I Was Wrong About the AI Bubble

Hinzugefügt: 2026-05-12

[00:00:00]I was wrong about the AI bubble.

[00:00:02]Allirds, a failing shoe company, is now New Birds AI. The stock went up 600% that same day. Museum is now Museum.ai.

[00:00:12]The stock went up 129% that same day.

[00:00:15]Tempest, a biotech stock that was started after the founder's wife was diagnosed with breast cancer, is now Tempest.ai. My wife was diagnosed with breast cancer almost 10 years ago and I was just perplexed at how little data was a part of her care.

[00:00:30]>> But everything has changed. We are absolutely now in an AI bubble bubble bubble bubble.

[00:00:35]>> And subsequent videos online saying it's the beginning of artificial intelligence taking our jobs.

[00:00:40]>> I'm a little bit more worried about the labor impact simply because it's happening so fast. People may not adapt fast enough.

[00:00:47]>> I think in the short term it will destroy a lot of jobs.

[00:00:50]>> My name is Dylan. I am an AI engineer and I have worked in technology for the better part of the last decade, working for some of the largest and oldest technology corporations to ever exist.

[00:00:59]And in this video, I would like to talk about what the AI bubble actually is, the scams in the AI industry, and how I could not have been more wrong. Stay tuned.

[00:01:13]I made a video a couple months ago detailing about how I didn't think AI was bubble and that everyone pushing this AI bubble narrative was completely wrong. But first, let's define what a bubble actually is. An economic bubble, also called a speculative bubble or financial bubble, is a period when current asset prices greatly exceed their intrinsic valuation. But that definition could be applied to every growth stock on the market. The price of stocks in the modern era is this simple formula. Price equals intrinsic value, what we just talked about, plus speculation plus hype vibes. I like Mly Fool's definition a lot better. A stock market bubble is a significant runup in stock prices without a corresponding increase in the value of the businesses they represent. A company's valuation should be determined by its business fundamentals, its profits, growth rate, and similar factors. In a bubble, speculation and euphoria take over.

[00:02:13]Pause the video now for all stages of a stock market bubble. Fred Wilson, a VC who funded many of the dot companies, said nothing important has ever been built without irrational exuberance. You need some of this mania to cause investors to open up their pocketbooks.

[00:02:32]But we need to figure out where we are at in this cute little graph. Let's take a quick look at three different companies. First, SpaceX, XAI, and Grock. Okay, that's all there is to talk about here. Open AAI. Open AAI at the end of March said it was generating revenue in the realm of $2 billion per month. But the problem is they are bleeding money to the one company that has been more successful in the AI race than anyone else.

[00:03:02]>> What is the path to profitability? When do you see the lines cross, if you will?

[00:03:08]>> And before I tell you about that company, if you like what we are doing here, we are halfway to 10,000 subscribers. So, please consider helping me and my team out a ton by hitting that subscribe button down below. Anthropic.

[00:03:18]In January of last year, they had annual revenue in the neighborhood of $1 billion. That's a lot, sure. But in the past week, they have sexupled to six times that in a single month. Today, Anthropic has a revenue run rate of $30 billion annually. That is two times the city budget of Los Angeles. But let's contrast that with a couple tech companies from the early 2000s during the dotcom bubble. VA Linux Systems IPOed at $30, first trade at $299. The first day gain of their IPO was 697%.

[00:03:50]You thought that Albert was bad, but what turns this from a bad stock to a steaming pile of dog was that their annual sales were only 17.8 8 million a year. That is a $533 times revenue multiple. That's more frothy than Palunteer, Tesla, and SpaceX all put together. ETY $7.7 billion valuation on Ibo day. A toy company that wasn't really doing tech became a unicorn in the early 2000s. I found a couple quotes about this company from an article in 1999. This is going to be one of the hottest deals to date. They've done just about everything right. I'll tell you in a moment how things turned out for etos. But let's take a look at one more company. The globe.com, the largest first day IPO in history, debuted in 1998. Priced at $9 a share, first trade at 84 bucks. They had a market cap of nearly a billion.

[00:04:54]But the Globe made no money. They lost $5.8 8 million in the first half of 1998. And for a brief moment, both of their founders were worth over $100 million.

[00:05:04]>> And that makes the young co-founders of the globe.com very rich men on paper at least >> until well by the end of the com bubble.

[00:05:11]VA Linux systems was down 97% from its all-time high. Etoys went bankrupt. The globe.com go ahead and put in the comments down below what you think happened to them. But the main point here is anthropic open AI, not you, Elon, have real revenue. $2 billion a month is nothing to scoff at. That is a far cry from ETY's $30 million. I wonder if our AI mania really looks anything like this. AI has revolutionized software engineering. You can spin up a deployed hosted production web application in seconds. Maybe we are in the displacement arc. Maybe the the boom arc. As AI gets more efficient and accessible, we will see its use skyrocket, turning it into a commodity that we just can't get enough of. This is called Javon's paradox. I use AI all the time. I use it to help me code. I use it to help me research the video you are watching right now. The AI companies are making tons of money. Set aside some of the circular dealing and strange financing for a moment, but Amazon wouldn't have invested billions in OpenAI otherwise. Hyperscalers are also spending from cash flow, not debt.

[00:06:20]Unlike the dot era, New York Times says the next economic bubble is here. CNBC doesn't know. The New Yorker says it isn't. But the moment I saw allirds museum and another company starting to make the transition to AI, it started to remind me of something.

[00:06:41]In 2017, the crypto altcoin craze was at an all-time high. Look no further than the stupidest cryptocurrency of all time, XRP. XRP skyrocketed from less than one penny to over $3 in a year. I decided, you know, in the spirit of this video, I'd ask Quad some questions about XRP. Hey Claude, does XRP generate any revenue? Claude, no. Hey Claude, does XRP have any intrinsic value that could justify a nearly $200 billion valuation?

[00:07:14]Claude, no. Final question. No revenue, no intrinsic value, no serious technology innovation. Hey, Quad. Based on everything I have just described, should XRP be worth less than a dollar?

[00:07:24]Quad. No. Wait a minute. During the same heroine rush of 2017, a tea company decided to start focusing on blockchain technology. Would you like the food ingredients stored in your fridge on the blockchain? NASDAQ delisted them the following year when the crypto bubble burst. The.com era had the exact same parallel. Companies simply adding.com to their name and getting huge pops in their valuation. It is here now that I would like to announce that Dylan John the channel is now becoming Dylan John AI. It's really a pricing narrative. The market is pricing the word AI the same way it used to price blockchain. The market is pricing the word AI the same way it used to price.com. Now, it might be a lot easier to understand why Allirds, the shoe brand, says it's now an AI compute company, or Alphaton Capital, a failing biotech, crypto, and Telegram play, is now Alpha Compute.

[00:08:22]>> Welcome back. Alphabet shares have surged more than 140% in the last year in large part due to its AI strategy. I don't think the pivot indicates a bubble. But I think the fact that the pivot resulted in a stock price explosion indicates a bubble. Stock prices increasing means higher demand means a lot of investors saw all birds AI and thought it was a good buy. That is wild. But it's not just the rebrands.

[00:08:48]The deception, lies, and euphoria has become more obvious, more loweffort, and more gross with each subsequent bubble.

[00:08:55]strangely resembles the dot hysteria where every business would slap a website to get funding and the market supported that behavior. It feels exactly like crypto. But the moment I saw that there's an AI for that, I knew generate romantic couple photos with AI.

[00:09:12]The first autonomous AI marketer that executes while you sleep. God knows we need more AI generated content on the internet. I think we can move ourselves up that chart now from certainly AI displacement to AI boom. The final nail in the coffin, AI capex. AI capital expenditure contributed more to US GDP growth in the past two quarters than all consumer spending combined. This is the first time that has ever happened.

[00:09:40]Hyperscaler Capex is expected to be 2% of the US GDP, nearly $646 billion. More than the cost of every NFL, MLB, and NBA team combined. How extreme is AI data center spending? It is to the point that even Chinese president Xiinping is warning about it. President Xiinping has issued a rare and direct rebuke to local government across China, questioning their rush into the same emerging sectors, AI, computing power. With more than 250 data centers under construction in his country this last week, he cautioned against the spending rush threatening to intervene. Our economy might just be three AI data centers in a trench coat. As a result, hyperscaler cash flow is imploding. Amazon, Google, Meta project a 45% reduction in free cash. AI related capex is approaching 30% of net sales, three times more than historic norms. Oracle's 5-year credit default swap spreads nearly doubled since September 2025. The bond market is pricing in meaningful credit risk at a company that was previously investment grade. We've spiraled out of control.

[00:10:51]Bloomberg took it a step further and they coined the term AI circular deals.

[00:10:56]The basic loop is a cloud giant like Microsoft, Google or Amazon invest in an AI lab, but the investment is largely paid with AI cloud credits, not cash.

[00:11:07]The AI lab spends those credits on the cloud giants infrastructure to train the latest and greatest AI models. But the cloud company, of course, books that as revenue, which increases the stock price, which justifies the next round of investment. Microsoft gives OpenAI credits which then those credits are spent marked as revenue back to Microsoft. It doesn't really make any sense. Microsoft committed $13 billion to OpenAI with a significant portion reportedly flowing as Azure cloud credits. Google and Amazon have both invested in Anthropic under similar structures. Nvidia is playing the same but uh slightly different game profiting from GPU sales. It reinvests in AI infrastructure. A company like Cororeweave. After Nvidia invests in Cororeweave, where do you think that newfound capital goes to? Coreweave needs to buy GPUs for their AI infrastructure. They head to Nvidia. The money is flying around in a circle doom spiral just like this video. Now, the message isn't that these companies are too big to fail, as we've heard in the past with all of the banks and then all of the oil companies during the Enron fiasco. It's that they are too entangled to fail. Now, we've already seen an example of what happens when even a tiny little piece of this fractures, the deepseek moment. Just over a year ago, China released a model that was comparable to Chad GPT. And Nvidia proceeded to lose over a half a trillion dollars in wealth creation in a single trading day. Shares in the advanced computer chip maker Nvidia plunged overnight after investors became concerned by a cheap Chinese artificial intelligence program called Deep Seek.

[00:12:59]>> That's about Elon Musk's net worth. Now, the successful dotcoms of the late '9s and early 2000s had a few things in common. They all vowed to change the world. They had crazy high valuations, and they were wildly unprofitable. That is us now. That is me now. The people who got wiped out in the dot were also young. They were also inside the industry. They were also software engineers, but they couldn't read the label from inside the bottle. I think I think I've learned my lesson now.

[00:13:30]Sam Alman in his own words, "When bubbles happen, smart people get over excited about a kernel of truth. Are we in a phase where investors as a whole are overexited about AI?" My opinion is yes. Daario Amode, we are near the >> the end of the exponential.

[00:13:46]>> Data Bricks CEO Alli himself called his own $62 billion valuation peak bubble.

[00:13:52]AI bubble. Yes.

[00:13:54]>> Now me, we're definitely in a bubble.

[00:13:56]We've reached peak euphoria. The question is how long can we stay there before everything implodes? In 1997,998 and 99 people were all trying to predict the future. The question is where in the euphoria are we? No, not the TV show. I think AI being here to stay, a great product with millions of customers can also exist with the fact that some of these companies are insanely overvalued.

[00:14:21]Anthropic trades at $600 billion on private markets right now, but Quad is still awesome. Chat GPT trades at $800 billion on private markets, nearly a trillion dollar, but Chat GPT isn't bad.

[00:14:35]Can something be both real and a bubble at the same time? I think so. The dot bubble was was real. It happened. But the internet did end up changing everything. The bubble doesn't care if the tech is real. The bubble cares about the money, not about the product. AI isn't fake. The money surrounding it has just completely lost all contact with reality. And maybe maybe that's the whole point. Quick shout out and thank you to my channel members, John W. and Ryan. Join the other channel members down below. And if you like this video, consider watching my previous video on why AI is completely making CEOs delusional. I will see you in the next