ANNOUNCER: TD Asset Management welcomes you to this week's podcast. As a reminder, this podcast cannot be distributed without the prior written consent of TD Asset Management.
INGRID MACINTOSH: Hello and welcome to this week's edition of TDAM Talks. I'm Ingrid Macintosh, your host here at TD Asset Management. And this week we're going to be talking about artificial intelligence (AI). So now the concept of AI is not new. In fact, the early its earliest references date back to the 1950s. However, 2023 has brought a meaningful acceleration in the use cases for AI that go beyond what we've seen in this incremental trajectory.
And most recently we've seen some of the pioneers of modern AI ringing alarm bells about the pace of transformation we're witnessing. Today, we're going to talk about the new age of artificial intelligence and specifically the impact it could have on markets, sectors and economies overall.
I have the pleasure of being joined by Vitaly Mossounov, co-head of fundamental equity research here at TD Asset Management, along with Julien Nono-Womdim and Juliana Faircloth, lead analysts in our technology, semiconductor and industrials sectors respectively. Welcome, everybody.
VITALI MOSSOUNOV: Good to be here.
INGRID MACINTOSH: Thanks. Okay. So this is such an exciting and important topic and it's all over the headlines and I really want to start with you, Vitali. Like, what is it that we're witnessing and why might this be different?
VITALI MOSSOUNOV: Yeah, well, at first we almost want to start with exercising a degree of caution because the last few years have seen so many false starts. And in the context of innovation over the period of COVID (COVID-19 Pandemic), we heard about IOT (Internet of Things), worried about cryptocurrency, about 5G. And we've often heard promises that we're about to enter a new age of innovation.
And as we all learned, those turned out to be incremental technologies more than anything. And then in some cases outright bad AI. We think very broadly defined, is different. Now we'll get the podcast, but Julien and Juliana, as to why that's the case. But broadly speaking, we think we are entering a period of a new cycle of innovation and something akin to what we saw.
And initially 200 years ago, in the early stages of the Industrial Revolution with the steam engine, what that did for transportation and industry at large and in fact the railroad, a public railroad, was founded just about 200 years ago. But then, of course, other cycles, electricity. In the late 1800s, what that did for telepathy and other technologies. After that course, the platform of computing with the transistors before that with vacuum tubes and whatnot.
So exciting times. But yes, a period we think where we're going to launch into something new and bigger.
INGRID MACINTOSH: So Julian, I'm going to turn to you as you focus your research around the technology in semiconductor spaces, why now? What's different? As I said, AI has been around for, you know, 50, 60, 70 years in its earliest instances. But what's changing now and why this acceleration and also, why this concern?
JULIEN NONO-WOMDIM: Yeah, I think it's important to sort of think about Silicon Valley, yet modeling simplistically, semiconductor is like the (inaudible) the power of the global economy. Some are inexpensive, some like the ones that power I'd like to see and we've been through this throughout the year, many of you know in the industry, and we now reached a point where the compute power is sufficient, especially sophisticated algorithms.
And these algorithms are solving increasingly sophisticated problems. So it's been an evolution, as you say, and we're reaching an inflection point where it's an exciting technology and some opportunities to do what.
INGRID MACINTOSH: We've talked a little bit and, you know, I've talked a little bit about, you know, some of the incremental changes we saw, whether it was autonomous driving, etc.. But now when we think about the Chat GPTs and this real transformation, what is it that is so concerning about really this acceleration of what's different about the way that AI is being used now?
JULIEN NONO-WOMDIM: Yeah, and bizarrely, that's much earlier. We liked that AI is not an experiment, as you say, and that's in the past. There's this there's been this effort to support the digital world, the physical world, i.e., autonomous. Yeah, that's a very complex problem and one that you're still trying to solve. And I'm actually optimistic that we will solve it over time.
But what happened in the last 10, 20, 30 years that we see in the personal lives, a lot of our a lot of our day to day life is beginning with digital portal. I mean, but certainly strategies both of and as a result of that it's not cheap to generate API is one of the sort of I guess the only instance where we're dealing with a digital only world that that's fairly easy for computers to understand. The fact that our lives are more digital, that information is train digitally makes it about life easier than it would otherwise be in a digital to physical realm.
So I think that's the distinct shift with respect why we've seen this surge and why it's taken on like, yeah.
INGRID MACINTOSH: It can really replicate to your point so much more of what we do day to day or what we've historically done can now be done by AI.
VITALI MOSSOUNOV: That's right. That's right. That's exactly Yeah.
JULIANA FAIRCLOTH: It's I might add too, I think part of the difference is, you know, there's been automation in robotics which had been in existence for a long time. Those are replicated.
VITALI MOSSOUNOV: Small physical.
JULIANA FAIRCLOTH: Repetitive tasks.
Robotic arms that lift up a piece in the factory, put it down somewhere else. This is really getting a lot closer to replicating human decision-making human opinion, taking in information and regurgitating something not that replicate the human mind more than a physical movement.
INGRID MACINTOSH: I love that expression. And to your point, you know, we talk about, you know, the type of innovation you first spoke about, you know, like whether it's in an auto facility, etc., it's like replacing what people can do physically. So it's a productivity boost as opposed to almost like a replacement or an override is what we're looking at.
So let's expand on that. Juliana, because we talked a little bit about, you know, the technology in the semiconductor space specifically. But when you look at industrials or even more broadly, other sectors give us examples of how that might play out, like what might be different going forward.
JULIANA FAIRCLOTH: Sure. So, I mean, there's tons of examples to think about. I think there, you know, a little bit less flashy. Let's say the activity by robotics have been involved in manufacturing for quite a long time. There's a lot of development now around machine vision that will allow for heavy machinery to be operated entirely remotely. AI Machine will be able to do replicate human vision and identify and differentiate between different things within your path.
Part of the big concern that people have with AI is the amount of jobs that might be …
INGRID MACINTOSH: It's what can't be intermediated, right?
JULIANA FAIRCLOTH: At the same time there are certain jobs that maybe should be displaced by machine. There's jobs that are dangerous working in mines
There are jobs that are dangerous to humans working on the railroads. Yet that ought to be it. Why shouldn't it be? That would be a great step for people and humanity and society to be able to move people out of harm's way more frequently. There's some really interesting examples at a company like John Deere, which, you know, doesn't necessarily see that exciting. But they have tractors now that use machine vision to spray pesticides only on weeds rather than rain all over the field do impacts crop yield credible are help. So there's a lot of opportunity to go on a tangent about many examples, but I'll leave it there for now.
INGRID MACINTOSH: But I think I think it's important to go on these sorts of tangents because again, we think about this as a technology story and a story, a human capital story, but how would it affect financials? How would it affect other sectors? You know, maybe talk a little bit about that work.
JULIANA FAIRCLOTH: I mean, I think the opportunity with AI, there's a lot of applications for AI and machine learning and the ability to highlight data that every company had. Data mining engineer data, consumer company update at every company has data. Right now, it's difficult to analyze that data efficiently and effectively and in a holistic way. There's applications for AI beating the financials, detecting fraud and helping banks determine who's a creditworthy person or business to lend to.
JULIANA FAIRCLOTH: That can all be augmented through AI. Again, there's let's not do that, particularly with the analysis of data. Where do we need to worry about artificial intelligence? But in the bank, more deeply embedded biases that may already be within the system. I know they're concerned that people need to think about and look again, that dark side of the AI where it doesn't help improve the operations of the bank. It can improve people's access to credit potentially. But is there a concern that biases right.
INGRID MACINTOSH: Yeah. Amplified. Accelerated. Yeah. And I think about a different use cases. So, you know, whether it's medical research, you know, trying to find cures for diseases or even, quite frankly, investment research, you know, it's fundamental analysts. You have patterns of what you look for. Theoretically, I can gather information for you, but it also could potentially reveal different patterns that you didn't even know that you were looking for different signals.
I know - I really want us to focus this conversation on what's happening in the markets. But when you think about things like fundamental research in the use of the use of AI and big data, help support that research. Could you talk a little bit about that?
VITALI MOSSOUNOV: Sure. As I think Julianna and Julie attest to be, they're very rigorous jobs that demand long hours of work and we utilize a lot of a lot of tools, especially the AI, the human human right.
INGRID MACINTOSH: The intuition. Yeah. Your intuition and your memory and your gut and your experience. All those things play.
VITALI MOSSOUNOV: In and there's a lot of pattern recognition, as you said. Yeah. Um, but ultimately, we are forecasting the future and so far, but we see AI has become very good at doing is taking, institutionalizing and processing historical information, present information, and helping us eliminate some of the, some of the mundane work, the drudgery of work we're already seeing. There are tools that can be applied by fundamental analysts to perhaps even cut down the number of hours that we're all working just to make us much more efficient at getting those insights, getting those conclusions right, instead of really reading thousand pages to find that one insight. Having AI steer you in the right direction and bring it up to the surface?
And that's exciting because it eliminates the work that you don't want to do, and it leads you to sit there, sit back and think and actually broadcast and try to read it - insights and the best recommendation.
INGRID MACINTOSH: So I promised at the beginning we'd have a little conversation about impact on markets, impact on sectors, impact on sort of the economy and the nature of work. So first of all, if you look forward, you know, I think time horizons, the speech that we're talking about, inflection point, if we're looking forward, this evolution or acceleration in AI, how do you see that playing out generally in markets that we'll touch a little bit on sectors specifically?
And then I think we'll go back to that. That bigger question.
VITALI MOSSOUNOV: So you need good evolution in the AI you're setting up for labor.
INGRID MACINTOSH: I know that we're not looking at a crystal ball, but you know, what do you think? And then may dig a little bit deeper into some of the sectors that might be most impacted. Let's start there.
VITALI MOSSOUNOV: That's a great question. We do try to forecast the future. And so it's a perfectly valid question. I look at a high level. We're looking at a technology that is going to be an enabler of automation, productivity. And with those two factors, we are potentially looking at higher margins and greater degree of profitability within, say, the large businesses that comprise the S&P 500.
Those are of course positives for earnings and revenues and typically stock prices - now there's a lot more that goes into that. Now, generally speaking, these are going to be a tailwind that play out over the long run. The big question, of course, is sector by sector. How exactly does that look? How does the market perceive these changes? Disruptive? And therefore the market may be willing to pay less for student earnings because it's unsure of the future? Or will the market become euphoric? And it's been seen in past periods of euphoria, pay excessive amounts of money, presumably. So all these questions are still up in the air. But I think at a high level, these are positive tailwinds to what really matters to business valuations, and that's earnings.
INGRID MACINTOSH: I want to sort of double down on that new opened up at the very beginning of the podcast, talking about some of the trends and themes we've seen in recent years. And again, headlines drive investor behavior at the margin, increased market volatility. So how do we know where to look for the true themes emerging versus the noise? And maybe, Gillian, let's talk about the technology and semiconductor space, right? There'll be a lot of headlines around this, but how do we talk about our focus on quality and what we look for here?
JULIEN NONO-WOMDIM: Yeah, there's certainly going to be a lot of headlines that, you know, as we talked about earlier in the discussion, semiconductor is an industry that you know, the audio's not as historically faster than the penetration of technology is still funny and up at any given point in time span, we have to know what it's going to need to provide that logic that somehow something happens in this particular instance instead of the result of that, we want to be better finishing evidence and then like we did and the results from the company show that the demand for it, the chips that power these AI engines, is extremely strong. And so for us it's recognizing the demand environment for the technology, it's recognizing that some companies are going to be beneficiaries. But at the end of the day, what we've spotted, it's a process of analyzing the future free cash flow of those businesses, determining certain bands of confidence about the reliability of those cash flows. And that materialized and ascribing some kind of risk to that.
I think it's very exciting to see technology change. But the process of investing is kind of unchanged, Vitali, do you think otherwise?
VITALI MOSSOUNOV: I think you made a great point and I think this is one that we should really underline here yes, because ultimately we have a large research team of analysts that do readers fundamental work and we own the best companies in the world in our investment portfolios as a result. And over the past few years we started with this podcast talking about the various hype cycles around technology like currency and blockchain. Thanks to the process that we've cultivated at TD Asset Management, we have entirely avoided all of these pockets of hype and eventually complete blow ups.
With AI, we are applying the same rigor to investing so there are going to be many stocks from here Ingrid that double, triple, quadruple and go to the moon without any merit. And eventually there'll be a price to pay on the end of that. We don't intend to participate on the upside or the downside of such businesses, but there are others that have real business opportunities and what we're seeing in this period of discontinuous innovation and those are the holding that our investors can look forward to, to benefiting from.
INGRID MACINTOSH: Yeah, I think that's critically important to understand that focus on research, the focus on quality. I just wanted to double back on the semiconductor conversation. Julien, the huge demand for semiconductors, how does that play out from a commodity perspective or is or commodity sectors that are impacted in terms of the ability to, to create there?
JULIEN NONO-WOMDIM: There are a lot of commodities that go into semiconductors.
It's a complex supply chain. But again, it's an area where the industry has built in a very at the outset of the crisis, there were some concerns around shortages in email. Yes. That those that are operating now abated the industry. But now the way is to recycle. And ultimately we talk about things like ESG and in fact, AI is helping the manufacturing process, the design process of semi-conductors becoming more efficient.
So it's a bit circular in that way. And so as much as I like to say that commodity risk is present, I don't know.
INGRID MACINTOSH: Yeah, it's not like EVs and lithium. It's not like that. It's there's abundance of what we need.
JULIEN NONO-WOMDIM: Essentially. Yes, there is abundance. And in the event that there isn't an abundance innovation will find a way to circumvent that.
JULIANA FAIRCLOTH: I think that since you bring up EVs it reminds me of the link that there is to industrials in the industrial economy – perhaps there is not a clear commodity play on semi-conductors, but from a capacity perspective, we don't necessarily have enough capacity. The CHIPS act in the US is trying to build that up more holistically throughout North America, but there is a lot of investment and spending that needs to go into support the massive amounts of semi-conductors that we need.
There's a ton of spending that needs to be done on data centers and that whole infrastructure. So there's a flow through to the economy right now, commodities specific, but there's a lot of investment in it.
JULIEN NONO-WOMDIM: Right. And I think the other piece to consider is when we talk about certainly jobs being automated away, that's a negative. On the flip side, there are jobs that are being created to build the infrastructure associated with those investments. But I think it's important to move thinking around the benefits that it disadvantages of an AI build out.
INGRID MACINTOSH: And also the agility or people's capability to change their skill sets to evolve with the landscape. And so before we before we sort of wind down, Juliana, let's talk a little bit more about other sectors that you think will be most impacted by this acceleration of this revolution.
JULIANA FAIRCLOTH: I mean, I see really an impact to almost all sectors. Just think about kind of a generic business model, the opportunity to use automation and artificial intelligence to improve a company's profitability and their margin by automating jobs. But I think it can be factors that blur that would be sort of a widespread way across many industries.
I think it will be interesting for companies to observe, or perhaps an analyst to observe how different companies choose to invest in and innovate on their own products and offerings to leverage, See Perspective. There's a lot of opportunity for different companies to be on kind of the cutting edge of different things.
I was reading about all sort of companies in preparing for this podcast. But it was interesting that Carlsberg is using artificial intelligence and sensors to take beer and select stock for people, and it allows them to come up with beers 30% faster than they ever would have come up with beer before.
VITALI MOSSOUNOV: I'll take that job, I don't want AI to take it.
JULIANA FAIRCLOTH: So now, of course, it's not a groundbreaking, critical innovation to society, but….
INGRID MACINTOSH: It's indicative of the thought process at a leadership table.
JULIANA FAIRCLOTH: Exactly. Yeah. And there's there's opportunity across the board, even in, you know, a beer company. And part of that can be automated. They don't need beer tasters anymore, Sorry Vitali.
VITALI MOSSOUNOV: Next time, I'll find something else.
JULIEN NONO-WOMDIM: It's fascinating because even in the culinary arts, like I did a number of years ago, IBM partnered with a number of chefs once in a while to sort of determine ingredients that pair well with one another, but like trading standpoint and you have your combinations that are just like incredible, like a arugula and green melon!
INGRID MACINTOSH: That nobody would have started the day saying, try this on. Yeah, but the data takes you there. That's fascinating. I think, you know, we've talked about the different sector levels. We talked about all the positives, margins, productivity, but we keep talking about, you know, people's jobs being disrupted. So there's got to be a downside in here to the economy totally like how do we think about that?
VITALI MOSSOUNOV: Well it depends on how controversial you like to be, but it's a topic of debate that I think the team engages in all the time. The typical answer is that through every cycle of innovation and disruption that past you have had jobs eliminated. It's been a painful period for those have lost gainful employment, but new jobs were created in and typically more so. In 1900 England, there were over 100,000 telegraphists and typists. That's jobs that, of course don't exist. Didn't exist 20, 30 years yet. Yeah, here we are. So going forward, I tend to lean a bit in the bearish camp unemployment here and we're talking decades out but yeah but big picture this is the first time we are confronted with a technology that is an automation tool at its heart and we talked about earlier about profitability and productivity gains for companies.
What that means is potentially less employment. And we already see companies signaling that, again, IBM, one of them that we've seen some about back office margins, for example, 30% less employment. As we can make out existing staff far more productive. So in the long run, you could say I'm bullish on profits, but a little bearish on employment which of course creates the need for another podcast about all the societal issues that this….
INGRID MACINTOSH: Well, I think about. And as we have this conversation, I think about what are all the things that we need more of in our society that you can't automate. So you can certainly automate medical research, you can't automate nursing at its core, you can't automate elder care, you can't automate not a number of these other things. So, you know, I think a lot of good conversations still to come in terms of the impact of the AI acceleration, the productivity acceleration and how it impacts.
I would love to welcome you all back in future podcast to continue the conversation, if that's okay.
VITALI MOSSOUNOV: That would be lovely.
INGRID MACINTOSH: Any closing thoughts or, you know, things that you would hope our listeners would take from this this point in history.
VITALI MOSSOUNOV: You know, hopefully what your listeners have been able to hear today, and especially with that, particularly Juliana and Julien, is the thoughtfulness that we're putting behind the investment research and investment opportunities. Julien speaking about semiconductors and, and. Julien it's speaking about the foundational technologies and architecture of what's happening, what they're really hinting at and that they didn't say explicitly act for the purpose that your audience, but they're going to be different investment cycles that happen as we go through this AI evolution - first that's the so-called shovels to the miners in this gold rush and that's where semiconductors play a role. And Julien had talked about the need to increase capacity in many areas of industrials, but then it will be other types of infrastructure and platforms and operating systems and applications. And why I say all that is A) again, so to pat these folks in the back and say what tremendous work they're doing, but that's to remind investors that we're in unprecedented times here in terms of the complexity of business, the speed, the rate of disruption and the complexity of investment research and analysis.
And we're certainly making investments on our end to make sure we're ready for it.
INGRID MACINTOSH: I think that's critically important. And I think about future questions that we really want to have. And one of the things I'm taking from this conversation is A.I. and the growth of it isn't just an investment theme, it's actually a lens through which you look at all investments. Right? To your point, Julianna. So I think something that we want to circle back on. Thank you all so much for for joining me today. Thank you. And for our listeners, you can find our recently published commentary and perspectives on the TD Asset Management site, along with more of our latest thought, leadership and commentary. You can always follow us on Twitter @tdam_canada and on LinkedIn at TD Asset Management. Thanks, everybody and have a great day.
ANNOUNCER:
The information contained herein has been provided by TD Asset Management and is for information purposes only. The information has been drawn from sources believed to be reliable. The information does not provide financial, legal, tax, or investment advice. Particular investment tax or trading strategies should be evaluated relative to each individual's objectives and risk tolerance.
Certain statements in this podcast may contain forward-looking statements that are predictive in nature and may include words such as expects, anticipates, intends, believes, estimates, and similar forward-looking expressions or negative versions thereof. Forward-looking statements are based on current expectations and projections about future, general, economic, political, and relevant market factors, such as interest and foreign exchange rates, equity and capital markets, the general business environment, assuming no changes to tax or other laws or government regulation or catastrophic events. Expectations and projections about future events are inherently subject to risks and uncertainties, which may be unforeseeable.
TD Asset Management operates through TD Asset Management Inc. in Canada and through Epoch Investment Partners, Inc. in the United States. Both are wholly-owned subsidiaries of The Toronto-Dominion Bank.
® The TD logo and other TD trademarks are the property of The Toronto-Dominion Bank or its subsidiaries.
Canada
