Is The AI Bubble About To Pop? Here's The Real Answer

Right now, the single biggest winner in this entire market is sitting right on the edge of a $6 trillion price tag. One chip company worth more than most entire economies on Earth, and it recently announced the largest stock buyback in history at $150 billion. Everybody already knows that headline. So, here's the one that actually keeps me up at night. The loudest bear on Wall Street, the same guy who called the 2008 housing collapse, just put real money down that this whole AI run is about to crack. If you're the type who panics and dumps everything on a scary headline, this video might not be for you. But if you want to see the real risk clearly, stick with me. And this goes way past that one famous chip name. There's a whole hidden layer of companies quietly collecting a toll on every AI data center that gets built, and almost nobody's talking about them. Watching your AI stocks bleed red on a scary headline while you just sit there hoping is exhausting. And trust me, I've been there, too. And if you stick around past the wrap-up today, I left a nice reward waiting for you at the very end, strictly for the real ones who stay until the screen goes black. I'm an investor, not a financial adviser, so do your own research. And the honest research keeps landing on one question that outweighs any single price target. Is the AI boom actually real? Or is it the same hot air that blew up in everybody's face back in 2000? Because those are two completely different problems with two completely different moves. A fake boom means you sell everything and run for the hills. A real boom that just got priced too rich means you hold the right names and wait for your spot. So knowing which one you're in drives every decision you make from here. Start with the demand because this is where the bubble crowd gets lazy. The biggest cloud companies are on track to spend over a trillion dollars next year building this stuff out. And that's cold hard cash going out the door, not some press release. The orders are real. The revenue is already booked. and the chips are sold out before they even come off the line. Here's the part almost nobody talks about. What's holding AI back right now has nothing to do with weak demand. They physically cannot build it fast enough to fill the orders already sitting on the table. And that is not what the top of a bubble usually looks like. But let's be honest about what happens next. A scary headline drops. Some famous investor like Michael Bur calls the whole thing a bubble. And suddenly you can't even remember why you own the stock in the first place. We've all felt that pull sitting on a position you can't defend in one clean sentence. So the second it dips, you're already reaching for the sell button. And going it alone makes it 10 times worse. There's nobody to talk you off the ledge. So a long-term position you swore you'd hold for 5 years turns into a panic sale on some random Tuesday. The pull is real. And I like to do these videos where I present both sides. Let me take you back to 1999 for a second. The internet was absolutely real. It flat out changed the world and every bull back then was dead right about where things were heading. The NASDAQ still fell almost 80% and plenty of real companies went straight to zero on the way down. So being right about the technology and being right about the stock price are two totally different games. The internet winning did not save you one bit if you paid a dumb price for the wrong name at the wrong time. That exact trap is sitting right underneath this entire AI rally today. So, if demand is real and supply is tight, then where's the actual danger? It's buried in two spots the crowd barely bothers to look. And the first one is an accounting line so boring that most people scroll right past it. It goes by the name depreciation. Picture a trucking company that buys a fleet of brand new rigs. If they claim those trucks will last 10 years when they're really worn out in six, their profits look fantastic today and flatout brutal later when every truck dies early. That's the exact game the bears say is being played with AI chips right now. Some of the biggest spenders already stretch the assumed life of their chips from four years out to six. Stretch that timeline and today's profits magically puff up bigger even though the chips still wear out on the old clock. If those chips really age faster than the books claim, a big chunk of these AI profits isn't nearly as solid as it looks. And that's what Michael Bur is rightfully questioning. The second crack is how all of this gets paid for in the first place. A whole lot of this buildout is running on borrowed money and private credit, not just spare cash lying around. Borrowed money works like towing a heavy trailer up a steep grade. And when the baseline cost of money already sits near 5%, these data center loans can run 8% or higher. Once you stack on the risk and that drag only gets heavier the second anything slows down. And real quick, if you're getting value from this breakdown, hit the like button and subscribe to the channel. And if you want my full deep dive analysis or direct Q&A access, come join the community on Patreon. All right, let's get back to it. So, tune out all the yelling about whether this is an AI bubble or not. That's a lazy yes or no question, and it won't protect a single dollar in your account. The real question is a lot sharper, and it boils down to two simple tests you can run yourself from your couch. These two tells are the whole difference between a boom that keeps paying you and one that quietly rots from the inside while you're still standing there clapping. And the best part, you don't need some fancy Wall Street terminal to watch either one. Test one is the depreciation honesty check. Keep an eye on whether these companies keep stretching how long they claim their chips last and whether the older chips actually keep earning good money after the shiny new ones ship. As long as the used gear stays busy and those schedules hold steady, the profits are real. Test two is simply following the money. Is this buildout getting paid for with real cash flow or with mountains of debt and vendors quietly financing their own customers? And the single cleanest tell for both. Watch the earnings estimates because as long as the numbers keep getting revised up, the boom still has fuel in the tank. And the day they start getting cut is the day the whole story changes. Now, let me put my own cards face up on the table. I own a lot of the biggest names in this whole space, Nvidia, and I'm still adding to it. So, I've got real skin in this game. That is exactly why I refuse to lie to myself about the risks. Because if I'm dead wrong here, it's my own money that eats the loss. So, we deal in nothing but cold math. And if you've been following my channel for very long, you know I've been in Nvidia for years and stuck with it even when people were bailing. Fast forward and Nvidia has been hitting record highs, just like my portfolio. The bullcase is clean and it's strong. Demand is real, supply is tight, the margins are some of the fattest in the history of business. And even after this monster run, the leader trades around 17 times next year's expected earnings. Back in 2000, the big internet darling traded closer to 100 times earnings. So, we're nowhere near that same valuation insanity. The bare case has real teeth, too. And I'm not about to wave it away. Michael Bur, the guy from the big short, argues that depreciation games mean reported AI profits are padded and all that debtfueled building could unwind in a hurry. He's betting into 2027 that the crack shows up sooner than anyone expects. And the honest truth is he could end up being right. Michael Bur has built a reputation as the boy who cried wolf. To his credit, he made an immense amount of money when his prediction about the housing market crash actually panned out. However, he has historically been wrong far more often than he's been right, proving that even a broken clock is right twice a day. So, where do I actually land? The boom is real. That's my base case, and that's where my money sits today. But Bur isn't wrong to question the depreciation, and he's got his scope locked on the one spot that genuinely counts. So, I'm watching those two tells like a hawk. Here's how you tell a true fortress apart from a fad when the whole sector suddenly gets scary. Nvidia is pulling gross margins near 75%, operating margins around 65%, and it threw off roughly 127 billion in free cash flow over the past year, while revenue grew more than 80%. You simply do not print numbers like that at the top of some fragile bubble. I like to see gross margins holding right at the top of their class and cash flow strong enough that the company needs permission from absolutely nobody to write out a rough patch. Go ahead and name me one competitor that could actually replace these category kings like Nvidia by tomorrow morning. There are smaller names out there with more upside, no doubt, but they have exponentially more risk. And betting on these giants is a lot like betting on the house. Over a long enough stretch of time, the house always wins. These mega caps are earning billions just in interest on their cash piles while the weaker players are drowning under debt they can barely even service. The rich get richer. And in a high rate world, that gap turns into a straightup canyon. So, let's tie this all the way back to where we started. AI is out of the box for good. And this whole thing is really about which companies survive and compound versus which ones quietly bleed out. You're walking out of here with two checks you can run yourself. Depreciation and financing, plus the one tell that front runs them both, the estimate revisions. But knowing the setup is maybe 10% of the actual battle. Running these checks week after week, then actually holding your line when the ugly headline hits and the market opens blood red. That takes a full operating system, not a one-time video. Watching from the sidelines is easy. Doing it live with real money on the line is a completely different animal. A couple of sharp tools in your pocket won't save you if you've got no daily framework and nobody to pressure test your thinking right when you're about to do something dumb. That right there is the gap between knowing the play and actually winning with it. So, here's the twist I really want to stick with you. The fact that this boom is real makes it more dangerous, not less. Real booms are exactly where the heaviest leverage piles in right near the very top because everybody can see the story is true and nobody can stand the thought of being left out. The.com builders were dead right about the internet and many still got completely wiped out. So, the question stops being whether AI is real. it becomes whether you'll still be standing when the first real stress test of this thing finally slams into the market. And that test is coming whether we're ready for it or not. Here's what keeps you standing when that test finally comes. You hold what you truly understand. Enough to sit tight when the screen turns red. That is what my Patreon community is for. I share every investment I make and the thinking behind it from the first buy through the trims. Take Big Doss. He rolled an old 401k into an IRA. Put it in a simple basket of AI and tech funds and held while it grew from $96,700 past $124,80. He never touched an option. He bought a thesis he believed in and held. The community costs about a buck and change a day, but I don't pitch it hard because I just ask people to look at the numbers we just talked about. How many investments do you need to get right for that fee to dissolve into insignificance? For most members, that fee is completely covered with a fraction of just one investment that pays for years. If you're serious about your money, you already know how that math works. The link is down below. If you stayed through this video, you got the twist nobody says out loud. Drop dangerously real in the comments. Tribe check. The realest booms pull the most leverage to the top, and that is what makes them bite. For those of you who stuck around to the very end, here's the special reward. Everybody out there is busy fighting over the big chip stocks, but the smarter money is looking one full layer down at the physical choke points every single AI data center has to pay a toll to cross. Three of them barely get any airtime at all. The first one is power, and one name to watch there is constellation energy, ticker CE. These data centers need firm aroundthe-clock electricity, and you flat out cannot run a server farm on a cloudy afternoon's worth of solar. Constellation owns the nuclear fleet that giants like Google and Amazon are racing to lock up under 20-year deals right now. And there's only so much reliable base load power to go around. The second one is heat. And that's Vertif, ticker VRT. Pack that many chips into a single rack and they will literally cook themselves because plain old air cooling just can't keep up anymore. Vertive pumps liquid right across the chips to keep them alive. And its order backlog just blew past 15 billion, which tells you exactly how badly this stuff is needed. Third is the one factory the entire world leans on. Taiwan Semiconductors, ticker TSM. And full disclosure here, I own a lot of TSMC. Every advanced AI chip, no matter whose logo ends up stamped on it, has to get built and packaged in their fabs. and they're already running flat out at max capacity. When a single company is the only road across the canyon, that is the ultimate toll booth. And that's exactly where I want to be standing. These are the three choke points to throw on your watch list and really dig into. Follow the toll, not the hype, and you'll start seeing this entire buildout the way the big players already see it. That's your reward for staying to the very end. Now go do the