Mel Mattison: Why AI Demand Won't Collapse
Transcript — TFTC
Article: https://www.tftc.io/mel-mattison-ai-demand-bull-case
Transcript page: https://www.tftc.io/mel-mattison-ai-demand-bull-case-transcript
Published: 2026-08-05
Machine transcription, lightly cleaned; may contain errors.
========================================================================

[0:01] Marty Bent: Mill Madison, welcome back to the show, sir.

[0:04] Mel Mattison: Well, thank you for having me, Marty. I appreciate it.

[0:07] Marty Bent: Yeah, well we're gonna talk about a lot of things today. AI, is it real? Is it hype? Is it a mixture of the two? We've got the Fed, obviously Warsh has taken the helm since we last spoke. we we've got some micro stuff to talk about in terms of individual memory stocks that that you're following, and then we can get into the fiscal side of things, how how much debt the US federal government is and how they can sustain that debt and all the entitlement programs that they've built up and then probably ending it with Scopuscent's sort of call for a new economic regime at the New York Economic Club a couple of weeks ago, which has been a hot topic in my circles.

[1:04] Mel Mattison: Well AI is definitely one of two main forces that I see driving markets right now. I think it's AI, and then what I wanna just kind of encapsulate because a lot there's a lot of interplay, is just kind of geo macro stuff. And you know, d that's Iran, that's Ukraine, that's the Fed, that's fiscal deficits, and those are things that The the reason I categorize those together is because they're not related to like earnings, right? So another simple way to say it is what's going on in the broad market, like broad macro, and what's going on with earnings in company specific. And earnings are all about AI.

[3:29] Mel Mattison: downside, then you can't invest in anything. So I will grant you there could be some big memory breakthrough. But absent that, you have companies like Micron signing multi-year agreements, trading at six, seven times earnings. You have capacity growing at two, three hundred percent a year, not capacity, demand of what people want growing at two, three hundred percent a year. while capacity is only growing at twenty, thirty percent a year, and you have companies like Apple saying we've got to raise prices, you've got I mean, if if if there's not truly a a memory shr shortage, if there's an Ed Zitron, you know, false narrative out there that this is all BS, then why is Apple raising prices?

[5:47] Mel Mattison: like the last forty eight months. So I think everything just keeps getting cheaper and people just want to say you gotta sell because it's all a big bubble. And I just don't see it.

[5:57] Marty Bent: Yeah, I mean the the Ed Zitron call out is funny 'cause I've seen his commentary and him do the sort of the the cycle of going through C N B C Bloomberg and getting clipped out on X and he's saying there's nothing there, nobody wants these things. And then I'm juxtaposing that to my usage of it, which has gone up I think personally exponentially if you're just looking at all the agents and sub agents we're running here. And then you project that forward of more people adopting this, particularly agents, and we're definitely not even at the one percent ad adoption threshold of of everyday people. And then you think of like robotics and self driving cars and all the compute that's gonna be necessary.

[6:37] Mel Mattison: No. W and the memory demand there is like exponentially more. Like it's it's it's ridiculous. And and Ed, I I've listened to I'll be honest with you, I've probably listened to thirty to forty hours of him on podcasts because I do not want to listen to a bunch of bulls tell me how great AI is and just buy micron. I want to listen to the people that are saying, Here's the problem with the story. And he has certain points. And but the thing about him is one one thing I'll say about him, he's one of the cleverest speakers I've ever heard. Like he is so quick and fast with the remark, but he's so quick and fast with with the remark that often you'll listen to him and you're not really processing what he's logically saying.

[8:58] Mel Mattison: That went public in like the 80s. Microsoft, same thing. These things went through that. They had huge market caps. They came out of it. You also had companies like Google that weren't even public until after the dot-com boom. Tesla, same thing. And so, like the internet and everything that happened there was real and it was a huge thing, and it became a big deal. Did we have a huge blow-up top and then a collapse? And then an even a slow but steady rise to even greater market. I mean, what was the Nasdaq when it collapsed? I think it was like four thousand. It's like thirty thousand today.

[11:24] Mel Mattison: I think you're just missing the big picture.

[11:26] Marty Bent: Yeah. I mean before we get into Trump accounts, which I know you want to cover with those passive flows that many people aren't talking about. And I've said this multiple times over the last six months. Like this seems like a there's like an equities market, not like a bailout, but like a a mechanism to produce passive flows. But before we get to that, stick Yeah. Yeah, that's a great one. Yeah. But like bringing it back to like Zitron and some of his criticisms, like one of which being a lot of these companies, particularly

[11:42] Mel Mattison: De Ponzi needs more sources. And and that's a good one. And they're gonna keep doing it. So you wanna fight it, go ahead.

[11:55] Marty Bent: The hyperscalers have gone from buying back stock to hitting the ATM to issuing stock, raising equity, and even going into more debt. And they're highlighting that as like, this is the end. I've been thinking about this a lot too, because you do want to check your priors and be like, okay, where where am I maybe having a blind spot here? But like when I see that, it's like you could take that signal that way, which is like these companies are desperate. They're diluting equity holders to raise cash to keep. Piling money into a loser is what Zitron would say. But the other one is like, well, maybe there was just a period of time post two thousand eight where the allocation of capital warranted the sort of risk matrix that these companies have warranted, like, hey, we just buy back stock, try to juice EPS.

[13:11] Mel Mattison: I mean, everybody used to complain when they were doing all these buybacks like they have nothing to invest in and it's a bad sign because they're buying back stock. And and now it's like, no, well they're not buying back stock anymore. And so look, I mean Apple to me just hitting new highs today and it's up over fifty percent in the last year. I mean, like that's something I wanna stay away from. Like we we just keep going through these micro rotations and it's like, okay, Apple's gonna have to raise prices. It's up over fifty percent in the last twelve months. It's at an all-time high.

[15:40] Mel Mattison: And now they have no more AI spend to do, they have no more debt to pay, they have no interest expense, and that's if AI fails. So if Zitron's right and AI is this big nothing burger, and the big everybody's gonna he's he's said things like, we're gonna look back in a few years and say, What the hell were we doing building all this compute? Well then number one, you gotta take away the threat, right? You gotta say, well, there's no threat to Google, because people might say, Well, Google's No, if AI is a nothing burger and it's going away, then Google's gonna still have its moat, Amazon's gonna have its moat, Microsoft's gonna have its enterprise software moat.

[17:58] Mel Mattison: And so do I think like you can just buy Micron and hold it for the next 40 years and retire? No. But do I think Micron's got a longer window of stellar returns than six months to a year? Yes. I think it's got a multi-year time horizon for outperformance. And I think a lot of these AI names do. I think a lot of these chip stocks do. I think if you look at the Korean market, which has been selling off, it does. But you look at the Korean market, it's 50% of the Korean market is two stocks. It's SK Heinek and Samsung. The average stock in the SP 500 trades within a 50% band every year.

[20:24] Mel Mattison: It's like, my gosh, there's open claw. my gosh, there's, you know, anthropic is cash flow or heading into profitability. But no, they're not because actually, you know, Colossus, they gave I I mean, all this stuff keeps going on. And it's just meanwhile, the valuations keep going up, the the profits keep going up, and I think waiting and trying to call the top. has just been the dumbest thing you could possibly do because it it's it it is going to top and and I've had some thoughts that that's probably not too far away. I don't think it's necessarily five years away. Could be two, three years away.

[22:43] Mel Mattison: I think a million dollars is a I I mean five hundred thousand dollars is a pretty cheap price to pay. And that's what people are paying. I mean that i I I think some of the cheapest stocks I've ever seen in my life are these Korean memory names. I mean, it's just I just don't understand w how like three times earnings Growing at sixty-five, seventy percent, and that's what I mean they've been growing well over that, but this is what's forecast for for fiscal year twenty-seven. These peg ratios are ridiculous. And people are selling it and saying it's a bubble because they believe what? That somebody's gonna come out and there's gonna be a memory ferry that figures out or look, anthropic and open AI can blow up.

[25:01] Mel Mattison: I mean, like literally, like he's talking about NVIDIA possibly going he said it was unlikely, but he said it's possible NVIDIA is gonna go bankrupt. I mean, Marty, th this amount of fear mongering and then it it's played into you listen to the kids booing at college. Like there's this AI a lot of people have talked about Sam Altman and Dario Modai being like the you know, it's like Mussolini and Hitler. Like they're they're very bad spokesmen. Like nobody likes these guys. They they come across as arrogant, they come across as condescending, they they just they're not relatable and the stuff that they say a lot of the times, it just puts people off the wrong way.

[27:17] Marty Bent: That's a good segue into the next topic, which is the Fed in macro, which we were discussing before we hit record. A lot of people are beginning to price in rate hikes later this month. I think it's at fifty fifty right now. If you look at if you look at the the probability of there being a rate hike or them holding it steady, what's your take on this, Warsh again coming in between now and our last interview? And how how are you viewing all of this, Warsh? At the helm and

[27:47] Mel Mattison: I I do not think there's gonna be any rate hikes. I think there's gonna be rate cuts this year. somebody who you know I respect, I've referenced him many times is Jordy Visser. He's talked about a potential for Warsh to hike actually in July and do a surprise hike. This would essentially like assert Fender independence, but then talk it down and say that's it. so like a twenty-five basis point hike possibly. But Nobody is even talking about like a major hiking cycle. And if you listen to Walsh, I think he comes off a little conceited sometimes too. I think he probably needs a better haircut, but he just does not strike me as someone who's oblivious to the broader macro concerns.

[30:11] Mel Mattison: You know, rents are I think rents are over five thousand dollars a month right now for an average one bedroom in Manhattan. you know, yeah, prices have gone up in places like that. But if you live like where I do in North Carolina, I mean, you know, the a house has not moved up or down more or less in the last four years. You know, you gotta go back to the last World Cup, you know, to see you know, house prices cheaper or more expensive than they are right now. I mean, like housing just has not moved. And yet at the same time, during that period, earnings, nominal earnings, and I'm using nominal because housing prices haven't moved, are up like over twenty percent.

[32:41] Mel Mattison: That's absolutely separate from whatever you know the two year yield is at. And so all he's gonna do is hurt the average person, and also stymy bank lending, which I think he wants to stimulate. and that and that goes into Besson and Besson's speech about you know, tokenization, about deregulation, and so there's this big strategic narrative. That I think the Trump administration has right. And then at the same time, we're fighting this big tactical narrative that I think the Trump administration has wrong. So they keep making these tactical mistakes, these fumbles, but their long-term strategic vision of embracing digital assets, embracing deregulation, letting AI, you know, letting America compete, bringing back manufacturing.

[34:08] Marty Bent: Yeah. Well let's talk about the big strategy from Bissent and Trump administration, what they're getting right long term, and then follow that up with what they're getting wrong and how they can maybe write course there.

[34:22] Mel Mattison: Yeah, I think you know, what like I said, what they're getting right is, you know, deregulation. I think one thing that has not been understood enough is the relationship of the balance sheet to interest rates and what Warsh has said. So I think this is where the market has been getting it wrong. The market looked at Warsh, they see him as a hawk. They think he's gonna like decrease the balance sheet. That's gonna decrease money supply. Look, inflation and inflationary aspects are primarily driven by money supply. Money supply does not necessarily need to be increased by the Fed balance sheet. It can be increased by bank lending, it can be increased by the velocity of money.

[36:45] Mel Mattison: We had so much money being created through private credit. That's essentially you know, g putting a ceiling on M2, right? And so we don't need the Fed's balance sheet to expand if commercial banks start lending more. This is what Warsh the deregulation, this is all part of the best and plan. this is, you know, We don't need to raise interest rates every time growth comes up. We d we don't need to have a huge Fed balance sheet. You know, we can absorb treasury supply through stable coins or tokenization. We can so there is this bit broad strategic plan that I believe is out there that I think is going to be necessary because if you do not address and we're already at over eight hundred billion dollars in net interest expense this fiscal year and we've still got the rest of July, August, and September.

[39:10] Mel Mattison: The oldest boomers are just hitting 80. And if you look at fiscal projections for Medicare and health spending, it's gonna blow up. And so we're getting interest expense blowing up. At the same time, we're gonna have healthcare spending blow up because boomers are hitting what I would call peak medical care years. And that's not gonna end anytime soon. And it's just gonna get worse. And so like if you look at the we're gonna hit forty trillion dollars in in US debt in the coming months, we're gonna hit over a trillion dollars in interest expense year to date in a couple months. how you can look at the US dollar and not say, I wanna own gold, I wanna own Bitcoin, I wanna own silver in the same way that people thought was so smart twelve months ago.

[41:38] Mel Mattison: To a regulator, you were like, you know, it was it you did you just didn't do it. I started three different broker dealers: SecFi Securities, Equity B Securities, Vested Securities, they're all broker check. People want to start them as the CEO of them. I went through the whole FINRA application process with all three of them. Two as new broker dealers, one as a continuing application, which is where you buy a broker dealer and then you appro apply for another broker dealer license. Every single time, and we had the best Manhattan attorneys that advised us, we were VC funded, and we had these fancy guys from a firm called Lowenstein Sandler.

[43:53] Mel Mattison: Jordy Visser, he likes to talk about agents, the a the agentic economy. It's gonna be, you know, they're not gonna be using dollars. once people start to realize this, they're gonna say, my gosh, I need to be there. Just like a year ago, people were like, my gosh, debasement trade, I need to be there in gold and silver. Then it ramps up, you know, 150%, sells off, everybody hates it for six months. And then and and then it starts again. And so I think the same thing hap is happening right now with some of these chip stocks. And what was hated, some of the hyperscalers, like I said, I mean they're having amazing returns.

[45:52] Marty Bent: Yeah, so let's jump into that. How how I mean what should we follow there in terms of adoption to then try to port to the flows that you just described? Like what wha what 'cause everybody has access to a Trump account, you have to physically go out and get them for your kids. w how's this gonna play out in your mind?

[46:17] Mel Mattison: I think it's being way underestimated by the markets because I don't think people fully understand the tax advantage of these accounts. And it's not just the thousand dollars that the government's gonna give. It's the ability for anybody to give to a kid that and and it's five thousand dollars a year and it's available to every single person under eighteen years of age. So Every American under 18 can go create or their parents can go create a Trump account. Not ever if if you're sixteen, you're not gonna get a thousand dollars. So it's like you have to be born, I think, since Trump took office or something like that.

[48:42] Mel Mattison: Einstein talking about compounding the most powerful force in the universe and you're like, you know, is that really right? Like you put ninety thousand dollars into some kid's account and he's gonna have thirteen million? Well, yeah, because of compounding, because it's that many years. It's sixty years. We're starting from year zero when the kid's born to fifty nine and a half. So it's called sixty years. Nominally SP's up ten, eleven percent a year. You're putting that money in And so whether that's going to happen or not, that can be up for debate. What I don't think is going to be up for debate is that there's going to be a lot of well-off grandparents putting money into their grandkids' accounts.

[51:03] Mel Mattison: And he's just gonna start hiking rates because inflation's at three point two percent, which I wouldn't be surprised if tomorrow, and it's always risky making a call less than twenty-four hours before a number comes out. I wouldn't be surprised if we're well below expectations tomorrow on CPI and even more low next month. And I think we could have some deflationary prints, like literally negative numbers. especially if the Trump administration would get out of its way a little bit with the SirAN stuff, because those are the tactical headwinds I talked about. But when you had oil go from a hundred and twenty dollars a barrel down to six handle in a matter of weeks, I mean that's gonna be good for reducing inflation.

[53:05] Marty Bent: You've always been very bullish on the show. And it seems like it's I mean, the AI stuff, like when he when I hear Ed Zitron I have to wonder, like, is there some sort of nefarious incentive driving it? 'Cause or is he not using it or is he not using it the right way? 'Cause you you can

[53:22] Mel Mattison: No. He just sells his research. He makes money on that.

[53:27] Marty Bent: Yeah. But is the research good? Like 'cause I'm using the stuff. We've like basically automated a ton of the back end stuff, a TFTC, and like I said, my token usage is going up exponentially, it feels, particularly with the new frontier models launching. We're hopping on them right away. And it's undeniable. Like objectively, they're just looking at my micro situation, looking at my business, like it it is helping us be crazy productive. Like it is a multiple and productivity. the cost are manageable for us as a small team. Maybe that's different for enterprises. And it does seem like enterprises are trying to figure out how to manage token spend and what the right mix of frontier models and open source models is. But I think to flatly say that this is a bubble built on idiotic technology that that hallucinates and

[54:19] Mel Mattison: Nothing. Yeah. He he claims it's nothing. He he doesn't even make the claim that like open AI and anthropic are, you know, insecure business models. Like he literally makes the claim that we're gonna look back on this period and say this was the biggest waste ever and we're gonna be like massively oversupplied with compute. And I would I like will grant him That there's no guarantee, like I said, that open AI becomes a ten trillion dollar company one day. In fact, I kinda doubt it ever does. But that does not mean that AI, that data setter builds outs are worthless. That that you know, meta like I I made this argument months ago where I was basically like, Look, everybody's talking about shortages of compute.

[56:35] Mel Mattison: When people are willing to draw out, like where's SpaceX gonna be in fifteen years? Like who the heck knows where SpaceX is gonna be in fifteen years? And so that deserves a pretty heavy discount, even if you grant that it's gonna access this twenty-four trillion dollar TAM and it's a like granted it all, but you gotta discount the hell out of that. but was it a big leap once the automobile was invented to say automate Autos are gonna be a big thing. Didn't know if Ford was gonna be great, Studebaker, a lot of companies went bankrupt, a lot of railroad roads went bankrupt, a lot of dot coms went bankrupt.

[59:01] Mel Mattison: VHS, then they got betas. My dad didn't like that stuff. There were people that got cable. My dad didn't like that stuff. there were kids that got video games. I mentioned the Commodore 64, the Atari 2600. My dad didn't like that stuff. He bought a new video game console called the Atari 400, which even though it was less than the Atari 2600, was actually more expensive, but it was supposed to be a better computer, but nobody liked it. Everybody liked the 2600. So like All this stuff happened. All this stuff happened and these companies came and went, but they have literally just transformed the world over what is a relatively short period of time.

[1:01:30] Mel Mattison: to France, to the United States, is to even China. Just swamped in it. And the basement trade's not hot right now, that doesn't mean it won't be, you know, three or six months from

[1:01:43] Marty Bent: Yeah, when it's not hot, it's the best time to allocate typically.

[1:01:46] Mel Mattison: It is. It's it every single time people hate the hyperscalers. It's ridiculous how many times when everybody thinks it's the sm smartest thing to do, that's when everybody's invested in it and it's got nowhere to go but down. Like about four weeks ago, everybody's like, you gotta bet on the the the builders, you know, you don't bet on the spenders, get out of the hyperscalers. Everybody was saying that. Hyperscalers knocking the socks off of the you know quote unquote, you know, the the picks and shovel plays in the last three or four weeks. Because it just gets overdone and the narrative gets old, and now you've got micron down a few hundred bucks from the high and you've got meta up eighty bucks in the last couple of trading days, and people are saying, What the hell happened?

[1:04:11] Mel Mattison: And you wanna take short term bets, take short term bets, but I do, but I do it with a small percentage of the portfolio. Most of it is a long term diversified portfolio.

[1:04:24] Marty Bent: I that's sage advice. Any predictions between now and the next time we meet at after Q three?

[1:04:29] Mel Mattison: what what's a let me see if I can come up with a shocker for you, Marty. You know you're my favorite person to talk to. I don't know if I have anything that's actually too out of consensus or crazy. I d I do think that with gold right around four thousand, Bitcoin not too far away from sixty, that the next time we meet, I would I would wager that gold and Bitcoin are going to be outperformers. relative to equities and and even some high flying equities. I I would much rather invest in Bitcoin or gold right now than I would Micron or SpaceX. It doesn't mean I don't think they're they're they're not they're they're they're not good.

[1:06:53] Marty Bent: Yeah. Yeah, it is funny. The four year cycle does seem to be repeating. And no, it's always good to zoom out with you, Mel. I think this was like the most dense high signal sort of we ran through all this in an hour and eight minutes here now. So thank you for coming on. Let's catch up in the fall when Q three is is over and I'm always looking forward to our end of year. retrospective and predictions for the following year. So this is gonna be a fun one.

[1:07:25] Mel Mattison: Yeah. I'm too. I and I've got a few that I'm I'm actually thinking could still come to fruition that nobody thought. One of them was Mag Seven outperforms S P five hundred. I had Bitcoin outperforming gold. You know, I I I had some things that didn't seem to make sense back in January, that I that I predicted at the end of the year, and they still don't make sense to be honest with you. But I think there's still plenty of time left in the year for some of those things to come full circle. And I think a lot of people that kinda gave up on the Mag Seven early in the year or gave up on Bitcoin and thought gold was the B's knees as it hit fifty six hundred and there's no way Bitcoin's gonna outperform it. I still think Bitcoin's gonna outperform gold for twenty twenty six and I still think Mag Seven's gonna outperform S and B.

[1:08:20] Marty Bent: Clip it and send it to Peter Schiff freaks.

[1:08:23] Mel Mattison: Mm.

[1:08:25] Marty Bent: Mel, we'll we'll catch up in the fall, brother. Enjoy the rest of your summer.

[1:08:29] Mel Mattison: All right, thanks, Maureen.

[1:08:31] Marty Bent: Peace love freaks. Tiki.