Loading summary
Transcript34 lines
- [00:00]
A
On this episode of Full Signal, I am rejoined by Kai Wu of Sparkline Capital. He has a new research report out that's about the AI buildout and the next phase of AI adoption. He has a ton of proprietary data research and charts. We get into valuations, what to do in the next leg of the AI trade and much more. This is a fantastic conversation. I think you're going to love it. Kai, it's great to see you again. We're getting into the adoption phase of the AI boom, according to your latest research. Can you walk us through what that means?
- [00:33]
B
Yeah. So if you study the history of past technological diffusions, whether it's the electricity, the Internet, railroad, cell phone, there's this S curve pattern that happens every single time. And in the very beginning, the most foundational thing that happens is you see in stage one, what I call the infrastructure phase, there needs to be a build out of the underlying rails in order to support this new technology, whatever it might be. That would be the fiber optic cables. Today, it's AI data centers. Then at some point in this diffusion process, there's a switch where you move from phase one to phase two, which is what I call the adoption phase. That's when the questions for the investor class and for CEOs switch from can we build it? To will they buy it? In the very beginning of the AI boom, the question was, can we put together trillion dollar deals to build all these data centers? Can we actually make these GPUs chain together and work in a coherent way in these massive clusters? Can we continue to achieve improvements on our large language models as we go from GPT 3 to 4 to 5 and so on and so forth? I think at that point, at this point, it's pretty much the answer is, yes, this is happening. We are now in the process of spending trillions of dollars building out these data centers. So the next question becomes, okay, great, so now this is happening. Will they buy it? Will the end users of these products, whether it's small businesses, enterprises, or consumers, actually derive enough ROI from their investments in these, in these tools to then drive enough demand to justify the expenses that these companies are putting forth? And what you find if you look at the historical episodes, is this pattern of where returns match where you are in the cycle. So in the very beginning, infrastructure stocks outperform. The telecoms did the best in the early 90s when they were the hot stocks building out the Internet. Then at some point there's a leadership shift where these stocks start to roll over or lag And a new crop of winners come up and ultimately are the long term successors. We talked about this last time in October with the last piece. If you look at the history of past technological booms, rarely if ever is it the actual builders of infrastructure who are the long term win the cycle. Usually it's not the railroads themselves which all went bankrupt. It's the businesses like the retailers that ship goods across the country on those rails. It's not WorldCom or AT&T that won the Internet era. It was Netflix and Meta who built businesses on the back of the utilities that these telecoms became. So I think that's a useful framing for thinking about where we might be in the current cycle with. If you think about the past few months, we started to see this shift happen where for the first time past couple years, it was only up. Oracle would announce a new deal to OpenAI. Their stock would pop 30% in one day. The MAG7 continued to roll Coreweave. All these stocks did really, really well until around October, November of last year. For the past four months, these stocks have actually underperformed. Oracle is now 50% below where it was when that deal was announced. And a lot of the Mag 7 have started to lag the broader market. That's the question which is is this a signal now that we're starting to tip from the initial infrastructure phase to the adoption phase?
- [03:57]
A
That's a lot to unpack. Right. And one piece of this I know you're watching, that I think is maybe a more granular way to track it is what companies are actually talking about. So we talked about last time mentions of AI on earnings calls. But what you've done more recently is you're tracking not only mentions of AI, but but companies that talk about actual productivity and economic gains from AI. How does that play into your thesis?
- [04:26]
B
Yeah. So the question before the house, based on what I just said, is are companies deriving real ROI from their AI investments? If the answer is yes, that's good, that's bullish. That means that maybe we won't get there as quickly as many of the AI infrastructure firms believe, but we'll get there eventually. The, the technology is improving and these firms are ultimately getting value from these investments. If the answer is no, that's very, very concerning. And so this is kind of the trillion dollar question. You could look at survey data, which is what most people do. They look at the MIT survey, the McKinsey survey. A lot of these companies have come up with these data sets, but the problem is that they're very biased. They're not representing the samples. It's a small number of CEOs or CIOs being asked. So the approach I wanted to take was to say, what if we looked more broadly across all public company earnings calls.
- [05:15]
A
What if.
- [05:15]
B
Which is more or less a representative sample of the corporate sector. And to your point, we talked about this in our last discussion, that it's interesting because you can just look for companies that talk about AI on their earnings calls historically, and those stocks have outperformed. What's really interesting though is that is, of course, I think over time what's happened is in the really early years, those were the firms that were really cutting edge. But it's become so consensus that at this point, know these CEOs are talking about AI on earnings calls because they know they have to. Right. And so what you have to do when you parse through all this data is to kind of cut through some of this theater, some of the window dressing that's occurring in the corporate speak. And so what I did for this analysis here was I said, let's create this taxonomy where there's going to be three buckets of mentions of AI. The broadest bucket is just, did you talk about AI and you're using your business. The second medium category is did you not only talk about that, but did you also talk about and were able to point to a numerical quantified gain in productivity or cost savings? Like, oh yeah, we were able to reduce costs by 20% due to this AI investment, or we were able to increase asset utilization by 15%. Those sorts of mentions that were numerical, not just vague. And then the best category was not just that, but also roi. So in other words, being able to link the revenue gain relative to how much you're spending on the AI investment itself. Because if you can make 20%, you can make $20 million on an AI investment, but the AI investment costs $100 million. That's not good ROI. Right. And so we were able to use actually a large language model to parse all the different earnings calls through time and ask the question of how do we categorize dimensions in each calls into each of the three buckets.
- [07:06]
A
Yeah, I mean, it's incredible. I was reading through your report again this morning. The Companies that mentioned AI driven ROI, if I'm reading the chart right, 5.2% outperformance against the market, and then the companies that mentioned just economic gains, 4.8% and then companies that just mentioned AI, 3.2%. So there is A tiered performance here. Maybe the question is for investors, just find the companies that are deriving economic gains from, from AI and then bet on those, is that right?
- [07:42]
B
Yeah. So yes, I think the answer is yes. If you step back and ask the question of AI is a disruptive technology, it's changing the way we do work. It stands to reason that companies that are more front footed with regards to implementing and deploying the technology stand to have an advantage over their laggard peers. Right? There's going to be a separation here and you do see that in the stock price. If you simply said, I want to look at, I'm going to listen to all the earnings calls this quarter and buy. The companies that are not are able to point a specific ROI on their AI investments relative to the ones that are not that historically would have worked. I think you can do better. Right? Because I think that these earnings calls are a helpful barometer for where we are in the adoption curve. We saw that in the very beginning of the sample, maybe like 10 years ago, very little mention of AI at all, let alone ROI. And then as you kind of roll forward through time, obviously the chatter about AI increases. But in addition, the percentage of companies that can report, you know, tangible gains from their AI investments has also increased as a share of overall AI mentions. But I want to say today about a third of mentions of AI come along or linked to a tangible gain, like oh yeah, we were able to improve call center productivity by 16% or something like that. So you're seeing this kind of macro trend. But I think what you also want to do is you want to get in front of that. These earning calls have some limitations. First of all, many times companies are investing in AI but have yet to report the gains because of course enterprise adoption is a multi year process. So maybe it's still too early to really be comfortable to talk about this. Or it could be the case that companies are doing stuff and don't want to talk about it for competitive reasons. They don't want to clue in their competitors on what they're doing. So I think a better approach is to look even beforehand and say before companies are even talking about the gains they achieve. Can we look at the underlying data of adoption? And so what we look at is we track like the patents, the trademarks, the hiring patterns of these companies, kind of what they're doing under the hood. And we're finding that it is the case that when companies are investing more on our metrics in AI and then you kind of look at the next year's, you know, earnings calls, they tend to increase the. The increases the probability that they'll then go out and talk about this. And so these things are kind of highly linked, but they're predictive, right? So it's kind of a leading indicator, so to speak, of, you know, when companies publicly go out and announce that they've been getting gains from their investments.
- [10:12]
A
Some of you may not have heard this, but our partners at Public just launched something called generated assets. It brings AI into investing in a way I've honestly not seen before. Here's how it works. You type in an idea like AI powered supply chain companies with positive free cash flow, or something like defense tech companies, growing revenue over 25% year over year. Publix AI then dispatches a swarm of agents that can scan every single stock, evaluate them, and instantly build a custom index around your thesis. What really stands out is how clearly it explains why each stock is included. And before you invest, you can even backtest your idea against the S&P 500. So you're making decisions with real context and not just guessing. Beyond generated assets, Public lets you invest in stocks, bonds, options, crypto, all in one place. They'll even give you an uncapped 1% match when you transfer your investments over from another platform. If you want to build a portfolio that actually reflects your thesis, visit public.com openingbell that's public.com openingbell now let's get back to the conversation. As independent investors, retail investors, is there any way to access that proprietary data besides, let's say, reading your research? Like, can we just go find that stuff out there?
- [11:29]
B
I mean, that's the good news, right? Everything here is, is for the most part, public. So I'm not like, you know, doing surveys where, you know, I have to be an MIT professor to be able to do it. You know, earnings calls are all available online. As you can kind of go on Google, look at who works at different companies. Based on LinkedIn, you can. The patent and trademark data is by definition a public good. So all this data is available. I mean, obviously, kind of piecing it together and mapping it correctly to the companies is work. But yes, in theory, anyone, you know, anyone has access to this information, it's all publicly available.
- [12:03]
A
Okay, that makes sense. With your framework on the roi, economic gains, and mentions of AI, you have a second framework which you also wrote about, which there's three categories you identified. AI laggards, AI infrastructure firms, and then early adopters. Can you walk us through each of those? And then Maybe some examples of companies that fall into each bucket.
- [12:28]
B
Yeah. So let me talk about how this is constructed first. So basically what I did was I said there are obviously lots of different parts of the AI ecosystem. There's the chip producers, the memory, the storage, the hyperscalers. But to make things simple, I said let's just define two categories. AI infrastructure, the picks and shovels, and then everyone else. So infrastructure contains all these companies. So again, it's not just hardware, it's also software. So I would consider like Microsoft, Coreweave, Oracle as being amongst these names. OpenAI, Anthropic, of course, these are the names that are kind of the producers of AI that then sell it to the customers, the users of AI. The other half are the users and within that category. So first of all, about 10% of the stock universe is in this broad infrastructure category, and that's a broad definition. The other 90% can then be split. What I do is I use my adoption score to say we know from low to high, this is a continuous measure how much different firms are investing in AI. Now we're going to arbitrarily define a cutoff at some point to say guys who are below that threshold are considered laggards and those who are above are considered early adopters. Again, you can calibrate it differently, but it intuitively feels right. The fact that the early adopter category ends up being about 10% of the universe. So you have a 10% of the universe is infrastructure, 10% is early adopters. The other 80% are laggards. Which kind of feels right to where we are in the cycle. If you look at the US Census, they have a survey that they do for businesses. Around 10% of businesses are adopting AI in their production of goods and services. Kind of lines up more or less with these numbers too. So that kind of feels like a ballpark. Correct. Way to segment the universe.
- [14:17]
A
And then thinking about it, as an investor, how are you, let's say, how do you weight your portfolio and how do you find these names that go in each bucket?
- [14:27]
B
Yeah, it's really interesting because there's kind of two ways to think about it. One is you can think about things on the industry level. Are there sectors that are more exposed or stand to benefit more than others? You can also look at the intra industry level. So within industries, are there leaders versus laggards? That I think is actually kind of the more interesting angle. It's really difficult predict whether or not software stocks will ultimately rebound or not. Right. There's a lot of things going on. And it's not like a monolithic asset sector, but I think it's a lot easier to say within a given sector. These few companies are really leaning into AI. They're really forward thinking and they're deploying these technologies or at least making strides towards that direction. And these other guys, they're clearly doing nothing. And I actually have an interesting chart where I show on a per sector basis the distribution of AI adoption scores. And in most cases it looks very right tail, very skewed, where you can see a huge clustering of companies at the zero bound where they're just not doing anything and then a small number of leaders who have really separated from the pack that are aggressively pursuing the opportunity. I think all else equal, you'd expect that those aggressive AI adopters should eventually separate from the pack to the extent that AI does indeed turn out to be a transformative technology.
- [15:45]
A
Okay, I think I'm going to have to pull up that chart. You had one sort of table that you were writing about and in the infrastructure camp you had Nvidia. Does that mean we are, let's say, moving past the Nvidia stage of the cycle?
- [16:04]
B
Well, I mean, I don't want to speak specifically to Nvidia. I mean, I will say that the outperformance of the AI infrastructure category, including Nvidia as well as some of their competitors and other parts of the AI value chain over the past few years has been nothing short of incredible. Remember, Nvidia has historically been a massively cyclical business. Semiconductors just boom and bust over and over again. Now we're obviously in a upswing in the cycle, but these things never last forever. And I'm not going to say that I can predict timing of these things, but it does feel like if you think about the narrative in this paper, I quote Satya Nadella from two weeks ago, where he talks about how the big focus is now on not so much the capital spending that these businesses are doing, but more on will this technology diffuse through the Will a pharmaceutical business be able to use AI to accelerate clinical trials? If the answer is yes, that's when we know we're not really in a bubble and we're actually seeing kind of the real adoption of this technology. The fact that the CEOs and investor class are now shifting towards that frame of mind, I think does to me suggest that the shift is underway. I don't know exactly when this will happen, but it does feel like things are starting to change a little bit.
- [17:20]
A
I agree it does Feel like there's a deep movement in sentiment right now, or at least the first month of the year. This week we saw Cellophane Financial services stocks, we had a AI tax product come out, LPL Financial, Charles Schwab, a few other names sold off, and this is just a week after software sold off because I think anthropic or maybe some other tools came out that made people really nervous about software valuations. If we take those sectors as examples, can we apply your framework here where if we just let's say the knee jerk reaction was too extreme and then winners and losers emerge? Can we take your framework of find the companies that mentioned economic roi, AI driven gains as maybe a predictor of which ones will come back versus which ones will stay flat?
- [18:16]
B
Yeah, I think it's so interesting what's happened the past couple weeks. AI has been around for a while. ChatGPT came out over three years ago. And for anyone who's been close to the technology, playing around with it, seeing its progress on a week by week, month by month basis, it's kind of obvious to most people that this thing is the real deal, this is happening. So it's so interesting that these stocks, whether it's the financial stocks or tech stocks, well basically they're basically unchanged. The investors don't really care at all. And then suddenly overnight something happens and they're down 10, 20%. It's just kind of an interesting anecdote around how the market works gradually, then suddenly. I will say that what this points to is the fact that the market is waking up to this idea that AI is likely going to be a disruptive force, as was the cell phone, as was cloud computing, as was the Internet. Now the question is less will it enhance productivity? I think the answer is obviously yes. The question is more what parts of the economy, what companies will accrue that value? Who will capture the value that is being created, the economic value being created by these technologies? Maybe it's none of the companies, maybe it's just the workers. I don't think that will be the case, unfortunately. This time maybe it's the individuals, the users, maybe that will be the case. But as investors, we need to focus on within the corporate sector. Amongst all the publicly listed stocks, which companies do we expect to get the most benefit? And which ones do we think will be the losers? Because that's always what happens. These periods of disruption are really interesting for investors. They create dispersion, they create winners and losers, they create the opportunity to have. Taken your position in terms of which companies you think will do well or not. And then when the dispersion arrives, that's when the separation really occurs. And so far you mentioned software, like it's interesting. You see Salesforce, Adobe, Figma, all these companies, many of which are totally different businesses, all go down together, right? Like are we putting, you know, LPL and Schwab in the same category and S and P Global? I guess. Right. But like it's just so interesting how the market is kind of these, as you mentioned, knee jerks, a good term, selling off these broad swaths of the market pretty indiscriminately. Yet like it's pretty clear that both from a business model standpoint and from an AI adoption standpoint, many of these companies are on totally different dimensions, totally different playing fields. And so likely for active investors, it does create good opportunities to buy quality tech businesses or financial businesses at a discount.
- [20:46]
A
Well, when you see something like Microsoft sell off so much, it's not like Microsoft will stop making money. It's a massive, massive corporation. They have a gazillion dollars on their balance sheet. And same with other major software companies with the starts of the year being pretty rocky for tech. How are you thinking about the Mag 7 right now?
- [21:07]
B
Yeah, I mean, I think I made my position somewhat clear in October when I wrote that letter on AI capex. Basically my view then and continues to be the case is that These businesses, the MAG7, have historically been basically the best businesses ever created. If you were in business school and you were kind of crafting and sketching the perfect company, you'd craft Google. And for years and years they just massively outperformed the stock market. And investors have become highly conditioned to think of these companies as these kind of cash flow juggernauts with super high returns on invested capital, asset light businesses like monopolies, network effects, so on and so forth. I think what's happened is over the past year really, and it's accelerated more recently with the last quarter, that these companies are basically transitioning their business models from asset light to asset heavy. They are turning away from the model that led to so much success in the past and are now migrating more towards this utility like platform where they want to invest the most capex in building out the biggest AI cloud, acquire the most GPUs, that's what they think will win. Now the thing is that if any single company does that, that is fine. But collectively, if everyone's trying to run that same play, you create this issue of overcapacity. Too much competition and kind of commoditize your own business, which I fear is what's happening. You have folks like the Google CEO saying I'd rather go bankrupt than lose this race. When people are saying what they're trying to do, that is a bit of a concern. I've generally been less excited about the Mag 7 now and to be clear, they've been big holdings in my funds. I've enjoyed, like many people who own indices and such, the wealth that's being created by these companies. And I highly respect many of them. But I think given how they're valued and Mr. Stool Condition, to think of these companies as these asset light companies and given the business that they're now getting into in the competition, competition having increased, I think that's less exciting. But conversely, that doesn't mean that on a relative basis the rest of the market is a lot more appealing. And we've seen that, we've seen this year this rotation away from the Mag 7, away from these AI infrastructure stocks towards a broadening that includes value stocks or companies in other industries. Tech has been the worst sector in the S and P the past few months. So I think we have started to see this play out. Will it continue? I expect it to, but you never know. There's so many factors that move markets on a day to day basis.
- [23:40]
A
Real quick, we'll get right back to the interview. Just wanted to pop in and say if you like this content, I write a newsletter every single morning called Opening Bell Daily. I cover macro, the stock market, asset prices, why things are going up, why they're going down, and if you want to get that for free, you can sign up at the link in the description. Let's get back to the interview. Well, speaking of valuations, you had this other great chart that essentially shows the AI infrastructure valuations are so much higher than the early adopters and the AI laggards, to me, that, you know, screams opportunity. I hadn't seen it broken out like that before and I thought it was really smart. Is it that irrational just to pile into all the laggards right now? I mean the early adopters. Because if they're so much cheaper compared to the infrastructure plays that pretty much make every single headline. And no one's talking about the early adopters. Maybe it's don't overthink it, don't try to be clever about it. Just find these early adopters.
- [24:44]
B
Yeah, I mean, let me frame it this way, which is investors have this kind of dilemma where on one hand they are forced to say Am I an AI bull? If I'm an AI bull, I want to buy Nvidia and all the AI infrastructure stocks. That's the obvious way that people have played this trade historically. Or are you an AI bear? In which case you want to underweight, short them and buy the laggards, buy the value stocks, European stocks, small caps, the things that have no AI exposure. But that's not a satisfactory answer, right? Because that's like saying, oh yeah, I think the Internet's a real thing, but I'm just going to not buy any Internet stocks. Then you miss out on the entire productivity gains. Right. And that's why I think that there's actually a third option that people are underestimating, which is these early adopters. Because the early adopters are an interesting category of companies that they stand to benefit from AI. If AI indeed increases your productivity by 10x as a coder, great. And you're a software company, you can actually use that if it will make your legal department much more efficient. Amazing. So you could easily conceive of how the corporate sector could expand their margins and take market share. The early adopters could, based on AI. So that makes sense. Yet the good thing is that you're not wagering as much to do. So, first of all, these companies are still capital light. They're not spending all the money to build these servers. They're offloading that to the hyperscalers. Second is, as you point out, their valuations have not gone up. So whereas Nvidia and those stocks, their valuations have pretty much doubled or more in some cases over the past few years. With the AI, you know, hype building, the early adopters valuations are still pretty much in line with the laggards. So in other words, the market is not rewarding companies for their AI investments. On the adoption side, they're rewarding you if you're Oracle and you're making investments in AI data centers, but not if you're on the early adopter side. So, yeah, to summarize, I mean, the early adopters are interesting because they stand to benefit from AI becoming, you know, more and more powerful over time, yet they don't have the same nosebleed valuations or capital requirements that their infrastructure peers have.
- [26:45]
A
Can you share a couple names that you think are interesting in the early adopters category?
- [26:51]
B
Yeah, so actually there's a whole exhibit we should put this up actually, where I went through the earnings calls and kind of pulled out interesting examples. Right. So you have examples like Target talking about how they can Increase their, the picking of their, in their factories by I think it was 40 something percent. You have, you know, C.H. robinson, which is a logistics company, talking about how they use AI to win more business. You have companies in pharmaceuticals, in defense and in real estate. I think public storage gets a mention there, insurance and healthcare. So kind of sectors across the economy, names that you wouldn't normally think of as being AI companies are saying publicly and again it would be fraud if they were lying saying that they're getting these gains in different ways from AI. And that could be. We actually do an analysis of this where we show it's both revenue expansion, so taking market share. It's also cost reduction, headcount reduction, more efficient utilization of their human capital and industrial assets as well as some kind of more long tail things. But yeah, so it does appear that we're starting to see real ROI from AI adoption at the enterprise level. Now to be fair, it's that from a very low base. So you know, when I describe the chart of how the mentions of these gains are looking, it's like started very low and it's getting much better and it's happening quickly, but it's still a relatively small percentage of companies that are doing this. And in many cases it's not scaled right. So it's like at this point these companies are getting ROI on their AI pilots. But what we still need to see is can this scale to a higher degree such that it's really starting to move their bottom line. And I think we're still some companies are doing so CH Robinson being a good example, are actually tying their good earnings to AI investments. But in many cases it's like, yeah, we're getting good roi, but it's not as clear on how that will actually move the needle for these big companies. But again, it's only a matter of time. It's still. While it has been three years for enterprise adoption, that's still early innings.
- [29:02]
A
Okay, Kai, I want to get your thoughts on something. I was talking to an investor recently and he told me if you're so optimistic and so bullish on the technology, that almost means you have to be bearish on the stock prices. Because if the technology gets so good, these companies will essentially, you know, costs will fall so dramatically even if demand is so high that their share prices will suffer. Is that a way that makes sense to think about this or, you know, is this guy off his rocker?
- [29:37]
B
Look, I think there's a lot of second and third order consequences of how this technology can flow through. I Think the first thing you need to think about is I was sent a chart this morning by an investor that showed the labor share of profits, corporate profits versus the shareholders, the capital share. And it's been on this declining trend through time. So back in the day when unions were more powerful and companies were less monopolistic and technology was less prevalent, it was the case that labor commanded a significant share of corporate profits over time. That's just been in secular decline the past, say, five decades, accelerated by globalization, technology, the weakening of unions, monopolies, so on and so forth. So that's one question which is first of all, will it grow the pie? I think the answer is yes. The second question is will it change the distribution of the pie adversely for shareholders? Unfortunately, I think the answer is no. I think this will only accelerate the demise of of labor at the expense of capital. Now, within each sector there's going to be winners and losers. As I mentioned, it could be the case. I guess I'm trying to play out this thesis that there's one mega company that just takes all the profits and the 99.99% of companies just go bankrupt. I guess there's a distributional thing and if you don't own that one company, you're kind of out of luck. But I think there's going to be reshuffling. I think if you own the index, you're generally going to be insulated because if you own all companies and you're going to own that one company as well as the losers and things will be fine. So that's fine. But I do think, going back to my point earlier, that if you're an active manager and you are able to kind of see a little bit ahead, you don't have to be a prophet, but just see a little bit ahead. Which companies here are trying to stay ahead of the trends, which are trying to stay on top of the technology and implement it for their business, they stand to win and take market share from the laggards. So I think the opportunities today are less like trying to make a bet on going long and go short. The market based on is AI good or bad. I think the opportunities more are on a relative value standpoint, trying to find companies and sectors, but more interestingly, within each sector which companies are early adopters as opposed to laggards in the AI race.
- [31:51]
A
That's a great way to think about it. Kai, I'm going to try to tell you my understanding of your paper and your thesis here, and this kind of falls in line with what we've been seeing in the market, a broader rotation in the market out of, let's say, big tech mega caps into small mid cap and sort of other laggards of the last couple years. It's not a story of an AI bubble popping, it's more just a very gradual rotation out of infrastructure into the second and third order winners of AI. Is that the right way to think about what your view is?
- [32:28]
B
Yeah, I think that's exactly right. That's what I expect to continue to happen. And we're starting to see. I mean, I guess I'm not going to say that it's all played out yet, but how you're describing what's happened I think accurately describes the past, say, two, three months. And again, I think it's still early. We'll see whether or not this continues and if it does, that'll kind of validate this thesis that we're shifting now from this infrastructure phase to the adoption phase. But I do think that the early signs do seem encouraging that this thesis is going to play out.
- [32:58]
A
If I had to put you on the spot here to poke a hole in your own thesis, how would you start?
- [33:05]
B
I mean, I think it's pretty obvious if AI is truly the last invention and what we being anthropic and open AI, what they build is just so much better than anything that we can conceive of. And yeah, I mean, we're a bunch of financial, I don't know, people kind of squabbling over DCFS and CapEx and these things that don't matter because you invent godlike intelligence. We're done here. Throw this all out the window. In that state of the world, of course, we'd have a lot of things to worry about that are not financially related. It's more a question of can we beg our AI overlords for some ubi?
- [33:45]
A
Geez, Kai, where can people find your work online?
- [33:49]
B
Yeah, just go to my website, sparklinecapital.com I have a bunch of stuff posted there.
- [33:54]
A
Okay, I. I read all your papers. It's excellent work. Everyone should do it too. I love having you on the show. We're gonna do this again very soon.
- [34:02]
B
We'd love to be back. Thank you.