Remember Pets.com? Sock-puppet, no profits, all hype. 2000’s cautionary tale. Back then, everyone chased clicks, not cashflow. Today’s AI mania rhymes — but it doesn’t repeat.

Microsoft, Nvidia, Google, Amazon — these aren’t vapor startups. They’re profit engines spending record cash on compute. The question isn’t if AI changes the economy, it’s when the market stops pretending it won’t wobble first.

The Spend vs. the Return

Big Tech will pour roughly $400 billion into AI this year. MIT says only 5% of projects deliver measurable gains. That’s not fraud — it’s friction. Scaling AI takes time, data, and compute discipline.

Still, the hype’s hot. Nvidia trades north of 30× forward earnings, and the street’s split — some call it unsustainable, others call it the new oil. I’m somewhere in the middle: this is a cash-flow story, not a fantasy one.

This Time, the Players Have Balance Sheets

In 2000, dot-coms borrowed from the future. In 2025, Big Tech funds the future out of free cashflow. Microsoft and Google can burn billions on GPUs and still out-earn half the S&P. That’s not bubble territory — that’s strategic reinvestment.

Even the so-called “circular” deals — Nvidia investing in OpenAI, Meta borrowing for data centres — look risky on paper, but they’re building capacity, not just headlines. It’s not tulips; it’s infrastructure.

Would I Buy Nvidia? Absolutely.

If AI valuations correct, I’m not panicking — I’m doubling. Nvidia’s grip on compute is monopoly-grade, its margins obscene, and the demand curve still points up. Cloud firms can’t train without it. If that stock drops 20-30%, that’s not fear — that’s opportunity.

Same goes for Microsoft. Every enterprise AI layer sits on Azure. You’re buying the pipes, not the posters.

The Real Risk

The risk isn’t that AI collapses — it’s that investors confuse volatility with failure. AI adoption will be uneven. There’ll be write-downs, dud projects, even layoffs. But the long-term capital cycle is intact.

When Pets.com fell, the web didn’t die — it matured. AI will do the same. And those who kept dry powder through the noise ended up owning Amazon at $6.

Kyri’s Take

This isn’t the dot-com collapse. It’s the build phase of the next infrastructure wave. If AI stocks dip on weak earnings or “bubble” talk, I’m not trimming — I’m scaling in.

Because bubbles only matter when they burst before the revenue arrives. This time, the revenue’s already flowing.

Takeaway:
It’s not about dodging hype. It’s about knowing when hype meets cashflow — and being ready when it does.