode or closed captioning on your television, those are standards that CTA helped develop. And now we're focused on standards and use of AI in healthcare, cybersecurity standards and labeling on IoT devices for consumers. And then we have a robust market research department, which is where I work, and our job is to keep a robust market research department, which is where I work, and our job is to keep a pulse of the consumer, to understand what's going on from a research perspective in the regulatory field and then also understand what's going on in various financial markets in terms of what's the investment looking like in a variety of technology spaces. That's CTA, but people probably know us best for CES, formerly the Consumer Electronics Show, but CES is the most influential technology trade show every year in Las Vegas.
Ryan:So you say formerly, what is it now? It's just CES.
Brian:It's kind of like MTV. It's like well, I think it's a representation of how the industry has really changed. About 40% of our exhibitors at the last show this past January, 40% showed CASE Enterprise Solutions. So that's where it's like. Is it consumer? Not really, it's every part of technology and I'm sure we'll talk about it. But consumer and enterprise as a sort of a line and delineation is blurring more and more.
Brian:And that's a result, really, when you think about the post pandemic period, where hybrid work, hybrid school and all of this it sort of blurs that line Interesting.
Ryan:So I think of technology and it seems, and it feels like technology is accelerating, and I was thinking about this the other day and that's just within the last 100 years or so. I'm sure it felt like technology was accelerating amazingly. Or you think of all these other innovations that have taken place over time, but it's still. It feels like we're in a period where there's like a step change. I don't know if that's accurate or not. What do you think?
Brian:I think it is accelerating and I think it's also a reflection of there's disruption. Economics 101. Christopher Freeman, a British economist. He always would say that in times of disruption, or I should say in times of downturn or social and economic stress, innovation actually accelerates after bunching up. And you could argue that that's going on in the post-pandemic era, especially with the ongoing AI revolution.
Brian:And if you want more historic examples 2008, you have the global recession. You see Airbnb, uber, lyft, credit, karma and a variety of other sort of online-based services all either launch or IPO. In that time you can go back even further SARS epidemic or pandemic, really, you saw JDcom and Alibaba become the multinational and multi industry firms that they are. They went from selling just hardware at JDcom online to what we know them now, which is not just an e commerce company but an AI company, a search company and all these other facets. So usually it starts to accelerate and I think that's what you're seeing, where there's this pivot going on with those hyperscalers like your Microsofts, your Googles and your Amazons, where they're trying to grapple with. Well, what's the next step in AI?
Ryan:Yeah, and it really builds on. You know, every technology builds on the technologies that came before, and I think that's true with the hyperscalers as well, because you couldn't get to whatever the next stage of technological development is, without that as sort of the foundation right.
Brian:Exactly. I mean, if you think about it, right, when you think about things like cybersecurity, cloud and AI, these are all very interconnected technologies. That's why we talk about them a lot and when you think about it, cybersecurity is probably the most mature of those three, right, really, since we've been online or even before, then you either started with hardware, where you have to secure that and make sure that it can't be breached, can't be hacked to. Now we're at software. Now we're at, then, cloud, which starts to really pop up in 2008. And now we're at AI, like these three all work interconnected with one another, because one offers you security, one offers you scalability in cloud.
Brian:And that last part is the third S, what we call simulation. So how do you simulate human productivity? Because we produce more data on a daily basis than we could process if we filled every job. So something has to fill up that gap, whether that's automation through software, like AI, or automation through robotics, like AI, or automation through robotics, but either way, it's the scale of maturity. And what are these three key digital utility or modern technological necessities for any sort of enterprise and even, to a degree, consumer living, because AI is underpinning I mean, I'm wearing a smartwatch. It's underpinning that, it's already going into your phone, it's on your TV and cloud is propping up all those services and you need security. So you already have cyber on your phone as well.
Ryan:So, yeah, I want to unpack some of those a bit more in our conversation today, and I think cybersecurity is a great place to start. I think of cybersecurity as, being kind of to your point, it's just linked with all these other digital innovations. Anything that is digital, anything that's online is you know, requires cybersecurity. You can't do without it. It's not a discretionary expense. This is something that everyone has to have, and so, as I think about that, you know. My question for you is which types of industries, which types of companies are the most? I don't know if the most vulnerable or the most attacked who should really be sure that they're investing in cybersecurity?
Brian:Yeah, of course. So the ones that jump right to mind, there's two it's healthcare and finance. When you think about right, it's your literal life and then your money, which funds your life. Healthcare is a big one because there's a lot of privacy elements into it as well, as we've found in research over the years. And that's secondary as well as primary is, if consumers feel like that especially if they're adopting, like a digital health solution, so a diabetes management app with a connected device or even a wearable if they feel like that that's gonna be hacked and that data is gonna go out, they will likely not stay with that device. They will look for something else because, right, it really is.
Brian:I mean, we talk about HIPAA a lot, right, in the last few years, there is an overall concern of okay, here's this data that is literally valuable to me.
Brian:That's the first one, and then you have the financial markets, and that's another one where it's just comes down to it's securing money and the transaction of that, and that's where you need to really ensure that you have these layers of security that can protect not just your own investments but, in a lot of ways, the investments of entire municipalities, because ransomware really likes to attack those systems, and that's where you're seeing those attacks going. So the UK and the US have been crippled by Russian state backed ransomware firms that have hit whole hospital chains. And where you encounter this is what you understand when this happens is you have surgeries delayed, you have healthcare outcomes effective, and so it's not just, oh, we're ransoming off money for you to pay us to get your data back. We're also pretty much forcing you to say, oh, we got to pay it because people might die because of this. So there's the pylon effect, and that's really what makes cybersecurity so important is it's the downstream impacts of it. I mean, we're still dealing with the repercussions of the CrowdStrike app.
Brian:Yeah, I want to talk more about that as well, of the.
Ryan:CrowdStrike app. Yeah, I want to talk more about that as well. So you know, to your point about the hackers, the Russian hackers, who's doing the hacking typically? Is it foreign actors? Is it domestic? Is there a mix? Where are the most important threats coming from typically?
Brian:Yeah, of course, and so it's about 75% of attacks and this is from a recent Verizon reports Usually about 71 to 75% are coming from financially motivated actors, so those are usually your independent groups. Ransomware as a service is actually one of the most lucrative illicit economic ventures Ransomware as a service.
Ryan:It sounds like one of the cloud categories.
Brian:It might as well be because what you're doing is you're a private entity or maybe a state actor behind it and you say I want you to hack in and hold this target's data for ransom. You can make a lot of money off of that and, in fact, like the average cost right now of a breach is about 4.9 million US dollars. That's a 10 percent increase from last year, so these costs are only growing. So, as I said, 71% to 75% is financial. What's the remainder? A lot of that is state-backed actors, and Russia is a major player in this space. But you do see Iran, you see North Korea, and I'm sure that major other global powers, maybe including our own, probably engage in that to a degree, and I think it's recognizing and this is something the US government has recognized, having been a consultant in the past of the US Coast Guard cyberspace is considered that fourth theater of war. After you know you think, lansing and air, people often forget cyber is where a lot of that's going to be occurring now.
Brian:Look at the invasion of Ukraine, for example. It started with a cyber attack. That's immediately where Russia went, before ground invasions, because you're disrupting the ability to communicate, the ability to coordinate and beyond.
Ryan:Interesting. So you mentioned layers of cybersecurity. Can you explain that a little bit?
Brian:Yeah. So when we think about how the area has changed, there's just this tapestry of protection. You have endpoint security, you have cloud-based security All these terms I could throw at you. At the end of the day, I think about there's like five key layers. The first is hardware, so securing those actual devices themselves, how are you and this is something that companies like Cisco do really well, which is how do you make sure that the device itself on a hardware side can't really be breached or hacked? That's the first one. Then you have middleware, and middleware is the types of software and hardware solutions that literally sit in the middle between software and hardware, and so how do those work and get those secured up? Akamai is pretty extensive in this space as well, working with those traditional hardware players to really make sure that you're securing this layer.
Brian:Then you get into the last, which are kind of two more layers which are on the software side, which is one is your typical antivirus, so think I mean they're now Norton LifeLock, but Avast was a major player in it. Norton LifeLock still is. These are the typical stuff that you see. If you're a consumer just opening up your laptop and you see a service running on it, that's usually your antivirus. The last are specialized software services on the software side, which is your CrowdStrike, your Palo Alto networks, your dark traces. These are the specialized services designed to deal with a specific problem like cloud security, securing an endpoint which is simply just a login terminal or entry point.
Brian:And then the fifth is not so much a layer created by middleware software services. This is kind of what we would call the managed security service provider layer. So these are these companies, like your Booz Allen's, your BAE's, your Accenture's Accenture is getting back into it after some time. But think about kind of all those contractors. They have to create custom solution layers, usually for the intelligence community or other government clients that require not just a custom need hardware, software, middleware solutions. But I actually need people, whether I hire internally or I go to a contractor and say this is the type of security that we need, and I think that's an important layer that I think gets often overlooked.
Ryan:So you mentioned CrowdStrike again and I love to get your thoughts on what exactly transpired and how it's possible that a software update is pushed out to worldwide customers and that ends up seizing up computers, seizing up the airline industry. Basically, any thoughts or reflections on what happened there?
Brian:Yeah, I think a lot of it comes down to is you're seeing the interplay of cybersecurity and cloud at play here, right? So the error was caused, as we've learned, really by the pushing of an update into Microsoft products and solutions, whether it's Microsoft Office, which a lot of the world runs on, and this also was illuminating too, where a lot of Americans, I think, they saw the travel impact on airlines. They may have seen some disruption work-wise, but we're pretty immune overall in some ways because Apple is pretty prominent in terms of using their software and operating services in the US, but globally you take that step out, it is more of a Windows Android world, so naturally, you saw hospitals impacted across the world. This comes into where there's a few things at play, which is I push back against the argument that well, is it because we're leaning on one security service too much? Crowdstrike is impressive because they offer more of a platform-based approach. They have AI solutions. Where it's you have this problem, we have a software service that you can go to and grab and we can implement it. They are so popular because they are very effective and strong in what they do. They have a strong expertise in that space.
Brian:What I think is going on. Here is something that I think needs to be buttoned up in a bigger focus is cloud security in general, which is you have a lot of misconfigurations in the cloud, in the cloud. If you look at EU regulation where Microsoft was kind of had their hands slapped by EU regulators about, you can't be the only one with this type of cloud arrangement. We need more competitors. As a result, you've actually opened up the window to more misconfigurations in the cloud ecosystem, where it's maybe a wellentioned regulation, but the downstream impacts say well, we didn't allow this expertise, and about 90% of cloud deployments have a misconfiguration in them, and so that all goes down to human error rather than software error, and I say this in the lens of.
Brian:I think we'll be dealing with those repercussions in the long run for a few more years, especially if a government is trying to figure out well, how do we secure our own systems? Now? A good example is there was the NotPetya attack in 2019. People really will remember this one as the WannaCry ransomware attack that hit Maersk, the shipping company, and Merck, the pharmaceutical company.
Brian:I got to see this attack through a few different lenses and I think this is what informs my crash view is merc was a client of mine at my first job, so I saw and them get impacted directly. They couldn't email me for two weeks and it all started with someone opened the wrong email. Mersk shipping container coast guard oversees that. So when I was consulting with coast guard in 2020 uh, or I should say this 2017 the attack in 2019. They're still dealing with creating policy for it two years after the fact which, if you look at cyber security attacks, innovate all the time, and so now, looking at it in hindsight, in this third role we're still talking about not pet yet, because it was indicative of we're almost especially from the government side at times behind the innovation curve on the threat landscape. You have to continue to update with that. So I think CrowdStrike and Microsoft have lessons to take from it, but I do think that collectively it's we're not taking cloud security seriously enough in the implementation level.
Ryan:So well-intentioned regulation. It seems like regulation's always got some sort of good intention behind it, or at least from someone's perspective it does. But oftentimes the free market can work faster and I think that's a really interesting perspective, because the hackers don't have to go through a process and committees. And it doesn't take them years, it takes them as quickly as they can do it, as quickly as they can make adjustments. So that mismatch between regulation and the speed at which hackers can kind of change their tactics is really interesting.
Brian:It is, and I think that's why CTA. One of the things that I'm really proud of is that we really try to use our work as a not-for-profit trade association to basically coordinate private enterprise and those companies from the startup level to the publicly traded mega caps, working with us and the US government to come up with sensible regulation on it. A really good example of this and one of our analysts that works with me on our thematic index work. He helped create the US Cyber Trustmark, which is a new labeling program for IoT devices to think about. Almost like you can see health information on the back of food packaging that lets you know okay, this is the calorie, this is maybe the intake that we're getting. The idea is it's a voluntary program, but companies can put this is the cyber risks associated with this. This is kind of like cyber hygiene is what they tell you to practice in the government sector as well as in the private sector increasingly, but this is in the process of being deployed.
Brian:It's been approved by the FCC, but that was pioneered by us convening with an AST and our companies and member companies and getting them to say you know, this is an important standard member companies and getting them to say you know, this is an important standard and that's the issue at the end of the day is we also have a sprawl in the United States government, whereas you have 50 states oftentimes competing with one another on regulation. That's why, when you have something more of a national framework, you are in a better position. That's where, if the EU regulation of Microsoft, while well-intentioned it, may be overstepping, the one thing you can say in a lot of ways is it's coordinated between countries. It's not individual countries pushing back on one another. So there's always that element to consider when looking at well, how do we build frameworks? How do we build guidelines? We take that approach to AI as well as cloud.
Ryan:Yeah, so looking forward, one of the things that we've talked about in the past, brian, has been that, I think, is just really both scary and fascinating is quantum computing and what the impact of quantum computing will be on cybersecurity specifically. Could you talk a little bit about? Well, first, what is quantum computing and why that's significant for companies to think about with cybersecurity.
Brian:Yeah, of course. So quantum computing is? I'm not a quantum physicist, I'm not an expert on that.
Brian:Let me know, believe it or not. No, my background is actually originally in international relations, funny enough. But when you think about quantum, I think the big one is it's moving from the. If you're looking at traditional computing, it's, you know, zeros ones. You kind of are in a binary state flipping between one. But I think ibm put quantum computing best, which is when you spin a coin, right like you have heads or tails, a binary option on coin.
Brian:Quantum community is if you were to spin the coin and you're constantly operating between heads or tails, opening up billions of possibilities of what the outcome could be. That is where you're getting to the point of okay, it's not going down this or that, it's all the could-bes, the what-ifs in computing. That's how I would describe it, or at least how IBM describes it, and I like that description. So when you get into that world, that means you're processing faster, you're moving through sort of encryptions faster, and this is where you start to get to the cyber risk, which is well, it maybe takes a hacker's you know at this point minutes, hours maybe to breach a device. If you have quantum, where you have the ability to process through billions more combinations at a faster pace we're talking seconds, and so this idea comes into play that, when you think about the future state of cybersecurity, there's a concept called post-quantum cryptography PQC. The idea is, quantum computing is not really here yet. To be very clear, I think, if the 2020s are going to be the race between AI and GLP-1 inhibitors aka we Go Beyond Zempik of who will win the decade in terms of maybe transforming society, 2030s are probably the quantum decade, where quantum computers would become a lot more, not just commercial, but more maybe even publicly available. We saw the first commercial quantum computer at CES, by the way, at 2019 from IBM, and when you think about it, though, is 2030 is only six years away.
Brian:You need to start preparing for the fact that, if attackers don't have to worry about regulation, they continue to innovate. They're gonna be simulating scenarios of how can we use quantum computing to hack in, so now an AST has already had competitions asking for vendors and private enterprise to say can you create something that, theoretically, could not be hackable in quantum? They put that challenge out two years ago. There have been two winners since. Obviously, those winners are a moving target, so that might be static in time, but either way, you're already seeing those solutions come into play, and a lot of that testing for quantum is taking place on the cloud and that's Microsoft Azure. You have AWS. I think it's either Azure bracket or AWS bracket. Either way, microsoft and Amazon, their ecosystems, with cloud and using AI, can simulate all these quantum scenarios to ensure that you can build up a more secure future. And that's really, at the end of the day, like cybersecurity, while older than maybe cloud and AI has to be the most future looking, I think, in some ways, because the landscape is always evolving.
Ryan:So I'm sure there are people watching or listening to the podcast that have heard the term cloud, cloud computing, and they have some idea what we're talking about when we say cloud computing. But on the other hand, they you know it's one of those terms that's been used for so long now. Maybe they're not even sure what it means and they're afraid to ask. So can you talk a little bit about what? When we're talking about cloud computing, what's the difference between your? You know what we were doing 20 years ago with computing.
Brian:Yeah, so 20 years ago we would be using what's called on-premise computing. So the idea is you have your big data servers at all times in your office, usually filing away and putting your systems. In fact, we used to have a one at CTA called Glass House where a lot of our secure data would literally live inside this like classroom, and that's where it would be. You can't have that file leave.
Brian:Now, with cloud, people are familiar with data centers which are externalized. We are in Virginia right now recording this. We are miles away from many of the data centers that power the modern internet, because what you're doing is those data. That data can be stored there but then transmitted via software, via digital infrastructures that allow people to be able to host their data somewhere else. That way you don't have the costly maintenance fees for on-premise computing, so there's a cost effectiveness. With cloud, you also create a lot more unlimited storage capacity versus on-premise, so in that way you have a scalability.
Brian:So this is the difference of why maybe 20 honestly like maybe probably 25 or 30 years ago where, if you're like a mom and pop store, you really are probably going to be a mom and pop store for a long time where you have your customers around you. You're storing the data and files and information about them on your on-premise computer. Nowadays, a mom-pop store can be a global company, and that is offered through. They can use cloud tools like your Shopify's and others that really allow you to unlock a bunch of software to maintain customer profiles, outreach services, as well as e-commerce and shipping solutions. Amazon might get still half their revenue from Amazoncom, but really the power of Amazon is their cloud product behind it, including the infrastructures that they're selling to everyone. So cloud at this point is really a modern staple of enterprise if you want to be able to compete nowadays, because someone's always creating those solutions. That's where I think that's the importance of cloud. There's a sustainability element to it too.
Ryan:So the scalability, the productivity enhancement comes from essentially being able to outsource some of your computing power, your storage, the infrastructure, the software, and I think that's a development that's been happening for 10, 15 years 2008,.
Brian:I think in a functional point, it's kind of the start of the smartphone, and that's where cloud becomes very important. And so when I was consulting with the US Coast Guard, it was on digital transformation policy, and that is a term that I'm unfortunately still hearing, because it's been 16 years. Are we transforming or are we coexisting with these technologies? And that's a really important trend. But yes, exactly, it's about the idea of offloading those materials. It's a reliability, a cost efficiency, scalability, flexibility.
Ryan:So, in terms of how mature that theme is, um, can you make some comments on on your perspective on on the maturity of that as a theme, Because it has been around for a while? Are we saturated? Is there still a lot of potential growth ahead? What do you think?
Brian:I don't think we're saturated. I think, if you recognize that cloud is a necessity, like cybersecurity at this point, right for that, scalability, cost efficiency and reliability you realize, though, that one size fits all isn't necessarily what works anymore. I think the last 16, or even maybe kind of 2018, was all right, let's just adopt, let's move our companies into the cloud, let's just start there, let's get Azure, let's get AWS to have our infrastructure, and then we start buying software and services that specifically meet our needs. We're in a phase called cloud right-sizing, which is a recognition that companies don't want a one-size-fits-all approach because one that might not be as cost efficient anymore, and so what you're looking at is what's called multi-cloud solutions. So there's a lot of sites that it's like about 76% or around 80% of large organizations are multi-cloud, which means that they're using AWS, they're using Azure, google Cloud, or maybe looking at companies like Fastly, which also provide important infrastructures, recognizing that some infrastructures work better with some software solutions than others. That's just the nature of when you're interfacing between or communicating in general, whether that's as humans, right, like some people work on better teams than others, same with software, believe it or not, and so this is where you see companies say, okay, we're going to have an Azure instance and we're going to pay for this certain amount of hosting abilities because we want to put our client retention software here, we're going to put our lead gen software here, maybe all of our marketing. Then we're going to use Google Cloud because this is where actually, maybe, payroll works better, and so we're putting all of our accounting and financing here, and so there's a specialization that's occurring here.
Brian:When you have, though, these multi-clouds working between one another, how is that data moving between one another, right? How do you ensure that everything gets packaged? So we are in a little bit of a revolution of essentially like data siloing, and how do you turn what's called containers of data, so data A through F is in container one and then G through Z is in another container? How do you get those to seamlessly mesh between different sort of infrastructures? There's a whole solution world of companies like Docker that focus on this process, called it's very clever containerization, and that's how do you move it in between other. It's a bit niche, but I think it's overlooked, because I think a lot of people tend to go oh, I mean clouds, as you asked it.
Brian:You said saturated, and it's not really. Maybe you have infrastructure services really becoming a little bit more on this front. You have the players that you know, as I've mentioned Azure, google Cloud. There's probably not going to be a lot more entrance on the infrastructure side, but the platform as a service, this layer that sits in between software and infrastructure to coordinate all those software options you have. That's where the big play is going to be over the next few years, and those are companies like MongoDB as well as others. Even Dell has gotten into that game. Right Hardware to software side.
Ryan:So there's also a linkage between cloud computing and artificial intelligence, AI. Can you talk a bit about that?
Brian:Yeah, of course, and so when you think about, maybe, what's what does the cloud computing ecosystem look like? I've talked about infrastructure maybe becoming a little bit more. All right, you know who the players are. It's the hyperscalers, your Googles, your Amazons and Microsofts. What's the advantage that they have then? Infrastructures allow for some of the best training environments for data, and you can't have artificial intelligence advance without a lot of data Data. One of my colleagues calls it the brain food of AI, or I like to say, the language of AI. That's how it speaks. This is where data is processed, sequenced, held and beyond, and if you can't effectively sequence or basically create a logic to your data, then you can't really train an AI solution very well or an AI product.
Ryan:When we think about NVIDIA.
Brian:Why do they do so well? You have Intel, AMD and a variety of other players on the hardware side building all the chips that power this revolution. Nvidia does that very well. We know that their GPUs are great. Everyone wants their processing power on the hardware side. Building all the chips that power this revolution. Nvidia does that very well. We know that their GPUs are great. Everyone wants their processing power on the hardware side.
Brian:What they also do well in comparison to these other companies is they do the software side too. They have been doing the cloud packaging of data for a long time. They have been working to sequence and process that data, creating a variety of tools. They actually started that mostly in the gaming space, which is funny. Like a lot of people overlook that, the gaming industry usually has required a lot of advanced cloud hosting solutions. When you have developer teams of hundreds to thousands that sit across 15 plus countries at a time, developing on tight deadlines a variety of packaging software, you need to make sure your data is really well held. So that's where they've built this customer base, which includes companies like Microsoft, who do AI, who do cloud, who do gaming, and Nvidia has really captured those markets and that's why they really operate quite well in that space.
Ryan:You mentioned sustainability a moment ago, and one of the things that I've been thinking about and observing is when it comes to the demand for electricity that's come from not just the hyperscalers and cloud computing, but what's coming down the road and demand for AI, data centers, those focused on AI, and all the need to process and pay for the electricity and generate the electricity and added to that, you've got mining for Bitcoin and all the demand for electricity coming from that.
Ryan:I mean, when we think about that step change in demand for how, how did these companies deal with that?
Brian:Yeah, I think that's honestly. There's a reason it's being signaled as a major problem. Uh, to be frank, especially when you think about it in this way too Um, we are back in um under this admin at least. We are back in the Paris climate accords, which have 2030 targets. They were implemented or launched in 2018. That means we're halfway to it now.
Brian:So think about the pressures of you have a lot of not just governments signing up to these deals, but think about all the companies that have made 2030, 2040, 2050, where zero emissions, where net zero carbon neutral, whatever phrase they want to use. They've made a lot of commitments, but, as you put, mining for Bitcoin, AI requires a lot of power, and so a lot of this is going to come down to is how do we innovate? This is where innovation and product accelerate, because whenever you have a challenge, I like to think necessity is the mother of invention. That's a very old phrase for a reason. The idea is how do you innovate data centers themselves to become more sustainable? So we saw at CS 2020 or 2021, 2021, our all remote CS during pandemic, Microsoft was talking about underwater data centers. That's one approach. You've seen some of that work. Microsoft actually had major successes with it but then has actually shut it down. I'm waiting to hear why. I think it's one of those. Maybe they didn't have investor interest at the time.
Brian:But the big one is you've probably seen that there have been deals and hookups between OpenAI and Amazon with nuclear energy, specifically nuclear fusion startups, recognizing that nuclear energy is, while there might be carbon, that goes into uranium mining and some of the other materials mining, and there is the question of where do you put the waste from it. It is a zero carbon emitter and it is much more energy dense than a variety of other technologies. That's going to be an area where I think you're going to see a lot of innovation. So Amazon I know I think it's Talon Energy is a nuclear energy company that they have partnered with and bought space, and then Helion Energy is a nuclear fusion startup that I believe Sam Altman backs, and they're now going to be working with OpenAI to create that power, Because, at the end of the day, we're seeing rapid advancements and this is something we're watching with some of our work with climate tech and sustainability.
Brian:That appears with some of our members as well as at CES, is solar and wind have moved forward. Battery storage and energy storage solutions have gotten better and they are being able to do a certain capacity, but if there is going to be this exponential increase in demand, you have to start looking to other solutions. And maybe I'm biased, I'm probably a nuclear hawk, but at the end of the day I look at that and I go why not look at that and this is opinion something that is zero carbon, that can meet your goals, goals but is energy dense to meet that gap? Because that's innovation at the end of the day, and so I know microsoft's even looking at small modular reactors yeah, yeah, and there's.
Ryan:Whatever the source of power generation is, there's the infrastructure build up that needs to occur with that to connect these things, to make sure that the technology, um, you know, has a steady flow of electricity. I mean, I would imagine it's more difficult to run a AI data center if you've got an intermittent load of power that comes and goes. That would be pretty difficult.
Brian:Exactly Like you need to ensure to like and this goes out to cybersecurity. Our grid, especially in the United States, really needs a lot more just coordination, security and just critical infrastructure needs to be looked at, and this might be the best time to do it, when you think about it, and I think it meets all the agendas of AI, building more data centers, how do you build them more efficiently to the grid, but also how do you do it more green, or greener if you will, and I think all those don't have to compete against one another. There is place. That's what technology is supposed to do. It's supposed to help try to achieve and solve some of these problems in a coordinated fashion.
Ryan:Yeah. So AI has been what's on everyone's mind for the last couple of years now. There's certainly a lot of enthusiasm. Some would call it hype. Do you think there's? I mean, where are we in the hype cycle? Maybe one way to ask it Is there what's behind all the enthusiasm? Is this something that's going to change our lives tomorrow? Is this going to be further down the road? What do you think about AI?
Brian:Well, I think AI has already changed lives. I guess that's the thing. I talk about this with my parents actually too, because they asked me when I was visiting them earlier this year. They're like, why do we keep hearing about AI? And right now I'm like, well, you already use it every day. About 90% of your content recommendations in Netflix, that's AI. Almost a little over half of Amazon's e-commerce revenue can be attributed to the very old AI and machine learning algorithms behind the content recommendation drivers and even on healthcare outcomes.
Brian:I know, with this, this has learned a lot of my preferences on the Fitbit of. This is the workout that I'll probably do. This is the recommendations in terms of just better health outcomes from a diet perspective. I know, for me, I live and breathe by my app and so it's already doing a lot of things for us. I think that's important to remember at the end of the day, because what we're talking about when we're talking about hype cycles is probably we're talking about generative AI, which is newer. Large language models, while posited, I think, in 2016 or 2017 so actually older than we think by google, really started taking off with chat, gpt and open ai's advancements, and I think that there's going to be a lot of investment into this space, they're saying. I know there's estimates of eventually reach seven trillion dollars of investment.
Brian:The question with the hype cycle on generative AI is you do have to ask what's the $7 trillion problem, and I think a lot of this is is it customer service? It might be. You have CVS already has employed on their pharmacies a fully generative, ai driven customer service chatbot for prescription, phil. So you're not talking to the pharmacist If you call in. It's going to be an AI that gets the prescription in. Um, my mom works at is a pharmacy tech at cvs and she says it's been a game changer as a worker because they have more time to focus on accurately filling out the medicine than being on the phone with customers and they've had minimal complaints, at least in the stores that open operated it.
Brian:But the big one I think about with when you think about AI as a whole is generative AI is just one slice. You have digital twin technology. You have multimodal AI, which is the ability to process different types of information at once, whether that's visual text or audio. There's a lot of different angles to be approached. So when we think about it, maybe generative AI is possibly about maybe is at the top of the type cycle now, but there's such a larger ecosystem in the AI world where you have NVIDIA announcing we have a digital twin on the entire earth that can do weather prediction for a variety of countries as they grapple, maybe, with climate change. That solves a lot of problems right now for the billions of dollars.
Brian:And I think the healthcare story is ultimately also the big one for AI you already have. We saw this innovation from a Dutch company at CES. It's one of my favorites is they could use AI to look at your healthcare records and, again, making sure it stays private to understand that if you have surgery. Usually it takes about something like five to six days to see if you're going to get an infection after surgery and that's pretty critical time when you think about it right. They can analyze your healthcare records to predict and they have more accurately. They've reduced that window to three days. That's two days of potentially a life-saving infection prevented. Or another one is Moderna.
Brian:When you think about their mRNA vaccines that they've been developing. They used to have something about. Basically a candidate a day is usually maybe what they could pull up each month. So you know, that's 30. With generative AI solutions being applied to sequencing and trying to create combinations of mRNA, they're getting about 1,000, plus therapeutics, a month, and so that's the acceleration, and fortunately, regulatory approval pipelines can be timely in the healthcare sector. But that's still a benefit that will probably, by the end of the decade, you might be seeing some vaccines or some conditions that you never thought possible because of AI.
Ryan:Yeah, I've always been fascinated always the last several years by there's DeepMind, deep minds alpha fold project, which really essentially solved the protein folding problem, which ables you know it provides so much more information on things that researchers and scientists didn't know about because it was really difficult to to determine the shape of a protein. But if you have the shape of a protein, you can develop treatments and cures for things that maybe you couldn't before. So that's again. It's fascinating to see what sort of outcomes will potentially improve in healthcare because of some of what's gone on with AI, machine learning and big data analytics and all these things, and there's a thing called one of my favorite innovations also saw at CES is called the Living Heart.
Brian:It's from Dassault, the French company they created using AI. This is actually where you get into whether you call it the metaverse spatial computing enterprise, xr this idea of just essentially a digital representation of a physical object. They were able to use AI to create a digital representation of the human heart and by doing that, they can practice surgical techniques that otherwise, like you, got to practice somewhere. Right, and they've already found, including one of the executives at Dassault that was looking at it. His own child was saved by a surgery technique that was pioneered in a digital atmosphere before being done on pediatric cardiological patients. So you already have I think it's at this point they're saying, like hundreds of heart patients, children whose lives have been saved because of AI powering the metaverse, believe it or not, still having, uh, an impact, and that's something that's been happening for the last few years, and they're already working to try and do it for the brain now.
Ryan:So there's always been this linkage between AI and robotics. One I think of as being AI is the cognitive, the intelligence, the analytic part, and the robotics is sort of the physical outgrowth of that, One of the things that I've heard more about. Maybe it's because of Elon Musk and Optimus and Tesla, but the idea of personalized robots Is that something you hear about from the CES or any of the companies that you're talking to?
Brian:Personalized robots- yeah, humanoid robots, home robots there's a bunch of different terms for it. We've seen that at CES. Samsung actually has been piloting it for years. They have something called the Bali and I think I've seen at this point, my first ces that I attended in person was 2020, so right before the pandemic um, and it was massive and that's where I think I saw it for the first time and I've seen the iterations over time. What's been kind of cool is it's literally it's like a tiny ball. Reminds me of bb8 from the star wars Wars, the new sequel movies, opinions aside on how good those movies are, but BB-8, right, is adorable. No-transcript, you're getting almost this concept this is a term I think I first encountered about a decade ago is cobots, this idea of. I think there's a lot of fear with robots and, and certainly there has been displacement I mean, anyone who's watched the terminator movies has a little bit of yeah everyone's gonna have a little bit of fear.
Ryan:We're also gonna come for us and thinking about, like the matrix maybe maybe, maybe, oh god, hopefully not the matrix.
Brian:That's terrifying.
Brian:Um, I don't want to be in the pod um but, uh, when you think about, um, kind of like the idea of displacement of, maybe on, like the industrial line, where, especially like automotives, where there are some things that can that robots have been able to do from an efficient, as you put it, like automated process, this is more the idea of how is the robot working with you to each day? Because I mentioned before and the stat is we produce 2.5 quintillion bytes of data every day. That is 2.5 followed by 18 zeros, literally impossible for one. The human brain really can't actually quantify that. Like you, you hear 18 zeros. You can't quantify that, but you also can't process that data in a day with everyone, as I said, filled out in a job role.
Brian:So how does maybe not just AI, but how does a hardware expression of human productivity like robots, come into play here? And that's where, if AI is put on board, you get what's called self-learning robots, robots that go from being kind of input A output B. A good example is your robot vacuum Roomba. Right, would like eat up socks all the time and then you would have to hard code it like recognize a sock with the computer vision on board. Now, the idea of self-learning is it can learn to avoid certain objects on its own, so that the input is A but then the output is B to infinity.
Brian:And so I think humanoid robots is a major focus. I know for Tesla, I think that you see Samsung and others that are trying to get into that space. I do think that that is a bit more future state, maybe end of the decade, where you might see them on the consumer side. I think it will be enterprises where the adoption goes first. That usually is when you think about industrial or hard robotics, but I think you can't get there without a lot of AI advances we're seeing today.
Ryan:So, in terms of future use cases AI and robotics what are you most excited about?
Brian:I mean, I always go back. I think I always go back to the healthcare story. I think this idea of um a really good example is like. You have companies like Medtronic, johnson and Johnson Stryker um, that have been in the medical robotics space for years, right, um, a lot of your knee replacements was your hip replacements are actually being done by robots now. Um, but um, what I was reading about is exciting is I think it's medtronic they are trying to enter more into I think it's the hip and knee uh side, and part of what they're doing is they got approval for a new robot that has learning capabilities on board, because the idea is as much as that sounds like this is an old space where, oh, yeah, yeah, knee replacement isn't just knee replacement. We can just use a robot to do it. Each patient's different At the end of the day like every human body for the most part, is pretty different.
Brian:So the idea is the next step is personalization. Um, that's a big term that they use in the digital health sector, about like why do people get apps, why do people wear Fitbits, why do they wear smart rings and whatnot? Is because they're looking for personalized. I am empowered for my healthcare outcome on my own. Well, what if the robot can learn really well what your body is using the data that you have on board? That is already happening is already being implemented and that I don't know. That, to me, is extremely exciting.
Brian:The other one, too, too, is like is I always think it's really cool to see like the hard industrial labor in like extreme environments, so like mining, when you think about it, like how dangerous that could be. That's where you've already seen we have caterpillars come to cs and brings their like giant autonomous excavator. How do you get it so that in some of the high intense heat environments or danger environments that may be instead where you still will need human labor and human guidance, how does it maybe have a robotic avatar that's in there doing it? No, that's interesting, yeah.
Ryan:Yeah, I mean there's so much more we could talk about. This episode of the podcast has flown by. I'm looking at the time We'll have to have you back on again at some point in the future, hopefully to talk more about this, Because again, I have like a dozen more questions to ask you. One of the questions that I always end episodes on is you know what, as you kind of go about your schedule, what are you reading these days? Is there anything that's really interesting, that's caught your eye? Any books that are on the Brian Comiskey book list? Sure?
Brian:So I'll do two really quick. The first one is going to be very quick. Is I finally read on the fiction side of the Wheel of Time series? Brilliant, absolutely, incredible. It's 15 books and it's well worth the time. It's better than the Amazon Prime show airing right now. So always, the book usually is better than the movie 15 books.
Ryan:That seems like an audio book.
Brian:It was yeah, it was a two-year commitment, so that's that's the fiction side, but non-fiction, um, so I am, I'm firmly a millennial, um, my wife and I both actually game. Uh, it's a bigger industry than music and movies combined right now. But, um, I wanted to understand how do those games get made, who is working on those games? And, um, I really recommend blood, sweat and pixels by jason schreier. It is 10 chapters, so it's a little bit on the on the short side. Maybe it's less audiobook um side, but, uh, each chapter is about a particular video game and want. Some of them are the massive ones you may have heard of in space and some of them are smaller and it's all about the story and literally the blood, sweat and pixels that go into the development of each. And it was really eye-opening when you understand when people are really critical and reviews of games especially.
Brian:Sometimes gamers can be a little bit hostile yeah, on it is understanding the labor, the work, the care, the creativity and then the humanity that goes into each of these development. Um, I think it's one of the best books I've ever read, really, and I highly recommend. He has a few others one coming out on the Activision Blizzard history including the merger with Microsoft. So highly recommend Blood Sweat and Pixels by Jason Schreier.
Ryan:Yeah, video games have come a long way since I got my first Atari 2600 when I was a kid. It's amazing looking at some of the graphics and just the detail and all that goes into some of the games that my kids are playing now. It's just, I mean, worlds apart. Well, thank you for the book recommendation. Again, thanks for joining us on this episode and hopefully you'll come back again at some point in the future. And thanks to all of you for joining us on this episode of the First Trust ROI podcast. We'll see you next time.
