The AI Investment Frenzy Looks Exactly Like the Dot-Com Bubble — And $4 Trillion Is on the Line

I've been thinking about this all week and I can't shake it. Alphabet just completed an $80 billion equity raise. SoftBank is blowing $52 billion on European data centers. Anthropic quietly filed for an IPO that could value them near $1 trillion. OpenAI and SpaceX are also eyeing trillion-dollar listings. Analysts are now projecting these three IPOs alone could add $4 trillion to US equity markets.

And someone finally said what a lot of us have been thinking: this looks a lot like 1999.

The Numbers That Should Make You Pause

Let's just stack the figures up and stare at them for a second.

Anthropic's revenue run-rate hit roughly $47 billion in May 2026 — up from about $10 billion the year before. That's nearly 5x growth in 12 months. Impressive? Absolutely. But its valuation is sitting at ~$965 billion. That's a price-to-revenue multiple of about 20x on a run-rate basis. For a company that isn't yet public. That is burning money. That is operating in a market where its primary competitors — OpenAI, Google DeepMind, Meta AI — are backed by entities with effectively unlimited capital.

Meanwhile, SoftBank is writing a $52 billion check for European data center infrastructure. Not to buy AI companies. Just to build the buildings the AI companies will run on. That is a bet of almost incomprehensible scale on the assumption that AI demand will not just sustain, but exponentially grow for the next decade.

And Alphabet — one of the most profitable companies in history — felt it needed an $80 billion equity raise. That's not a company struggling for cash. That's a company signaling that the capital requirements of the AI race are so vast, even Google needs to go back to the market.

Why the Dot-Com Comparison Actually Holds Up

The dot-com crash wasn't caused by the internet being fake. The internet was real. The technology was real. The transformation was real. What caused the crash was that investors priced in 30 years of growth in about 3 years, funded companies with zero path to profitability, and ignored the basic question: when does this actually make money?

Sound familiar?

To be clear, AI is real. The productivity gains are real. I use these tools every day and I've seen what they can do. But there's a difference between "this technology will change the world" and "this specific company at this specific valuation is a good investment right now." In 1999, people confused those two statements constantly. There are signs we're doing it again.

Quantinuum — Honeywell's quantum spinoff — is seeking a valuation of up to $14.3 billion in an expanded IPO, planning to raise $1.46 billion. Quantum computing is promising, but commercially viable quantum at scale is still years away. The market is pricing in a future that may be real but is being priced as if it's certain.

What's Different This Time (And Why It Might Matter)

Here's the honest counterargument: the companies leading this wave are not Pets.com. OpenAI is generating real revenue at scale. Anthropic's $47 billion run-rate is not imaginary. Google's AI integration into Search, Workspace, and Cloud is already driving measurable returns. The infrastructure build-out — chips, data centers, power grids — is creating genuine, durable assets that will have value regardless of which AI model "wins."

Also, the Fed isn't running zero interest rates anymore. Cheap money isn't flooding into speculative bets the way it was in 2020-2021. The companies going public are doing so at mature revenue scales, not on vibes and PowerPoints.

But none of that means the market can't get ahead of itself. And right now, the market is pricing in a scenario where AI revenue grows to justify $4 trillion in new equity value — on top of valuations that are already stretched by any historical standard.

My Take

I think AI is the most transformative technology since the internet. I think the companies building it will create enormous value over the next 20 years. I also think a significant portion of the current investor excitement is pricing in scenarios that are optimistic even by bull case standards.

The dot-com bubble didn't destroy the internet. Amazon survived. Google survived. But investors who bought Webvan, eToys, or Boo.com at peak valuations lost everything. The technology won. Most of the bets didn't.

We're in an AI bubble. That doesn't mean it pops tomorrow. It doesn't mean the underlying technology is overhyped. But it does mean you should be thinking carefully about which companies are actually building durable value — and which ones are just surfing the wave.

What's your experience? Drop a comment below! 👇

Do you think we're in an AI bubble, or is this time genuinely different? And which AI company do you think survives a potential correction — and which ones don't?