WHAT DOT-COM BUBBLE?
Everyone wants to know if AI is the next dot-com bubble. But here’s a better question: what dot-com bubble?
I know, I know. We all “know” the story. Irrational exuberance. Pets.com sock puppets. Investors throwing money at anything with “.com” in the name. The NASDAQ soaring to 5,048 before crashing 78%. The cautionary tale we’ve been teaching for 25 years about the dangers of ignoring fundamentals.
Except here’s the thing: the NASDAQ today sits at over 22,500. That’s 4.5 times the supposed “bubble peak” and 30 times when the “bubble” started in 1995.
So either we’re in an even bigger bubble now - or maybe the dot-com era wasn’t actually a bubble. And many individual companies did spectacularly even if you bought them at the height of the “bubble.”
What Actually Makes a Bubble
A bubble is something that when it collapses, there’s nothing there - or what’s there can never satisfy its original hype. What happens when a soap bubble pops? That’s a bubble.
Dutch tulip bulbs in 1637: prices soared to a year’s wages per bulb. When it collapsed, you had... bulbs worth pennies. Tulips could never justify those prices because anyone with dirt could grow more tulips.
The dot-com era? When it “collapsed,” we still had Google, Facebook, Amazon and $500 billion in fiber optic infrastructure, productivity growth that transformed the economy, and the foundation for cloud computing, streaming video, and mobile internet.
Here’s where people get confused: there’s a popular misconception about the word “risk” in investing. If you invest in one stock, that’s true risk of permanent loss - the company can go to zero and your money is gone forever. But when talking about stocks in general or broad indexes, “risk” usually means volatility - prices moving up and down around a generally upward trend line, their true value.
The dot-com era had massive volatility. Individual companies went to zero. But the broad market was bouncing around an upward trajectory because the underlying value creation was real.
The Gold Rush Wasn’t a Bubble Either
During the California Gold Rush, thousands rushed west. Most individual prospectors failed or barely broke even. But nobody calls it a “gold rush bubble.” Why? Because the gold was real, the fortunes were real, and the infrastructure built - railroads, banks, supply chains - created lasting value far beyond the gold mined.
In the early 1900s, over 2,000 companies tried to build automobiles. More than 99% failed. But nobody calls it the “auto bubble.” We recognize it as genuine transformation with brutal competition and concentrated winners.
The dot-com era followed the same pattern. Hundreds of search engines, dozens of social networks, countless e-commerce sites. Most failed. But Google, Facebook, and Amazon created trillions in value and transformed how humans live and work.
This is what genuine transformation looks like. During transformational times, there’s usually not much question about the tremendous future impact of something. The internet was clearly going to be huge. The question wasn’t “will this matter?” but “which of these hundreds of companies trying to get through the same door will survive and how much will it all be worth in the end?”
That’s not a bubble. That’s brutal competition during genuine transformation. If you want to be in early, you have to be a good judge of what’s happening. But the opportunity itself is real.
Knowing that the internet was going to be transformational did not give anyone free license to think that any company touching that space would be a goose laying golden eggs.
If you think that knowing something that tens of millions of people reading the newspaper and watching cable TV already know is a sure path to making you rich, then you are who they are talking about when they say “a fool and his money are soon parted.”
The Numbers Don’t Lie
Let’s look at what actually happened to the NASDAQ:
From 1995 (bubble start): 750 → 22,500 = 30x gain From 2000 (bubble peak): 5,048 → 22,500 = 4.5x gain
If it was a real bubble, the NASDAQ should have collapsed back toward the 750 starting point and stayed there, like tulip bulbs returning to pennies. Instead, we’re 30 times higher than when the “bubble” started and still 4.5 times above the “peak.”
But those are broad indexes. What about individual companies? Many did fantastic, even if you bought way into the “bubble period”:
Amazon: $1,000 invested in 2000 would be worth roughly $50,000 today. Yes, the stock fell 95% from 1999 to 2001 and didn’t return to its 1999 high until 2009. But even buying at the peak, you’re up 50x.
Microsoft: $1,000 invested in 2000 would be worth about $12,400 today with dividends reinvested. The stock fell 65% in the first two and a half years and took 14 years to regain its 2000 peak price. Still up 12x from the bubble peak.
Apple: $1,000 invested in late 2000 - when Apple traded below its cash value at less than 1x earnings - would be worth over $213,000 today.
These aren’t bubble collapses. The volatility was painful. The opportunity was real.
When Bubbles Are Actually Bubbles
Not everything is a transformation in disguise. Some things really are bubbles. Let me show you what an actual bubble looks like.
Cryptocurrency’s market cap currently sits around $3.27 trillion. When this bubble collapses - and it will - what will be left? Several trillion dollars will just go “poof” and disappear into the air. It’s just a fiat currency with no government backing or regulation.
It won’t be worth the electricity to read its ledger.
Does Anyone Think AI Won’t Be Transformational?
It’s clear it will be. How fast? What will be the end outcome? There are many questions and so there is no certain way to valuate the opportunity.
Most likely 80% of the jobs that exist today will vanish. Everything will be different 5 or 10 years from now.
Existing companies will have tremendous savings in terms of salaries and increases in productivity. That will certainly drive their profits through the roof. Companies that don’t adapt will disappear. Where is Best Buy and Sears and many other places that failed to adapt?
Transformation Doesn’t Wait For Certainty
Here’s the uncomfortable truth: if you need to know it’s a sure thing before you invest, you’ve already lost. By the time it’s obvious, the returns are gone.
A fair number of forward-thinking investors made fortunes in the dot-com era and are making them again in AI. They weren’t lucky gamblers riding irrational exuberance. They correctly identified transformation early, understood the uncertainty was fundamental rather than informational, and bet accordingly. Many of them are the same people both times.
The stock market is not a bank account with predictable interest. It’s a mechanism for allocating capital to an uncertain future. Genuine transformation creates genuine uncertainty - about timing, about winners, about the trajectory of change itself. If you could predict it with certainty, the returns would be Treasury-bond level. The spectacular gains come precisely from the spectacular uncertainty.
So you have three choices:
Try to pick the winners. You’ll be spectacularly wrong often - dozens of Friendsters before Facebook, scores of AltaVistas before Google. But if you’re right about the transformation itself, being right about some individual picks can pay off enormously.
Try to avoid turbulence and bad timing and wait for reasonable valuations. You’ll be “right” about the crashes and feel validated when Pets.com fails. You’ll also miss the transformation entirely.
Buy the S&P index and accept that you’ll own both winners and losers. This is what Warren Buffett tells everyone else to do. It’s boring, it requires no skill, and historically it’s captured every genuine transformation while surviving real loss and excessive volatility.
Warren Buffett understood this trade-off. He stayed out of Google, Microsoft, and Amazon not because he thought they were overvalued bubbles, but because he knew he didn’t understand them well enough to value their business and to sleep at night. That’s honest self-awareness. But the cost was missing some of the greatest wealth creation in modern history. He publicly regrets not investing in Google since he understood the business.
The Real Lesson We Missed
We taught ourselves that the dot-com era was about irrational exuberance and ignoring fundamentals. We’ve spent 25 years telling investors to avoid “bubbles” and wait for “reasonable valuations.”
The actual lesson is the opposite: genuinely transformative technology creates fundamental uncertainty that looks like irrationality. The winners create spectacular value but they don’t wait for traditional metrics to catch up. By the time Amazon had predictable cash flows, you’d missed 50x. By the time Google fit into value-investing frameworks, you’d missed 30x.
So Is AI a Bubble like the Dot-com bubble?
Which brings us to today’s question about AI. But we’re asking the wrong question. We’re comparing AI to something that wasn’t actually a bubble - it was a genuine transformation that markets couldn’t efficiently price because the uncertainty was fundamental, not just informational.
I’m not predicting how the stock market will look in 10 years in regards to AI or whether you should be jumping in.
Meta, Google, Microsoft, and Amazon aren’t irrationally dumping hundreds of billions into AI. They learned the real lesson of the dot-com era: it wasn’t a bubble, it was a sorting mechanism. And if you’re not willing to bet big on potential transformation, you get sorted out.
Here is the real irony: while we are flooded in the news with hand wringing over the AI bubble, cryptocurrency hardly is a footnote and is even championed by our current administration.
The cautious approach that avoided the “bubble”? It’s produced index-level returns for 20 years. The “overvalued” index that owned all that supposedly-crazy tech? It captured one of the greatest wealth-creation periods in human history.