Making sense of the AI rout. January 31, 2025

26:04That's what you can afford to repay. You say, oh, okay, cool. How much house will that buy me? And the bank manager does the maths and says, well, okay, the current rate is 6.5%. I've got to assess you with plus 3%. So at 9.5%, your$3 ,000 a month you can afford to repay buys you, I don't know what the numbers are. I'm going to say a million dollar house. I don't know what the numbers are, right? So you can afford to borrow a million bucks on your repayment ability at the current rate. And you go, beauty, thanks. When rates drop, you don't go to the bank manager. No one goes to the bank manager and says, hey, I only still want a million dollars.

26:34How much less can I pay off? You go to the bank manager and say, how much can I afford to repay? And the bank manager says, oh, still three grand. Oh, beauty, great. How much can I borrow? Oh, that's now$1.1 million. Oh, so I can go and bid another$100 ,000 at auction on top of what I was going to to get the house I really want. Yeah, you can. In fact, I really like you to because I'm in the business of making loans and I get paid based on how much profit I make on those loans. So yes, you could have had a million dollars yesterday. Today, after rates get cut, you can now borrow$1 million. Go and knock yourself out at auction.

27:03And that's kind of the worked example of what you were saying before, mate, about when we get an extra bit of a rate cut. It's not like people say, oh, great, instead of paying three grand a month, I'll pay$2 ,800 a month on my million dollar mortgage. My new purchase, the mortgage itself, they will, but a new purchase, it'll get capitalized in a price. Because no one goes to the bank manager and says, how little can I borrow? You go to the bank manager and say, how much can I borrow? And the bank manager whacks the numbers in the spreadsheet and says, actually this much, which is exactly your point.

27:28I'll move on to the solution on my previous solution. Do you want to add to that? No, you articulate it perfectly. So here's the thing I mentioned. Game theory in action. Game theory in action. And it's completely natural. Well, I say natural. In a market where housing is scarce, in a market where there's a cultural expectation that prices go up, and in a market where we all, frankly, we're humans, we all want the best we can afford, particularly when it comes to housing. Especially when a million dollars buys you a dog box, right, these days. We say it as if, ooh, a million dollars, which when we were kids wasn't like, wow, right?

28:00Now it's like, oh. We were on a lot when we were kids. These days you win a lot, you know, the million dollars lotto and you still couldn't pay your house up. You know? Anyway, so yeah. Yeah. So that's going to happen almost certainly because we all, market dynamics and game theory and just the fact we want the best we can afford. That's just what Hugh wants to do. Of course. I have other solutions in terms of supply and demand. That aside though for now, the Treasurer has a really huge opportunity to do something. I want to really bang the table on this one, Ram. The Treasurer has a huge opportunity to fix that problem, to stop it happening.

28:30And it's really, really, really, really, really stupidly simple if he was of a mind, and by the way, the opposition is not suggesting this either, so it's not a partisan comment. If he was of a mind of saying, you know what, I really don't want to make things harder for first-home buyers. I really don't want them to get screwed when rates drop. I would like them to be able to still afford the house they want to buy and actually maybe, to your point, put a bit of money back into the economy rather than just capitalising it in their house prices. How do you do that? Well, it's really simple. The APRA buffer I mentioned is currently 3%.

28:57On top of 6.5%, that makes it 9.5%. You're qualified for the loan. 6.5 % is arbitrary, but that's kind of the bank average variable rate at the moment. If rates – I'm going to make big numbers, mate. If rates would have dropped to 5.5%, because the RBA cuts rates over the next 12 months, absent anything else, house prices go up. All the treasurer has to say to APRA, the banking regulator, is, hey, when rates come down, increase your buffer. So in that very simplistic scenario, APRA could simply say, hey, rates have gone from 6.5 % to 5.5%. Our buffer's gone from 3 % to 4%. What does that do? It means the borrower still is assessed at 9.5%.

29:37Which means what? It means there are not borrowers out there out-competing each other with a$1.1 million,$1.2 million loan. They're also getting a million dollar loan. So what does that do? Macro prudential controls is what you're talking about. And it's stupidly, stupidly simple. and I've never ever had anyone, other than ideologues who say government shouldn't decide who borrows money, and I kind of get that to some degree, there is no economic or kind of policy suggestion I've ever heard as to why that would be problematic. I'll give you one. You want prices to go up. You want prices to go up.

30:06Yeah. And literally that's it, right? But if you care a bit about first-home buyers, if you care, frankly, the rest of the economy, not as it helped first-home buyers, to your very point at the very beginning, Ram, it means they're putting more money into the economy. If they could afford to pay$3 ,000 a week at the current rate, they'd be repaying$2 ,800 at the new rate, still buying the same amount of money. They're going to have more money in their own back pockets. They can't put into prices because the buffer's gone up. Guess what? It gets spent. So it's good for them. It's good for the rest of the economy.

30:36It's good for everybody except those who want to see their house price go. Yeah. It's really, really straightforward. forward. Oh, by the way, after we nearly destroyed the entire global economy back in 2008, one of the, you know, when people were sort of looking for, you know, where to point the finger and stuff and some of the solutions and gosh, let's never let that happening. Was that really bad? Yeah, that was really bad. Let's not do that. Yeah. Okay. What do we do? Let's do these counter-cyclical buffers. That was exactly what happened. That was exactly what was discussed. All the experts said what you just said.

31:09Yeah, it's a really good idea. We should do that. And we did it. And then because we've got the memory of a goldfish as a society, we recently have gone, no, let's not do that anymore. You know, it's like, what? Like, are you joking me? So, and, and, and here's the, here's the other political calculus. So I said before, this is rough, but pretty close. One third of people own outright. One third of people have a mortgage and one third of people are renting. Of the third that are renting, most of them are pretty young and are not very politically influential. Yeah. So if you're a politician and two thirds of the populace want house prices to go up, and the one third that doesn't want it to go up are just young people who you don't care about anyway because they don't donate as much and they're just not as politically influential and numerically they're not as significant.

31:57You know, it's just sort of like, I'm just going to go to where the most votes are, right? And everything you say might be true. But it's not going to help prices go up. And for all the people out there going property doubles every whatever it is, seven years or something. It's sort of like they've been told that by it's not just like your weird uncle at a barbecue. It's sort of like financial planners, accountants, the banks, you know, leading economists, you know, a billion idiot influences on social media. That is the religion of our time. And it's almost as close to a social contract as you can get.

32:38And politicians are dumb in a lot of ways, but they're super smart in one. And that is that they can read the room better than anyone else. And if you're a politician who's worth their salt, you go, I am not doing anything that is going to make even the possibility of prices not going up. So that's the answer to your question, which I know you know. Yes. Because in every other sane respect, you're 100 % right. Correct. And you and I have said before, if you're not going to sell your house, it doesn't matter. What's the difference? It'll make zero difference. Not only that, if it does double and then you sell it, you're going to buy something.

33:12Literally, unless you're in a nursing home or dying, you don't get out of this one, right? Now, if you're on an investment property, you care about that, so there's that group. But that should be a relatively small proportion. It shouldn't be hard to say to people, hey, you own your home or you're paying a mortgage. It doesn't matter to you whether it goes up or down. By the way, you've probably got kids or your siblings have got kids or there are kids down the street who are going to be screwed when the house prices go up and that's going to be on you. You're right. and it's a political problem.

33:42I shouldn't be surprised. I shouldn't be angry about it. I should be more resigned to it. That's how you get to be you rather than me. Come on over to this side. It's really great. You still rant a lot, so I'm not sure whether it's actually that much better, but it lets me more realistic. I still, I just, it beggars belief that, and I know this sounds stupidly naive, you're treasurer of the country. You, in theory, want to be in politics to make a difference. You know, you, in theory, have some sort of view about what sort of country you want to leave. You want to walk away and say, my legacy was X.

34:11I just honestly, I don't know how, either it's a complete disconnect from reality or you just have to tell yourself so often that you stop thinking about it. I don't know how you go to bed at night and say, I'm going to enact a policy or to allow a policy to remain in place, which is going to screw over a whole generation of kids. And you're like, yeah, that's just going to have to be the way it is. I honestly might have, that's why I'm never in politics, right, because I never climbed the grizzly pole. But I honestly don't know how you don't sit there and go, to your point about Phil Lowe, none of our pollies are poor, right?

34:40How you go there and say, I'm going to sit in this chair for another three years, enjoy the trappings of power, screw a whole lot of kids, and then leave. And that's my contribution. Yeah, but you don't think that, though. I mean, look, the soldiers marching people into Auschwitz were convinced that they were good people, right? And that's probably not a very palatable analogy. I apologise for that. But it is like no one in there when they look in the mirror thinks that they're evil. You will do whatever you can to rationalize things, no matter how smart you are. And the treasurer of any strike will be there going, yeah, look, it's a tough decision.

35:13There are trade-offs, but on net, this is the best thing. And you could hook them up to a perfect lie detector. And they're like, yeah, no, they believe that. Because you have to believe that. You have to believe it. Because otherwise you have to confront the reality that you are a force for, evil's not the right word, but you know, yeah, in the world. And humans as a general rule just don't do that. So you've got to accept the world as it is and not as you would have it. This is a bit of lesson that I'm still trying to reconcile. I mean, and then here's the other dimension and not to make it, I mean, we're going to be using the Trump word a lot this year and next year, the year after.

35:56But you have to, because it is the by far the, or not by far, but it's certainly the biggest economy in the world and the most influential in the world. And you've got a commander in chief out there who's very much jawboning and saying everything, like the political reality of that situation is very much going to do everything that can be done to make it stimulatory. So asset prices are going to go to the moon in the next few years. You know, I really think they are. And I think it's going to be a fake growth because I think in real terms it won't be nearly as good and I think it's going to be a very uneven growth and so it's not going to be a growth that's going to be shared very well.

36:34But you're going to do it because when you pump a bunch of cash into the situation, things that are genuinely scarce go up in value a lot and it's hard to think of much more than leading companies and good quality properties that just, you know. So Michelle Bullock is there with her various council of elders in the temple deciding what to do. That is very much a reality that they have to confront. Look at the Aussie dollar, right? You mentioned before the stability of the currency. So what's happened to the Aussie dollar lately? It's tanked. Is that too strong a word? Well, so again, so tanked compared to the US dollar, and it's like almost a 20-year low.

37:22Yeah. It's a little bit, it's not misleading, but the full picture is it's down a lot less against other currencies. So apparently it's down about 4 % trade weight in index terms. Yeah. But down a lot against the US, and largely that's US dollar strength on the back of, effectively, the Trump tariffs that we talked about before. And as you say, we'll mention Trump's name a lot. The tariff in expectation pushed the US dollar up a lot. And I have a nice... currencies and we're so we're doing this uh thursday morning the 30th of january so overnight uh uh pal guy over speech and basically said no actually i don't know if i'm going to be cutting rates anymore yeah the madness continues i'm doing this no i'm not i'm going to do this no i think this is going to happen actually now i changed my mind what do you think i think this so the usual circus sort of goes on but then the latest performance is is uh performance art piece is that and and what did the aussie dollar do because again it's a relative game so so all i'm trying to say is that there is there is what's happening here in terms of the economic landscape in terms of what the uh intellectual framework ideological framework that goes around that in terms of how people think at the reserve bank there's the australian political reality and then there's the american political reality on top of that as well and it just like to me just coming back to accepting the world as it is and not as i would have it i just i just and i've been saying it for ages.

38:41I see that cans continue to get kicked down the road. I think everything and anything is done to avoid any kind of reckoning in any way, shape or form. And I think we run hot for a while. It's not going to get anywhere near the peak of inflation that we saw during COVID. But I think it's going to be super, super difficult for the US to get to 2%, which is their target. I think it's going to be very hard for us to get to even the mid-range of ours. And I think that's exactly if I was a uh an architect of all of this given the political reality that's the best you can hope for because a little bit hotter than normal is just like enough that is I just need enough so that people aren't walking down the street with pitchforks and nooses that's that's the kind of map because it will inflate this debt away which is the elephant in the room here sure comes back to you right it's like there is a bucket load of debt that is out there public sector, private sector, corporate sector.

39:33And it's just sort of like, we need it. We need inflation, you know? Well, actually, no, we don't need inflation, but that's, that's, that's what people think. And it is, it is. So, so, so, so again, just to bring this back to the real world of like, okay, guys, what the hell do I do with my investments? Stay invested is what I would say. And, and, and look, I would, that's a bit of a cop out because I tend to say that all the time. And I stand by that, actually, because it's sort of like - Well, it's always been the right decision. It's always been the right decision. That's the other thing for all of the macro.

40:03I say regularly, the macro matters a lot for policy decisions and actually for our standards of living and quality of life. Yeah. It tends to matter a whole lot less for our investing. And so it's why we kind of want to talk, not most of us in our mouths, but just on two different planes. Because there is the, hey, what's the right thing for the country? What's the right thing for the society? What's the right thing for individuals and for groups of people? And then there's how does it impact our investing? And historically, there's been very, very, very few times there have been genuinely, objectively and foreseeably bad times to invest.

40:33Absolutely. And so it's just it's going to be, look, again, we said before we don't make predictions. So this is just for fun. Right. But I really get the impression that this is going to be a scenario where it's kind of a stagflation kind of environment, not a lot of real growth, a little bit of inflation. people seeking people being rational a economic agents going well i can't preserve my my um my earnings in money why would that work so i need to preserve it so like ask anyone in the world like would you save your money in cash like i don't know maybe grandma in the nursing home who came from a different era might everyone else in the world buys a property or buys some shares because money sucks right like it's terrible even a high interest savings account is woeful and probably negative in real terms.

41:23And so you get to a situation where even though on a fundamental basis, even though on a mean reversion basis compared to sort of traditional metrics in markets and the rest of it, I think it's more likely than not things will have to run a little bit hot. Now, again, maybe I'm wrong. I mean, it wouldn't be the first time. But either way, it's sort of like I only have a limited set of options in terms of where I preserve my wealth And when I look across the landscape, cash is always at the bottom of the, unless, unless I have an immediate or new term requirement to spend it, because it is the medium of exchange, then other than that, it's not going, it's not going into cash.

42:06And if it's not going into cash, it's going into houses, it's going into equities, it's going to those kind of, you know, meme coins, for goodness sake. And, and, and, and, and that is why I think a lot of the hard-nosed value investors have had their face rubbed in it. Not because they're wrong. Well, they're wrong. It's relatively right. Yeah, exactly. They're wrong. That's what you mean, though. I know what you mean. But their arguments are pretty solid, assuming you think the world operates in a sane way. But once you realise it doesn't, then it's like, well, actually, yeah. Yeah. These valuations are a little bit silly, but stay with it.

42:36It might even melt up higher. Anything's possible. Let's take a 90-degree turn, mate. We kind of referenced AI at the very beginning of the pod, and it's taken us 43 minutes to get here. So thank you for spending a bit of time or letting us rant about some other things.

42:52Deep seek. Wow, I was trying to set this one up. So, I mean, look, and it gets really technical really quickly. So we're going to try and keep it a little bit accessible if we can. AI, we talked about last week. We'll talk about probably next week and most weeks, or at least not often enough. Trump, interest rates, inflation, and AI. Exactly. That should be the new name of the podcast. There's your agenda for the next four years. so look yeah so AI it's been huge AI's been huge we're still in the first innings of AI right we are still so stupidly early despite how great it is already and the the expectations of the market the expectations of investors just frankly consumers people businesses continue to evolve it's kind of like the internet right it's like kind of you know AltaVista turns up as a search engine kids you won't have any idea what this is and it's like oh wow we can search the web that's amazing this is going to be great AltaVista's fantastic and then Ask Jeeves turns up.

43:45Then Yahoo turns up. Then Google turns up. And frankly, AI might be the next generation of search, frankly, in an analogy. But either way, we are an AltaVista stage. We are at, I can log onto this internet thing using a dial-up modem. That's kind of cool. Back in the day, you had to enter an address. You did. And if you didn't know the address, you didn't know how to get it. You can look it up. The SMH only had their technology section online because who wants to read anything else online? Fortunately, there was only like 12 websites in the entire world, so you could remember them all. That was easier.

44:13Anyway, Anyway, it's a long way to set up the fact that we saw massive, massive, massive news this week. Not even necessarily in terms of the news itself, but in terms of the impact of that news. So Ram's already mentioned DeepSeek. This is a new Chinese-based AI. I want to say new. They've been around for a while, but they released a new model and a massive claim. Now, the claim is, and I say claim because I think there's a bit of uncertainty about how generous or genuine this is. This is all in the last two days. Right? This is super fresh. So, actually, I'll take a half a step back. Like, NVIDIA is a chip maker, computer chip maker.

44:45They make very specific chips that were originally used for games, still are, and then the Bitcoin miners kind of got on with it. And then the AI people realised they could use these chips specifically and better than the other chips that existed for training and answering AI or dealing with, suggesting, creating AI. And that sent NVIDIA's shares to the moon. They're up 20-fold in the last five years alone. Biggest company in the world. Right? Not anymore. And that's true. But at one point they were. Look, they're top two, three. Yeah, top half thousand at least. Massive. They were worth$3.5 trillion.

45:20So that's the setup because when DeepSeek comes out and says, hey, guess what? With this new AI model, it's actually every bit as good as ChatGPT. By the way, ChatGPT probably, no one knows for sure, estimates are massively big, by the way, between$100 million and$1 billion to train something like ChatGPT's most recent model. so that's a 100 million to a billion keep that number in your head deep sea come out and say we've got something that's good by the way we didn't use anywhere near as many chips as we used to and the whole thing only cost us six million dollars to trade that is 94 less than the bottom end of the range claimed for other ai models to teach well for other ai creators to teach their models how to deal with the questions and comments and requests that we make of ai that is an astonishing saving, literally 94 % cheaper, a whole lot less chips.

46:12And NVIDIA's shares fell by 17 % in a single day, which is a lot in anyone's language. That equated to$600 billion US dollars in one day. And that fall is larger than all but 13 of the 13 largest US-listed companies. In other words, NVIDIA lost more value than the 14th largest US company is worth in total. So that's kind of shocking, right? And by the way, the figure I heard was like, it's the size of like all the major banks and BHP or something. I'm going to get it wrong. And just for context, given the market cap, take every single, there's about 2 ,200 companies on the Australian Stock Exchange, add them all together.

46:55Yep. And they're about a trillion dollars short of what NVIDIA is worth. It's crazy, isn't it? You know, that's all I say when, you know, America matters, right? Like it's like when it sneezes, the world catches a cold. We are such a tiny, tiny tadpole. But yes. I'm still saying that I will let you jump in. So that's enough, right? That's big in and of itself. The reality is that that sort of fall happened because the market had assumed a certain thing about AI, had assumed that they would always use NVIDIA's chips. They'd always use a lot of NVIDIA's chips. There's going to be many, many more of them created.

47:30And so this company was worth$3.5 trillion because the future was so bright for the business. And the deep sea comes out and goes, yeah, maybe it's not right. Now, I will say there are plenty of people, including our mate Elmo, Elon Musk, who has said, no, it's all fake. They didn't really mean that. They got the chips through the black market. Backstory is China's companies can't buy these Nvidia chips because the US has banned them from having them. So some idea that they got a backdoor through Singapore. They're using more than they said. This is all they were making numbers up. They're trying to screw with the Yanks.

48:03we don't know whether it's real or not. It's almost not the point there. I mean, it is in one sense, but we're so stupidly early. I used an allergy on Lyme, which apparently is wrong. I said it's like looking at a group of kindy kids and trying to bet which one's going to be the tallest by year 12. And someone said apparently kindy heights are apparently reasonably indicting to be the 12 heights. I was going to say, what's wrong with that? Right? But that kind of idea, again, take the spirit of the analogy rather than the actual one. Looking at something in kindergarten and saying, you know, try to pick the ducks of the school in kindergarten or pick whatever you want, right?

48:33Or you're in the maternity ward. There's a whole bunch of little babies wrapped, coddled up. You know, which one's going to be the tallest at 21? Right, right. Or the smartest, all over the world. And so that's kind of, you know, we are so early. We're so stupidly early. So I'll stop there. I've got some thoughts, Matt. But having set it up in a way too long a ramble, just some instant reflections on the week's news. Yeah. I mean, I'm still trying to make sense of it. That's good. I'm still trying to distill my thoughts and I'll form an opinion and then I'll change it. Someone said on Twitter, they're like, oh, you've really changed.

49:06And I was like, well, grow or die, right? I think, thank you. That is such a compliment. I appreciate that. I don't think it was meant as a compliment, but I took it as one, right? And yeah, any other position is insanity to my mind. But yeah, so at this current state of play, I think it was reasonable to assume that these mega companies, the open AIs, the Anthropics, the Gemini, or the Googles, whatever. They had such a potent combination of scale, first mover advantage, technological know-how that the domain of AI was really their domain. In other words, some gigabrain in their mum's basement is never going to compete, right?

49:52You can't. It doesn't matter. You could have 10 Albert Einsteins, the Albert Einsteins of AI and programming in a room. It doesn't matter how smart they are. They just don't have the resources. They don't have the capital. They can't be caught up. And under that framework, when you look at NVIDIA, you look at some of these valuations, it's actually kind of reasonable, right? Like they should be pretty highly valued given the potential, given the future, given the league, given the competitive advantage. And what we don't know, so who knows, but what DeepSeek has fired a shot across the bow of is that no, this is a tiny, relatively tiny company that's trained it really fast, really cheaply, and not only kind of like is almost as good, but is actually exceeding open AIs on a whole bunch of benchmarks.

50:42That's a big deal. It's a big deal because it just basically, as we would talk about in any company and the competitive advantages and moats, it's just like, huh, maybe that moat isn't as big as I thought. And now look at it from a, what does that mean in terms of dollars and valuations? Well, let's say OpenAI, just keep it simple. It's the only one that has a decent model. Well, they can turn around and say it's a hundred bucks a month. Yeah. And in fact, I would pay more than that. And I think a lot of people would. It's like, actually, like it's us and there's daylight and there's some dinky other model over there.

51:15It's a thousand dollars a month. It's like, okay, I will pay it because it's still, you know, how much would it cost me to hire 100 PhD students? Well, that's a lot. Or I can pay$1 ,000 a month and get a million PhD level kind of, you know, it's a no brainer. Now, when all of a sudden competition out there in the marketplace says, oh, no, I can do it for like five bucks a month. That pricing power goes away. Right. And like that's the whole point of a moat. The whole point of a competitive advantage is to give you pricing power, is to give you better margins. In fact, we've talked about repeatedly on this podcast that that is really that is the indicator that there is a moat in place is that there are really high margins.

51:54Because in any domain where there's no competitive barriers to entry, you have a margin that's just an inch above the cost of production, which is great, which is capitalism doing its thing. And if it's not like that, there is a moat that is there. So if NVIDIA, OpenAI and all of these tech companies do not have that moat, there's two angles to this. There is one is like, oh, so it changes everything for AI. This is actually good. They've made actually the way that they did it is a breakthrough. And in fact, OpenAI can use some of these approaches too and combine that with their scale. This is wonderful for the sector.

52:33This is wonderful for the technology. It is wonderful for the end users. However, in terms of is it good for the market valuations of these companies, no, it's not. And 17 % fall, while a knee-jerk reaction, wasn't crazy. Given what we know, it might look crazy in a month's time when we're listening back to this. But, you know, you shoot first, ask questions later, and you're sitting there going, this is an unassailable company here. Nothing can touch it. Oh, you're going to sell, right? Or maybe if you're not, you know, such a flighty investor, neither of us are, but a lot of people are going to do that.

53:08So I actually think that was a perfectly rational response from the market. Kind of. I mean, I think it's absolutely rational. I think, and so we talked about kind of this, there's the news itself, the AI news itself. The market, I think you're right that it's totally rational. In fact, I would argue maybe it wasn't rational enough. Yeah. And I think, so Nvidia is trading on 47 times earnings. And that's a lot in anyone's language for a small, medium-sized company. When you order the largest companies in the world and trading at 47 times earnings, a lot is expected. Just in sheer dollar growth.

53:44I mean, you know, if you're a Woolworths and you go from one supermarket to two, it's easy to double, right? If you're Nvidia, you're already doing billions of dollars a year in chip sales. Doubling that and doubling it again to get to a reasonable PE, that's not so easy. That means, just to put that in everyday language, that means if they paid out to their shareholders every single last cent of net profit every single year, it would take you 47 years to get your money back. Exactly. Unless they grow. Unless they, yes. That's what the market's betting on. what i and we just finished talking about the uncertainty the innovation the pace and everything else and i just look there's a really seen video shareholders listening and and i don't have a strong view on the stock what i would say is the market is putting a lot of faith because the price is evidence of that in the idea that nvidia is and will continue winning for extended periods of time And I've got to say, mate, I love high-margin businesses.

54:43But I also am becoming, or have been for a while, more fearful than I used to be of high-margin businesses. And it's the Jeff Bezos line from Amazon. Your margin is my opportunity. So here's the thing. We used to kind of think about businesses growing. I mean, take Woolies, an easy example, right? It sells a bit more every year. Marginers are roughly the same each year. So it kind of grows profits. And you can assume because they're repeat purchases, they'll keep buying more every year. That sounds good, right? You're going to buy more baked beans from the retail that offers them. You buy Woolies, Coles.

55:08You might go to Aldo or Costco. ago but you know you're probably gonna buy it from Woolies about the same number maybe a little bit more than last year it's kind of relatively easy to at least extrapolate if not predict if you're a chip maker and I again there's a massive range of outcomes here but you're a chip maker you sell chips this year they're gone next year people are going to come back and buy more chips from you to get growth now firstly the chips you sold last year don't go away they'll be replaced at some point but they don't go away in 12 months so you're trying to find new applications secondly you at the moment are the only game in town you got a 90 market share and you've got a 90 gross margin which is beautiful except if in the year's time or five years time or 10 years time someone turns up says hey first of all that we don't need as many chips as we used to so a chip that sounds gonna fall next year well that kind of sucks even at that same market share and margin less volume is less volume or someone else comes out says um we've invented this chip and look it's not quite as good as nvidia but it's half the price so what you can do is you can actually afford to buy twice as many and your total computing power is higher.

56:09Now, all of a sudden, Nvidia has got to do one of two things. Either keep its price, its margins, and lose market share, or try and maintain market share by dropping its price and probably dramatically. Again, in either of those scenarios, not only does sales might, let's say sales halves, profits are going to halve, profits are going to fall by even more than that. And again, you're paying 47 times earnings for a company that requires it to continue to grow for years. Now, the other side of this is, in 10 years' time, there is no competitor. AI use has gone up tenfold, twentyfold, a hundredfold.

56:40Nvidia is still the only game in town. This is a$30 trillion company, not a$3 trillion company. Yeah, it's cheap as chips at the moment. And that's entirely possible. So I'm not here to poo-poo Nvidia or poo-poo AI or anything else, other than to say, if you're investing in the future, which is what we all do, the confidence, the imputed confidence that the market has, that Nvidia will and will always be spectacularly big, powerful and profitable, that's a heroic bet in my view. I'm not saying don't own any shares. I'm not saying you can't even buy them at the current price, but I do think you need to recognise what you're buying, which is all the good news is built in.

57:13It is literally, as we say, price for perfection. And that's why we saw the big fall. All of a sudden investors went, huh, you mean there can be competition? You mean the dynamics of the industry might actually change? You mean there's not this straight line which says number of chips required by NVIDIA will go up for this rate forever? and that's kind of what I just want to make people aware of is, you know, well, this isn't the most exciting, I don't know about this, by the way, not the most exciting business, but you can reasonably extrapolate its future with a, you know, with a relatively small range of outcomes and a relatively high probability of success.

57:42It doesn't mean you should buy the shares, the price matters. With Nvidia, I mean, you know, in one version of the future, this is a business making 15 % margins with a 30 % market share. In other, it's making 95 % margins with 90 % market share of the market 10 times the size. So, but just know that you're doing that. And I guess my - That's the question. That is the question you have to answer for yourself. And that's what I found a little bit weird about the 17 % fall was the people who on Monday night, or Monday day, I think it was our time, went, NVIDIA's going to be fantastic. And then Tuesday night went, oh, you mean this could be competition?

58:17And it's weird to me either way, because either you should have said, I think it's going to be great. There'll be competition, but I figure NVIDIA's got a great future. So I'll just hang on anyway. Or I'm never going to pay that price because it assumes perfection and you're not going to buy it. It's always the fall that really makes me wonder who owned those shares and what they were thinking in both of those days.

58:38Yeah. You know what strikes me? I agree with everything you said, by the way. But. No, well, I'll just put some comparisons out there because you've said that 47 times is high for a very large company or not, depending on how much growth you expect. But it's certainly high if the growth isn't as expected. Let me bring you back home for a second. These are much smaller companies with much smaller moats and much less runway. WiseTech is on a PE of 127. Xero is on a PE of 116. REA Group. Yes. You know, the real estate listing company. That one. Yep. Two, sorry, 68. These are blended PEs too. So it's actually got a fork.

59:29So it's a little bit forward looking. So it's already a banking growth that hasn't and may not happen yet. Technology One, PE of 75. Pro Medic is 272. Like, anyway, as I said, things will probably get even crazier from here, but there is a smoking gun there in terms of excess liquidity in the system. But yeah, this is why, and we said this with quantum computing and we've said it before with AI. As techno nerds, we love it. We're here for it. Game on. As an investor, I think both of us have been far more reserved in jumping on the hype train for exactly this kind of reason. Because it is evolving at such rapid rates.

1:00:14And by the way, next time we talk next week, the world could be turned upside down five times. Well, it was Ali Baba AI model released overnight, speaking of it. So again, yeah, you're right. Things will change. That's the point. That's the nursery or the kindy kids or whatever example you want to use. That's literally where we are, right? And it's too, we just, there's a lot of good stories that are out there. And there's a lot of very plausible stories that are, or narratives, I should say, that are sort of, that are out there. But we just don't know at this stage. there's you know sam altman was asked on it recently and you know he's he's a wonderful salesman i'm yet to work out how legit he is um but gosh he tells a good story and that's the kind of person you want when you're trying to raise you know god ungodly amounts of of money but he's basically sort of saying well isn't that cute uh we've got something under wraps here it's gonna blow your socks well you know maybe he's maybe he's full of hot air or maybe he's not right And maybe we're going, oh, wow, DeepSeek, it's almost as good as the current, you know, ChatGPT model.

1:01:15And next Tuesday, you know, AGI is released and we have a silicon god who controls it. I mean, I'm being hyperbolic, but you know what I mean? Like, that's... Yeah. It's in the realm of possibility. And it'll come as a surprise when it happens. We're not going to get six months of notice on it. It's going to be all of a sudden, hey, guess what we've just done. Or someone at UTS goes, oh, yeah, we did it. And we did it in completely different... We took a turn that no one thought was worthwhile, and it turns out that this was the – that unstopped the bottleneck that led to exponential advancement that others can copy.

1:01:52It just changes everything. So it's sort of – when you're talking about supermarkets, to use your example, there's innovation that's possible, right? There absolutely is. Maybe the robots do some of the sorting, and maybe there's more improved IT systems. Maybe there's better logistics – all of that kind of stuff. But it's not. It's evolution, not revolution. Yeah, exactly. That is the exact way to say it. And whenever you're in a field where things can be turned upside down in a heartbeat, that is exciting from a consumer standpoint. But it is not, to my mind, it's not investable. And the other point I will make, and it's been said by others, is that, and I'm more and more convinced of this, and I think this latest news kind of sort of solidifies this view, is that the AI models themselves are really commodities.

1:02:42The value creation will be at higher levels. It will be how are they used and integrated into existing products. That's where the margins are going to be above commodity type. I think, I think, notwithstanding everything I just said. I'm not even as positive as you. I suspect that it will be a tool like the internet. And the internet doesn't give anyone any advantage. No. I mean, not using it's bad, and we'll talk about one very briefly in a second. But the internet didn't give anyone any advantage. It actually brought the cost of delivery down an absolute truckload. Who gets the benefit? Consumers.

1:03:18We get the benefit. That win is just a massive, massive opportunity. it really gives us a great outcome. And I suspect, if you think about the way AI is used, I don't know for a fact, but I suspect it'll probably end up something like the internet. So I don't even know. I mean, some will use it better than others and some will benefit from it more than others. We mentioned last week, the Amazon.com and the Walmart.com kind of example, right? They're both in internet businesses at some point. It was done better by someone. So it's an inflection point. It's potentially a game changer. is potentially something that can make a massive difference if you get it right and i think that's kind of for me the biggest uh expectation i have is will there be one or two winners from ai as a group maybe i mean maybe it's nvidia maybe it's some other model maybe it's something else uh if you compare that though with the likes of um i'm trying to think of the internet examples but you know if you kind of roll it back to you know who won who won from the internet and at different times it was different companies it was well maybe it's the um the switch makers you know cisco's of the world uh maybe it's maybe it's maybe it's the telcos uh the cable companies in the us and it wasn't those guys where did the value accrue i mean you know because each of those things was to your point commoditized so much so that no one actually made the extra money out of it in fact any benefit kind of flowed right through the value chain and it's a little bit like i think we've said this before it's the it's the the flyers versus the airlines versus the airports?

1:04:47Where did the value accrue? Well, a decent amount to the airports, none to the airlines, and a heap to us because we get to fly for a tenth of the price. And that in and of itself is a really, really, really big example. So yeah, I take your point. I think the opportunity for success is there. Whether that's enough for more than that, I guess that's the open question. And look, the other thing I will say is despite all that, despite all the expectations, despite all the positivity, despite all the I think, I wish, I hope, I believe, we don't know. So back to the kidney thing, which is if you – are you going to go into the nursery?

1:05:22I'll use your example. It's a better one. Are you going to go to the nursery and say, I'm going to bet on that kid being the NBA basketballer or am I going to bet on that kid being the Ducks of the year? I'm going to – might you be roughly right? Yeah. Is there a chance you're very wrong? Yeah, particularly if you're making a single bet. So I think knowing what you're investing in matters, I think if you think NVIDIA is a winner forever, then ignore the 17 % drop. The shares shouldn't have dropped if you held them for the right reasons, right? But to imagine that there was some future that was turned on its head by almost one fifth of the entire value of the company because of one announcement by one company somewhere, either it mattered and you shouldn't have paid as much in the first place or it doesn't matter and you shouldn't be freaking out about the share price drop.

1:06:05And knowing which one of those is which is always why it's hard. And sometimes that's why things go in the too hard pile. now by the way listeners i'm going to choose to take this personally uh in the middle of that rant andrew page disappeared he said he's around he's okay he's around i've spoken to him his internet dropped out which if you've listened for the past couple of weeks is uh may portend something we're not exactly sure what uh but yeah he's disappeared so we're all at the end of the podcast anyway we're going to talk about mosaic brands uh which was the uh katies and noni b uh they're finally given up The administrator just said, we can't sell this thing.

1:06:39We're going to liquidate the lot. It's a story of retail. I mentioned kind of online, and I was going to get to that point of who uses the internet better, for example. A great example of needing to adapt and stay up to date. The companies that don't use AI or don't use it well will be at a massive disadvantage. Whether that means the AI companies themselves do well, that's a different question. But if there is a technology out there that other people are using and benefiting from, you better be careful. It's not the only problem that Mosaic had. They had issues with, you know, as brands, they weren't relevant.

1:07:09Fashion is a stupidly difficult business to do. So, yeah, we're going to talk about that a little bit, but we probably got to over an hour, so I think we're probably done. I might wrap this one up here. On Andrew's behalf, I will say cheers, because that's what Andrew says at the end of this podcast. We will get him back, I hope, by hook or by crook. I am going to leave you on a cliffhanger. Maybe we'll be back on Sunday. Maybe we won't. Pretty good chance we'll find a way. But if we don't, we'll talk to you next Friday. Otherwise, we won't know what Andrew's been doing to keep myself fit on the weekends.

1:07:36Let's try and find that out. If we do rejoin you, thanks for listening. We hope you've enjoyed this particular one. I'll get him back somehow. I'll give him a call now. All right. Have a great weekend. Fool on. The Motley Fool and people appearing in this program may have positions in the companies mentioned. General advice only. Please speak to your financial professional to understand how it may pertain to your situation. Subscribe to the free newsletter at fool.com.au forward slash listener. The Motley Fool operates under Financial Services Licence 400691.