Invest Like the Best with Patrick O'Shaughnessy
· Patrick O'Shaughnessy
[00:00:00] S0: If you think about the last couple of years, probably the best thing you could have done is just be long the AI build-out in all its various forms, and maybe that will remain true going forward. But it seems like the market a little bit is starting to think now ahead to the other implications of AI. Software – we're talking the week after software got absolutely decimated in the market, and everyone thinks that because of Claude Code and the amazing experiences that they're having with Claude Code, software businesses are just screwed. I'm curious how you're starting to think now beyond just the AI build. Okay, it seems like, here's the thing, like it's gonna be here. Now the rest of the world has to start to absorb this technology. How are you thinking through that? I'm just super curious what you think about the software sell-off. But even more broadly, the real economy now has to start to swallow this new technology. I'm so curious how you think that's going to happen.
· Dan Sundheim
[00:00:00] S1: It is incredibly difficult to know, and I don't think that's because I don't have perfect information. I think it's just these models are improving at a rate which is exponential, and understanding how that makes its way into the real economy and the implications, is difficult. I think that you probably want to use a few frameworks. It really comes down to like, which companies do you think will have a moat in most circumstances. It's fairly straightforward to identify moats that are protected from digital LLMs, like just the proliferation of digital intelligence. Once you get into robotics and other areas, you start to have to question the moats around some other traditional industrial companies. And also, globally, how do countries that were arbitrating labor do relative to a developed economy? So there's gonna be phases of this, where the first phase is software. And that's really because Claude Code entered the zeitgeist and it's like, all of a sudden, people have seen Claude Code, and all of a sudden they just see on Twitter that people are saying like, “Oh, you know, I created a CRM system in like a day.” And I was like, “Oh my god, this isn't good.” That's kind of where people are now. We wrote in our letter at the end of year, I said, “Look, the build-out is still gonna be a thing in terms of like, places to invest in the public markets. But it's increasingly going to become which companies are affected, and it's gonna become – there haven't been any shorts in AI. There were like, basically no shorts prior to 2026, really. If you wanted to just say, “I'm gonna short something because of AI,” you didn't make a lot of money. In our letter, I said, “There are going to be a lot of shorts – some longs – because of AI.” Software's the first one. I think the market tends to swing to extremes. My guess is that software will have to evolve, will probably be a worse business model going forward. But I think the same way that Walmart evolved with e-commerce – and yes, would they have, all else equal, preferred that e-commerce never happened? Probably, at least at the beginning. It required enormous amount of investment, their margins took a hit, they had new competitors. I think that'll be the case with software too, where companies that have really great distribution and great business models and are systems of record for companies.
· Dan Sundheim, Patrick O'Shaughnessy
One of the things I did is like I asked the LLMs, I said, “Are you designing your own ERP system?” And they said, “No, we're buying a new ERP system from this company.” [00:00:00] S0: Quite telling. [00:00:00] S1: If they're not – at least you're protected at least for a few years if they're not doing it yet. So I think systems of record are going to be difficult to displace. I think companies, while it's neat to create software for small productivity enhancements, if you really want to run your entire business on something like an ERP system or a CRM system, I think it's gonna be quite a while before people are just gonna be vibe coding an ERP system. But I don't think that you can just sit back as a software company and say, “We're a system of record. We'll be fine.” You're gonna have to integrate AI and find ways the same way Walmart integrated e-commerce into their business model. And it was painful for a long time and they’re probably on the other side of it. But this is like, I'd say fairly low conviction, because everything about AI's impact on the economy is inherently low conviction. Because I think everyone is likely underestimating how much these models are going to improve. And to really think about what's going to happen, you have to almost not think like an investor, you have to think like somebody who's into science fiction.
· Dan Sundheim, Patrick O'Shaughnessy
[00:00:00] S1: If they're not – at least you're protected at least for a few years if they're not doing it yet. So I think systems of record are going to be difficult to displace. I think companies, while it's neat to create software for small productivity enhancements, if you really want to run your entire business on something like an ERP system or a CRM system, I think it's gonna be quite a while before people are just gonna be vibe coding an ERP system. But I don't think that you can just sit back as a software company and say, “We're a system of record. We'll be fine.” You're gonna have to integrate AI and find ways the same way Walmart integrated e-commerce into their business model. And it was painful for a long time and they’re probably on the other side of it. But this is like, I'd say fairly low conviction, because everything about AI's impact on the economy is inherently low conviction. Because I think everyone is likely underestimating how much these models are going to improve. And to really think about what's going to happen, you have to almost not think like an investor, you have to think like somebody who's into science fiction. [00:00:00] S0: Can you imagine a version of this story where this is all just overblown? Is there any coherent potential future where five years from now we're just like, “Actually these things weren't that big of a deal.” Or they were much less of a big deal than we thought they were going to be, sitting here today?
· Patrick O'Shaughnessy, Dan Sundheim
[00:00:00] S0: Can you imagine a version of this story where this is all just overblown? Is there any coherent potential future where five years from now we're just like, “Actually these things weren't that big of a deal.” Or they were much less of a big deal than we thought they were going to be, sitting here today? [00:00:00] S1: I think the only way that would be the case is – and even in this I think that argument wouldn't hold – would be if scaling laws just totally stopped. But even if scaling laws stopped, even if these models got no better, I think you probably have three years of people learning how to incorporate AI into their daily life or their companies. Certainly that wouldn't be good for the businesses if scaling laws stopped, but I still think you'd have pretty profound changes within the economy. And betting that scaling laws are going to stop is a really low probability assumption. I mean, there's just nothing that suggests that's the case. In fact, everything suggests the opposite. I think it's difficult to really get your arms around what that means. Because we went from “This is like an interesting chat bot that's like Google” to like, “Oh my God, these are going to be solving problems that humans can't do.” We're already almost there. [00:00:00] S0: I have a 12-year-old son who's interested in investing. I think your son's interested in investing, we've talked about it before as well. What do you tell him about the future of this profession? Given these tools, surely it applies to us too, and we may be smart now.
· Patrick O'Shaughnessy, Dan Sundheim
[00:00:00] S0: I have a 12-year-old son who's interested in investing. I think your son's interested in investing, we've talked about it before as well. What do you tell him about the future of this profession? Given these tools, surely it applies to us too, and we may be smart now. [00:00:00] S1: Elon Musk says, I think a line he's used is, “It's better to go through life being an optimist and be proved wrong than a pessimist and be proved right.” To be young and to be interested in something and be dissuaded because… [00:00:00] S0: You're just obsolete, yeah.