ED MINN: Thank you Alex.
ALEX: Ed, before we get into AI, I wanted to talk about portfolio construction for a minute, because at Weatherbie you’re really involved in two quite different portfolios. You’re co-portfolio manager along with the Weatherbie team of the Weatherbie Specialized Growth Strategy, which invests primarily in small and mid cap companies. But you’re also the sole portfolio manager of Select 15, which is something of a different animal. Can you talk about what it’s like to work on those two portfolios?
ED: Well, Specialized Growth, we seek to invest in quality growth companies that we believe will be long term winners and the large caps of tomorrow. And that is a 50 stock strategy, and in Select 15 that is essentially a best ideas version of Specialized Growth, where I take what I view as the most compelling 15 ideas from that 50 stock universe. And I use five key factors to boil it down from 50 to the 15 holdings.
Number one, growth. The more the better. Number two, estimates. Number three, valuation. Companies that can at least ideally hold their valuation multiple. Number four, catalyst. And number five, diversification. And essentially Select 15 was born out of our highest conviction ideas.
ALEX: Great. So, Ed, as I was getting ready for this podcast, I was taking a look at the piece we did way back in May 2021,
The Race to Embrace Artificial Intelligence, which people can still find on Alger.com. Tell me, what drove you to be so focused on AI even back then?
ED: Well, simply put, I view AI as one of, if not the most important mega-trend that will span my career. And so, it’s critical for me to think about this in depth in terms of how to pick the long-term winners and position the portfolio for success.
ALEX: And you believe that generative AI is actually comparative to earlier transformative technologies like electricity and the Internet?
ED: I think it is up there. I think that economists cannot detect the improvement in productivity within the first five plus years of major innovations like the Internet, or the steam engine. It takes time for the ecosystem and regulations to develop to really leverage these major innovations. But I do believe that artificial intelligence is going to get there and have as big a transformative impact as the bulls expect and hope.
ALEX: How do you think the adoption of generative AI takes place?
ED: I think about it in three phases. There’s kind of the low hanging fruit phase, which we’re in now, with generative AI, as we see large, sophisticated tech companies add features or add on skews to their existing products that are powered by generative AI. Companies like Adobe, Microsoft, Salesforce, some of which already have products, and generative AI power features out there. And I think we’ll see so many more over the next six to 18 months. And I think it is fairly simple for enterprises and consumers, in the case of Microsoft being able to incorporate these into their lives or businesses. And so, this is sort of the low-hanging fruit phase.
And I think, this initial phase of capturing the low-hanging fruit will see some disappointment, and I think that expectations are overheated in certain areas. And I think of the next phase of generative AI as being when enterprises more fully incorporate generative AI across their processes and begin to customize the models, versus the off-the-shelf model that they can readily access now. And for a variety of reasons, that is very difficult, complex, time consuming, and there are a host of security and privacy considerations that we’re increasingly learning will make this a very multi-year effort.
And then in the last phase of generative AI, I think the hope and expectation is that everyone will have a co-pilot, or virtual assistant of sorts, that will help them take care of simple tasks, both at work and in their personal lives. Former Google CEO Eric Schmidt has posited that we’ll have co-pilots maybe five years from now. And he said that in February of this year. My own view, which I humbly submit, would be more like 10 plus years from now. But I think we’ll get there and it’s very exciting.
ALEX: But it sounds like you’re trying to keep a fairly healthy skepticism about it?
ED: Yes. Indeed. You know, nine plus articles out of ten I read are super bullish on the promise of a generative AI, and I don’t think I’m hearing enough about the many hurdles that it must go through beyond this kind of low-hanging fruit stage, where Microsoft Office 365 adds some generative AI features to it. Sure, that will be easy. But to really embed it in the processes of companies and customize the large language models, that’s a whole different ballgame.
And I am skeptical of the pace because I talk to a lot of companies in the real world, and I think if it’s a normal public company, they don’t really know what to do with generative AI. They may tell you that they’ll use it to increase some efficiencies in the back office, because they know that investors expect them to have something to say on generative AI. But really, I think they’re at best doing some experiments. This will come over time, but I think of it as more like three to five years out.
And then, the regulatory environment is not yet well defined, and potentially very onerous. You know, there are some obvious copyright issues that have to be hammered out.
ALEX: Do you think there are some current opportunities for faster adoption that are legitimate?
ED: Sure. I think certain industries just lend themselves to faster adoption of generative AI. One that comes to mind is education, specifically language instruction. I mean the whole concept of generative AI is that it can output natural language text. But it’s excellent at translating languages and the holy grail of foreign language instruction has been to have a virtual tutor that is at parry with a human tutor. Right now, we don’t have that. But I think it’s just a matter of time. And not surprisingly one of the launch partners of Chat GPT-4 was a prominent foreign language learning company, Duolingo.
ALEX: Ed, I wanted to ask about your portfolios. Can you talk about specific holdings that you’re already seeing effectively beginning to leverage this theme?
ED: Yes. In terms of how we’re positioned in the portfolio, first of all, we’re not getting ahead of our skis. You will not see us invest in an unproven generative AI pure play that just doesn’t fit our investment process.
I think where you will see us play is in the picks and shovels approach, to get exposure to generative AI. There’s a massive amount of high performance infrastructure networking equipment, compute power management devices, all sorts of equipment that is necessary for the success and adoption of generative AI. And I think we will also gain exposure to generative AI through the optionality of companies that are compelling on their own right. Be they existing holdings which we may add to as they show that they're getting some traction in that area, or new perspective holdings that have that optionality associated with them.
We think that long term winners in the generative AI space are likely those which have a dated advantage, meaning they have access to a large amount of unique data which will enable them to have a customized large language model, not an off-the-shelf generative AI functionality that everybody else can fairly easily and inexpensively have. I think a lot of the differentiation will come from the data, because the models themselves will be kind of largely similar in terms of performance in the end.
And so, one such company that comes to mind is Definitive Healthcare. It’s a cloud-based software company that sells healthcare intelligence solutions. So, this company, over more than 10 years, has amassed a huge and unique database of all things healthcare. Some of the data is proprietary, collected by the company itself through surveys or phone calls. Some of it is public data, and their special sauce, if you will, is how they’ve consolidated it, cleaned it up, and created interesting linkages across all this data, such that users, their enterprise clients, can glean insights from the data. And so, this is a company that’s not presently leveraging generative AI in a meaningful way. But three plus years down the road this unique store of data could be powerful, because their customers will be able to interact with Definitive Healthcare’s platform via a more natural language interface, thanks to generative AI.
A common use case is a pharmaceutical client has a new drug, and they want to figure out how to maximize and accelerate the sales ramp and peak sales of this drug, obviously. And so, they might come to Definitive Healthcare to figure out how many salespeople they should hire, how to divide up the sales territories, what the sales quotas should be, and even what doctors are most likely to prescribe this particular drug. That’s a fairly common use case where this company’s unique database can be extremely valuable to customers.
ALEX: Ed, final question. What message do you think investors should take away from this conversation?
ED: My overall view is that generative AI is super exciting and will be transformational in terms of productivity for workers and people in general. However, I think it’s more important than ever to apply a lot of human intelligence to the research process, to separate the hype from the reality, and identify the long term winners and losers.
ALEX: Ed, thank you so much for doing the podcast this afternoon.
ED: My pleasure Alex.
ALEX: And thank you for listening. For more information on Weatherbie Specialized Growth and the Alger Weatherbie Select 15 strategies and for more insights on artificial intelligence and investing, please visit www.alger.com.
