How to Think About Investing in AI (EP.242)

0:28We all need to make smart decisions with our money. pretty often. That first idea is that AI will change the world, something certainly we've all heard. And the second idea is that as a result, the most popular AI stock should outperform. If you've been listening to my past several guest episodes, this is a topic we certainly have touched on, but this is the first time I'm going to give you a quick framework for how I've been thinking about investing in AI. As always, you can find detailed show notes at thelongterminvestor.com and at the top of the episode page, whether you're listening in Apple Podcast or Spotify or wherever you get your podcast, you can find a link to sign up for my newsletter.

1:07It comes out every other Wednesday, has links to the stuff I'm reading, as well as exclusive content that you cannot get anywhere else. Okay, I think we need to start with the idea that AI will change the world is not the same as AI stocks will outperform. And the key thing to understand about that framework is that I'm thinking of AI as a general purpose technology, like electricity, railroads, or the internet. Those technologies don't show up as just a single product, especially when they're introduced to the world. They show up as a long, messy rollout where the real payoff comes from rewiring workflows, not just this one breakthrough moment.

1:47You can actually see that rollout happening in the reports of trillions of estimated dollars that will be spent on AI-related build-out. So it's understandable that people want to make sure that they're either invested in AI companies themselves or have exposure to the companies benefiting from the expanding infrastructure to support AI-related activities. But it's unfortunately not that simple. And for starters, the build-out is expensive and expensive build-outs can crush margins. So when industries are in an arms race, companies spend aggressively, sometimes faster than profits can follow. And this leads me to the second idea, which I think if you were around for the internet bubble, you saw this yourself.

2:30Creative destruction is undefeated. I mean, look, the internet changed the world. That is not a debate. But if you time traveled back to the 90s and bought the quote, internet winners, you learned a painful lesson. The technology itself, it can be inevitable, even if its leaders are not. And it's sort of like railroads. You saw a similar trend. I don't think that AI is going to be any different. And I think the winners of the buildout phase are not guaranteed to be the winners of the use phase. So the investing challenge becomes, how do you participate in the economic upside without pretending you can reliably pick the long-term winners at peak optimism?

3:13And so I think there's two buckets. There's the direct AI exposure, the builders, and these are the companies who are most associated with the AI build-out. The semiconductors, cloud platforms, data centers, infrastructure. The opportunity, if demand keeps rising for AI, is that these firms can grow fast, but the risk is expectations. At this point, the market's paying up for the future, and the future has a habit of arriving late or unevenly and sometimes with lower margins than people expect. Now, the second way to look at AI investing is the indirect exposure, the users, because historically, the biggest beneficiaries of a general purpose technology is going to be the broad adopters, the companies that use the new tool to reduce cost, to shorten cycle times, improve customer service, tighten inventory, and make workers more productive.

4:08Over time, that's where profits show up in deployment. And this users group, it's also where you're less dependent on guessing which platform wins or which model wins or which chip wins. You're simply owning the idea that AI becomes embedded in normal business life. So I think the big point is that the steadier bet may be that AI becomes a normal input and productivity spreads rather than trying to pick the winners. That kind of leads me to four notes that I jot down whenever I'm talking to clients about this topic. Four things as it pertains to portfolio construction. One is that we want AI exposure.

4:47I think that's fine, but we don't want AI concentration. And most diversified equity portfolios already have meaningful exposure to the AI build out through broad market holdings. The danger, if you go beyond that, is turning it into a narrow bet on a handful of names. Now, the second thing is that we don't want to equate economic transformation with tech stock outperformance. Because the economy can benefit while profits get competed away for a while, and markets can price in a lot of that optimism early. Third, we want to think about the adopters too. the companies most likely to turn AI into sustained productivity rather than just relying only on the builders.

5:29And number four, and this is really important, we must respect the possibility of AI disappointment. Not AI goes away, but more like the return on investment takes longer. Because the cost of capital matters, we know that competition is fierce, and markets often overshoot in both directions. I think the real thing that you have to remember whenever you say, hey, here's this obvious thing in front of me and I want to profit from it, is that a good financial plan doesn't need to make the right call on AI or on any sort of economic trend. It really just needs to make sure that it has the direct and indirect exposure to benefit from the earnings growth that comes from a general purpose technology adoption.

6:14That's all I have for you today. Again, you can sign up for my newsletter at the top of the episode description. Comes out every other Wednesday. Links to the stuff I'm reading, exclusive downloads, as well as links to my latest content. As always, thanks for listening. And until next time, to long-term investing.

6:45Peter Lazaroff is an employee of PlanCorp and BrightPlan. All opinions expressed by Peter and any podcast guests are solely their own opinions and do not reflect the opinions of PlanCorp or BrightPlan. This podcast is for informational purposes only and should not be relied upon as a basis for investment decisions. Clients of PlanCorp and BrightPlan may maintain positions in the securities discussed in this podcast.