Ask Us Anything: Is AI changing productivity expectations in business? + Are whole body MRIs getting more popular + Wes Maas investing in AI

0:00I'm Adam Schwab. I'm Adir Shifman. And this is The Contrarians with Adam and Adir.

0:08And we are back. Episode 178, Ask Us Anything, where you ask the questions and we will answer them. Ask us questions on business, big tech, whatever floats your boat. Mike, question number one. Hi guys. Question one this week comes in from Courtney Askew. Courtney asks, I'd be interested to hear your views on the preventative health screening space. In radiology specifically, do you expect whole body MRIs to become more popular and does broad non-targeted screening without clear clinical questions increase the risk of error and liability for providers? Amazing question. I think you're the expert doctor on the panel.

0:49I'll leave it this one to you. Such a great question. And I'll answer in a self-hypocritical manner. I'll do split personality answer. So the doctor answer to this would say you shouldn't go and get – so let's talk – so a whole body MRI, I'll just summarize this very briefly. So it's not an x-ray. It doesn't use like light waves. Basically it's agitating particles, magnetic resonance, and it gives you an amazing amount of detail. to look at slices through the body, much higher than a CAT scan, CT scan, which does use x-rays. So MRI is the gold standard of all these types of diagnosis. For seeing extreme detail, that's right.

1:33And so you're going to see all sorts of things when you do an MRI. And we should understand that I think probably like when a car comes, a Toyota comes off the production line, the Toyota next to it is essentially identical inside to the one that came off before. Absolute consistency. The human body is not that at all. There are every one of us has variations on a theme, even if we're so-called normal. And that could just be weird things and anatomically that are unusual, but don't have any effect. And we never know about. And so what I would have learned in medical school, and I think the general disposition of doctors is you shouldn't go and do unnecessary investigations because one is it's a waste of money and time and two is you'll find some of these variations of normal and then you need to decide what to do about them and what you'll probably do about them if you've got the money to do these expensive tests in the first place is you'll go and investigate them further and they'll almost certainly be nothing.

2:40And the consequences of doing that are that it will cause a very big spike in the healthcare spend and chew up resources and send you on a wild goose chase and maybe you'll even go and start doing things that carry a little bit of risk because lots of things carry a little bit of risk, right? And so that would be the medical view of this type of thing. now I would love that you that was my view for a long time but as I've gotten older my view has shifted and one of the reasons it's shifted it's especially shifted for my ironically my preparedness to do this stuff to my kids versus myself so there are some things that kill people suddenly there are some things that grow silently and then when you notice them it's too late and they're going to kill you as well.

3:31And those things you will pick up randomly by one of these screens before you've got any symptoms. And sometimes you hear about these stories. For example, a guy that plays footy, this is a true story, gets a head knock, goes to the hospital, gets an MRI of his head to make sure there's no bleed and they find a brain tumour in there in the early stages that they just would not have picked up before symptoms. And he ends up having brain surgery and surviving something that would have killed him if it wasn't picked up. So you hear those stories, they're rare. And so the flip side is, maybe you do it because you can pick up these things.

4:09And what I believe is this, we are going to increasingly see, this is called preventative medicine. I'm going to use that term, even though I think it might not be the right term. Maybe we can call it preemptive medicine. But you're going to see an increase in preventative medicine because its availability is increasing, the price is diminishing, and also this generation that I always joke about is prevent to do anything to try and live forever. They love this kind of stuff. And I think what's more important than saying we should or we shouldn't do it is just like when you make an acquisition for a company, you write down before you buy the company what you're expecting from the outcome on what you're going to do in different circumstances.

4:51I think before you get this preemptive MRI, you should sit with a doctor or some medical professional and say, this is how I'm going to treat different outcomes. For example, if I get something and people are 95 % sure it's a variation of normal and investigating it carries some risk, make a decision before it happens about what you're going to do because after you find it, you're going to be emotionally influenced by the fact that there's this weird thing. but you will find weird stuff like there's a thing in your brain you can get called an arteriovenous malformation that can end up bleeding at some point and like it's incredibly rare but how do I feel like I think about it more with regards to my kids because of how paranoid I am about something happening to them than myself like I don't really care if I die to be honest I don't say that flippantly like it's not really my problem if I die right like it would be great for the pod but i know like you might have to bring mike in for more commentary but like but like i'm i don't wander around thinking i want to live forever i'm terrified of dying like i'd rather not die because it won't be great for the people around me but i am terrified of something happening to my family and so that's more the context in which i think about it are you going to do it i would probably consider it i was actually speaking to mark mcconnell our friend yesterday and he's got he's invested in an interesting business that can do pancreatic cancer blood markers and they're getting to the point where it's about to be commercialized so something like that well that is he involved in that business yeah i mean you know they've already got that for some other diseases yeah i think um that that is and that will if that works and it seems like it does work that will be completely revolutionary because pancreatic cancer basically kills most people because it's found within a year so that's so i think things like that i'm 100 % down for the full body MRI.

6:36It's the false positives, obviously, that you're talking about that's the concern. But that's screening. Like you're saying, blood screening. We already do tons of that. Like when you turn, I don't know how old you are, 50? Like as in you're not 50, but when one turns 50, I think you do faecal occult blood to check for bowel cancer. I prefer getting scopes done, but there is a risk with the scope, by the way. I think it's one in 100 ,000 of a bowel perforation. So yeah, I think adding that to the screening, I've got no issue with that because like the consequences of a false positive are that you'll go for an MRI of your abdomen but it's different to and then they'll say yes or no right it's different to an MRI what what what will what will determine whether or not you go for this preemptive MRI probably getting around to it I think I probably will at some point but uh it's not something I it's not something I wouldn't do it's just something I haven't got around to doing more so moreover are you worried about anything or you're just doing it because like oh i just want to know if there's something in there yeah probably the latter and have you thought about what you'll do if they say oh this is a bit unusual maybe we should like cut this thing out i guess you'll deal with it when i guess depends on how equivocal the surgeon or doctor is on on what it is and what sort of obviously if you're using it like a pancreatic cancer as extreme example that's that kills you within a year you want to act on that ASAP versus something that has a 5%, like an aneurysm, for example, which an aneurysm tends to have a 1 % chance of killing you per year.

8:04So it comes down to a judgment call. Do you get the aneurysm clipped to call straight away or do you let it sort of continue on like most people or treating it like prostate cancer, which you may not do anything about if you're 75 because most people end up dying with it anyway. So I think it just depends on what the prognosis is. I think in the end, I'll capitulate and get it done. I don't say that. I say that a bit shyly as someone who comes out of a medical profession because I know that there'll be people that think that's the wrong advice to be giving people but I feel like this is the problem so now that it's available you've got a choice and like before you didn't have a choice you couldn't get the MRI done like no one's going to do like there was not enough machines now you've got a choice and so how are you going to feel if you're dying of something that you would have picked up on if you would have got this MRI scan that is how people are thinking that's how I think right?

8:53But I do think this obsession with personal health is just another, I don't know, part of the narcissism that's taken over the world, that part of it's social media, part of it's like just the fixation on self, part of it even bleeds into the don't hurt my feelings movement. Like I think that there is this unbelievable fixation on me, me, me that exists today. And I think this is largely an extension of that as well. What about something like, I've got these done regularly, so those Inverted Commons executive health checks. Epworth used to do one. Now there's a few companies that do it where they don't give you an MRI, but they give you a bunch of different tests.

9:32You do an echo on a treadmill. You can get ultrasound. You can get a bunch of different stuff. It isn't quite as full on as an MRI. What do you feel about that kind of level of confidence? Well, I tell you how that differs. Every one of those tests that they're doing is looking for something that is correlated to an adverse health outcome. It's scientifically correlated. So for example, an echo that is looking at the anatomy of your heart. And so you know that if there's an issue with the anatomy of your heart, like do they need to do something about that, for example? And so when you go on an exercise stress test, they know that if you struggle to perform in a particular type of way, that is linked to an increased risk of cardiac disease and sudden death.

10:16So in all cases, they're looking for something quite specific and screening you against something quite specific. And if you were 25, they wouldn't be getting you to do these things because the data is just not supportive of needing to do that kind of screening. I think there is a difference between this preemptive scan, which is apropos of nothing, I'm just getting older and I'm a bit worried and I want to know what's going on, versus as I get older, my risk factors for particular diseases are increasing. So let's measure correlates to those risks and adverse outcomes to see whether I need intervention.

10:53I don't see those as the same thing. Fair enough, Michael. Question number two. Question two this week from Philip Berryman comes in from LinkedIn. Is the use of AI changing the expectations of productivity in your business? Have we seen a tipping point where employees are expected to be using AI in certain areas? If so, how does this differ between different functions? Thank you, Phil. Well, we can say yes, yes to this question. Has AI changed your expectation of efficiency? Yes. Do your employees have to use AI? Yes. There's those two questions, especially if they're in tech development, but even if they're not, I mean, my God, if someone is going to work every day and they're not engaging with AI in sensible ways, then I want that company to be my competitor.

11:44Like that is unbelievably good news for me. And so then it's like how are different areas, how do different areas have different expectations? I think, you know, we have talked about coding 100 times. If you're a developer and you're not using AI, then I don't know what you're doing with your life. Like you better change quickly or, I don't know, learn how to lay bricks, I guess. But then this is the other thing that I found. I'm going to say something. You tell me if you think I'm... maniac for saying this line okay yes okay you mean the previous line i think i'm like significantly better at writing than what ai outputs for me so when i ask ai to write something almost always i'm significantly editing that writing myself because i think it sounds clunky and AI-ish.

12:37And also, what shocks me is the grammatical structure of what's coming out of these large language models is dreadful a lot of the time. And so I thought copywriting would just be wiped out by AI, but the copywriting is pretty bad, to be honest. Yeah, I agree. I don't think you can use AI for copywriting. Yeah, it's shockingly bad. Certainly not writing articles. I think that's probably a little bit harsh. I think it's definitely suboptimal. It just doesn't come out. I've tried writing articles with it and it's I think it's good for giving you some ideas yes and it's like it's effectively like another web search I went to Dubai recently wrote a couple articles for our magazine and I'll use AI to give me some ideas on what to write about but I think it's very different between and this is why travel writers are such great writers the real travel writers and then the great travel writers that they they can capture the emotion of a destination in a way that AI can never do and when I write so travel pieces I try and incorporate a business element to it and AI can't do that either.

13:34So I think it's fine for some elements of copywriting, but generally it's not going to be replacing copywriting. I think everybody thought copywriters go on day one. As it turns out, they're not. But I just can't believe how straightforward grammar is and how bad the large language models are. And like the number of times I'll say, I needed just a quick summary for this thing, like two lines, and something will be produced. and my prompt response will be, that is dreadful grammar. Fix the grammar and rewrite that and then it will generally make it better. So I think that's problematic. Strategy, so are you talking to AI models about strategic thoughts and planning, brainstorming with them or not?

14:17Not really. I'm using it for my board packs a bit, so I guess there's an element of strategy there, but I don't love it for that. I think it's a bit unintelligent sometimes. So I'll often say, I should say, there is a danger in feeding your information into a public large language model. Yeah. So you can make your own choices about that. But I'll often give a hypothetical to a model that I'm thinking about and I'll change, you know, like I won't make sure people aren't identified or companies aren't identified, but I'll try and give it a hypothetical. And then I'll say, tell me why this is wrong.

14:50And then I'll tell it the hypothetical. and I'll use that to challenge my thinking. Now, some of the stuff that comes back, because this is just a pattern recognition engine ultimately and so it's looking for patterns and it can find deeper patterns than we can find. It's got access to much more information and processes in a different way. A lot of the time what it comes back as objections are a bit dumb. Sometimes they're pretty smart and I thought, oh, that's an interesting way to think about that. And so I will use it for a devil's advocate as I'm trying to think through ideas. So I will use it like that.

15:24I think I'm still working out the right way to use it. I mean, I use it very extensively and I've used it for a long time. Coding is the most obvious way that it's useful. It's also useful for general workflows, but it's the more abstract ways. It definitely can't abstract think, but using it to support you in abstract thinking, I think is a really interesting emerging way of using it. Yeah, I think operationally is the obvious one where, and I think there's a lot of people talk about this, the thing that AI is the best for is where it's high repetition, low cognitive load. So when you've got those kind of roles, it's great.

15:54And so we use it for our team built a great fraud engine using AI. So we're fortunately not a business that gets a lot of fraud because it's pretty hard to defraud a travel experience because you've never got to rock up at the hotel or the tour. So eventually people realize. So the only time you can really do is short dated stuff before we get the charge back. So you can't really defraud something more than a month because we pick it up before the person travels. But it's been great to be able to help. We've had periods where I've had a lot of fraud and we can sort of smash that down. So they've done a great job creating a fraud engine because, again, pattern recognition, high repetition, low cognitive.

16:26Our chatbot obviously does a great job. So we deflect thousands and thousands of tickets with chatbots, getting high repetition. So I think it very much depends on what we're trying to do. Obviously, our developers use it a lot. Our product managers can now act as mini developers in some way, which is super handy. Obviously, they aren't developers, but they can do a lot more than they used to be able to do. So product managers, designers, developers are all kind of merging into one in some way. We still have specialists in each field, but they can sort of cross-pollinate across each other. So there definitely are elements of the business that it's great for, and there's elements of the business it's terrible for.

16:58We haven't had much luck in copywriting so far, which is surprising, but the case, voice AI still is challenging. I think we thought we'd be a lot further down the line on voice AI, certainly for customer service. It's not quite there yet, and I don't think anyone's really quite there yet. So I think there's a lot of areas that people thought by now we'd have sort of down pat with AI that hasn't been. I think the panic in some cases was overwrought and in other cases were underwrought. Clearly, it's made developers a lot more efficient. There's no question. That's been the highlight so far for us.

17:29Well, I think the question to ask yourself, oneself, whenever you're thinking about using AI is, do I think there might be a pattern to this? That's the question to ask. Fraud, there's certainly a pattern to fraud. And like the more data you feed into these things, the better they get. So fraud is great. And then I might say like code. So code is a language. Every language is entirely pattern-based. That is a language. That's how a language works. Creating a new novel or creating a painting or an image that's never been done before, there's no pattern to that. If you're going to create images based on patterns, you're going to get some variation of stuff that's already been done.

18:11Whereas if you're like let's say there's a young author who's a brand new style of voice in the way they write. So there's no pattern to that. So that's not something that's naturally intrinsic to AI. And so maybe I'd say with business, lots and lots of business and economics. Like economics definitely follows patterns. I think we're in the nascent stages of AI as part of the economics decision making. But I think that should become huge. Like economics is all patterns. And so that to me feels like a huge area of opportunity. On that note, we'll go to a super quick break. Back with our last question in a moment.

19:03And we're back. Question number three, Michael. Final question this week comes in from James Frank via email. James says, long-time listener of Contrarians. thanks for your hard work. He also says, I'm enjoying the deep dives with the Warren Buffett of Australia, aka Scott. A question, what do you think of Wes Mars and Mars Holdings selling off of hard construction assets to invest into AI and electrification? Firstly, I just want to correct something in the question. I don't work hard at this podcast at all. Thank you, James. Great question. Just a quick background. Wes Mars is an AFR Rich Lister who, just a couple of weeks ago sold his, that's a publicly listed business, sold his construction part of the business for$1.7 billion to an international business.

19:51And he's led at this ride on AI investment. So that's a really interesting play idea. What do you think of Wes's big bet? So this is a pretty, I mean, I read one AFR article about this and I read it pretty quickly because my eyes were rolling. But basically it's pretty amazing. It's an ex-athlete, right? Who's built this construct. He was a rugby league player who never quite made sort of the really high level. He did play at NRL level, but wasn't a star. But he obviously has had an incredible career as a business person though. And he went and built some construction business with$15 in his pocket or something and then just built out.

20:32Yeah. And it became a$2 billion business listed on the ASX in 2020. So it's been a ridiculous career. And so whatever I say about this, we should start off by saying this guy definitely deserves incredible respect for what he has created with his team in terms of this business. And ultimately, he got cash for this, right? Well, so what happened is, so a publicly listed business was bought by German-based global giant Heidelberg, formerly Heidelberg Cement. The mass construction business had 1140 staff, 40 quarries, 22 concrete plants, two asphalt operations, and about half the overall group. hasn't sold the whole thing.

21:10The shares issued at$2 in 2020 and hit almost$6 last week and then plunged 27 % after this announcement. He's putting$100 million into our friends at Thermos, Oliver Curtis's business, a data center business. All right, we'll get to that. But he sold it for what? Almost$2 billion. He sold these assets in cash. And so he just goes and spends 100 mil on whatever, right? And so the only reason the share price can drop, well, there could only be two reasons. One is people think he sold the asset too cheaply. I don't think that's the case. Or two, people are worried he's going to go and burn the$1.9 billion on AI and have nothing to show for it.

21:48And I think also people invest in a construction business, not an AI business. So you sort of invest in one thing and now you're getting something completely different. And so my view on this is that the long-term future of the construction and property part of AI, which is what this guy is basically doing, wants to do data centers, presumably. Is that what he's all about? He's investing in firmness. Yeah. Okay. So I'm very bearish on the long term of those things. And so I don't think this is a good plan based on my bearishness on that. I think the biggest shock about all of this is that no one would have expected this to drop 27 % when he jumps in and becomes a meme stock.

22:28I mean, he's taken a great construction company, sold an asset for a ton of money, Let's say he sold it for at least as much as it's worth. It's pretty incredible to liberate$2 billion of cash. And now he's tried to jump on the meme stock bandwagon and wiped out 20 % of the company's value. I think that maybe goes to speak to the fact that Australia has not really bought into this. That's kind of what you said a few weeks ago, that you don't think Australia's buying into this meme stock thing? He is. He inputted 100 million firmness. I remain a huge bear on firmness. but I think it will take a while for that to play out and I think in the short term, his$100 million might be valued up to$500 million or a billion dollars.

23:10Who knows before the chickens come home to us? Well, he's got a 1.7 % stake. So that values the business at almost$6 billion. It doesn't actually have any data centers yet. So it makes Iron look like a, looks like Berkshire Hathaway, this thing. So it's, I'm not sure how. And somehow Firmus has this cooling technology that CoreWeave, which is a real, I want to use the word business loosely, but a real enterprise. They haven't figured out this cooling technology, but somehow Firmus has. So if that's true, then Australia has become the most revolutionary, you know, tertiary or quaternary industrial player in the world.

23:50If you believe that, I've got some shorts I can sell you for$10 ,000 each. There's literally zero chance these guys. That could be true, but you know what? Once in a while, everybody ridicules incredible innovation and it turns out to be true. I just find it hard to believe it's true in this case and that all these other global players haven't figured out the same thing. Yeah. It's completely unthinkable. Do you think this is correct? That they've got some revolutionary cooling technology? Well, yeah, it's almost – it's 99 % implausible. 1 % or less could be true. Less. Maybe 0.01%. Yeah, but not zero.

24:31So where there's life, there's hope. Yeah, well, nothing's ever zero, but as close to zero as one can possibly get, you're bumping right up against the zero. The chances of me investing in firmness is zero. But I will say this. Why do you think he has done this? Do you think it's opportunistic and he's tried to jump onto the meme stock band, or do you think he fundamentally believes that this is the future of the world. I think it's probably the latter. I think he's obviously had some incredible success, whereas I'm just not sure. It's only$100 million. In the end of the day, it's a fraction of what he actually sold this part of the business for.

25:09Do you know what percentage he owns of that company? I think it's 50%. All right, so he should just pay a$1 billion dividend out of the one point whatever he sold, pocket$500 million of cash, and then who cares? Then if you want to invest in that as a public company investor, You've got this massive dividend and then you can decide if you want to stay in. I mean, I think the market's voted about whether they want to stay in the first instance. But like I don't really have a strong view about this. I think the guy can do whatever he wants. It's his company. You can back it or not back it. Well, that's the thing.

25:39When you buy a company that's 50 % owned by someone, you're taking that founder risk. You can get some incredible returns like Wes has generated for his investors. And you can also be sort of subject to this kind of madness. So I think you get the good and the bad. It's certainly a big bet he's taking. It's a bet, in fairness, he's taking on his own dime. So I've got no issue with him doing it. It's his right to do it. And I don't think investors can sort of throw their hands up in the air because they sort of got what they get. It's like someone who invests in satire and can play the share price is down 90%.

26:08Well, you kind of made your bed, guys. Time's asleep in it. Or someone invests in Harvey Norman, and then they're always arguing about Jerry Harvey. And it's like, listen, if you don't want it, don't buy the stock. You know what the deal is. He's going to buy all sorts of dairies and this and that because that's what he's going to do. and like you can like it or not but no one's holding there. I just, I cannot believe how much this is starting to feel like the dot-com boom. Starting? It is the dot-com boom. But I know it's so, this vibe is so reminiscent of the dot-com boom of like there are just these businesses and they don't do anything yet but they're the promise to do something.

26:44And like, you know, the words that you can hear, if you listen carefully, what you can hear is this time it's different. and like those are the words that are just the bubble of all. That's the words of a bubble, right? We've got a living environment where the US market, Australia not quite but not far off, the US market hit record highs a couple of weeks ago, like absolute records. You have got unemployment creeping up. You've got kids who can't get jobs. You've got inflation at risk. You've really got a record market. None of this makes sense. You've got terrorists about to smash the market.

27:17You've got immigration in the US down to zero, so you're not going to get GDP growth because there's no one coming in. So you look at all these underlying factors and everything points to bad and you look at the markets and almost every market's at close to record highs. I feel – I've said this before. I find the whole thing terrifying. I'm not sure we'll talk about this at some point, but like Australia's in bad shape. Like this economy in this country is in very bad shape and it doesn't feel like it's in bad shape. I look at you, Mike. You seem pretty happy. You got good stuff. I'm about to refer you another client.

27:54Like it feels great, but this country is in really bad shape economically and I would say socially as well. So what's this company worth, by the way? So it's currently worth$1.5 billion. It's sort of back to levels of a couple of years ago. How can it be worth that if it just sold something for$1.7 billion? That is a very good question. That's what the market's valuing. Is it debt? There probably isn't debt in there or something. It hit a high of 581 in January. It's now down to 416, but that was a really high high. It was very peaky. So it's only said it had been a level once. And it's not very different.