Lou Basenese, chief market strategist at Public Ventures LLC, spoke with Quartz for the latest installment of our “Smart Investing” video series.

Watch the interview above and check out the transcript below. The transcript of this conversation has been lightly edited for length and clarity.

ANDY MILLS (AM): Nvidia’s run. Is it over?

LOU BASENESE (LB): I hope so. Just so I can stop being embarrassed for missing out on the original run last year. In the first wave of hype and a new tech trend, you see a standout performer like Nvidia was last year. It’s continued that run this year, but I don’t believe it’s sustainable. You’re looking at a $2 trillion plus company. Can it go to 4 trillion? It’s less likely than finding other AI plays. And that’s what I think happens in the second phases of growth of a new tech trend, is you look at those first derivative plays, and by that I mean companies that are doing something similar that are benefiting from the AI tailwind. So you look at a company like Arm Holdings, I have no disclosures here. They’re benefiting, right? They have IP on all the leading chip technology that’s out there. They collect royalties from it. So any boost for AI or electric vehicle chips is gonna benefit Arm Holdings. You can look at the AI trend from a data center standpoint, right? The more AI computing power we have, the more data storage we need. So a, a REIT, again, no disclosures like Equinix, I think it’s like $80 billion. One of the biggest data center REITs out there is gonna benefit from the boom in AI. And then the other thing is, I think people have always looked at ChatGPT as like the quintessential epitome of what is AI and I think that’s really narrow minded because we haven’t seen practical applications, right? We just think it’s a way for teenagers to cheat or to look for quick answers.

AM: And adults!

LB: [laughs] And adults. Yeah. I didn’t wanna accuse my peer group, but yeah, it happens. But now we’re starting to see practical applications. So we actually backed a company that is using AI to, with its medical device to help detect heart attacks early. And it’s equivalent to a 12 lead EKG that you would get in the hospital. But it’s actually portable. Disclosure there. We bank them, I don’t own it. My firm does own the company, named HeartBeam. But it’s a real world application of how AI could improve our lives instead of just help us cheat in life. So I think you’re gonna start seeing more of those innovations happen where real technologies are now using AI as a way to improve their product and improve outcomes in clinical healthcare settings, in drug discovery, in a lot of areas that are very useful to humanity and in turn to investors as a result.

Read more: The AI craze is no dot-com bubble. Here’s why

AM: Nvidia has been compared to Cisco systems in the late-90s, early-2000s because of the hardware issue of it. Like once you get enough AI hardware out there, then it’s up to everybody else to make use of it. And that’s where the profit will be in the next phase of it. 

LB: Correct. Yeah, and I would look at, in the current lineup of big tech companies, Nvidia is the pick and shovel supplier, right? Anyone doing AI, Hey, you need our chips. Whereas you look at a Microsoft, Microsoft can actually propel AI usage by introducing it into their products, right? So I look at a Microsoft as a more compelling AI play. I don’t have any disclosures there. But because it can drive business using AI, whereas Nvidia’s waiting around for everyone else to get into AI and just selling them the tools to enable it. Microsoft and companies like it are developing actual products that incorporate AI and that will generate revenue and profits, not just selling the supplies.