The Canadian Investor - The $1 Trillion AI Spending Boom & the Race to Make It Profitable with Braden Dennis

Episode Date: October 3, 2026

The AI spending boom is reshaping much more than the technology sector. In this episode of the Canadian Macro Show, Simon, Dan and Brayden discuss why massive AI infrastructure spending is increasingly competing for capital, how that could be contributing to higher bond yields and borrowing costs, and why hyperscalers are increasingly turning to debt to fund the buildout. They also look at how dramatically the business models of companies like Microsoft, Alphabet, Amazon, Meta and Oracle have changed as capital expenditures surge.  They then debate the biggest unanswered question in AI: where do the profits ultimately come from? The conversation covers consumer adoption, AI agents, enterprise use cases, advertising, Chinese competition and whether today's spending can eventually generate adequate returns—or whether parts of the boom are headed for a major reset.  Plus, they discuss Canada's potential advantage from abundant energy and natural resources, what rising bond yields could mean for mortgages and housing, and why Oracle may be one of the more vulnerable players if the AI buildout runs into physical or financial constraints. Watch the full video on Our New Youtube Channel! Check out our portfolio by going to Jointci.com Our Website Canadian Investor Podcast Network Twitter: @cdn_investing Simon’s twitter: @Fiat_Iceberg Braden’s twitter: @BradoCapital Dan’s Twitter: @stocktrades_ca Want to learn more about Real Estate Investing? Check out the Canadian Real Estate Investor Podcast! Apple Podcast - The Canadian Real Estate Investor  Spotify - The Canadian Real Estate Investor  Web player - The Canadian Real Estate Investor Asset Allocation ETFs | BMO Global Asset Management Sign up for Fiscal.ai for free to get easy access to global stock coverage and powerful AI investing tools. Register for EQ Bank, the seamless digital banking experience with better rates and no nonsense.See omnystudio.com/listener for privacy information.

If you're a DIY investor ready to take control of your portfolio, BMO all-in-one ETFs simplify the process. Whether you're new to investing or looking to streamline your existing holdings, these all-in-one solutions are designed to help you invest smarter. With management fees on popular portfolios now reduced to 0.15%, you get professional diversification at a lower cost. Check out the asset allocation ETFs at BMO ETFs at BMO ETF. When have you seen such speculative, chasing, anthropic growth speculation, plus like, hey, I can make pretty good risk-free rate at the same time.

It feels like the consumers on life support already in both Canada and the US. And in Canada, that factor, obviously pushing bond yields up. You know, that renewal wall that was looking pretty soft recently is now looking pretty violent for a lot of people. I think one of the issues, One of the issues I have with the AI trade is you listen to those conference call, and they do not give you an answer on where they think RYC is going.

They use very vague terms. I think they don't know. I think they think it will be profitable. Still today and through the rest of 26. No product market fit for personal agents. Zero. The product market fit today is still coding. All right.

We are back. for the Canadian Macro show. You're going to be talking a lot about AI today. We're joined by a friendly face, the Chief Vibs Officer. Is that what that says right there? Sure does. Incredible.

Yeah. I was going to say the AI chief, but... I didn't know what to put my title in on this new platform. Chief Vives is good, yeah. Chief Fives feels about right. Braden, it's been a while. You said you had to dust off your mic there for a sec? I moved to offices in August,

so I had to go into a box, dust off the mic, make sure all the chords are working, but it's good to be back, boys. I haven't been on the show in a while. And Simone, we're recording another episode in two weeks on the original feed as well, too.

So I'm back. I'm not going anywhere, you can't. No, he's married, wedding in the rear view mirror, so now he's got all the time in the world, right? Yeah, yeah, yeah, exactly. So what are we talking about today, boys?

I think you own the agenda. Yeah, so you want to go over it then? Yeah, you're the, You're the second AI chief. Yeah, I think we're going to talk a lot about the AI trade. I mean, there's a lot of talk of a bubble. I feel like people aren't prepared enough for it to not be a bubble, by the way. Like, what if it's, what if it works?

And so I want to talk a little bit about that. I think maybe before we get into the AI stuff, I know you guys wanted to talk a little bit about the tenure. And I think that it ties in relatively well what's happening in the bond markets, because, you know, it looks like the AI CAPEX financing is very derivative of the costs that are associated with, you know, with the risk-free rate. And so I think that it would be, you know, remiss not to start maybe with what's happening in the capital cost environment.

And then we'll, we'll jump more into the AI stuff. But AI, I think we're going to talk a little bit about like, is it a bubble? And then, you know, like, I think a lot of the agentic personal assistant market that's been been created as a result of these companies realizing that that might actually be one of the most efficient ways to drive consumer AI spend, which I agree with them. I just don't know if the numbers are going to be as big as anyone might hope. So we'll get into that. Plus, you know, the actual smart people on the call.

We'll talk a little bit about the financing systems and stuff and how these guys actually make money and raise money and where the risk might lie for those businesses. But Sima, you already got the chart up here. Everybody's favorite chart to look at in the last couple of the days. So what's going on in Bond world? It's kind of, it's interesting having Braden here. Because like, you know, in the past,

I feel like you weren't that interested in the tenure. And you're the one that brought it up in the chat. I was like, guys, we need to talk about the tenure. You know it's serious now that even Braden's interested in the tenure. Well, I'll just make one comment and then I want to hear from Simone.

But the, I'm not a macro guy. Don't care. But, but, Warren Buffett over here. When have you seen such speculative chasing and thrice?

anthropic growth speculation plus like, hey, I can make a pretty good risk-free rate at the same time. I don't know if I've seen that personally in a long time. So what happens in the next three months is the really interesting question with those asset prices. Yeah. And I think, and I'm also curious to hear from Samal because I feel like he says these a lot better, these things a lot better than me, but is it like at what point does this become breakage? Like you saw a run up in bonds in the 90s recession. And I'm not trying to be the Dumer guy, but like I just think it's it's kind of unignorable that this is a bit of a problematic situation for anybody borrowing money.

Well, Sima, can you zoom that chart out to 10 years? Sure. Yeah. There you go. Yeah. So it's the highest it's been since I think if we go all time, you have to go back. you have to go back to early 2000s to get the tenure at that rate. Yeah, around that.

I think I saw something. It was like 2003 the last time I was at that right. But let's just say mid-2000s. So 2005 around that time was pretty close to 5.3. That I think it hit last night. And now it's gone down a little bit on softer inflation reading for PCU, which is the Fed that is their preferred metric in the U.S. So I think that's part of it.

But still, I mean, it's run up. And it's not a big drop. The two euros drop more, and the two euros tends to be a reflection more of where the market things the Fed is going, kind of short to shorter term. That one has dropped a bit more. But I think this is likely a combination. And the Fed and the Bank of Canada said it, like they are competing with financing from the AI buildup. Like that is one part that they're attributing to higher yields.

Obviously, I think there's the other part where, you know, people lending money to the U.S. government are looking at this and they're like, we're not stupid. Your debt is $40 trillion. Like, when will it reach 50? Like, you have these large institutions who are just looking at this and saying, like, you know what? We need to be compensated if we're lending you money over 10 or 30 years. Because we don't think you're going to get that spending under control. And I agree with them. Like, you know, wasn't there? And I can go buy Microsoft corporate paper.

Yeah, you can go and buy those hyperscalers. And I had like an interesting article I shared with you guys. Like Pimco had a really good article on this and I'll just share this here. But what they said in their article, essentially you're seeing a little bit more, you're obviously seeing more issuance from hypers because now they're no longer relying on cash flow and I think had that in your slides right so they're using debt to finance the AI build out as well but what the kind of disprove is that there's still demand for that boring utility that is very profitable low risk and one thing that they were trying to show is just like look yes

spreads are rising for AI in general and you're seeing this here so you're seeing like especially the slope so versus the 10 and 30 years. So you're seeing market like investors into institutional investors saying, okay, like if we're going to lend to hypers, longer term, we want to be more compensated. But they said it's more a reflection of supply, right? There's more supply of this debt coming on the market. But it's not necessarily affecting those boring utilities. That's the other thing that's interesting is that you're not seeing that supply affect other parts of the market. So I thought that was interesting to see because I think a lot of investors are seeing what's there is that yes, they may want some AI exposure, but they also probably realize

that there is some inherent risk with that, some uncertainty. And those boring credit investment great businesses are not necessarily feeling the pain just now. Because I've heard people saying that it's going to make financing costs more expensive for the rest of the market. And so far you're not seeing that, at least on the corporate side, you can make the argument otherwise for the U.S. government, but there's, I think, different forces at play there. Yeah, there was a question in the comments there. Is copper and electricity not the cleanest AI play? And I think, you know, to add to what you're saying about the boring businesses, I think

that there's an argument to be made that, you know, there are some unsexy ways to get exposure that are way less, that feel way less risky. And I think that like I've been, I'm not a stocks guy, right? But like I love the ring of fire. I have real estate exposure in the ring of fire in Ontario. And I think when I look at, like I have been playing a little bit in, you know, I obviously max out all my registered accounts and I have, I can't really do real estate deals with that capital in a functionally effective way.

So I'm like, I like copper. I like electricity. been a big Enbridge guy, but obviously that's a little bit different. But I think that there's a lot of derivative plays that excite me actually more than the, I like all these AI companies. Like honestly, as somebody who works in the space and uses this stuff daily, like it is life-changing technology. I just don't know if it's priced appropriately based on how much their money,

they're incinerating to make their product financially accessible to the consumer. But any thoughts on that, Brato? Well, consumers and businesses will greatly benefit from this very intense competition that's happening right now among these players. And, you know, they've been, the capital allocators that these companies have been indicated from late stage private

that you have an infinite piggy bank. Go win this thing. And, you know, and so the guys with actual piggy banks are willing to go into negative free cash to fund it. And then the ones that don't are willing to get late stage private forever. So we'll see how that plays out. You know, of course, the most obvious thing I'll say on the entire call, which is groundbreaking to no one,

is the next 10 years will have more demand for compute and more demand for electrons. The problem is, bringing electrons online is really slow. I used to work in it in power generation utility builds. Like how long it takes to bring on new power gen, it's insane. Like, let alone bring on a new nuclear license. Even if I can stand up a quick SMR, it doesn't matter. I got to wait for my nuclear license to come in.

That doesn't happen very quick. So, yeah, it's not very quick to just bring on electrons. doesn't that make Canada like actually very well positioned from a from a market perspective for this AI KAPX like you know we are starting to see some more data center project I think the biggest private investment in Canadian history was announced with the bell thing in Saskatchewan but like we have a ton of hydroelectric you have like flaring natural gas like they literally like unless I'm mistaken they literally light natural gas just to get rid of the excess natural gas that they could otherwise use to generate electricity in in Alberta right so could could we could.

we not and we have cold climate i don't know if that way like these these things get hot and they need to canada was born on third base for this for this trade for everything yeah okay that's true yeah that's a good point but especially this trade i mean it it's like it's like a double has been hit and i'm still and i'm still standing on third base over the last 10 years But like a double's been hit, just walk to the plate, man. Like, that's, yes, absolutely, Dan, absolutely. Having cash on hand is essential for any business. Traditional business accounts hit you with high fees

while paying little to no interest on the cash you need for day-to-day operations. That was our experience too, until we switched to the new EQ Bank business account. Now, every dollar earns high interest with no monthly. fees and no minimum balance. You also get free everyday transactions like EFTs, bill payments, mobile check deposits, and 50 outgoing and 100 incoming free interackey transfers. And to sign up, quick and fully online, no branch visits because, let's be honest, no business owner has time for that. We use it for our own business and it's the first account that actually helps our money work harder while keeping operations simple. Check it out.

today at EQBank.ca slash business. I've been thinking about doing a Quebec city trip where I bring my mountain bike and the family tags along. I could spend part of the day riding while my wife and my daughter explore the city, walk around the old streets, stop for ice cream, and burn that energy in a beautiful park. Then at night, we'd all meet back up, go out for dinner, and make it a family trip. And while we're doing that, our home in Ottawa would just be sitting empty. That's when it hit me. Listing our home on Airbnb just makes sense. Someone else could enjoy our place while we're away, whether they're here for a summer weekend, visiting family, or coming in for a big event in this city. They could explore the Rideo Canal, take the kids to the water park, catch a CFL game downtown, or just use our place as a home base to experience Ottawa like a local. The extra income could go right back toward our own trip, and we can choose to host a only when it works for us.

Your home might be worth more than you think. Find out how much at Airbnb.ca slash host. So are you confident that we'll be capable of capitalizing on this to the scale that we should and could? And it doesn't have to be, it's not a political question, but like what would it take for you to feel that way also? Oof.

I mean, that's, I guess that's what's the bets that they're making right now and we'll see about execution, right? It's all about execution from here. We can have some really exciting meetings in downtown Toronto and get important people at the table. And I think that that's a start. But let's see how the next six, 12 months goes. Right.

Yeah. Simul, any thoughts there? I'm pretty positive when it comes to that. I think a bit like Braden, I think Canada is very well positioned there. And I think as much uncertainty we're seeing on the global scene right now, I think one thing that is it's doing, it's opening a lot of eyes for Canadians in general. You're seeing a lot more support for controversial projects now that you have in the past. I think people are just realizing that sure, in the grand scheme of things, whether it's pushing the ESG agenda or whatever it is, people are like realizing, okay, like that's nice, but we also need to be like prepared if things.

you know, don't go as plan if the world is changing. And I think, I mean, you'd have to put in, you know, be living under Iraq to not realize that the world is very different today than it was 10 years ago. Yeah. And I think it will continue. I think we will continue to have to expect the unexpected. I mean, with a thought in March or like February or March,

within the span of what, like a month and a half, two months, the U.S. went and captured the head of state in Venezuela, then turned around and started a war in Iran with Israel. Like, would you, and now looking back and were still in a mess in the Middle East, like, who would have thought that just a year ago? Probably not a lot of people had that on their bingo bingo card. So I think it's just, I think, you know, the positive from that for Canada from a standpoint is just that I think it's making us realize that we have a really. big advantage and I think it's just creating support for that. Before we talk about CAPEX, Dan, what do you, what's your read through on this run in the treasury on mortgage rates next 12 months? Well, like the U.S. housing market, so my actual read is that

something's going to break and that we're probably more likely in a lower rate environment, maybe not in 12 months, but I mean, even even as a, early as 12 months. Like it feels like this just like you have so many factors that are going to break the consumer, you know, cost, push inflation from oil. You know, Simone and I've been talking a lot about like Costco earnings, Walmart earnings, and it feels like the consumer's on life support already in both Canada and the U.S. And in Canada, you know, that factor obviously pushing bond yields up, a lot of, you know, that renewal wall that was looking pretty soft recently is now looking pretty violent for a lot of people or more violent than it did a couple of

a month ago. You're seeing consumers continue to pile into variable rates instead of fixed to try, because pricing's cheaper and maybe try and capture some downside in rates. So like let's say the consumers already, it's just, it's a ticking time bomb, right? It's just a matter of time. How long can, can we, you know, keep going at this current pace? But then you look at the US and like, what is it, 50% of consumption is the top 20% of people. So you still have like, you know, mathematically, if the rich keep getting richer, the balance could still, you know, you could still technically prop up the consumer economy and your lower 80% it would suck for them, but, you know, they don't really have an opportunity to fix it on an administrative perspective for another

two years, maybe at midterms. So the question becomes, does something break in the, in the, in the, the wealthy, like in the top end of that K-shaped recovery. And, and that's, is this money starting to get sucked out of, of your, you know, your AI CapEx, et cetera. And now that this huge economic engine starts to, to break down a little bit. That's where I think like that, that's going to determine the whole path here. But all of that to say, the U.S. housing market was already on life support. Like, you know, they're, some of the lowest pending home sales in, since, since Redfin started recording it. Highest number of sellers relative to buyers,

so like the deepest buyers market. Prices aren't really reflective of that. They are in a couple of those like COVID-darling markets, like Florida and Texas, but like hasn't really come down a lot. Canada, obviously, we know what's happening in the housing market here. It's not exceptionally good. But we do have,

we still have that shift that people can make and they're visibly doing it into a lower price product to try and kind of like keep things going. But the idea that any housing market recovery and volume, like people are going to start buying more houses at any point and that if that's going to help support residential investment, which is real estate commissions and you know, what you have up on screen here, Home Depot, like, you know, the contractors, etc. Just I don't know what the case is for residential investment to be contributing to the economy

in the next two years regardless of what happens with the bond market. So I think rates are going to suck for a bit and then eventually something's going to break and probably will be in kind of like recession-era stimulus on both the rate side and the spending side. I don't know what your thoughts are. Average ticket has actually been resilient post-COVID. Transactions net net on a comp sales growth look slightly negative and now flat recently. So pretty resilient actually. But inflation adjusted, that tickets down. That ticket down a lot. Yeah.

This is the challenge, right? It's like does the central banks become being comfortable running inflation a little bit hotter than maybe they they ought to keep pushing both the bond market up, but also just keep bleeding out like all businesses on on the fact that they're, you know, if you're marking it to real returns, are we actually really growing or are we just really just seeing inflation kind of push everything up? Yeah. Is this the McDonald's shirt? Yeah, the McDonald's chart. That one has been, and McDonald has been just struggling this year. I think they're having a lot of issues.

Consumers are just finding that it's becoming less and less affordable. And I think people are just starting to make the calculation in their head. They're not just comparing McDonald's with other affordable restaurants or comparing it with eating from home, right? So it's becoming more and more than in comparable. Are they? I mean, not everyone. What do they do otherwise? I don't know.

I mean, what's a good question? What are your options, right? I think they're just having trouble offering that kind of value. I think McDonald's has a... I thought Braden was going to say, OZemPEC, to be honest. No, no. Zempeg problem.

Maybe there's that to like a headwind as well. But McDonald's lost itself as a brand, man. Like, it's not a... Like private equity slop on the restaurants. Like, I know you get so hard to find it. It's so hard to find one with a play place now. Yeah, like the play place is an example. Like, I feel like, you know, you know those images where you see like things looking cool in the 90s?

And then now they just look so sterile in the 2020s. Private equity slop. McDonald's is like the poster child of that, in my opinion. Like, it's just got no, it's just got no aura anymore, man. I don't know. They did. Did you see that thing that they put out about the? a new like see-through restaurant that they're going to release that's like all glass i can't see

that doing well i'll put it up on the screen but i saw on um on twitter it was doing making the rounds there but basically like this is and it has a play place in it as you see that like red box there um oh okay yeah like but i feel like that a lot of this is just like at least that's got a fun factor to it yeah and i mean it it is pretty cool if it like in a glass it and an all glass macdonald but It does create more value though. Like, Brayden, you're not there yet, but Dan, I'm sure you can relate with that. Like, I'd be more interest in going at McDonald's. If I know my daughter, we can spend a couple hours and kill some time.

And she, you know, we just go and have fun and we'll keep an eye on you. And it's, we don't have to pay. A place to go is half the business. Yeah. Yeah. Yeah. Yeah. Man, like, during the pandemic, straight up, like, that was like me and one of my

buddies who I do business with like we would literally just go and it was like the only time we would get be able to get to chat like when you have the kids home all the time in the pandemic and you would just like let them go rip and like now there's like all these small businesses starting in in the suburbs I'm not sure if you have these around you Simone but like where they're they're starting businesses that are like you pay 15 bucks or 20 bucks to go and play in a indoor playground in the winter right yeah yeah we have some we have even like there is not far from my place There's like a coffee shop but made for like essentially parents where you can bring your kid and there's some like a play area. They can go and play.

I mean, parents will like that creates value for me because let's time I'm paying like I'm going to McDonald. We spend 35, 40 bucks for food total. My wife and my daughter. And then she goes to the playground. Well, that $40 is that's it. Whereas if they didn't have that, I might not choose McDonald because we'll still have to go buy some food. somewhere or maybe we'll eat from home and then I'm going to have to pay 25, 30, 40 bucks to go to jungle gym, right? So I'll have to find something else to do and most of the time those things

cost money unless it's a park or something like that in the summer. So I do think that's a, that's a good idea for them to bring that bag because it does bring value for families for sure. When I saw that photo, I was like, I've, I know that. I've been there. And then I read Chicago, my hotel, I stayed in some really bad hotel and it was across the street from that exact McDonald's. Boys, let's talk about some of the change in fundamentals for these hyperscalers, shall we? Yeah.

Okay, I'm going to chart on here. Yeah. Okay, put this together for a presentation last week on, you know, kind of the players behind the seat. Like, where's the money going? That's the question. Like, where's the money going? I know they're spending a ton on infrastructure.

Who's benefiting? And of course, we know in video, we know people are buying GPUs, but what else, right? So five companies, Microsoft Alphabet, Amazon, Meta, Oracle. In fiscal 19, they spent $71 billion in capital expenditures in their financial statements. LTM, trailing 12 months, they have spent 586. billion. So, you know, we're up 8x during that period, really, really steep curve starting in fiscal 24 through through now. And I look at this chart and it looks like this is what will happen in the future. It almost looks like some bold, grandiose estimate. No, these are historical numbers.

This is not forward looking. And so what is forward looking like that, you know, what does that curve ramp up to next year. Well, we're getting a readthrough of that now with quarterly earnings coming out here shortly. 36% of cash flow. So this is Amazon's CAPEX here, 36% to 83%. So it's absurd. It's really a change in their statements at a very severe level. Google's, free cash flow. I mean, can you think of a company that used to just gush free cash? Simone, I asked you two years ago, which company gushes cash flow? What comes to mind? Yeah, Google, Microsoft, yeah, Apple. So, yeah, those asset light companies, maybe in Apple a bit less, but still, they were generating cash flow, their asset light. And they've basically shifted their

business model within a year or two. 100%. Yeah, 100%. And, and, and, you know, there's obviously argument to be made. Like there's there's only three real hyper scalers. How do you catch up? There's way too much capex required. The ramp on Google Cloud is so compelling to run these workloads. It's a pretty good product. I think AWS, GCP have kind of the best products. And so how do you not invest? How do you not invest? And they've made the statement that, hey, we will actually issue equity. We will raise debt and we will quadruple down moving forward. And, you know, where's it going, right? I think Micron

has been the poster boy of the trade in the last 12 months. And for good reason, I mean, have you seen a chart look quite like this? I mean, it's pretty ridiculous. Or have you, have you tried buying a PC recently and getting some memory? Yes, because I buy a lot of computers. It's expensive. I buy a lot of computers, right? Like I have 70 employees now, right? And I buy a lot of computers. The memory trade is obviously there.

You see gross margins from Vivida go. I actually have a slide on this. Where did gross margins go for? Nvidia is the orange line. Gross margins in 2000. That's way too far back. 2005. Let's go sooner.

Okay, gross margins on Nvidia in 2020 was 62%. now it's 75%. That's a very sharp increase in gross margin, right? This is not something that is gyrating constantly. That is a big increase in pricing power. Returns on capital for these guys are just ridiculous. So where's it all going, right? This is, this is why there's been a generational run on the companies that have been beneficiaries of this chart. Yeah. So, is there is there and i'm starting to hear like projections for 2027 that uh capex could be close to one trillion for the the five names that we know yeah and why they're going to keep running it right yeah you think they'll keep running it if it like on the assumption like at what point does

the market say we need to start seeing some what what's the the turn or r i c right on this yeah yeah i I actually charted that on fiscal because their return on invested capital has been going down. So I think you shared with me a chart, Dan. So I just figured I just put it on fiscal.a.i. So here you have, I'm trying to pull the companies. Yeah, there you go. So you have Google, Amazon, Microsoft, Oracle, meta. And look, I think the trend is clear.

If you go back to five years, essentially, you can see for the most part, it's not linear, but for the most part, R.C is going down. And I think one of the issues I have with the AI trade is you listen to those conference calls and they do not give you an answer on where they think RYC is going. They use very vague terms.

I think they don't know. I think they think it will be profitable. They think it, I mean, why would they be spending? Yeah, I mean, why would they be spending that much if they didn't think it would pay dividends down the line? but the reality is they're not sure. You listen to the call. But for Google, it's a return on incremental GCP revenue acceleration.

Yeah. RIC is completely useless in that sort of calculation, right? Like you're going to obviously get a negative return. But if I can get 80% year-over-year GCP rev growth at pretty good and increasing operating margins on that, business, what do you do, right? Like it's, it's got to be the toughest CFO job on earth, right? Yeah. And is it just a long term? Like, like, I guess the challenge, like, the challenge that I see is that there's two, there's two sort of things that that don't work, right? Like, one is these AI companies

might have overestimated how much compute demand, they're actually going to be able to, like, get consumers interested in doing it. And I think TPBN, had that, I think I sent you that video, Samo, which is like, the guy's argument was basically like, consumers don't want to be productive, you know? So like, they don't, like, I thought that was an excellent argument. I'm like people don't real, like people like the buying random crap online and like booking your flights and whatever is actually like leisure time for a lot of people. I'm not one of them, but like in a business context, I love. Ask my wife. Yeah. Yeah. She likes to do shopping. She likes to do it on her own. Yeah. She actually. You're going to get an

AI agent to take away the one 15 minute thing I enjoy every day? Yeah. So, but this is the challenge, right? It's like, so now we have that we're in this era of, of personal AI.

So, so we have problem number one is that what you guys described that. There's not actually a return on all of the compute that's actually being processed anyways. And then number two is, is there, I think that there is a pretty significant upside to, as people start to learn how to integrate AI into their lives in ways that's like,

actually enjoyable and media. and gets rid of stuff that they don't want to do, especially in a business context, that'll drive a lot of productivity. Like in my industry, I would say one of the highest adoption industries of AI in real estate and people aren't really using it right. But now that we're starting to see them use it right, they, it's really meaningfully changing their lives. And I think as small businesses start to adopt that, which I think muse is actually a really,

really nice like surface area for that because most small businesses, their digital storefront is Instagram, right? it's not a website anymore. And they're using one of their primary methods of contact is WhatsApp, right? So you now have this AI agent that plugs into both of those things. I think that to me, the bullcase for AI, like way more token spending consumption is automation of business processes. Very, like many businesses, most of the businesses that are on Instagram as their digital

storefront, very simple businesses, right? Being a realtor, moving paperwork from place to place. Or like, I'm a roofer and I just need, you know, a lead to come in and then I get it booked. Like, AI can do that stuff. But it's still like, I think we're still. very early to that being a compelling thing to make an investment thesis off of. Like, you know, what's your paid adoption of AI is, what, 3% of the population? Even if that doubles by end of next year, it's not going to solve the problems that any

of these companies are dealing with, in my opinion. I don't know. Thoughts? Well, well, consumer always goes ad spend funded every time. And I don't see a world where this breaks the frame on that business. There's always going to be some consumer paid subscription aspect of all of these businesses, but you live long enough to become an advertising play. Yeah.

Yeah. Having cash on hand is essential for any business. Traditional business accounts hit you with high fees while paying little to no interest on the cash you need for day-to-day operations. That was our experience too, until we switched to the new EQ Bank business account. Now, every dollar earns high interest with no monthly fees and no minimum balance. You also get free everyday transactions like EFTs, bill payments, mobile check deposits, and 50 outgoing and 100 incoming free interracket transfers. And to sign up, quick and fully online, no branch visits because, let's be honest,

no business owner has time for that. We use it for our own business and it's the first account that actually helps our money work harder while keeping operations simple. Check it out today at eCubank.ca slash business. I've been thinking about doing a Quebec City trip where I bring my mountain bike and the family tags along. I could spend part of the day riding while my wife and my daughter explore the city, walk around the old streets, stop for ice cream, and burn that energy in a beautiful park. Then at night, we'd all meet back up, go out for dinner, and make it a family trip. And while we're doing that, our home in Ottawa would just be sitting empty.

That's when it hit me. Listing our home on Airbnb just makes sense. Someone else could enjoy our place while we're away, whether they're here for a summer weekend, visiting family, or coming in for a big event in this city. They could explore the Rideau Canal, take the kids to the water park, catch a CFL game downtown, or just use our place as a home base to experience Ottawa like a local. The extra income could go right back toward our own trip, and we can choose to host only when it works for us. Your home might be worth more than you think.

Find out how much at Airbnb.ca slash host. So I had a birthday party to go to for my daughter last weekend, and I just knowing that parents, you know, there's a lot of government workers, but some corporate workers too. And I ask about, like, nine or ten of them. I'm like, oh, so, you know, do you use AI at work? personal life. The ones that do for personal life, it's just mostly an enhanced research tool responding to email, stuff like that. And then I asked for work and it was pretty much the same thing as people like, they just use it for very basic things. Research tools summarize saying draft email, draft PowerPoint presentation. And even one guy said he works in a, for a corporation. He said,

or sales team, they would blow through credits. They would like blow through Claude because some of the sales agents were using the most advanced model to just create basic PowerPoint presentation instead of using older models. But it's just kind of funny that you talk to people. And for most of it, like they just don't really maximize AI for what it can do. and they're also not willing to pay for it for their personal usage. I agree with Brayda. I think it's going to have to be either ad funded or selling the data that you provide them somehow.

Yeah. So that makes Google and meta like interesting plays, right? Because they have two of the biggest ad service areas on the internet. And like meta now, you know, we thought they had intimate targeting data on us before we started using their AI assistant. Now it knows what I'm spending my money on all the stuff I intend to buy. I mean, so, I mean, both of those two actually stand to benefit quite a bit from my lens. If they could, if they could really, the problem is they suck at generating models. We'll see how Google's model, the new one that they're, whatever it's called, the Argoner, like Gemini 4.

It sounds compelling, but I mean, they always sound compelling before Google drops a model and then it sucks, right? So they've been throwing out big flop since like the original 2.0 flash in my opinion. Yeah. Yeah. Meta stands to benefit hugely. I think you've seen that huge rally in the stock, like a couple of hundreds of billion in market cap whip back up because if there's anyone who knows how to monetize attention, look no further. You know, so there's obviously something there. Now, I have a hot take for you guys, which I might get absolutely roasted for. is something we get clipped for in three years and looks like absolutely.

Yeah. Everyone gets roasted. Still today and through the rest of 26. No product market fit for personal agents. Zero. The product market fit today is still coding. But then that market's tapped out, right? Like isn't that all of these labs pivoting towards personal agents

an admission that they think that that market, like they've maximized the tam on that market? I mean, if going from zero to 100 billion in ARR in two years is not sufficient, then I guess they might have to go get a new tan. Yeah, but it's still the, when agents are a piece of software and they can create more software and the non-developer folks are getting the benefits of the developer folks. We haven't even touched that market because as Simone said, at work. Personalized software? At work, the people who are using AI for those basic functions is like caveman compared to how developers are using it and building coaching agents. We have not transcended to everyone using it in that way.

And that is a 100x from here and in compute a thousand X, right? So the personal agent is useful. I think it's cool. But what it's really going to make people want is a human concierge. Yeah, 100%. Who does things the way they want. Yeah, yeah. But how did they become profitable?

Like, I've not heard a compelling reason, especially, let's talk about like, let's say, Chad GPT, Open AI and Entropic, right? Like, how do they become profitable? That's the one thing. The model, yeah. Like, when you have pressure coming from the Chinese side of things, it's putting some cost pressure in. From an operating margin and gross margin perspective, I think the labs are super compelling.

I, you know, I think the gross margins that are going to come out of, Anthropics, you know, next few quarters are going to blow people away. It's those big KAPX training runs that absolutely melt them. How long does that go for? Yeah, I guess they know they have a model that, like, until we reach the model that everybody can agree is like the final product. But like, because we are seeing these guys developing models now where they're getting a better ROI on the, you know, like they're spending less tokens per result, let's say,

or like per, you know, cost per query. And you're starting to get, like, I think everybody's kind of accepted that, you know, after Fable, like, we don't really need to pursue the frontier of like super intelligence that much. Like nobody, there's not really a market for super intelligence, right? Like most business functions are,

are pretty simple or, you know, like most things that most people need to do. So like let's stop spending insanely on, on the top end and really actually focus on optimizing and making these more, like, you know, less token spend, better outcomes. more reliable and it seems like i mean muse was really the first one who kind of headed in that direction like with i think it was meuse spark 3.1 grok kind of did the same thing and they're good

models to use like they're not again they're not brilliant but i don't like i'm in real estate man i do not really need brilliant right like no offense in my profession but yeah so when do we get to where they've settled on good enough models we don't actually need to do that that that huge capex and and simon just uh on the one point you mentioned on china like china's a subsidizing these models as well, right? So there's an end point to both, like to both of the, the parties in the space race running at a capital to be able to subsidize these costs, in my opinion. I would say the Chinese government or the CCP has some pretty deep pockets. Like it could be part of their plan. If I put, I know Braden likes to put his tinfoil on from time to time. But if I put that on, am I like if I'm the CCP, do I not just try to make dosi, US model as like money, like the biggest money losers as possible or from a strategic perspective. Yeah, I mean, they're not stupid. They're seeing how much GDP growth is coming from the AI buildout.

If you start having an important player like Open AI or Entropi starting to struggle because they just need more and more capital to keep going, it could be a strategic advantage for them because if they're really lowering the amount of revenue that they can get, they're accelerating the losses for those models. I mean, from a geopolitical standpoint, it could be beneficial for China if they want to keep up in that race, right? So it could be a way for them. Yeah, like it could be a way for them to do that.

I think it's a, you know, hey, let's commoditize this. Yeah. Play, right? Yeah. But, you know, who knows what happens in a few years? I'm on the ground making pretty big, you know, multi-million dollar compute run decisions about which models to use. And so is my co-founders, of course, who are very close to this.

When Kimmy came out, it was framebreaking for about three days. Yeah. It was a K. Kimmy K-3 came out and his frame breaking for about three days. And, oh, why would you use Fable? you know, why would you use GPT6 Astro or whatever? Like you just use Kimmy, it's just as smart. It uses, you know, one-tenth of the cost or whatever it was.

And you know what that model did? Was hungry for about 20x the tokens. So go, okay, I have an offer for you. Each token, half the cost. But you need 20 times more of them. Right. So less efficient for you guys, huh? Way less efficient.

Horrible unit economics, right? So it does what happens to open source over time, there's obviously something there. It could commoditize the need for, you know, frontier intelligence on a five-year trip. But follow the market. You know, anthropic and open air and going from zero to 100 billion of ARR

if it's truly commoditized. Like the market, the market is the market, right? these buyers are not are not stupid people are people are running big huge compute runs at these big organizations and they're not they're not dumb the same way that we're making that math decision as well too so i have you know as someone who's model agnostic i hope costs keep coming down i hope i hope it's a very competitive race but i have not seen personally that there's an actual optimal open source route where I'm going to save myself money today. Well, to use open weights, to use an open weight, you have to, like, I was looking at what would

it cost me to run like deep seek for and I'd need like 150,000 in like, and compute in my house to be able to do that. Right. So it's like, well, I'm just going to keep consuming it at $100 a month on a subsidized plan then. But and and I guess the challenge. Like when you're using compute at that scale, like are you guys paying on API or yeah. So like this is the other challenge is when you look at the way businesses consume on API, they are paying extreme costs and they're they're at the upside. That's businesses who are paying market costs for these things. And it looks like we've already reached the point of diminishing returns on on their

willingness to pay for that based on the results that they're getting. So we know that there's a huge market for consumer plans, and that seems to be the appropriate pricing that the average human being is willing to pay for, and average human being making business decisions is willing to pay for compute. But they can't offer that without, what is it, 70x subsidy? I think we pulled up the chart a couple of episodes ago. I'll try and find it. But I think Open AI is 70X subsidy on their 200 a month plan.

It's like we have 14 grand in compute equivalent on API costs. at $200 a month. So like when does that become viable? Like, or does it never? I think that's the biggest question, right? I think Braden has a point. Like, the revenues wouldn't be growing that quickly if there wasn't demand for what they're offering.

Like, clearly, you know, they're doing something right from that standpoint. But the other side of it is, does it ever become profitable? I think that's the other side of the equation is like you can think of it being maybe their costs do go down over time. They're able to just keep the pricing around the same price. Or maybe they're able to increase pricing over time. They just realize that users, power users like you, Dan, that's paying $150, 200 bucks a month. They realize that, you know what?

You're still willing to pay $5, $600, $800, whatever it is. They're able to increase it and then they become profitable that way. It just I just don't know. I think that's my biggest issue with these companies. is just I still don't understand what their path to profitability is. There are some ways that they do, but I still, I think it's very murky. I think you need to open your eyes to the idea that they can remain unprofitable and growthy for much, much, much longer than any of us.

We got to tweet that as a quote. The labs can remain unprofitable longer than you can stay solvent. Yes. No, it's good point. I think it's an excellent point. Yeah. I mean, the entire U.S.

economy is riding on it. It's frame breaking how much their investors don't give a shit about what we are talking. When they go, when they IPO and retail capital, doesn't that change? Or are we in the world of like YOLO people betting on like Elon Musk's ability to like, is every, do we have five Teslas on the market now, right?

Maybe. Because, you know, you can underwrite the story of how that $2 trillion company is a $10 trillion dollar company in three years. And but see, I still have a hard time. Like the only way I can see them doing it is if they, you know,

have you, have you guys heard the, I can't remember the guy's name, but from, um, from Palantir talking about like, he's just dunking on these labs like on those rants on TV. And he's talking about like basically that how they can exercise like massive corporate

espionage because like people are just dumping stuff into these models. And, you know, I mean, you see it with Anthropic with like, the bio, like the bio lab, right? Like a wet lab that they have in, in San Francisco,

it's like if they started doing, and Dario has always talked about this idea where like they'd be the only company left in the world, right? I mean, but if all of these businesses are using, and like right now it's a landing to say, by the way. Yeah, yeah, he's a, he's an intro.

I don't know, they got to stop giving that guy a microphone, but it is entertaining. It's super entertaining. But the, you know, if they, if they see all these inputs from their users who are not unchecking the box that the labs can train on their models.

And now their models have all of this like proprietary information about, you know, really, really good businesses. Can they not just start spinning off and creating drug company, right? Like this drug company that they're creating or like do they not just start being a derivative? And they actually are the AI that is operating all of these because it's not, you know, at a certain point it becomes no longer hard for them to just run the business. And maybe they start, you know, they almost like go like private equity roll up style or startup style.

or startup style and start picking off actual cash generating businesses, and they're using their own compute at relatively discounted rates. I'm just theorizing here. I'm trying to create a bullcase for the counterpoint to Simol's thing. But I think it's either they make models way more efficient where consumers actually can afford them, which we're trending in that direction, just not fast enough. And Jev, like we haven't talked about Jeff, but I think Jev as like a classifier AI is a big step in that.

Nobody's really using it a lot. But I think as those labs start to adopt their own kind of like classifier AI principles, they'll probably actually see a lot of of compute reduction or cost reduction. But the other one is that they actually just become like these like antitrust busting like monolithic entities that just start absorbing everything that they can possibly improve efficiencies on with AI. I mean, they could happen. Yeah.

Yeah. I mean, it's not impossible. I guess for me where I have a little bit of trouble is just historically when you have like a great new technology like this, no matter how great. the technology is, there's usually a bubble that there's a correction at some point. So I think that, and there's going to be casualties, there's going to be probably one major business that right now we're like, okay, it's just crushing it or it's at the forefront of this trade that.

You think that's Oracle, right? I mean, if I had to bed, I know it is, I mean, Oracle is, I mean, I think it's the most vulnerable. You're seeing that in the credit's default swamps or the highest. of the hypers. Like, it's not even close. You have Oracle and then you have the rest. And rightfully so, I mean, they're just bleeding money.

We saw the force major being announced for that New Mexico project. And that's starting to emerge as a risk, right? For this trade is you're getting these massive investments into building AI data center. Some of the hyper scalers will own some of the building. They'll lease some of the building. So I think most of them have a bit of a hybrid approach there. But in the case of Oracle and that project itself, you have Blue Owl that is essentially the landlord, but Oracle that is the tenant, but also responsible to making sure that there's power that gets to the project. And that's what they invoke for the false major is they're trying to essentially pay a lower lease until it becomes operational.

And because of some regulatory issues, it's actually being pushed back at least six months from what they said. but I think it is something that's worth noting because then you have another problem that could happen. If you start seeing these problems happening, whether it's just not enough power or power able to get to the actual side, or you have NIMBs and you have states or municipalities kind of putting a stop to them because they want to review the process, if you start delaying that, one of the issues they might start facing is you have an Oracle buying computing equipment, right? Nvidia chips, they are buying HBM, high band, end with memory. Well, if that's been purchased and is being stored and not used, sure, it's not, there's no usage, at least it's not being like, you know, in use, so it's still new.

But if there's a delay for a year, who knows what kind of chips Nvidia will have in a year from now. So that's the other issue is they may have less of a useful life for those chips or not be able to capitalize on the first year where they can probably get the highest. revenue from that infrastructure by, you know, selling it, selling the compute to wherever it is, Open AI Entropic, you name it. I think the other, like, question is, I mean, you look at the credit default swap chart that you mentioned and I pulled it up, but everybody's seen like almost a two, or like a, you know,

a 50% increase in their capital cost, right? So the other bare case is that maybe nothing, nothing breaks massively. Like, you know, Oracle could actually just end up being, you know, they get pushed back down to earth earliest in the cycle. And that actually ends up being a saving grace for them. Again, I don't know enough about this, but this is just like a hypothesis. And somebody else who's kind of given a little bit more room to spiral out of control ends up being one of the ones that could be or sort of the one that dies, your Lehman Brothers moment or whatever it is, where, you know, these guys, if my cap, as a real estate investor, If my capital costs go up 50% on a renewal or on my next project based on my original projections, things start, like I have to stop growing at the same rate that I was growing at originally.

And I start scaling back the growth of my portfolio or my ability to generate more income. And maybe they just kind of are forced by the market, because I think this is the market telling them, right, that, hey, you know, maybe we should sober up a little bit. party, you know, let's not, let's not make the hangover too bad here. Let's, you know, start drinking some water or whatever. And they, and they sober up and actually start behaving properly. And this is the bond market kind of trying to force that move today rather than deal with the catastrophe that could happen if, if they, if they don't. I don't know what your thoughts are there. But it could just be a, a softer kind of like plateau and roll over into a, you know, a typical bare market where valuations start to gradually correct and don't just collapse.

like, you know, you see this 01 style.com bubble blow off or whatever. I don't know. Like, but historical precedent would tell me about, you know, your valuations aren't dot com yet. We're not railways yet. We're not like great depression yet. But we're not even close. Yeah.

We're like, you know, it's, it's, it's high enough to be like, yeah, this is not good. Is it though? Is it though? I feel like, I think so. Like based on is it for, I don't know what is it for P is like the one that that they use most often. For the things that crumbled. we're not even on the same universe in terms of bubbly from dot com.

I'll pull up the chart. No, I agree. Well, you're not far. I guess it depends on the metric. Let me see if I can find the cross section. So by what metrics would you say that it's not as bad? Just the things that are tied to the AI trade infrastructure-wise that, you know, get taken as the poster boys for.

So there's more derivative, like there's, and there's more derivative economic growth. Like dot com bubble was like a bunch of stuff where basically nobody had any idea and these guys were like cheering on. Extreme speculation. Yeah, yeah.

Extreme speculation. Like, you know, easy, easy one to call on Viti as, you know, largest company in the world. In their July ending quarter, it's 63 billion of operating income. They're the picks and shovels play, though. Like, InVDia doesn't worry me, right? Like, they're fine.

I think they're fine. I'm more worried about, like, labs who don't have the path to profitability, IPOing. That's the capital that gets sucked out of the market in other plays. You know, so you have, like, you have, what, $5 trillion in IPOs, like, more IPOs in these three businesses than like the last 70 or something combined. And that's like, you know, capital is going to start getting moved around. I don't know. Quick question.

I don't know if you guys have a look on this. When dot com bubble crashed, I was six years old, so forgive me. But were those companies as hated as the ones going public right now in the public eye? Like, I think the only thing we're missing in this conversation moving forward is how hated these companies are and how unpopular. Do you think that's actually the case? Like I feel like most people I do. Like the AI is like the models or? Yes, the models.

Yes, the models in the, you know, the average person you talk to that's not in our little bubble. Those companies as well as just like people creating data centers. There's a lot of fake news about how it just sucks up all the earth's water to power these things. Yeah, we just did a whole episode on that. Yeah, from the real estate angle. And I mean, it's not actually like, did you see the post from Creamy, on Twitter about like how it actually does well I mean you know like if I'm sure you've seen the almond farm and golf course thing and whatever but yeah it's the same way of like saying

that a nuclear power facility consumes a ton of water it's like well you don't want the water touching the radioactive rods the same way you don't want water touching computers like we're talking about closed loop cooling I think it's different if I go Back, I was around 15 back then. I think right now, I think what you're seeing is especially combined. So you're seeing this part of the economy doing really well. And I think where we get back in talking about the K-shaped economy, but I think what's happening is there's not a lot of people that are like benefiting

from the AI build out, right? The ones that are benefiting tremendously. So a lot of people are just saying, you know what? And I think it's especially true in the U.S. says, okay, my power bill is going up. There was this massive data center that was built a year ago and they're equating kind of, you know, one plus one is equals two. They're like, you know what? This is part of the issue here. So I think that's why I think there's a little bit of maybe a different sentiment than back in the 2000s where 2000s it was or late 1990s. It was something brand new, right?

It was very novel. It was just a new technology. And this is new, but it's almost a continuation of the dot com, right? So it's almost like the next evolution. So I think that's probably the biggest difference where, yes, there's some newness to it. But for a lot of people, I mean, you ask him how different their life is from now, like in their day-to-day stuff from now versus 2022. And they'll probably be like, you know, it's pretty similar in terms of my day.

day to day stuff. The biggest change though is that I'm being squeezed for a lot more and I'm, they're trying to put the blame on something or some part of the economy. So I think, and the data center build out is a, it's an easy, I mean, wonderful scapegoat. It's a, yeah, it's a very easy scapegoat. I think also the AI industries, the prominent voices, sometimes they're not helping themselves. Yeah. And. Oh, yeah. Yeah. Yeah. And it's probably not helping in the world saying none of you guys are going to need a job. I'm just going to do all of them. Yeah. Or you're, you know, you're all going to die because of AI or whatever.

Like two weeks ago, they told us that we need to hit the brakes. Too dangerous. Right. So that wasn't that long ago. That was two weeks ago. But can you blame the population for being scared and not being super on board with it? No, their marketing is making, like, is designed to make people scared. So, yeah, it's like, you know, it's like, yeah.

The big short clip, it's like they're not, they're not something they're bragging, you know. Yeah. Yeah. Yeah. Yeah. Yeah. Yeah. Yeah.

Yeah. They're actually bragging. Or you have like Trump tweeting or truiting essentially like, you know, I don't care that your power bills are going higher. We need to win the AI race for the grade of the country. Like people are hearing. I do agree. Like, I agree that it is like an existential point for, for the, for the, for the, US. But if that's the case, then they need to nationalize these companies, which the AI labs are,

and I'm not a call me, by the way. Like, I don't think it should be nationalized. But like, if that's, if it's actually an existential national security threat, then it needs to be positioned as such, which, I mean, there's an argument that that's the case, like you were saying about China. Like if China can outspend them indefinitely and, I mean, what we haven't really necessarily talked about, this is a real surveillance play. Like, these companies are exceptionally good at getting data from human beings and not talking about their feelings and pictures of the growth on their toes

and all this stuff, right? Like, would you prefer that that's the US? I don't, like, I don't necessarily know. I would prefer it's no government, but it's going to end up in the hands of some government or private company. Or would you prefer that that's China who owns all of that data for the world?

I don't know the answer the question, but that's kind of the lens at which I would think they would approach it that way. And that's where you get Dario and Sam sort of saying, I think they also know that they're going to need some support, at that scale to stay alive, the labs, and if they want to keep progressing at the rate that they're progressing.

How much do you think people actually cares not the right word? Everyone cares about their data privacy. I care. Everyone should care. You guys care. How much do people actually think that there's any sort of protection with their data and have already accepted our fate with this stuff? I think you're right. I think most people are just like it is what it is.

Dude, and this happened in months, man. Like, because I, the easiest way to compare that is is open claw versus MetaMews. Like I was really early to the agentic AI stuff. And I was like all it like, like, I call it like post, you know, post, living in the post privacy world. Right. So like, yes. Exactly. Yes. I think most people have, have been living in that world for a long time without realizing it.

You know, in Zuckerberg knows all this stuff about you from your Instagram. If he wanted to, you know. And, and I'm sure the. We know the stories of the NSA and what they use for ad targeting and whatever. But when I was using OpenCla, everyone's like, how are you giving it all this access to your personal data and guardrails and blah, blah, blah. And I'm like, you guys have like, this person is asking me this question on Instagram. You know, like, and you don't think that you're not like in a, if not a worse position.

Like at least I have that all in custody on a Mac mini sitting somewhere. It's not in some server at a massive global corporation, right? And then Muse comes out and all of a sudden all the realtors are like, oh, we use. Like, you know, it's amazing. And it's like, did you notice that you connected to it with your Instagram account? And then you added your what's that? And because that company has had just a tremendous track record.

This is the funny part, right? So I think that the vast, anybody who's using like, and I think it's everybody, right, who's on their phone all day. and using all of these platforms, like they've accepted without, you know, explicitly accepting that they live in a post-privacy world. Well, many of them haven't, you know, it's been explicit when they sign the terms and conditions.

Yeah. Yeah. Yeah. Yeah. Anybody read an end user license agreement? Oh, well, this is like people, people have, you'd have to spend 40 years just spending time reading all of the terms and conditions that you have.

agree to on an annual basis. Yeah, yeah, yeah. Literally not enough time on planet Earth to lead them. Yeah. Yeah, so I don't think that, I don't think privacy is a headwind. No, I think for one of the biggest headwind is still people are, like the vast majority of people. And I think sometimes we are probably guilty, all three of us of like having, you know,

living in an environment where we talk to a lot of people that are into this technology, that use it a lot. But if you start talking to the general population, it's. the sense I get is like how is it making my life better. Yes. Oh, 100%. I think it's more hated than any of us about willing. So that's kind of a, well, see, I disagree on the hated part, but we'll set that aside.

Because like, like I actually think that because as somebody who says like, I said I was going to set it aside, but I'm actually not going to lean into it. But then I'll get to the other piece, which is I think they have an adoption and a product problem, right? But the, you know, like I say stuff. I try and shoot right down the middle and like say stuff the way that it is. And I see people yelling super loud on the left and the right or on the anti-data centers or pro-AI argument. And I think that like those voices, like anybody who really no disrespect to all the people in my comments and all the people on the chat right now. But like anybody who really is meaningfully like angry and like upset about this stuff and just using like Instagram comment threads to make their point, their opinion really doesn't matter all that much. Like no offense.

Like what matters is the 80% of people that are in between those, the silent majority. And I think the silent majority doesn't have anything to say about this. They're still like, I'm just going to, I'm going to watch it progress. And I don't really care. Like I don't, I haven't seen a data center. You have five people who live in the rural area two hours from me have seen it and they're pissed off about it. But like I use Netflix and I use Instagram and Twitter and and I like open A.

I like chat GPT. It helps me solve a lot of problems and whatever. And now all of a sudden they're like, this is, I don't really give a shit, right? And so I think most people are like- It has immense consumer penetration. Right. So I think most people are apathetic. Human beings generally are apathetic.

And if it's like if this thing can make me, most people are NPCs, they just don't want to suffer their whole life. And if this thing can make their, and I feel bad saying that, but it's true. Like that's how most people want to live their life. They just want to not suffer, right? If this thing can make my life slightly better, I'm okay with it, right? Whatever the consequences, water, blah, blah, blah.

The same, and that's the same reason why, I think personal agents don't have as much product market fit as people think it does. Not many people care about being productive. Yeah, so I think we're going to, I think I totally agree with you. I think in your personal life to be clear. Most people want to be not. People care about being productive at work. Yeah, or like or maybe like productive in such that like can I do less work and get paid

the same amount, right? Like more money, less work is like the best, you know. And that's the use case to me. It's like, but yeah, I agree with you. Most people in their in their daily lives, they want to make leisure of the, of the, mundane day-to-day tasks that we get to do when we're not at work. You know, I think chat QBT, because you mentioned about the consumer penetration that they have, like them releasing dots is going to be really interesting to me.

Like, Muse was a step, I think, and it's a step, I think is a big step in small business and all of this. But chat GPT is already sort of like your consumer, like AI curious consumer and it's web-based. Like the hard part when you get like go grok-buby. as an example. Let's go back or like open cloth. So it's not easy to set that up. Grockbot a little bit of improvement, but I still like to get the most out of an to treat my AI model as a harness, I have to download it on my computer and it's not constantly telling me how to do that. Now we've got what's overwhelming for a lot of people. Yeah. But now we've got web based and it's

already in the hands of a lot of these AI curious consumers. And it's probably only a matter of time before Claude comes out with some sort of derivative version of this. And you know, oh yeah, it's summarized my emails. Right. Like this is, the joke of like everyone's like what's the use case for open clause. It's going to summarize my emails. I'm like I send like thousands of outbound emails a day to, you know, like real estate professionals that I'm trying to recruit people we want to get on as customers, sponsors, whatever. Like there's real business productivity cases for me. I think that the next challenge becomes do most people actually have things in their life like business ambitions or things in the goals

that this thing would even help them progress towards? Or as it just, kind of like taking slight things off of my plate. That's why I think it's fake product market fit. Like, yeah, you know, I'll stand by this. I think, I think personal agents are as fake product market fit right now. And, you know, until until it reaches a point where it's like an expert human to human concierge service, it's not, it's not making your life better today. I don't believe it does.

And especially in consumer day-to-day life. Yeah. But it is that. Like I think it already is that. Like for me, it does what you just described. But it takes a lot of work. So what's an example?

You have to be a really, like I can have to go book a flight or like a trip for me or like, you know, call to book a showing for a property or go in and book a showing for a property. I can have it gather all of the information about a, yeah, a deal that I'm working on in an email, text message or voice format.

But it takes quite a bit of a lift for me to get there. But I think the other really, really, really important part. that nobody really thinks about is that in order to be good at using AI, you would also have to be a good manager, right? You would also have to be good at managing human beings. And I don't have that skill, by the way. Like I can't tolerate employees.

Like I, that's the very crude way of saying it, but I'm not very good at managing people. But AI, I don't have to be nice. I'm actually, it's actually better if I'm not. And I don't have to really tolerate crap and whatever. And so it's been, I have leadership skills, but I don't have like the other components that I think make it necessary.

But a lot of people I don't think have, have the the ability to delegate properly. Like even in my industry, I see this really well. There's people who are self-employed and they basically just go and find the next customer or the next boss to tell them how to do what jobs they need done. And then there's people who actually run businesses and have systems and standard operating procedures and the ability to delegate tasks properly. So even if there was product market fit, I don't think most people are capable of maximizing

because they don't know how to actually like request work and define. a definition of done and systematize things. And that's the next problem. Because people want it to know everything that they're out trying to ask it to do and read my mind. Whenever we get there, yeah, okay, it's real. Yeah. I think it needs a physical element too

where it's like some sort of humanoid is now using the tools that were made for us a la a vacuum, a broom, a dishwasher, a washing machine that's when there's like extreme product market fit because

robot like the four what do they say 40K dude like that's I'm done like I'm gonna have a hundred of them yeah like it's a kind of

that because that is where consumers will pay to do the shit they don't want to do yeah this is a challenge

to do online shop yeah they want to actually booked that flight. Dude, for a vacation, if you're middle class, going through of booking that flight

and doing that planning. It's aspirational. Is actually half of the journey. It's the same way that when you buy a Rolex, you don't buy... I don't have a Rolex on, by the way.

I just have a cheap same thing. Raises his arm. Yeah, raised by like shitty sake. It's actually a beautiful watch. I love it. They're not buying, you know, the watch. they're buying the actual process of how they make you wait in line,

prove that you're not a nobody. The dopamine. Yeah, and that buying process is actually what that luxury brand sells. And for a lot of flight booking is aspirational as well too. I agree with that, 100%. If you're flying for work all the time, it's not aspirational.

You probably have an assistant who does it for you. Yeah. But everybody, so then what is it? consumers just use consumer plans at work to do their job because I feel like that is actually probably happening at scale and and like kind of the real deal could be I feel like that's I think there is a I have I have a thesis around a huge market opening up for luxury human concierge services off yeah I saw that on I saw that on X too I feel like that's oh did you I thought this was a No, somebody else was talking about that.

I thought you were going somewhere else with that. I fully thought this was a novel thought I've been happening. But I think that that is going to be a huge business off the back of this. And it could make for great content marketing around like dealing with rich people's problems too. Well, this guy's thesis was more that there would be more like more of these luxury experiences were boomer marketed. whereas nobody's really built that experience for millennials. So it's different than yours. But yeah, like similar kind of same kind of vein.

Yeah. Boys, I got to drop off here once. Yeah, I think we all do. We ran pretty long here. Amazing. That was good. We'll have to have you back whenever Oracle goes bust. And we'll chat.

And shout out to Stuart here, Grand Seiko over Roll flags. They're both fantastic watches. Both beautiful. Yeah. I'll stick with my own watch. watch guy, man. Yeah, there you go. Always tells me my next appointment.

That's all I need in life. I'm going to buy you guys. I'm going to buy you guys mechanical watches for Christmas. No, you can get me, get me, no, just get me the, the nice Apple Watch. We're good. Yeah, yeah, yeah, that's a good watch. It's just so big. I feel like a quarterback or like Spy Kids or something, you know.

I wear it for most days and it is a insane piece of technology. Okay, I'm going to go through. I'm cold. Done. All right. Later, boys. Later. Thank you, Bray.

The Canadian Investor podcast should not be construed as investment or financial advice. The host and guest featured may own securities or assets discussed on this podcast. Always do your own due diligence or consult with a financial professional before making any financial or investment decisions.