The Rundown 6/23/26: The AI Hype Machine, Canada’s Capital Gap, and the Return of Hard Tech

Speaker A: That moment there really said to me the same thing that Satcha said. Do not build your business so that somebody will come calling, particularly where they start to become monopolistic.

Speaker B: Welcome back, everyone, to another electrifying episode of Tank Talks. I'm your host, Matt Cohen, and joining me, as always, is the one and only John Ruffalo. Get ready. As we dive headfirst into the hottest news and most exciting headlines at the intersection of business, innovation and politics. John and I have known each other for quite some time, and during COVID we were pretty bored. And I started a podcast called Tank Talks. And I kind of duped John into showing up once in a while to talk about things going on in ecosystem and politics and business and startups. And then sure enough, he just kept showing up and never knowing the questions, never expecting anything in return, just kept showing up and still doesn't know the questions I'm about to ask him today, which makes this really exciting for me. But maybe, John, just so our audience understands what you're up to, you know, your quick background and how you come into knowing so much about everything.

Speaker A: Oh, yeah. Well, first of all, it's a pleasure, uh, to be here. When you first had asked me, I do remember you saying, would you mind being my first guest? And I was like, oh, of course, I'll do a favor for you. And then the next week it's like, hey, you know what, that went so well the first time. You might just do it one more time. And then I started to see a series of emails going, hey, dude, like you're asking me every single week.

Speaker B: So I actually just didn't even ask him anymore. I just went right to his calendar and just inserted into his calendar.

Speaker A: Yeah, exactly. So, no, it's a pleasure to be here. As some of you may know, I just been in the innovation space really for a very long time, going back to, on a full time basis, back to 1992, uh, which aside from making me very old, I get to see, and I have seen all of the hills and the valleys and the repetitive themes that happen all the time. And I love it when somebody tells me this time it's different. And I like, oh, yeah, sure it is. And it never ever is. But it is exciting and interesting and one of the things that, you know, I'm most pleased about is I believed back then that innovation was going to be the path for the future of this country. I believe that entrepreneurship was really needed to be at the top of the apex. So still, you know, 20 or 30, but 36 years later, it's uh, you know, you're still seeing the fruition and I see you all here and it's exciting to see that uh, you folks uh, are really the, the future of this country. So uh, thank you.

Speaker B: Well, let's let double click on this time is different because I think this time is different. I know you and I argue about this, but this time it really feels different. This week we just saw the largest IPO of all time. After 20 years going public, staying private and then going public already up. You know, I think it's like 80% since the IPO. It's trading around $215, 2.8 trillion market cap. The headlines are everywhere. First off, are you surprised at where it's trading?

Speaker A: No, because people are generally stupid as investors and uh, these are the hype machines. I think the Starlink business is a great business.

Speaker B: $20 billion revenue, 60% of revenues.

Speaker A: Yeah, yeah, yeah. You know, now, you know, you start to look at absolute fundamentals and fundamentals do show up over the course of time. Business is valued on its discounted cash flows. No one factors in capex. In fact, the Starlink Capex is not particularly egregious.

Speaker B: No, not at all. It supports everything else.

Speaker A: It's the other stuff that is absolutely egregious. And the math doesn't. Math. And you know, you get these hype bubbles and over time, you know, these bubbles always burst. The question really is, is. And it was really on a post Covid basis, the world was awash in liquidity. It's no longer awash in liquidity. And one of the things that I had mentioned to you, and you're starting to see it, you know, we got three massive IPOs coming out. It's not like cash just magically appears. It has to come from someplace and either people are going to move from bonds, which I'm not sure. Uh, that's not what you're really hearing because people are also having anxiety about where interest rates and inflation are. They're moving it from other assets that are largely technology related assets.

Speaker B: I mean the Dow made new time highs today though, so it doesn't feel like they're sucking it out from the other Max 7 players just yet.

Speaker A: Yeah, but that's a Mag 7 but, but they're starting to say wait, they said that there was enough capacity for SpaceX. The question is ad anthropic and open AI onto that and there's not enough capacity on that.

Speaker B: Well, everyone thought there would be around 3 trillion of IPO like market cap that was going to come out between the three of them. SpaceX itself is 3 trillion right now, Right? Right. So what does that mean for Anthropic and OpenAI?

Speaker A: Well, but you see what they're all doing though, they're only floating. A very, very tiny piece of float

Speaker B: is around 150 billion. It's tiny.

Speaker A: Yeah, it's tiny. And it creates artificial demand as well too. So all I would just say is that you'll see what the real values are, where the market really sees it. But right now it's on a hype.

Speaker B: I'll tell you what's interesting. There were, I counted it this morning, I think there was nine ETFs listed today on top of the options that were trading that were tied just to SpaceX stock. Nine ETFs. These are like the bull bear inverse 2x3x inverse, uh, trading ETFs. That's crazy. Amount of triple layer traction, like trades on the volatility of this thing. So we'll have to wait and see. I want to talk about the Sachin Nadella post that came out the other day. It's quite interesting. If people in the audience have not read it, I highly recommend you read it because it kind of counteracts everything that they at Microsoft said OpenAI would provide. What he said is not to rely on, on these frontier models to build your businesses because they will become competitive, they will eat away at all the knowledge you put into it. And we now need to start creating more local models that run with infinite loops and reinforcement and they keep that human knowledge back into the enterprise. How do you think about that when you're investing today? Because a lot of people still think that the next anthropic or OpenAI update is totally going to disrupt your business. And you should not be building your business on top of these models as a wrapper is what they used to call it.

Speaker A: Yeah, I mean, this is one of the reasons why we haven't still made an investment on the application layer. So if you look down the AI stack, right, start from the application layer, go all the way down through the operating layer to the data center, down to the chip sets, all the way down to the earth and the substrate of the materials. Canada really has nothing, unfortunately. And I think the issue that you're describing is where is your choke point? Where is your moat? And one of the concerns that we have had, at least today, is most application layer businesses are so reliant upon the underlying LLM now they compensate that by saying, well, we can switch in, we can switch out, which you know, which is fine. But are you still really reliant upon it? I will tell you we had a near death experience with one of my large investments over a decade ago at Omer's called hootsuite, which is still around and thriving. Hootsuite's fundamental mistake which we knew at the time of making the investment it was the single biggest risk but we put into remote category is it was built on top of the social media platforms. Very similar analogy. And because there was a number of different social media platforms we thought that distributed model. And it uh, de risked it. It de risked it. And then what started to happen, the consolidation started happening and the first, you know, the taxman cometh and it was Twitter at the time and Twitter basically said uh, we're going, you want access to our APIs, you'll pay a 30% gross receipts tax on the revenues. The margin wasn't 30%. And now we ended up doing a workaround on the solution that was very difficult. It degraded the product. But that moment there really said to me the same thing that just said do not build your business so that somebody will come calling particularly where they start to become monopolistic. And the unusual thing in this situation on the LLMs is and you and I had talked about this, you know three or four years ago when there is digital infrastructure and it's magically always this rule of three almost always applies. There's three telecom companies, you know, three automakers, railway companies, railways like airlines, et cetera. And we said there was going to be three LLMs. And at the time the funniest part was I said it was Microsoft OpenAI. I thought number two was going to be Google before it was called Gemini. And number three was a big question mark. Is it anthropic? Is it uh, mistral? Was it cohere, what have you. But what's interesting about it is that I said Microsoft/OpenAI because Microsoft really was the one running the show. So I find it absolutely fascinating at the moment that divorce has really occurred.

Speaker B: It is interesting.

Speaker A: Now he shoots them.

Speaker B: Well it's interesting because Salesforce just acquired Intercom fin, which for people don't know, Intercom was this customer service agent that you can talk to through websites like

Speaker A: Ada, like but they went through a

Speaker B: uh, quick and hard pivot what you just said, where they did not want to be tied to one model provider. What they ended up doing was they went open source and they started using different models for pre programmed responses so that their inference costs would go down. You're talking about Kimmy and Quen and these Chinese models and that's why Salesforce bought them because they're not reliant on all these one model providers. And I think your lesson from hootsuite is actually being played out now because startups are getting smarter by not using inference for full suite inference. What they're doing is they're living on the edge and they're doing pre programmed problem solving on the edge of compute so that they don't need to send all the packets of data over to the servers to get the full inference back when they only need 10% solved by frontier models because 90% of the workload right now is can be solved with these smaller models that live on the edge and that saves 85% of inference costs.

Speaker A: And you're not forwarding data. And you're not forwarding data.

Speaker B: So that question then begs what's going to happen to these inference model companies? What's going to happen to their high revenue businesses when people don't need them for all the things they offer?

Speaker A: Well, I'd say high revenue businesses that make no dollars burning endlessly. The unit economics are horrible. The unit economics like this is why I just said I was very, very negative on the LLMs. I still am. They need a business that actually drives their unit economics, their businesses. The pricing is all subsidized by venture capitalists. What is the true pricing? That somebody's actually I think we saw

Speaker B: it this week with Fable. I mean when Fable was out for a hot second, everybody's usage went up like crazy, ours included. In fact there was an article that said a startup did not turn off their usage limits and they had a $500 million overusage bill. 500 million. Half a billion dollars in overusage charges because they didn't put limits on their individual accounts for their employees. Now we hit our limits very quickly and we just waited for them to reset. But what we figured out was if you switch to a max 20 plan, which is a um, subscription plan, ah, it works out way better than having the overcharges go through at the end of the day. So everyone was starting to switch to max 20 plans. So it's very interesting what uh, you say that the models don't, does not work from an ah, economic standpoint but from a value standpoint it is still very valuable.

Speaker A: Well absolutely, but the real question in my mind is they're trying to act monopolistic. So which means there might be Two or three players. I think there'll be the three. Anthropic is the one. The one that I was actually betting on originally, if you remember, was Llama for Meta.

Speaker B: What's Llama?

Speaker A: Exactly. Nobody remembers now. It's a new name.

Speaker B: It's a new name.

Speaker A: But what did they go after? They said they were going to be the only.

Speaker B: Yes. Open source. Only open source.

Speaker A: And they.

Speaker B: I do believe, though, when you see the US Government step in and intervene and shut something down, it does bring a lot more questions to bear about, like, one, the power of these things, which we believe is real. I think even we know mythos can do some pretty crazy stuff. How sovereign countries like Canada need to have their own models as well. So let me ask you that. Where do you think Canada sits right now in its own AI Buildout from the model infrastructure, energy supply chain side to. Because it feels like we're playing yesterday's game today while the US and the rest of the world is playing tomorrow's game.

Speaker A: Yeah, it's a really tough battle here. So I think there's only two countries in the world that can actually build the full sovereign stack. It's the US And China. And no one else can do it. You can search this. I have the document. It's China's strategy to build this stack. And it started about 20 years ago. It was quite fascinating. Now, they didn't fully build it out originally, but they, uh, were very, very intentional right from the. Literally the substrate up in building a full stack because they anticipated that China and the United States will go to war, and they didn't want to be beholden to the United States, which leaves everybody else in the lurch. I think what the answer is going to be is, even though I talk about AI sovereignty, the truth of the matter is there is full sovereignty and partial sovereignty. And I don't get into that nuance. But what Canada needs to do is find its spots in the chain. And, uh, again, I mentioned to you, the truth of the matter is Canada does not have any choke points. There are some very limited choke points, particularly at the chipset level and at the mineral level. That's about it. And what Canada's going to have to do is align with our allies. And I use the term allies intentionally, because between countries, there's no such thing as friends, doesn't exist. We. We know. We feel hurt that the Americans are no longer our friends. Well, that's not true. No one ever is. But the allies have a strategic purpose together, like going to war. What we're going to have to see is various countries trying to focus in on the different elements of their stack and aligning standards so that there's interoperable standards and the collection of those countries is what's going to go up against the US or China and that's going to be the only way out.

Speaker B: Uh, you say going to war though. I mean, I think when you say the word war, you mean like actual defense war. I think we're in an economic war though.

Speaker A: Yeah, we're definitely in an economic.

Speaker B: So we should be playing pretty hard on our own strategy as well. So let's talk about the AI and

Speaker A: by the way, it could be a physical one.

Speaker B: It could be a physical.

Speaker A: But I didn't mean it studios.

Speaker B: Let's not talk about that stuff right now. Okay. Let's talk about the AI strategy though. That came out from the government recently. You've talked to Evan Solomon, you've talked to a lot of people in the government. What did you think about the AI strategy at first blush? Because BDC came out with a report that said Canadian small businesses, medium sized businesses who adopt AI could add 150 billion to the economy equivalent to a 6% boost to GDP.

Speaker A: Oh yeah, sure.

Speaker B: So I will not keep reading the rest of the stuff that they said because you and I both know the value of AI and what it can do. But the policy that came out in supporting of what it should be doing in Canada kind of felt like it missed the mark. We talked about this.

Speaker A: Yeah, yeah, I mean like I see what they were trying to do. So the good news is we finally got something really long overdue. There was some good objectives that they stated. Kind of unclear how we're going to get there. A lot of laundry listed. And this is where I started getting a little antsy is when you go through it, I recommend, uh, you reading it. There's lots of good stuff for everyone. Well, when you look to see the preamble, it's really about AI for everyone. And this is the challenge and the trade off that I think that they're really struggling with. I get it. As tech guys, you know, we want to go, you know, go fast and see where it takes us and step. Well, let's look at the other side of the coin. People are terrified of job losses. People are terrified at uh, privacy and surveillance and they're terrified of impact to their children and education. You know. And this is the one thing that I hate about what we do. Every single technology that we look like is always two sides of the coin, we love drones and yet drone warfare is what's killing people indiscriminately. And the real question on here, just like what you had said, if this technology is as powerful as it's being played out to be in the wrong hands, that could be horrible as well too. So when you read the document, they were really trying to appease everybody in there because that's the complicated part. The issue or the result though is that it looks like a laundry list of ideas. And I said to Evan, look, I don't dispute a number of the ideas, objectives, but which one are you putting the pedal on here? Because what will happen is when you're trying to do this, there's always tension between the two but you're going to have under stress, which pedal do you push?

Speaker B: Yeah, okay. I fundamentally disagree that in technology investments and investing or building technologies, it's always a double edged sword.

Speaker A: I will tell you every single one.

Speaker B: But on what percentage though that matter? Point. So no, no, like uh, if you have a technology like a car creation. Okay. Self driving cars.

Speaker A: Yep.

Speaker B: You're going to say that of course that car is going to kill someone. But if you look at the data, the percentage likelihood of that car killing someone versus all the drunk drivers and other human error out there as well means you'll never do anything. And I think that's the problem with these policies is that they are looking at everything as a double sided coin equally. 50 50. Where in fact it's like 1 to 2% on the negative side, 98% on the positive side and they give equal weighting to both.

Speaker A: Yeah, but here is the challenge in that analogy. I agree with that. But the 98 say on the good is fine. The 2% destroy the could destroy everything. That's the problem. Right. So when you have technologies like this, and again, who knows, I'm not saying

Speaker B: to ignore it, I'm just saying not to give it the same equivalency.

Speaker A: Yeah, well, I don't know because from the population perspective, and this is where the issue is going to be, we already have a polarization of income and wealth. The United States has hit its absolute all time peak. It was, I'm uh, going to get the stats wrong. 1% owns 60% of the wealth in the United States. Now you literally see this line. So what happens is, and this was one of the evolutions about three decades ago. I wrote this back when I was back in my early consulting days and it was around taxation. In fact it was the movement of dollars in Capital away from human labor to machine labor. And I started asking the question, how does your taxation keep up with that? Because all of our taxation is largely based on human labor.

Speaker B: It's going to tax the machine labor.

Speaker A: What happens. But then if you tax machine that's taxing capital, which is anti innovation, you're

Speaker B: taxing the revenue that generates from that machine so it's almost still being taxed.

Speaker A: Yeah, but what happens, but what happens is, is that the shift of the wealth moves from the worker to the owner of. So you end up in the situation where you have a trillionaire, a single trillionaire and let's just pretend what is better for the economy, a single trillionaire or a hundred thousand or a 10 million people? This is the dichotomy.

Speaker B: So let's jump into it because we have some questions from people in the audience that I have up here. So please continue submitting your questions on the slido. I'm going to bring them up now. One of the questions was how should the government address low trust and AI adoption in Canada beyond building AI, know how, which all like the jobs and stuff in this new AI strategy, Like how do you design the strategy to build trust in AI if you are doing it?

Speaker A: I think this was what was discussed and they used the word it needs to be trusted. Where you go too far is, you know, there's talk do you ban, you know, 16 year olds and under from using AI? Well that's a, I think that's a really dumb idea because how are they

Speaker B: going proficient at it?

Speaker A: Yes. Now uh, you know, the real questions are, and frankly the media industry did this. Let's just use media. So traditional media where you had television stations, etc. You had, you know, the regulations on things that you can and cannot do. The moment social media companies came in, they were very effective lobbyists saying no, no, we're not in the media business, we're in the technology business. And they try to avoid.

Speaker B: That worked well for 10 years and

Speaker A: they try to avoid regulation. And look at the shit show that uh, that's become. It's the same thing here is we need the guardrails on there on what is acceptable and not. But here is the problem now you see something like anthropic 3 years ago Dariel was fighting hand over fist to stay away the moment they get big enough now, yes, regulate us to ah,

Speaker B: regulatory capture is what they're going to

Speaker A: do because it's regulatory capture.

Speaker B: Okay, let's switch to the investing side, uh, on startups and what you're doing with your capital and what we're doing with our capital ripple. I think people want to hear that. So first question, if you can go back in time to the dot com boom with the current knowledge you have now, wouldn't that be nice? What overlooked business would you invested in and why?

Speaker A: Going back to the dot com that

Speaker B: you saw during the early days when you were investing, maybe not.comish, but even like Omer's days, you made the bets on, you know, Xanadu, which took a while to play out, but obviously turned out well.

Speaker A: Yeah.

Speaker B: What other bets did you see and miss out on that now in hindsight, 2020, you wish you did.

Speaker A: The one thing that I noticed and I wrote a lot about it, it was, it was a thesis that we didn't have back when we started in 2011 was actually the concept of trust. And you started to see peak trust to me was 2008 Google. Google search was spectacular in 2008 because you typed something in and it was magic.

Speaker B: It wasn't like filtered answers and stuff like that. Beautiful. Yeah.

Speaker A: Like it actually gave you the response. And I want to say I wrote this in 2016 was, you know, go to the third page because everything else is promoted and this concept of trust and that's what led my first investment to remember DuckDuckGo, of course.

Speaker B: Yeah, yeah.

Speaker A: DuckDuckGo was really Google 2008. That was the reason why we had done it and we did this six months before Cambridge had come up. So that was a fluke. But that whole idea of, you know, what I said to you about technologies, making sure that there was trust in the technologies, because I think that I was so enamored for a, uh, decade with the shiny new toys that I really didn't appreciate, as I do now, the negative ramifications of so many of these businesses.

Speaker B: There's a huge tailwind happening right now in dual, uh, use defense space. Obviously, uh, we both have investments in that area. What would you say to the founders in the room considering entering into that space from a startup building opportunity? And how would they position themselves to be able to see investors like us find their products or opportunities interesting?

Speaker A: Yeah, well, that's rather amusing, this space, because, uh, we selected this space back in 2019, even before we had started in our original white paper, we believed that the world was going to discombobulate. The world's going to be a much more dangerous place, and defense was going to come right back again. But it's amazing when a government announces billions of Dollars. And now everyone I see is a defense uh, company. You know what the positive side of the ledger is. So the government does want to procure Canadians or Canadian companies. One big problem, there virtually is none in the space. It's been eradicated. And hence they've bastardized the term of what a Canadian company is. And they defined it as a Canadian incorporated company with at least 300 Canadian employees which basically, you know, any subco of a sub of a, uh, multinational, uh, that qualifies. But I do think it's a great space. If you're serious about the space. I'd say right now, uh, what the government is telling me is they use the analogy is they use this, they said it's like we're a pile of manure and all of the flies are just swarming at us.

Speaker B: How do you filter for that? How do you filter for the flies going to the ship?

Speaker A: They don't know how to. That's.

Speaker B: I'll tell you. I filter for it. I asked them when's the last time they went to Ottawa. Yeah, when's the last time you met with someone in the D and D?

Speaker A: I actually asked them for what the product is and it's.

Speaker B: Well, I'm a bit earlier than that.

Speaker A: Yeah, well I'm getting it at growth levels where people uh, of PowerPoints no revenues and want to raise $100 million job.

Speaker B: So for us right now we're investing in what we would call like deeper tech or non disruptible by just the AI models out there, tech. So things that require deep domain expertise, going after really complicated problems, whether it's nuclear fission, you know, uranium exploration, freaking rockets that we're building. Are you seeing companies early on, even at Mavericks, which is later than us, still trying to figure out their product market fit. And how are you able to dissect that? Because you don't know what's coming around the corner from anthropic still.

Speaker A: Yeah, I mean the issue, say if it's based again on an LLM, trying to identify the moat and I gotta tell you nine outta ten times there is no moat. And if an anthropic or whomever decide that they want to go into that business, they can. So that's not very interesting to me. But what I actually really do love right now is the renaissance of deep tech in the broadest sense. Stuff that's really, really hard to do. You know, investors are typically lazy and they love these SaaS businesses where they can do all the math. But now they're actually having to start to actually invest in the technological moats. And I find that quite exciting. It's providing a bit of an industrial renaissance. Uh, we invested a year ago in Kepler and being there, you know, watching it launch and in launching the satellites in the sky, that wasn't normal five years ago. Right. And now people are starting to realize these are where the moats are and Canada's back to actually building stuff.

Speaker B: Tell us about the conversations you're having in the boardroom with some of the CEOs of the pension funds. You speak with CEOs of the banks and insurance companies on uh, where they are at in their AI adoption curve, how they think about working with startups or incumbents and how founders here in the audience should think about positioning themselves to get in front of these large institutions if they are willing to take risk on adopting technologies that are coming out of startups.

Speaker A: Yeah, I think that, I mean the large capital pools really only care about when the company is more mature. They're basically not interested in very early stage opportunities. Not because they don't necessarily believe in them, but they use. This analysis always drives me nuts because the amount of work that we got to do to make a. I think

Speaker B: that's changed for us, John. We have companies right now working with the biggest banks and private equity firms in the world. And these companies only launched four months ago.

Speaker A: Yeah, but, uh, what's the check size?

Speaker B: What do you mean?

Speaker A: What's the check size that they're asking

Speaker B: for a million bucks for a pilot for six months?

Speaker A: As an investor.

Speaker B: No, no, no. To just use the technology.

Speaker A: Oh, to use it. Yeah. I'm talking as an investor.

Speaker B: Investor, yeah, yeah, yeah. The investors are probably going to go through us or you to get access to this exposure before.

Speaker A: Yeah, yeah, that's exactly. Yeah, yeah, yeah. No, if you're talking about their own use, I think that they're still in the experimental stage for sure where they're just starting to hire AI related folks to look at improving their operations. M. I was talking more on their deployment side of the business. I would say it's still rather rudimentary. When you hear large capital pools talking about investing in Canada, ask the next question. The next question is specifically what asset classes you're talking about. And that's when you get the real.

Speaker B: Well, that's where this headline JP Morgan came out with, uh, saying they have 1.5 trillion to invest in Canada, which is not true. Yeah, they have 1.5 trillion to invest in several countries outside of the US to support the businesses. JP Morgan has a large presence here in Canada. But it is interesting to see JP Morgan, the largest bank in the world, saying that they are still going to be putting billions of dollars into the Canadian economy. Of the 1.5 trillion, I think I read they have only deployed around 300 billion. Uh, so they still have a ways to go. But what do you think the government and sort of institutions can do to attract more foreign direct investment into Canada? You know, we have this Carney summit coming up in September. You talk to a lot of people around the world. What are they saying about the opportunities they're seeing in Canada?

Speaker A: I would say again, specifically on the asset class. If you were to ask Carney and his team what do they mean by investments, they will jump to old economy infrastructure. Absolutely. It's ports, it's pipelines, it's bridges, it's tunnels, it's that. That's their AI strategy, but that's the investment strategy. And I was with them a couple weeks ago, as you know, and said, guys, where's all the digital economy infrastructure? And it was like, oh yeah, right, we, we never thought about that stuff.

Speaker B: What do you think about the data center build out problem here? Like we're building data centers. Do you think they're going to be needed when we have more of them in space soon? I mean, you're funding Kepler.

Speaker A: Yeah, yeah, I think, yeah.

Speaker B: You need both.

Speaker A: Yeah, you need both. You need both. Like the great thing on. So Kepler has 40 Nvidia GPUs spinning a lot. Not a lot. But it's more than zero. Yes, it's more than zero. And right now it's on other folks. And we already are computing the first global app right now on global wildfire protection. You can't do that on Earth.

Speaker B: Right.

Speaker A: So those applications that are unique to have the vantage point, Earth observation. There's a new category that you're going to hear. It's going to be the uh, vernacular. It's called ISR services and that means intelligence, surveillance, recognizance. It's only from satellite compute that you're going to see. All warfare is going to be based on this and you'll be able to detect anything, including drones moving from above the water up to I can't remember what certain limit in the sky. And it's. You can't do that on the ground.

Speaker B: It's incredible what Kepler's built. And you know, you introduced us to the founder, um mena, where we'd be able to see the offices and the satellite facility right here in the east end or the west end of Toronto. It's fantastic. Let's switch gears a bit to like the Canadian ecosystem here. John, you spent years building globally competitive businesses out of Canada. Be honest with this room in 2026. Is staying in Canada to build a uh, patriotic endeavor or is it a financial mistake?

Speaker A: You know I'm patriotic, but it doesn't mean I'm stupid financially. The right. I say it's more coincidental. I look to see our raw materials, our talent, our ecosystem, et cetera. I brought for a dinner up, it was interesting. He was the youngest CEO in Silicon Valley. For a long time he was a CEO of Cadence Design System. This guy is now, he's 70 years old now, but 50 years in the semi space. Everybody knows him. He was on the board of Adobe, like all of these like. But he's, he's one of the kings in there. And we had a, we do these entrepreneurial dinners and he was the only one from the valley. And I still opened up the conversation. I said, what do you think about Canada? And you know, the Toronto ecosystem? And he basically said, you guys are nuts. He goes, it is so spectacular here. Why you hate you guys complain. It's so good. The guy, he now has built a company in Ottawa and he flies here every two weeks and he goes like, knock it off. And these are the who's who of, you know, many of the entrepreneurs are in that room there and they were even taken aback. So I think that, you know, what we're really missing here honestly is confidence and ambition.

Speaker B: Sure.

Speaker A: And uh, we get that back on there.

Speaker B: Thank you.

Speaker A: The rest of. Yeah, thank you for this. All right,

Speaker B: I just want to elaborate on that point. So we just sold two companies in the last quarter. One to a large private equity US firm and one to a public company, Motorola. We're really good at starting great technology companies here. We're really bad at scaling them and keeping them here.

Speaker A: That's the issue.

Speaker B: And that's the issue.

Speaker A: That's the issue.

Speaker B: So should we just accept that we're a farm team for the U.S. see,

Speaker A: I don't want to accept it because the moment you start accepting that it's a self fulfilling prophecy and it doesn't lead to a path of productivity and uh, our standards that we have been accustomed to. So that to me is giving it all up and you know, you might as well move to the United States now. There are some people that will say it is more effective for me to move to the United States because all my customers are there. The capital in my space is there, blah, blah, blah, blah. You got to do what you have to do in maximizing the value of your company. I think though that a number of people, like, how many times have you heard this comment? It's always a classic comment. Those Canadian VCs are so conservative and blah, blah, blah. They wouldn't give me a dollar yet, dude. It's because you have a shitty business and they're going to go to the United States and they go and they don't get a dollar there as well. But you don't hear about that. I think there is some validity, particularly as you're seeking growth capital, is the reason why I formed Mavericks. The hole was once you started raising about $30 million, you were capping out.

Speaker B: This is like a pre seed round now.

Speaker A: But yeah, yeah, now it's gotten bizarre.

Speaker B: Real growth equity, but yeah, but real true growth.

Speaker A: We formed the fund on the basis that there was not a single player in Canada. That's a serious problem. And taking a minority stake where there's still maybe not the risk that you're taking, but there's still substantive risk on that for sure. And so the problem for a guy like me though is I'm reliant on getting capital from the large capital pools. Now if they're risk averse, we got a problem here. And I think this is the problem that we have. I think that I'm uh, telling you right, our pipeline is bursting at the seams. My team's getting a little stressed because we have multiple opportunities. I'm in the middle of closing our next fund, but already started seeking US Capital and my partner was out in Germany at super return. We starting to seek foreign capital yet. It's okay, we got, which is fine, but yet we got $2 trillion of pension capital here that has almost zero, not quite, but almost zero in venture and growth in this country. Yeah, uh, that's the part that really annoys me.

Speaker B: We have three companies right now in M and A discussions in our fund, all with US buyers. And it's hard to look the founder in the eye and tell them to go out and continue to raise capital, go on the journey when they haven't had a big win yet. And so what we often find is for the founders that have had those, you know, 30, $50 million exits, which is like the standard venture backed exit in Canada, they're more willing to go bigger on the next one, which is great. And if they build it in Canada, that's awesome. So I don't think we should block it in a way that's like, you know, advantageous to only us and not them. But I do think there is a part of the strategy around capital gains that you and I talk about, that reinvestment into Canadian companies, CCPCs, uh, after an exit happens or liquidity happens, will incentivize more people to continue to reinvest here.

Speaker A: I think that's fantastic. But you know what, the question that I would ask you though, is when somebody is building that business, are they building that business in Canada with the view that I might be capital constraint. So I'm not going for it all.

Speaker B: Yeah, basically, that's where we've done it to ourselves, John.

Speaker A: That's where we get hurt.

Speaker B: We've done it for ourselves by giving them less money at the beginning, lower valuations, more dilution.

Speaker A: It's, it's a so and so again. Uh, you know, you know the story. But say on Kepler, they were looking at only one piece to be a contract manufacturer like mda. Yeah. And we said to them, I don't like that business. I mean, it's, you know, it's. So why don't we build our own network? Own the network and then we create Cost Plus. And it was like, dude, you know how much money that's going to, uh, be? And I go, well, I'll, I'll cut a big check. And it ends up being they got to cut a massive check. But what's interesting, that fully verticalized strategy, and I got to tell you, there was two of the board members are Americans. They were getting frustrated and they were saying, hey, pretend that capital is not a constraint. What would you do? And then, well, of course they would do. This is. Okay, well, that's, I think that, that's what I say. Ambition and confidence.

Speaker B: Well, you're an immigrant from Italy. Right. There's a lot of immigrants who came to Canada here to build. Do you think Canada is giving newcomers founders a real shot at capital and customers? Are we still missing the mark a bit on that startup visa program? What would you do differently if you could?

Speaker A: Well, I speak to a number of the immigrants that come in. It's harder than you think to get into the right circles. Frankly, the innovation ecosystem is very clubby. I mean, you go to conferences, you see the same people in there. I think that when you're coming from another country where you don't have those natural networks, you are going to have to work harder. I mean, this is a great conference to have, but this is where it starts. But yes, Is it more difficult? Absolutely. It was difficult for my parents when they came over. They couldn't get a job because no one would hire them and all of my relatives are in uh, the trades because that's the only thing that they could do. And I think, you know, our level of education now has made it technically easier, but I think we can continue to do a better job. I think again, others say that the Toronto ecosystem is actually quite welcoming. Toronto Tech Week.

Speaker B: Incredible.

Speaker A: Was that not a uh, like it's incredible. Like that doesn't like we community driven event? Yeah, like no. Really no external speaker. So thank God that somebody came up with the idea of Toronto Tech Week because without that it was a little

Speaker B: bit harder for us to meet a golden and yeah, huge, huge kudos. We have a few minutes left. I have a bunch of questions on the slido, but it's just so hard to choose if someone wants to stay, stand up, introduce themselves and ask a question loudly for John and I to answer. We'd be happy to do that with a few minutes left. So anyone want to stand up and ask a question to John and I? Go ahead.

Speaker A: No questions about Italy not making the go ahead by the way.

Speaker B: Yeah, absolutely. The question was, are people, you know, she's from the U.S. she's saying there's a lot of confident, ambitious people. Do we have confident, ambitious people here in Canada? Absolutely, if I could swear I would. We have a company called Canada Rocket Co. John and I know the founder very well, Hugh Colias. He is building the SpaceX of Canada and when he called me to tell me he wanted to do this, I almost hung up the phone on him and told him to F off because could not believe he came with such confidence and arrogance into thinking he could build the SpaceX of Canada. And he said watch me. And so we gave him a little bit of time to think about that and he came back with his co founder who's a gentleman by the name of David Tenney, who was one of the early SpaceX employees who built the first Falcon engines to build this. And he said, this is my co founder, he wants to build this with me and he's going to bring all of his SpaceX friends back to Canada to come and build this company and they've already raised over $20 million, they've got a huge contract pending with the government, they've got launch space, all these amazing things and they only started the company in January. So the answer is yes, there are very, very confident people. And the great thing is there's talent that are coming back to Canada to build for the first time that I've ever seen, maybe even. John, how about you?

Speaker A: There is no shortage of ambition or confidence on the founders. In my view, there's plenty. Like I said, our pipeline is the deepest it's been since, uh, we had started. Where I get antsy on the ambition and confidence is really on the access to capital side. I think that we are largely on a risk off basis. So when you're hearing people wanting to invest in Canada, it's like, sure, you guarantee me a rate of return and I'll put money. And I was like, screw off, I'll do that myself.

Speaker B: This is a real estate.

Speaker A: Yeah, go ahead.

Speaker B: So family offices are quite new. The question was, are their family offices participating in the innovation ecosystem? Yes, there are. There are. A lot of them are our investors and John's investors. We know a lot of them. There's about 600 of the wealthiest families in Canada that are considered family offices. You got to remember one, they've all made their money in real estate and resources so very different than what John and I do. Two is they're only on maybe third generation, which is the generation that usually breaks. First generation makes it, second generation grows it. Third generation basically spends it all and loses it all. So there is this sort of dynamic here that we're seeing, but it's also an educational gap. There is a huge gap in understanding where we are in our innovation cycle and the time it takes to see startups get to the level like they are now with the outcomes we're seeing, you know, with the IPOs and stuff. And that time horizon is just not what family offices are used to. They're used to three five year hold periods, 15 to 25% IRR net returns, very, you know, private equity, real estate, fixed income structure. What John and I do is just not in that sort of like understanding for them. So it's an educational thing. We're doing the best we can. It's called the J curve. You know that J curve where you lose money for the first three to five years and then you have that explosive return at the end. We're still teaching them what a J looks like.

Speaker A: Yeah. And the other one is foundations and endowments in the United States. Huge, massive role. We got almost none of that.

Speaker B: Vicki is an incredible educational tool for family offices because it's that first in, last out, the multiplier effect. So Vicki is a great program that the government has for families to learn about it, but they still don't know much about how it works. But great question, uh, in the back there. The question I believe is how do family offices in other countries learn about our ecosystem and start to slowly deploy capital into the ecosystem? Is that the question you're asking? Got it. Do you want to take this one first? So we have a lot of foreign family offices that are looking to deploy capital into Canada and are already in our funds. What we have done is a bit of a different approach though. We have tried not to convince them that we are the best fund manager for them to invest in to get access to the ecosystem. What we actually get them to do is to look at, ah, the ecosystem as a whole and present them opportunities across the ecosystem from other funds as well. And we actually get them to meet the founders and explain to them what they have seen in this ecosystem, why they have chosen to build here and the reasons why international capital is coming into here. So we have investors from Hong Kong, Singapore, South America. Uh, I think that's it. Maybe South Africa is the last one I'd say. But for us it's more of an educational thing about how the ecosystem here works. We only have five major cities here for building of talent and ecosystem and startups, let's be honest. Right? So they're like, well, what's the difference between Vancouver and Toronto? There actually is a big difference between the ecosystems and we have to explain that it's New York, San Francisco, same thing. So it's more about them understanding the ecosystem at play rather than just trying to find a fund manager. Does that make sense? Question was what are founders doing right, to get answers for yes from us and what are they doing wrong?

Speaker A: People have asked me, uh, on the successful founders that I've invested in, is there any thread in there that pulls together? And there really is one. And it's really on their purpose driven mission of building the business. As you guys all know, building a business is extremely difficult. Uh, there's a lot of despair at many times, you know, the results don't really show up. Uh, and you need to be passionate about the business. I always use this example on uh, Shopify. Toby telling me, you know, in 2013 or 2012 that he thought it was unjust that the artisan who would build or make their product couldn't sell it easily and had to hire a programmer to build a website. He just thought that why does that even happen? Ask Toby that same question today and he'll give you that exact same answer. And it was his mission. And no matter what, you just knew that he was not going to sway from that mission. So it's not necessarily the smartest. It's really, uh. Although he is extremely smart, it's that persistence of purpose. Because the number of barriers that are going to fall in your way are enormous. And 99% of the time, something will go wrong and something will badly go wrong. If you're really just focused in on the money and the result, you're going to give up very, very easily. So I would say that's been the one thread that's very, very.

Speaker B: Yeah, for us at the early stage. Completely agree. The spikiness is what we call the founder has in terms of, like, how they think about their mission and vision and how they just almost ignore a lot of the truths out there, but also their ability to say no to things early on. We actually like to hear founders say no. It's like, why don't you go after this market? Or why don't you go after these customers? And when they say, no, I don't want to go after those markets because it's distraction to our mission early on. I respect that a lot. We call it like the red shiny ball syndrome. When they see a red shiny ball go across their face, they jump at it like a dog. We don't like that, actually, because then there's just a bunch of opportunities that are going to come at their way that they're going to go and chase. And so sometimes your early customers, if you choose them wrong, they can be a death blow to your business. If you choose them right, they can accelerate you to building the right product for the future. And I think that's what Toby was figuring out at Shopify. I think that we have time for one more. So the question was, I don't know if we said they don't have an edge, but let's just say we said it. Canada has no edge in AI and we have all these resources. Why aren't we just doubling down on the resources we have to get an edge in the AI race? That was the question.

Speaker A: Go ahead. I would say that it's spotty, but we're doing exactly in some of the areas that you're describing. I'll give you just two areas. So on the data center front, you're hearing about the opportunities leveraging, for example, our natural gas to power. These, uh, are natural gas that are really just being burnt off of the processing of or the refining of the oil. There's opportunities there, but I will give you a more specific opportunity. The fastest chip sets in the world is the, uh, actual wafer of which the chips are made. The most common use for AI inference today is silicon phosphide. Sorry, uh, you hear about silicon photonics. So photonics is using the speed of light to connect all of the, uh, chipsets together. The fastest chipset or wafer in the world today is indium phosphide. Indium is a rare earth byproduct of zinc. Canada's got lots of zinc. So we are looking right now at. And by the way, that indium, um, compound, it's a semiconductor compound, 98% of it is largely produced in China. Again, if there's a problem here, you can't make the actual substrate.

Speaker B: But the reason it's produced in China is just explained to the audience is not because they have more of it. No, it's because tools to actually refine.

Speaker A: And they had the willingness to do this because you actually don't make very much. Uh, and traditionally part of the challenge was environmental because the rare earth materials would be in these tailing ponds. Right. There's been new technologies now to mitigate.

Speaker B: Yeah. So what John and I are saying is that it's not that we don't believe we have an edge, it's that believe we don't believe we have the permission to gain the edge by a lot of the regulations that are put on us right now. And we don't have the capital willing to support these large projects that need to go out there and build before it actually becomes revenue generating or profitable, for God's sake. You know, that's the problem, I think we're saying. So it's a mixture of permission and a mixture of the capital willing to put the money in the ground and probably not see a return on it for 10 years potentially. And that's what we're trying to do, obviously at different stages. But great question. I think that wraps it up. John, thanks again for everyone joining us today. Really appreciate it.

Speaker A: Thank you very much. Foreign

Speaker B: thanks for tuning in to another episode of Tank Talks. We hope you found today's conversation as insightful as we did. If you're enjoying the show, we've got three quick things to ask of you. First, hit that subscribe button on your favorite podcast platform so you never miss an episode, whether that's Apple podcasts, Spotify, Google podcast or YouTube. Next. Follow us and stay up to date on upcoming episodes and behind the scenes content on social media with Twitter, LinkedIn and Instagram. And lastly, share the love. If you found value in today's episode share with a friend or colleague who'd benefit too. Your support helps us bring in more amazing guests and keeps the tank talks engine running. That's it for today. Until next time. Keep disrupting and innovating.