← Liberty's Highlights
Loading summary
Transcript61 lines
- [00:00]
A
Hey, David. Always good to talk to you. I think this is our sixth time doing this, this annual tradition. Somehow we just keep doing it. How are you?
- [00:09]
B
That's all right. I'm great. How are you doing?
- [00:11]
A
I'm good, thanks. So I know you've had a chance to listen to the podcast I did with nbi, the Trillion Dollar Club thing, and I didn't want you to feel left out not having a song, so I made you a little something.
- [00:27]
B
Awesome.
- [00:27]
C
It's that time again for the whiskey. David Kim and Liberty take their favorite seat once a year. They sit and talk the talk. It's that time again for the whiskey neat. David, Kim and Liberty take their favorite seat once a year. They sit and talk the talk.
- [00:57]
B
That's great. Old school. That's like flapper era.
- [01:01]
A
Yeah. Very, very different vibe. So this year we're using your annual letter as the agenda. I think it's a very unusually candid look at what's going on behind the curtain at Scuttle Blurb. So let's kind of use that as a foundation. I'll link it in the show, notes, people can read it, follow along, have that as context. My first question about Skullbloods more as a business, before we get to the market and all that is you talk about the decline in subs, right? How? Some of that is macro mode market, some of that is content mix, some of that is platform distribution changes. What are your current thoughts about all that?
- [01:34]
B
I've always looked at Scuttle Blurb as something of a lifestyle business. So I used to work as an analyst at a long, short equity fund and I quit about a decade ago to kind of go off and do my own thing. And this move was, you know, it was never about the money. It was more about the independence. I wanted the freedom to set my own research agenda. I wanted to learn about companies and industries that I found interesting, regardless of their immediate potential as actionable investments. And that's very hard to do at a fund. I think if your PM asks you what you're working on and you say, well, I'm just trying to learn more about industrial gas industry for no particular reason other than personal interest, that's probably not going to fly. But that's something you can obviously do when it's just you, right? Michael Lewis, he was on the acquired podcast some months ago and he was saying, you know, one of the great things about going off and doing your own creative thing instead of working for a company and having a boss and reporting shareholders is that you don't always have to follow financial incentives essentially. And that's one of the things I certainly appreciate about this work, is that I can kind of go off and research and write about something that no one's interested in reading about. And you know, I could write about companies like Trex and Booz Allen. Sexy that very few people read, but that I personally find interesting. But I also think, you know, I had kind of pushed this too far. I don't think you can ignore the marketplace completely if you intend to make a living off your writing. Since the beginning, I had treated this blog essentially as an investment diary, in a sense intended only for myself, with no consideration whatsoever for what anyone else was interested in reading about. And you know, quite frankly, that bit me in the ass. And you know, I lost probably a quarter of my subs over 4 years. I looked at that and just said, you know, if that trend continues, I'm just not going to have a business in five years. And you know, like, the line doesn't need to always go up. You know, I'm not optimizing for growth for its own sake. This doesn't have to be a million dollar enterprise or anything like that. But there is sort of a threshold amount of income that I need to bring in in order to sustain a comfortable life. Not a luxurious life, but a comfortable life for a family of five in Portland. So I put in place some changes last year to try to turn things around. And as for like why things had kind of soured as far as sub count goes, I can only speculate, I think maybe Twitter suppressing links to other sites, that certainly doesn't help. But I also suspect the issue goes a lot deeper than that. When I started this blog in I guess around 2017, there just weren't too many people focused on long form investment write ups on quality compounders. But since then there's just been an explosion of paid investment newsletters dedicated specifically to those exact kinds of companies. So there's just a lot more competition for attention today in this particular domain than it was back then. And it also maybe kind of feels like attention spans have gotten shorter. I mean like the idea that someone's going to kind of sit down for a solid 30 minutes and read a 7 to 10,000 word post on a company that isn't being hyped on Twitter or isn't making front page news, that just feels like very quaint these days. I mean, I don't know, it feels to me like people just seem more drawn to like short topical posts. This may not even be about attention span per se. It could just also be a function of the fact that like there's just so much more good content out there to choose from that you can find the exact right kind of content that's tailored to your personal preferences to fill like every five minute increment of attention, if that makes sense. So yeah, I think there's just kind of a supply issue when it comes
- [05:28]
A
to what I do as a generalist like everybody else. I guess these days you feel pulled towards tech and AI because it's the current thing that's swallowing everything else. Right? Everything is about that these days. How strong is the pull there? How strong is the tractor beam from the Death Star?
- [05:46]
B
Oh yeah, I mean, look, it hasn't caused me to divert my research efforts towards et al. I think I've stayed pretty true as far as just pursuing what's what interests me. But there's no question that AI and tech is kind of the current thing right now. And the appetite for tech and AI content is as insatiable as the appetite for GPUs. It's just, it is the current thing.
- [06:12]
A
In the letter you say that people don't just pay for ideas, they pay to be inform socially. Yeah, that's one of these like revealed preferences versus like have you changed anything because of that or is this something that you're trying to avoid falling into?
- [06:28]
B
Yeah, I mean, I have made some change. I guess what I laid out in the business update is there. There are kind of like three I think important tactical things that are important to newsletter paid newsletter success that I think I had maybe minimized in the past. The first thing is just frequency. Right? You often hear people say quality over quantity and it's just not true. I think the revealed preference here is that quantity is very important. And I think people do associate high frequency with value. And I sometimes joke that I could take the same write up and break it up in two pieces and publish each piece a few days apart and people would subconsciously feel like they're getting more value from that. And I also think frequency promotes habitual readership. It gives you more chances at conversion. So I think frequency is very important. I think the second thing is this idea of thematic consistency. And what I mean by that is it's basically, I think people's investment interests, they tend to cluster by industry or by style. So people who are into tech typically don't care about industrials or commodities, for example. So if you hop around from one industry to the next like I do, or you know, like, sometimes you're talking about compounders, and other times you're talking about turnarounds. You're not providing that consistent value to any given reader. And so what happens, and what I've observed is that a lot of subscribers will pay for a month subscription to read one thing you've published that they're interested in, then they'll turn off immediately, and then they'll do this over and over again. Right. And then, you know, that third thing we talked about was just the current thing. Right. We're social species. We. We want to be part of the conversation that everyone else is having. Right. Strathecary is kind of a good example of a newsletter that hits all three of these points. You know, Ben, he publishes several times a week. He writes exclusively about text, so he's thematically consistent. And there's obviously no topic that's more of the current thing right now than tech and AI. Right. And I don't know that you need to hit all three of these points. I think it helps if you're trying to push things to the max, but I think you need to have some combination of those three. And if you don't have all three, you need to compensate by excelling in one of the other two. I think very few people want to read a deep dive about business just for the sake of doing so. Right. I mean, I do, but nobody else does. I think there, for the most part, I think there needs to be a carrot, Right. Whether that be in the form of, like, an actionable investment idea or helping them know more about a company or industry that's related to something that they're already invested in. So anyway, yeah, like, all this got me thinking about where am I falling short and about areas where I'm willing and unwilling to adapt, and how do I stay true to myself and not fall into this trap of audience capture while also sustaining a viable business. Right. And so first, you know, I would say there's nothing I can do about thematic consistency. I have a broad range of interests. I'd like to cover a wide range of industries. And that's just not something I'm willing to sacrifice. I mean, AI is kind of the hot thing today. It could be something else tomorrow. I mean, who knows with this stuff? Part of the reason I quit my job was so that I could pursue my own research interests. And if I sacrifice that, then it's kind of like, what was the point? But I think I can do better on the other two, Right. Like, I think I can be more Proactive about finding places where my interests overlap with what's trendy. Right? So like, if I have five companies in my research pipeline, but one of those companies is buzzy right now, maybe I strike while the iron's hot and move that company to the front of the line. I was going to research that company anyway, so why not prioritize it? And so, yeah, so like a big change I made last year was I moved to higher frequency publishing cadence where I provide updates on stuff that I've covered in the past and try to just discuss things that are more, I guess, topical is the point. And so, yeah, because I know like one point of feedback I've gotten from subscribers is that they want more updates. So that's the other thing I'm trying to work on. I formally announced these changes last September and yeah, it's made an enormous difference so far. So net subscribers had been negative for like 16 consecutive months. And then almost as soon as I made the change to like higher frequency, my sub count began to grow again. So revenue from the substack version of the blog, that's grown a lot faster than the WordPress version. So last year it was like 20% growth on the Substack platform versus, you know, 6% on WordPress. But I've also noticed that like churn on the substack is much, much higher than it is on WordPress. Maybe that's a function of the fact that substack makes it super easy to subscribe. So like, so like my WordPress site, it's a little kludgy. There's more friction in the signup process than there is on substack. So maybe the signup hurdles filter out the less committed subs than on substack, where it's more like easy come, easy go. I'm not sure.
- [11:40]
A
Yeah, I think leveraging the archive right, with scuttlebits, which are the shorter updates. Great idea. Because you've been doing this for so long, you have like hundreds of companies that you know enough about to more easily do an update than if you were starting from zero.
- [11:55]
B
Yeah.
- [11:56]
A
The question I would ask there is that in theory, oh, short update, like 500 words. A thousand words, as you mentioned that. But they end up being almost as long as the full updates or not quite as long. But I still 3,4000 words. How has that been working out? How's that been fitting in the existing workflow? Are you just working more than before now because you're still doing the longer updates? Or is it somehow that you were doing this work, but it just wasn't going up on the website. But you were still kind of doing it for yourself anyway, so it's not that much additional work. How's that been behind the scenes for you?
- [12:28]
B
Yeah, I mean, I think once you've done a deep dive on the company and kind of set that foundation, it's a lot easier to do updates going forward. And so I find that the update pieces, even though they do tend to get quite long, like they can't be like three or four thousand words, they don't feel as onerous to research and write up. But you're right though, like when I first started the scuttlebits thing, I thought these would be kind of like, you know, 500 words, 6, 700 words max. But then I can't help myself. I just keep digging deeper and deeper and these end up coming almost like, you know, full length post and they run like three or four thousand words or sometimes longer than that. And so, yeah, it has required more work on my part. At the same time, like, we have this awesome new tool through AI now, and so that provides some productivity gains as well. So that's kind of a partial offset to the added investment that I have to put in to this. But yeah, so far it's working out well. I think from a time management standpoint.
- [13:28]
A
Another big change has been the price change in November 24, but these things take a while to work their way through with the annual renewals and all that. Can you tell me about, first of all, when you decided to do it, were you afraid it wouldn't work or there would be pushback? And then when you did it, how was it where people were just fine with it? Is it that your sub base is so professional investors anyway that it doesn't matter? The zero to one problem is harder just to get them to subscribe in the first place, but the actual amount doesn't matter that much. I'm kind of curious how that went.
- [14:01]
B
Yeah, it went better than I expected, actually. So, you know, I raised prices November 2024 basically because I had to, and I thought it was something I could get away with. My subscriber base kept shrinking and I needed to offset that somehow. You know, something like 80% of my paid subscribers are investment professionals and they pay for my blog through their corporate credit cards. And so I think I have some pricing power among that audience.
- [14:29]
A
Oh yeah. If they're making a multimillion dollar investment decision based on some of that info, like whether it's 20, 30, 40 bucks doesn't really move the needle, right?
- [14:39]
B
Yeah, exactly. I think it's pretty trivial even. I took the annual price from like 240 to 320. On a percentage basis, that's a pretty big increase. But on an absolute basis I'm sure it's kind of peanuts for many of these subscribers. And so nevertheless, I did expect a spike in churn, but actually there was no change there at all, which came as a welcome surprise. So the price increase was absorbed pretty well without incident.
- [15:05]
A
Another addition to the media empire has been the Never Sell podcast with mbi, which I really enjoy. Tell me more about that. The backstory. How have you evolved now being a professional podcaster? What's the story there?
- [15:18]
B
As you might be able to tell from this interview, podcasting isn't really my preferred medium. It's not something I consider myself all that good at. But mbi, he had kind of proposed that we do a podcast once a month. At first it was going to be we would talk about the write ups that we did in any given month, but then it kind of eventually moved to more of a free flowing conversation where we would just kind of. We have a Google Doc where we keep track of topics that come up throughout the month that might be interesting to discuss and then we just kind of freeform riff on those topics.
- [15:52]
A
I think I like that format better.
- [15:53]
B
Yeah, me too. I think MBI is a really good partner to have in this.
- [15:56]
A
But it's also a way like when you talked about the importance of the current thing, that's the way to do it for you guys in a way that maybe a better, a better medium for that than the writing in some way. Right. It's much easier to talk about the current thing there. Yeah, you kind of freeform it and then you get back to the research.
- [16:12]
B
That takes weeks. Yeah, that's right. Yeah, yeah. And there's just a larger audience for the current thing. Right. If you're using podcasting as kind of the, the top of funnel to draw people in, I think that's maybe where you want to be focused. I think NBI is a particularly good partner for this because he tends to have like a lot of things to say. Maybe relatively more reserved. Yeah, it works out, I think. And it's been going pretty well. Honestly, it's been less about just the commercial motives and about attracting subs than just about the fact that we just have a good time and I think that we just enjoy the conversation. So.
- [16:50]
A
Yeah, that feels very familiar. That's probably why I've had it on my podcast a few times. Moving on, the next thing I'd love to discuss with you is more on the investing market side. Let's look back at 2025 a year on market vibe type of thing, right. As an investor, I don't know about you, but me as an investor, when I think back on any random year, like if I think back to 2012, I don't remember all the details, but I still remember like what the general vibe of the year was.
- [17:16]
B
Right.
- [17:16]
A
So that's what I'm trying to extract from 2025. Maybe, maybe we're too close to it to be sure about it, but I'm curious what you think about it. Last time we talked the vibe was like a loop between like it's so over, we're so back, right? Volatility was starting. What's the 2025 version so far for you?
- [17:34]
B
Yeah, kind of a weird year. I think for much of the year it sort of felt like a lopsided market performance wise where with like perceived AI winners experiencing big gains and a lot of the other stuff being left behind. And then we also saw some big sell offs in names that have long been considered compounders. Right. And it's not like these things sold off for no reason. There's a narrative behind each of them. So with Copart, there's obviously fear of share losses to iaa. With Gartner and Constellation, that's AI, of course. And then you know, Old Dominion, Watscope, Pool core, that there's kind of macro stuff behind that. And it certainly didn't help that many of those stocks are trading at big multiples to start with. But I think like, you know, more broadly it kind of felt like small data points are being extrapolated out to infinity and it created these like big unexpected moves I think in both directions. I very often like found myself surprised by sort of the magnitude of the moves up and down like earnings reports that seem innocuous to me, creating like 20 30% declines and then what seemed like small positive data points creating huge moves up. And one of the weird things I've observed in this cycle is that the market seems to be extrapolating growth and earnings of certain cyclical companies. They look at Trex, right? Like traded at 65 times peak earnings at one point during the COVID boom. But then the COVID bloom went away, housing turned south, and then it traded all the way down to like high teens trough earnings, right. And then Pool Core, it's kind of the same deal 30 times peak 20 times trough. And I use the term trough loosely by the way. It's just weird because I don't think that's something you would have seen maybe 20 years ago. Like I could be misremembering this, but I'm pretty sure that in the 0506 peak of the housing bubble, like home builders traded at like single digit multiples because everyone recognized that the E was too high. But these days it kind of feels like investors are almost blind to cyclicality. They just extrapolate and it's a little bit weird to observe. So, so it felt like, you know, in some places and kind of the back half of last year you had like these beat down cyclicals where like you could maybe bet on earnings recovering eventually and then also getting the multiple expansion on top of that recovery. And so, so I was spending a lot of time there and like cyclical companies that were once steam compounders where you had moated well managed businesses that hit a macro air pocket and that saw kind of multiple contraction on top of the earnings decline. And then obviously AI and big tech has sucked all the air out of the room for understandable reasons. And there's just a lot of debate around depreciation schedules and return on CapEx and the limits of scaling. But you know, it's also important to remember that there are other areas of the market beyond the white hot center of AI and tech where the hurdles could be lower. I think maybe it's easy to forget that when the attention is so squarely placed on AI. I spent a lot of my time on these less exciting pockets of the market. Good companies that sold off for cyclical reasons. The LTL carriers, Trex, Watsco and yeah, I mean it's amazing how fast some of those names have recovered. I feel like the gains have been more broadly spread in recent months and now kind of the pain trade has rotated into software. And I don't know, that strikes me as like a tougher challenge in my opinion because you could kind of see AI being like an evergreen overhang for a long time. Even though I think like in a lot of cases those existential challenges are probably overstated and some of those names have traded off to interesting valuations. So I think I'm going to be spending a lot more time there. I started going down this path a bit last year with Gartner. The obvious bear case with Gartner is that AI is going to take the place of Gartner's research product. Why pay $20,000 a year for a research subscription when you have access to ChatGPT for $20 a month. I think it's probably fair to say that Gartner is a more discretionary product than it's sometimes pitched as. But I also think what the bear case misses is that for, for the kinds of questions that Gartner's clients are asking, you're not going to find a publicly available baseline or case studies on that. So something like, how should the head of IT at such and such global manufacturer evaluate zero trust security vendors in a legacy IT environment? A lot of that knowledge just lives in people's heads. It's not really written down or formalized anywhere. And Gartner, they're, they're kind of known for these reports, but the clients who just subscribe to the reports alone are only like 10% of revenue. The other 90% comes from people who want access to analysts and executives with domain knowledge. And then there's obviously like the CYA aspect, the fact that Gartner kind of provides this seal of approval for procurement decisions. It's seen as a credible source by board members. So I kind of feel like when I look at SaaS, I'm not saying that the entire space is going to be immune from AI related disruption, but I do think that there's a lot of baby getting thrown out with the bathwater. So yeah, the AI pain trade, it's spread further into SaaS and data providers like Factset, S&P, most recently Thomson Reuters. That's an interesting pocket to look. And I don't have too much exposure to software at the moment. I own one software stock that's Constellation. It's about 4%, maybe 3 or 4% position, but I get the sense that I need to be spending more time here.
- [23:10]
A
One thing about 2025 that I think I'll remember about it is how the index was telling me one thing, but everywhere I looked I saw something else.
- [23:19]
B
Right.
- [23:19]
A
Like in 2022, everything was going down the whole year and that felt a certain way. This year the index was down like 2, 3% from peak and everywhere I look, companies were down 40, 50, 60%. They were back to where they were like five years ago. And that felt very weird.
- [23:35]
B
Yes.
- [23:36]
A
I know that the math has to add up somehow, right. But I think the human brain is not very good with large numbers and like trillion billions. Like the big tech is so big, they are such a large part of the index that we forget how as long as they're not down too much, they can hide a lot of pain elsewhere.
- [23:53]
B
Yeah, yeah. And I think at least so far this year, it feels like the gains have been, we're kind of seeing something similar. Like sometimes I look at my portfolio, I'm like, geez, that's a lot of red. Oh, the market's flat. Yeah. So I feel like there's some of that going on where you know, there's a big sell off in SaaS and I don't even think big tech is doing all that well and then. But some of these industrials and energy and material stocks have just been railing pretty hard. One thing I will say though is like the software thing, these do kind of feel like harder bets to me all around compared to just like buying Trex or Old Dominion three or four months ago just because, you know, like there's this tension where you have to have conviction and some deep underlying theory about how things are going to play out. Otherwise you're going to get whipped around and head faked on quarterly data points. But at the same time you don't want to know, ignore the quarterly data and stay stubbornly attached to a thesis that isn't playing out. So there's a bit of a conflict there. And so let's say you're bullish on SAS and you think the AI threat is overrated. You buy the sas, they put up a good quarter and you see that as proof that the threat is over overrated. You buy more. But is there really like enough signal there? You know, like should that be confidence boosting at all? Not really. It shouldn't be all that surprising to see SaaS companies putting up good growth numbers today because AI adoption in the enterprise is still pretty early. If you have a holding period of five years, let's say it's 20 quarters and this is one quarter out of 20. And arguably the quarter that matters the least, the sell off, it reflects doubt about what the growth is going to look like five, seven year down. It's a statement about terminal value essentially. But nevertheless, I think there's a lot of interesting stuff there. I almost feel like I should rename my blog Fallen Angels because it does seem like I'm spending a lot of time writing about stocks that have just gotten really beat up.
- [25:53]
A
Speaking of what you wrote about, and this may be a hard question to answer, but in looking at it, I think you wrote about 21 companies in 2025. I'm seeing life science and tools, payment, fintech, HR, payroll, software, enterprise software, semiconductors, EDA, healthcare, building materials, energy services, defense, government kind of contractor stuff. So I'm curious, is there Something that stands out to you as research you're particularly proud of for whatever reason, right. Anything that you're looking back at the year and you're like, oh yeah, that's the one that if people have missed, they should check out. Because I'm putting that in my portfolio of writing basically.
- [26:32]
B
Yeah, I guess I'll say first, there were a number of posts that did well commercially for the obvious reason that they touched on the current thing, right? So like Gartner was one example, the post on shift 4's organic growth was another one. So there I just kind of broke down how shift 4 was calculating its organic growth and show how it deviated from the way that most other, you know, acquisitive companies do it. You know, in a nutshell, revenue from, you know, the, the acquired company in the, in the comparative year ago base and to compute organic growth. They include all the cross selling synergies on a gross basis, but they don't include, and I don't think many investors were aware of this and I've seen some sell side reports that compare Shift4's organic growth to those of payment peers as though they were apples to apples, which they're definitely not. And so I got a big wave of new subscribers from that post. But, but that's like a sugar high, right? Like almost everyone who subscribed, who subscribed because of that post did so on a monthly basis for that one post. And it wouldn't surprise me if most of them have canceled since then. Honestly, one write up that I like that almost no one read was the Booz Allen one. So you know, Booz Allen, it's a tech and consulting firm essentially that serves government clients primarily. It's got a long history dating all the way back to like World War I, you know, a little more than a decade ago they sort of pivoted towards hands on technical work where they essentially stitched together technologies from third party vendors into a solution that's tailored for specific use. Case for instance, they, they took, you know, a reinforcement learning model from DeepMind, they trained it on simulated missions and they put it on a YouTube reconnaissance plane. Right. And so you know, historically a lot of what Booz Allen's edge, it kind of came from like soft things like institutional knowledge that had been built up over 100 years relationships. So like a lot of Booz Allen's executives and board members, they used to be former high ranking government officials. Bloomberg once called Booz Allen the most profitable spy agency. That's telling you something there. And so but you know, also there's been like this trend over time of government contracts sort of bundling a large number of services, and that's benefited the larger players like Booze, who can staff all those spots. But I think like, the worry is that now you have these new primes with like Palantir and Anduril, who have what can be probably described as like a frenemy relationship with the legacy integrators. And I think that's a valid worry in my opinion. I think Palantir, they partner with Boos and others, but they also talk a lot about disintermediating them by embedding a lot of the integration work that used to be handled by the integrators into the software itself. And so, yeah, I mean, in a very simple sense, what Palantir does is it pulls data from a lot of different places and it ties it together and then it provides software analytics on top of that, which is all stuff that like Booz Allen does, right? And, and now a lot of that's being embedded in the actual end product. And so one example of Palantir's growing influence is that they won like a 10 year agreement with the US Army. And so this is a contract. It rolls up like 75 separate contracts into like one gigantic one valued at like $10 billion. And so I see like fragmentation as being a good thing for Booze because they can kind of pitch themselves as being at the center of the spider web, right? Like tying these different technologies together. So to the extent that more of these engagements get consolidated into product, essentially, that would seem like a headwind for Boos. It doesn't mean that Booze's involvement goes from like 100 to 0, but maybe in certain domains it goes from you know, like 100 to 90 or something. And so, yeah, I don't know. I think more generally these defense intelligence consulting firms, they don't strike me as particularly good businesses. Like, the margins are pretty terrible. They're dealing with a customer that obviously has immense bargaining power. And then, and then like in any given year there's going to be some area that's getting funded. And you know, these days it's the stuff at the tail end of the kill chain. It's like, you know, drones and UAVs and golden dome and you know, like, that stuff is sexy right now and I think if you have more exposure to it like Leidos and Khakidu, you're golden. But then, you know, who knows what the priorities of the next administration is going to be. And the one after that. And so, so I don't know. At the end of the day I put it in the too hard pile. I think booze once traded at 30 times earnings not too long ago, which was pretty silly relative to the quality of the business. I think it's at a more reasonable low to mid teens now. But I thought it was an interesting research project that I spent a lot of time on.
- [31:04]
A
No, that was a good write up. I enjoyed it.
- [31:06]
B
Okay, good. Yeah. Well, that's one person who read it.
- [31:10]
A
Another thing that when I think about 25 and it's not the only year was where this happened was kind of the culmination of a trend was that competitive boundaries between various companies and sectors and industries have become more blurry. Right. And, and we kind of saw it in big tech over time where all the big tech players used to be in their own lanes. Amazon was over there and Google was doing its thing and Apple over there. And over time they kind of started competing with each other a lot more. I think AI kind of was part of that. Right. But now it feels like whole industries that used to be more separate are getting invaded by someone else.
- [31:45]
B
Right.
- [31:45]
A
Like so AI is going into, I don't know, legal things and in government and then it's probably going to go in manufacturing over time with robotics. Do you feel like this is making it harder to understand like competitive dynamics and modes and advantages when people used to be on some island somewhere in their own industry, you know, fighting among each other now could be invaded by anthropic or whatever?
- [32:09]
B
I do, I think, yeah. NBI and I have discussed this in quite some depth in our podcast. Yeah. It certainly complicates the. You can point to each of these players and they were on their own island, as you said, with Amazon dominating retail. They each had like this one big thing that they were the best at and that didn't have much competitive intrusions upon. And now everyone's kind of stepping into each other's territory. I think that said, it does feel like Meta and Google are particularly well positioned because of. Because I just see advertising as just the best risk adjusted, most scalable use case for AI right now.
- [32:50]
A
Kind of a layer on top of the rest of the economy. Right. They're kind of like taking a cut of all kinds of other activity.
- [32:56]
B
Yeah, exactly. And so, yeah, and these models, they sort of run on engagement and targeting and conversion and it just seems like AI is just so well suited for that and now they're kind of diving into the generative aspects as well, like creating the ads and just becoming like end to end platforms. And so they seem very well positioned. I think the way I kind of see this is when you look at sort of the different and use cases for AI, you can sort of rank them as far as just how compelling they are on a risk adjusted basis. And I think like advertising is right there at the top. And then there's other kind of avenues as well, like These chatbots and SaaS and these other places that seem maybe a bit more speculative. One advantage that Meta has is that they can kind of turn the knobs behind the scenes. Right? They don't need to convince anyone to purchase a subscription or anything like that. People just subconsciously get sucked into the feed and I think that's one of the advantages that they've got.
- [34:00]
A
Speaking of AI tools and that's something we talked about in the past, both in these talks and in DMs. I'm curious about your use of these tools in your process. You mentioned using AI as a tutor for you get in life science or something very complex. Right? You can get a lot of the 80, 20 of it through that. But I'm curious these days, what does it look like your use of these tools Used to talk about Perplexity. I remember in the group DMs when O3 and deep research came out, it was kind of a freak out moment for everyone. It's like, are people even going to read newsletters anymore? And that's kind of calmed down. But still these things are evolving so fast that we shouldn't get used to anything. Anyway, I'm curious, what are you using these days? How do you differentiate between where it's useful and where it could be dangerous, where it could be like the, the wrong kind of shortcut to take.
- [34:47]
B
I use AI all the time and I use ChatGPT, I use Claude, I use Perplexity. So I use it for research purposes and I honestly think it's made my stuff better. I've found it incredibly helpful to get up to speed on various arcane things that I'm trying to understand. When people use AI to level up their work, I think that's terrific and I'm all for that. My concern would be, you know, using AI to take the easy way out, essentially to save time rather than to create better stuff. Yeah, AI, it lets you get to like a state of knowledge faster. But I also think that there's sort of a knowledge illusion that kicks in as well, where it's easy to Convince yourself that you understand something better than you do when an answer is given to you. Compared to, like, if you had to struggle and work your way to the answer yourself, and if you had to, you know, like, reconcile contradictions and deal with ambiguity, it's easy, after you've seen the answer, to think, oh, that was obvious, right? And then trick yourself into thinking that you could have come up with the answer on your own. I think AI deskilling. I think it's something to take seriously. Others have talked about, you know, the risk that, you know, at the same time that the robots get smarter, we become dumber by relying on them excessively. You know, that could add up to. To the surrender of something more serious. And even if the robots are better at us at all of those things, I would still say that these are desirable skills to have for their own sake, just for the purpose of being able to interact with our fellow human beings and create some kind of shared reality with them. And I do think that writing helps you develop those skills.
- [36:25]
A
One thing I'm curious about is if you have anything in the pipeline for next year that you are very excited to write about, to cover, to research. It could even be something that you're dying to write about, but you know, will get zero clicks, but you are excited about it. Anything like that on the docket?
- [36:42]
B
Yeah, yeah. As far as the research pipeline goes, I'm excited about all of it. I've got Viva Ryan, Specialty, Lamb Weston, U Haul, maybe S and P, Moody's, Thompson, Reuters. Yeah, I don't know how well those will do commercially. I don't think Lamb Weston and U Haul will do very well, but maybe some of these other ones, like S and P and Moody's and Top Storage, because those are kind of more in the headlines, I guess, these days. So we'll see.
- [37:09]
A
Awesome. I can't wait to read those before we go. And curious if you've read any great books, seen great films, any TV shows you love any great stuff like that these days.
- [37:22]
B
Oh, man. I was really excited about Pluribus, and I thought the first episode was really strong. And I still think the concept is very interesting, but Wheels kind of came off that one a bit as the season progressed. Part of the issue there was that, like, with Breaking Bad, you got sort of the perspective of several different characters right as part of the story. Whereas with Pluribus, it felt like you were kind of trapped in the room with this, like, relentlessly angry person
- [37:52]
A
for
- [37:52]
B
like, however many episodes. The concept itself is strong enough that it seems salvageable and maybe season two will be better.
- [38:03]
A
Yeah, I'm feeling the same. I'm ambivalent about it. I think it's a great start, great premise, but I feel like it's just part of a story. So it's very hard to judge. Until we know more, I'm kind of reserving judgment on it. But I'm definitely going to watch season two.
- [38:18]
B
Yeah, same.
- [38:20]
A
So I will put all of your links in the show notes. I encourage the listener to go check out your stuff to subscribe. I put the link to all of our past annual interviews. There's a lot there. Anything that we didn't talk about? Anything else you want to leave the listener with?
- [38:37]
B
No, I think we covered it.
- [38:39]
A
There's only one thing left to do. I'll play us out. Cool.
- [38:46]
C
It's that time again for the whiskey neat. David, Kim and Liberty take their favorite seat once a year. They sit and talk the talk. It's that time again for the whiskey neat. David, Kim and Liberty take their favorite seat once a year. They sit and talk the talk.
- [39:15]
A
Bye, everyone.