EP 28: AI with Brad Neuman, CFA - Alger

I think this Podcast is going to blow you away.  I was left speechless. 

The implications of AI are truly outstanding.  And we don’t know of any investment firm that is more of the forefront of this revolution than Alger. 

You are in for a treat, so tune in for a fascinating ride.

Transcript

00:07

Welcome back to the podcast, deconstructing Alpha. I am your host. Jeremy van Arkel with Frontier asset management, and today we have quite a treat. I've been looking forward to recording this podcast. This is a subject I think that is on the front of everybody's minds, but they really it's, it's, it feels so intangible, it's hard to get our arms around it. And so today, we're going to be doing a deep dive on AI and its applications in our lives, on its investment ramifications, on its economic ramifications, and maybe even how we should be looking to the future of AI. Our guest today is Brad Newman with Alger funds, if you're a listener of this podcast, Brad has been on here a couple of times before. He is a wealth of knowledge when it comes to technology. And he comes to us from Alger funds, where he is the director of market strategy. And Alger funds has long been our go to for technology innovation investing. They are a firm steeped in experience, pretty much always on the leading edge of change, and some of you might have noticed that they were just recently stated in the Wall Street Journal as being one of the top performing mutual funds in the last year. We're honored to have Brad back. We believe that Alger is on the forefront of AI, not only how to invest in AI, but also its implications and its its glide path of how it's going to unfold. Here, I believe this is going to be a fascinating interview for not only you, the audience, but it's going to be a fascinating interview for me, because I am not an AI expert and and with any new technology, it's a deep hole. It is a lot of local knowledge and understanding of the subject, and there are probably very few people or firms around that understand this topic better than Alger funds and Brad Newman. So I'm excited for this podcast. It's a little bit different than our normal podcast because it's really just on a specific subject, but I think it's going to be extremely helpful for our listeners to hear what's going to unfold here before we get started, please stay on the on the podcast till the end for important disclosures, and let's dive right in, ladies and gentlemen. Brad Newman with Alger funds, Brad, welcome to the podcast. I'm so glad to have you back. You are my technology go to and I'm pretty excited about this podcast, because there's some people out there in the investment sphere that are talking about this thing called AI, and I'm a bit of an old dog, and I'm slow to catch on and learn things. I hope that you can enlighten me and maybe some of our listeners on this AI revolution and how it's going to affect all of our lives, our investment horizons, and how we should think about it. So welcome to the podcast. Brad, yeah,

03:28

excited to be here. Thanks for having me. We will do our best. We've been talking about AI, not just since it was in fashion over the past 18 months or what have you, but for many, many years, so we have a long history and excited to talk about it.

03:44

Yeah, well, that's why, that's why you're my go to when you say something like, Hey, we've been talking about it for years. It feels like the marketplace and investors and the media and everything have only been talking about it for a few months, and it's taken off like a revolution. You guys are like, a lot of things, technology related, you get there first. So, so let's just go, take a big step back, and let's just start with just my lifetime. And I feel like there's been some several just major technological revolutions in my lifetime and in my generation and our generation. It all feels like it started with the laptop, the computer, then the laptop, and then Internet, and now AI. So how do you fit AI and characterize AI in this progression?

04:32

Yeah, well, I would even start before your lifetime. You know, technology, technological innovations have been occurring faster and faster for a couple of centuries now. So if you go back to the 1700s and 1800s it took over a century for an innovation like the stove or the steamship to reach 50% penetration of households. If you went back to the late 19th century, the telephone took 66 years to reach 50% Penetration of US households. And then things were accelerating in the 20th century, the washing machine, cable TV, took several decades, but were faster than the telephone. And then, as you mentioned, the Internet was much faster than that. It took only 14 years to reach 50% penetration of US households. And then, more recently, social media and tablets took under a decade, so it's certainly been accelerating, and I think part of that is because certain technologies create feedback loops, and those feedback loops are most prevalent in high tech. You know, good example of this is advances of computers help chip designers make faster chips. And so we've seen accelerating rates of innovation over time, because new ideas have built on older ideas, and that is happening faster in the digital economy in particular. Now generative AI in particular, in my view, will be by far the fastest ever, and that is because this is the first major technology that, in itself, is recursive, meaning that ultimately generative AI will be programming generative AI. And that's never been seen before with a major innovation.

06:22

So with these revolutions, these innovations that are, you know, relatively life changing, they're just enormous changes that occur. I feel like a lot of people, because that's happening so fast, and it's hard to grasp it, it's hard to understand how it'll affect us. You know, I feel like a lot of people get scared of this sort of thing, right? It's sort of almost mind blowing to when you start to dig in and think about computer systems that are training computer systems. And I think it's a bit overwhelming for some people and and I think that's sort of a bit of the stumbling block for maybe even me about why I want to grasp this concept and dig into it and and is that it's just sort of almost overwhelming. So, so let's talk about this age of conducted intelligence. I've heard you say this before. So what does that mean?

07:16

Well, we think that technological revolutions happen every several decades, starting from the Industrial Revolution of 1700s to steam and railways, steel, electricity, oil and automobile in the early 1900s and the last one that we all lived through, The Age of Information and telecommunications, and that started with the microprocessor and computers and, you know, and then made its way to smartphones and mobile internet. And today, we think that the next technological revolution has started very recently in the past few years, and we call it the age of connected intelligence, and there's a few building blocks of that. So the first thing is the Internet of Things, which is just devices that are connected to the internet. So, you know, in my house, I have more than 40 things connected to the internet, which is just, not just the PCs and phones that myself, my wife and my kids have, but you know, my my car is connected to the internet, my thermostat, cameras, you know, TVs, all sorts of different things. So all of those devices which, and there's many more devices connected to the internet now than there are people in the world, and they're growing at 15 to 20% annually, those devices connected to the internet. This Internet of Things is creating a lot of data, and that data we call big data, because it's it's so big that it can't really be processed by the human brain. But thankfully, another piece of this technological revolution can help us all process it, and that's cloud computing. In cloud computing, I would essentially just think of is like a utility for computing. So, you know, we all use utilities now for electricity. You know, we could all generate our own electricity. And I tried this when a hurricane came through my area in Connecticut a few years ago, and, you know, my generator kicked on, and I was able to generate electricity, but it was really expensive. It was not that reliable. Actually went out twice during the week that the power was out, and it's just not that efficient. And cloud computing is basically the same thing. It allows us all to essentially rent supercomputers without actually owning them, and have access to latest hardware and software. So it's actually affordable for small businesses and even individuals to take all this data from the Internet of Things and process it in the cloud, and that is when you apply that huge computation to all that data, you ultimately get artificial intelligence, because you can. Create these valuable insights from all this data, and that's essentially what we mean by the age of connected intelligence. It's all of these connected devices being processed by such a high levels of compute that we're able to get new levels of intelligence and new new valuable insights out of it that will help make all of our lives and businesses better, more productive.

10:26

So can I recap that a little bit? So that feels like you said, if, if everything's connected to the Internet, there needs to be a central source of all where all this data goes, and that's sort of the cloud. And then once you have all that data in the cloud, you then have computing capacity that can analyze all of that data and pick up on patterns and and to sort of hopefully help make our lives better, businesses more efficient, and etc, etc. So these pieces work together.

10:58

Yeah, yeah, you can't have artificial intelligence without the data, and you can't have it without the compute. So you need both those things. And I think all of those things are fueled, you know, but by having everything connected to each other. So yeah, it's all. It all has to work in harmony.

11:19

So that infrastructure of that, or the relationship of that, really actually makes sense to me. That's a very clear explanation of that. So, so can I jump back to just in the beginning? You, you, you mentioned that Alger has been on the forefront of AI technology and understanding and seeing how AI is progressing, and maybe thinking about it, maybe even investing in it for years, whereas I feel like it's just came out of nowhere. Can you comment on sort of the history of just the AI piece, like, where do we sort of begin that? Is it? Do we begin that with the Internet? Is that where that sort of began or,

12:01

well, so you know, Alger has been investing in leading edge of innovation since 1964 in fact, we're celebrating our 60th anniversary. And I don't know how much your listeners want to know, but you know we were investing in integrated circuits in the 1970s we own Intel in 1977 back then, just interestingly, their latest chip, when we own that stock in in the late 70s, their chip had 29,000 transistors in it. Today, if you're walking around with a modern cell phone in your pocket, you probably have 10 billion transistors on your chip. So it's been, you know, tremendous ride and productivity from all those increasing transistors. And you know, we were investors in personal computers in the 80s. We owned apple in 1987 when it was a split adjusted 38 cents, if you can believe that. So anyway we go, we go way back into innovation, technology investing, artificial intelligence? Well, artificial intelligence really started many, many decades ago. You know, was first kind of conceptualized by Alan Turing in the 50s, where he talked about what we now all refer to as the Turing test, where artificial intelligence would be able to pass a test where people couldn't figure, people couldn't differentiate it from a human. So it could, it could disguise itself, essentially, and be just as smart as a human, kind of a blind taste test, so to speak. You know, there was a lot of work on artificial intelligence in, you know, all the way up from the 60s, 70s. And then, kind of, there was kind of a winter for artificial intelligence where people got discouraged. And then, you know, you probably remember, or some of your older listeners will remember, you know, deep blue IBM's Deep Blue. And late 90s, defeated Kasparov and chess, and then IBM Watson won a Jeopardy in 2011 but it really wasn't until a program called AlphaGo zero in 2017 that there was, I would say, you know, a huge leap forward, which is when I started writing about AI again. And you know, we certainly have investments in semiconductors, etc. And then we got very bullish around the chat GPT era. We had owned some, some major semiconductor players prior to that, and we thought semiconductors would actually, were actually underappreciated for what they meant to the age of connected intelligence, you know, many, many years ago. And so we've been investing in semiconductor capital equipment and semiconductors themselves for many, many years. Well, I mentioned it goes all the way back to the 70s, but we thought there would be our applications artificial intelligence for many years. And then one. Chat GPT came out, was very clear that NVIDIA was leading a revolution in

15:08

GPU powered processing.

15:11

So you mentioned 2017 was there? Was there something that happened around that time that just enabled this to really come to the forefront? So if for decades, it was behind the scenes and slowly improving and just in not becoming something that was had the power to become mainstream, but then boom, it's on the scene. Now, was there some technological leap that enabled sort of such progress? So we

15:37

think about it as three big drivers, but one of them happened to come on the scene in 2017 so from for many, many years, chips are beginning more powerful. As I mentioned earlier. That example from the 70s, you know that we've just been adding transistors, and that makes chips more powerful. You know, Moore's law, number of chips doubles. Number of transistors, sorry, on a chip doubles every couple of years, and that's been making computation much more powerful. That's obviously very important to artificial intelligence, and in particular the graphic processing unit, which allows computers to think in parallel, meaning process a whole bunch of things at the same time, rather than sequential, like a CPU. So the evolution of chips over time, and the GPU, that's one piece. The second piece is more and better data. So, you know, if you think about AI as a rocket ship, you know, the rocket ship couldn't fly without the fuel, and the fuel is the data. And so the data is very important. As I mentioned, there's more and more data, you know, as the internet was built out to be able to fuel AI. So that's the second piece. And the third piece is a soft new software architecture that came out in 2017 that essentially allowed computers to think in parallel. Software to think in parallel. So in other words, the computer could sit the software architecture could allow the computer to look at a whole sentence and kind of figure out all those words in context, rather than sequentially, kind of calculating where a word what word should come next. So that was a big leap, and that led to deep minds, AlphaGo, zero around that time, which you know, competed or learned the game of Go, which is kind of like an ancient Asian game, and it wasn't given any gameplay instruction. It wasn't given any strategies. It was simply given the rules and it was allowed to play against itself. And just from that, in a few weeks, it was able to beat the best software programs ever developed in that game, and also able to be the greatest human at the game of Go. And essentially, it kind of marked a revolution in that it was the first time where computers were no longer constrained by their human developers.

18:08

I think that last comment about the computers in a matter of weeks, learning how to play the game of Go, which is an ancient game, and being better than any human is the is quite intimidating, right? That's part of the part where people go, wait, wait a minute. I'm a little bit scared of this technology, right? And so, so AI gets smarter every day, right? And so, what point does it sort of become smarter than a human? I think that's a common, or even replace kind of human thinking. So like, I mean, how are we on the progress scale of that?

18:51

So it so first of all, AI is, is accelerating very rapidly. So I mentioned Moore's Law and transistors a couple times already. We're doubling every 24 months, right? Well, artificial intelligence based on how much computation, how much training it's digesting, which is kind of the there's a pretty good relationship between how smart an AI program is and how much training it's been done on. So how much so the number of parameters, which is, you can think of it as kind of like data points or words that go into the training has been really increasing quickly. So GPT one in 2018 was trained on 117 million parameters. GPT two, just the next year, was trained on 1.5 billion parameters. GPT three was trained on 175 billion parameters in 2022 and 2023 GBT four was trained on 1.7 trillion parameters. So that's doubling roughly every four months. And I think it's hard for a human to think exponentially, or I know this for me, so it might be hard for your listeners too, but there's a very. A huge difference between doubling every four months and doubling every 24 months. And let me just give you an analogy. A six inch plant growing in Moore's law, which is doubling every 24 months, would be 16 feet in a decade. So, you know, a lot of progress,

20:16

but growing at the speed of AI, it would reach the moon.

20:21

Okay, that was, that was a lot.

20:24

Yeah, it's very, it's very hard to conceive of. I mean, you're talking about 20% monthly growth, instead of 3% or 821% a year, instead of 41% which in a decade is 4.4 billion times expansion at at AI speed, rather than 32 times at Moore's Law transistor speed.

20:46

So okay, I'm a little bit blown away.

20:52

So yeah, I don't so you know it means that we're going to have cheap computing technology more powerful in the human brain in our lifetimes. You many smart folks, you know, if your listeners want to read more about this and really get into it, there's a great book that just came out. It's on my desk as a kind of reference called the singularity. Is nearer by Ray Kurzweil, and I think he's predicting general artificial intelligence, which means smarter than the smartest human

21:23

by around the end of the decade, and then he

21:28

is anticipating kind of a singularity. He calls it, where AI just iterates on itself so quickly that essentially, humans merge with artificial intelligence at 2045 but there's going, to, suffice it to say, there's going to be a lot of change, more change I've ever seen in the next decade.

21:49

Yeah, and change can be scary or it can be an opportunity. So let's, let's just jump a little bit to like one of the other. You know, there's two schools of thought on what this does to how people work, right? And so the one school of thought which I generally try to attach myself to, because of all of history, this is what's happened, is that every technological revolution is like a balloon. You blow up the technology the balloon, and everybody benefits, right? And then there's another theory, or not theory, but I kind of sense that there's another way to look at it, and that that, that way to look at it is there's just winners and losers, and that, you know, if you were to think winners and losers, you would think, am I going to lose my job, right? And so, so does this boost the economy, or does this kind of net neutral, because it really helps some areas and other areas are just kind of blown

22:46

away. So I think there's going to be some displacement, as there always is. So, you know, we look to history for this, and, you know, we actually wrote, I think you can google and find it, the powerful gale of innovation. Is a paper I wrote some time ago where we looked at kind of what happens in big innovative periods, what happened to jobs. And obviously, some jobs go away, and new ones are created. And, you know, this has happened throughout American history, in the industrial revolution, many farmers, their jobs got automated by tractors well before, before the Industrial before they went to work in factories and They They essentially got a little bit better education and and went to work in factories. And so instead of having 40% like we did at the turn of century in agriculture, today, we have, you know, less than 2% working in agriculture. But those folks in the early 20th century went to work in factories. Their pace actually increased. And then, of course, manufacturing in the middle of the century, or later in the century, started to get automated, but also outsourced to other countries, particularly with China entering the WTO. And you know, services has been growing for the past many decades. And so the manufacturing base became whittled down, and folks found themselves able to get jobs in the services side, and they needed more education for that. And so I think this next step will be similar to that. And like I mentioned in this paper that we wrote the powerful gale of innovation, we mentioned that, you know, in the early 20th century, for example, there was, you know, ton of railroad employees. For example, there was over 2 million railroad employees. You know, today there's under 100,000 there was a bunch of cobblers and blacksmiths, you know, working on horseshoes and other things for horse. Is hundreds of 1000s today, you know, there's virtually zero of those folks. But back then in the early 20th century, there were no professional drivers. And today, about 3% of all Americans, you know, or several million folks, are professional drivers. And if AI really turns out the way we think it will, maybe there won't be any professional drivers in 20 years, but those people will find a new job, and that some of those jobs, or those jobs, haven't been invented. So ultimately, I think it'll be very positive for the economy. I think there'll be a lot of new jobs created, whereas people have to learn new skills. And the way we think about it is that there's a, you know, roughly a third of the workforce is knowledge based, and I think their productivity can increase by something like 25% with AI, I could talk about why we think, why we get to that 25% number. But essentially, if you put those two numbers together, the 30% of the workforce, that's knowledge base, times the 25% that that their productivity could increase, that gets you to about $11 trillion positive impact to the global economy by the end of the decade. It

26:13

it is easy to look at the current situation and think of bad outcomes. I think we're just humans. We try to detect, you know, areas that could cause us fear, because we're protectionist, right? We want to protect our own well being. But you can't deny that progress of technology driving revolutions, displacing things temporarily, but ultimately, the world goes on and gets more robust. It's just really hard to look at how how we live today compared to when they they had a lot of people that put horseshoes on horses and think, you know, life isn't better. I think it's probably better. So great, great way of summarizing that. Alright, let's take a quick jump over to Alger. Let's make it a little bit more about you guys. And so your Alger spectra Fund is a fund that has, you know, as you know, is has a lot of interest to us. And there was a nice write up in the Wall Street Journal about it just recently. And you guys have had a great year to date. And one year, I think, you've really been on the forefront of of this movement in the market that's driven so few securities, and you've done so well from that. And, and, you know, having listened to you talk, it's like, you know, you were one of the few that was like, No, this is real. This is, this is big and and so how is, how is Alger? How are you guys positioned now to really take advantage of AI, the continuation of this and how it unfolds and all that. How are you making any changes in your portfolio? How is your portfolio positioned and what you know? What aspects do you think are going to be things to look for in the future?

28:03

Yeah, yeah, you're right. By the way, the Wall Street Journal recently had an article where they looked at the 12, or the top 10 best performing mutual funds over the 12 months. Ended June 30, and Alger, including the spectra fund, had several funds on there, making the top 10. So we're really happy about that. We are we have, you know, a major theme in artificial intelligence, as we have for a while, within our funds. And we think about those investments in kind of two buckets, those AI investments. One bucket is we call artificial intelligence enablers, and the other one, we call artificial intelligence adopters. And so the enablers are those companies that breed the infrastructure for other companies to be able to utilize artificial intelligence. So you can think about that in two buckets, itself, infrastructure, hardware and infrastructure, software and services. So the hardware examples would be some of the poster Childs, probably for children for AI, like semiconductors. So that's everything from, you know, Nvidia, which makes the GPU processors, kind of the brain of artificial intelligence,

29:27

AMD and others

29:30

also. Memory is very important, you know, you can imagine, you know, if you're working on a big project with a bunch of paper spread out on your desk, Ram is very important for you to be able to access all those papers you can think about. You know, the the desk and the size of the desk being equivalent to how much RAM you have. You know, if you have RAM, more RAM in your computer, you have essentially a bigger desk to be able to recall more more papers and. Other companies, like Mike Brown are coming out with something called high bandwidth memory, which essentially like having a multi layer desk, to which so you can have everything close to you at hand. And that's very important for artificial intelligence to be able to have all that memory. So you've got processors memory. The companies engaged in the actual semiconductor production, like Taiwan Semiconductor, etc, the companies that help the data centers run, which have so data centers are kind of, you know, they're like the electric power plants for we were mentioning before, for computing, and they house all these servers or computers in there. And there's a lot of work that has to go into making these data centers run. They have to be cooled because they generate so much heat. So we own company like vertiv, that is engaged in the cooling process. Companies that build the data centers or run them, like Equinix, which is a real estate company, and all the way down to we think power is going to be in demand as these data centers suck up more and more power. So power generation and companies that help get the power to the data centers, we could talk about that. But anyway, those that would be kind of infrastructure hardware and infrastructure software would be the cloud platform and AI services like big cloud providers, Microsoft alphabet and Amazon, and then AI applications like Microsoft's copilot databases, which are really important because all this data is unstructured, which is a totally different you need a different way of storing unstructured data. You know, when you think about like the stuff that AI learns from it's, you know, blogs and web pages, social media, videos, all that we call that unstructured data, because some feet fit neatly into, you know, a particular category. And so there's databases, particularly for that, like MongoDB. Just we've owned some of our strategies for a while, network services, design, automation and and others for the infrastructure power surfaces. So that's one whole bucket. That's the AI enabler bucket. You want me to talk about the AI adopter bucket?

32:11

Sure, I'm I'm wrapped with attention. This is really just fascinating to me. So I would love to hear about the AI adopter bucket.

32:21

Okay, so the enabler bucket was creating infrastructure. So the adopter bucket is utilizing that infrastructure for, I would say, two purposes. So two buckets within the adopters one is to make their external product or service better for their clients. And the second area that adopters can utilize artificial intelligence is to make their own businesses better through internal productivity or margin improvement. And so you know, one example on, say, the external side, would be Pinterest. So Pinterest is a kind of like emerging social media platform where you go there say you're interested in remodeling your house, and you're looking for various ideas and different styles, and Pinterest is using artificial intelligence to help understand what would appeal to certain people as they, you know, like, what are they really looking for when they like a certain esthetic? Like, you know, is that a particular type of design and can, can the program serve more designs that the searcher the user will like, and to the extent that they can do that better, they would have better advertising revenue, because if users find what they like more, and they can click on some of those things to lead them to the website or to purchase them, then the advertiser will have a higher return on investment. If the advertiser has a higher return on investment, they will pay more for that ad. And Pinterest is seeing that improved monetization from using artificial intelligence, you know, then there's companies using AI for internal productivity. You know, we've heard a lot from like JPMorgan Chase and some large banks that are using AI to help automate various services within the bank, like protocols for they have to know their customer, KYC protocols and a lot of that's very time intensive, but with using AI, they can do it much faster, and they're seeing huge productivity benefits, which saves the company money. You

34:42

just touched on a whole swath of the economy, right? You've got energy, you got banks, you've got online or retail. You touched on the technology and the hardware side and the infrastructure side, and it just sounds and feels. Like, it really helps so many different areas, right? And, and it's, and so that was, like a really cool summary of the, you know, sort of the build out and then the people that are going to take advantage of it. Alright? So for my investment advisor listeners, that is my audience, mainly, how should they think about AI, and is there, have you been able to, I know this is an extremely specific question, but if you've been able to think about, anyways, maybe how the investment advisory community, I mean, you work with investment advisors, you work with RIAs and how they might be able to benefit from Ai.

35:35

You're talking about not, not, not in their not investing vehicles. You're talking about AI helping their own practices. That's what you

35:44

mean, yeah, yeah. That's kind of what I was thinking.

35:48

Well, you know, we're seeing some of our partners, some of the big wire houses use AI to basically make sense of unstructured data. So it's at your fingertips. So, you know, you can imagine, you have, you've had many conversations, obviously, with you know, your clients as an advisor, and you have a say, a call coming up with the client. And I'm sure, normally, to prepare for that call, you would go through their portfolio, what their goals are, what they've communicated to you recently, life changes, all those things, well, that's pretty time intensive. You can imagine a world in which you would have a co pilot. And I believe that everyone will have a co pilot in business and personally, looking at their emails and looking at everything they do, so we'll know their business and themselves better than anybody else, even you know, your parents, because it will see everything, and it will be able to recommend, well, first it will be able to, it will be, will give you a starting point. So I like to say that we'll never start from a blank page anymore. So AI will know that you have this meeting coming up, it will generate, you know, a an action plan and notes, and so you, of course, you may want to edit that, because the human touch is, is still important. We haven't been completely replaced yet. But you won't have to dig up all that information. And you'll, you'll, you're, you will have, you'll be much more productive, because you won't be starting from, you know, from zero, blank slate.

37:27

I like the co pilot. You know, putting a name to something really, kind of helps me visualize what it is, you know, I would, I would love a co pilot. I'm ready for a co pilot. So, so that's great. So, I got one final question for you. And you know, I think I already know your answer to this, but I'm going to ask it anyway. You know, there's, whenever there's something like this that comes along, there's kind of two schools of thought on investing. I don't think it's as cut and dry as value versus growth or anything like that. But, you know, one school says, Oh, this, this is incredible, but the prices are too high on these stocks and, and we've seen this before, and this is getting bubbly and, and maybe the market's overvalued, and it's overvalued for one specific theme, of which we don't know who the real winners are going to be. That's sort of one angle to looking at what's going on in the market today, maybe. And then the other school of thought would be, would be, have you really thought about what's actual technological changes, and what the benefit is, and and how revolutionary it is, and really it will drive way more growth, and that this is not a bubbly market. So what would you say to the Are we in a bubble? Because the AI driving a few stocks, driving the overall market higher, and that kind of smells like the internet times versus this is the beginnings of a new revolution, and it's going to change not only the way the economy works, but also change the prospects for the market,

39:06

right? Well, the answer is complicated, because, you know, so first we'll just economically speaking, I think the answer is more simple. We think we're at the very beginning stages of the economic implications and development of artificial intelligence. So, you know, in a nine inning game, you know, we're, we're just kind of coming out of the dugout, like the, you know, there's still a lot to go. Like you mentioned, you know, we don't have co pilots yet, you know, I'm not, I didn't use one to prepare for this podcast, but I certainly would have liked one. So it's very early on, and there's going to be tremendous amounts of productivity improvement. I mean, I highlighted my own estimate of, you know, many trillions of dollars over the next several years. So the opportunity is humongous, and we're extremely early. So. That said the stock market does begin to discount that, because there's very smart people investing in the stock market, and collectively, we're all extremely smart. So the stock market is a history of anticipating these events and ultimately misallocating over allocating, and that's kind of a great thing about capitalism, because when there is a huge innovation, it gets a lot of capital. And you know, if it wasn't flooded with capital, you wouldn't get that development ultimately. And that's been a centerpiece of all these technological revolutions that I've talked about in the past, you know, and you mentioned the internet, we got over allocation to many different types of companies and investments in the late 90s and early 2000s and only after that did we see the huge productivity enhancements. So I think this may be similar. I think that there's probably some stocks that are rich. I think one big difference with the internet, besides the fact that the valuations aren't quite as extreme yet, is that there's much more underlying earnings growth right now than there was in the internet. There was a lot of businesses searching for a business model in economics. I mean, certainly we all remember, you know, like, you know, the pets.com and all those companies where the business model really wasn't fine, had had difficulty earning a profit. But even even wonderful companies back then, like Google, they didn't really know yet how to monetize. Today, we have companies like Nvidia, where the earnings have nearly kept pace with the incredible stock price appreciation. There's been tremendous amount of value creation from an economics perspective already, and there's hundreds of billions of dollars being spent on this AI build out. So I think it's more of a real world phenomenon right now than the internet was. I think the earnings are more real. I do think that there is probably some misallocation. And I think investors may want to look at kind of adjacent ways to play artificial intelligence. You know, for example, power companies could be and related areas. Maybe one interesting way to play artificial intelligence that's less obvious and less bit up right now. So I think you have to be judicious. You have to know the valuations and the opportunities, the addressable markets. But I think big picture, we're still very early with some spots of overvaluation.

42:36

Brad, I think

42:40

you've answered all my questions, and you've answered all my questions in an incredibly Intel, incredible like the knowledge you're not the knowledge, the insight, your experience and intelligence in so many different areas of technology, again, once again, proven themselves. And I'm just so glad that you have taken the time to to participate in this podcast, and kind of now thinking I've got to learn some new stuff about AI, and it's going to be more work for me, but that's that's okay, I can do that so, but I really, you know your time just invaluable, like this is just been so eye opening to me. Is there anything that you would like to add? I mean, through what we talked about, I'm, you know, I'm guessing we need, like, four more podcasts on AI, even to get to the the to really get somewhere near a complete story on AI. But is there anything you'd like to add, like, just sort of, on a introductory podcast about AI, about what, you know, something we might have missed.

43:43

Well, I just, I think people should recognize, you know, in conversations with some clients I've had, I'm not sure that that, that people were fully on board with how revolutionary AI would be. I mean, I think they, they thought it was a big deal, but some people kind of thought it was almost like a parlor trick. And how, you know, cool is chat. GPT is just, you know, something you could ask questions to. But How different is it than, you know, Google search? And I would say that AI is really more kin. So everyone talks about GPT, but instead of thinking about GPT is the actual product, I would think of GPT for another acronym, which is general purpose technology, so much like electricity or the automobile or the computer, AI is another general purpose technology, which means it will impact many different areas of life. So for instance, electricity created factories and lighting, and ultimately, appliances, you know, in the refrigerator, and then you had Coca Cola that could take advantage of the refrigerator, you know. So it spawned many different industries. The automobile, yes, allowed us to travel further. And what did that do? It changed how we organized. People moved to suburbs. Shopping Centers were created like it transformed all of society. The computer, the digital revolution. And service jobs. So and, you know, an internet changed communication and social networking was became, you know, the way we communicate and share things, and the sharing economy allowed us to, you know, to utilize a company, you know, that has no cars. It's the largest transportation company in the world, and a company that has owns no real estate is the largest hotel company in the world. So I would say that AI is going to permeate everything you know, not just how you interact with information, but I think it's going to be in art. I think it's going to be in movies, music. It'll be in your business life or personal life. It'll be a confidant, you know, it'll be a lot of things. And so, you know, ultimately I would just, I would just close with a little over six years ago, the CEO of Google said that AI would have a more profound impact on society than fire or electricity. So if you really believe in that, then that tells you why I say we're so early in the revolution.

46:03

This has just been fantastic.

46:07

I'm I'm kind of speechless, so I don't know what to say. I don't know how to close this, other than just thank you. Thank you for sharing that. Thank you for helping our audience understand a little bit more about AI. And again, you're just like, I don't know who else to go to in the industry that would know more about this than Alger, and you have well represented that. And just thank you for your time.

46:35

All right, great. Yeah, it was. It was really awesome to be here, and I hope I can join you again soon.

46:40

Thanks. Man. Appreciate it.

46:42

This podcast is for informational purposes only. The information does not constitute advice or a recommendation of any specific investment, mutual fund or mutual fund company. Before making any investment, you should carefully seek independent, legal, tax and regulatory advice. In particular, you should seek the advice of a licensed financial advisor regarding the suitability of the investment product, taking into account your specific investment objectives, financial situation, any particular needs, and your ability to assume the risk and fees involved before Investing. This podcast and presentation are for informational purposes only. Frontier assumes no liability for any action taken in response to listening to this podcast. Frontier asset management is not affiliated with any specific Fund Company. The views and opinions expressed by each speaker are their own as of the date of the recording. Any such views are subject to change at any time based upon market and other conditions and frontier disclaims any responsibility to update such views you.