Been thinking about this a lot lately—everyone's worried AI will trigger another dot com crash like 2000, but honestly, I think they're looking at the wrong sector.

Let me break down what I'm seeing. Back in 2000, the dot com bubble was pure supply-side madness. Companies with zero profits, zero cash flow, yet their stock prices were going vertical. Meanwhile on the demand side? Almost nothing. No Douyin, no Alibaba, no WeChat. People were using OICQ and MSN for chat, downloading files with FlashGet, checking Yahoo for news. That was basically it. Most Chinese families didn't even own a computer yet—we're talking CRT monitors and floppy disks with 1.44MB capacity. The infrastructure just wasn't there.

Now look at AI. Yes, Nvidia and AMD's stock prices have surged, but these companies are actually profitable. They're selling real infrastructure that AI genuinely needs. On the demand side, we've got Gemini, Claude, GPT, Doubao, Perplexity, and all these AI applications already gaining real users. The infrastructure has to come first before applications can scale—that's just how it works.

But Web3? That's where I'm seeing the real echo of a dot com crash scenario.

Take a look at the actual numbers. ZKsync is pulling in $458 daily revenue but trading at $176M market cap—that's a PE ratio of 1052. Optimism doing $2,427 daily with $253M market cap? PE of 285. Sei at $3,564 daily revenue and $424M market cap? That's 327 years to break even. Blast is actually negative. These valuations make zero sense when you do the math.

And here's the thing that really gets me—what are people actually using Web3 for? DeFi, RWA, some meme platforms, prediction markets, Perp DEX. That's basically it. Everything else we talked about—metaverse, blockchain games, inscriptions, social platforms—they're all ghost towns now. Compare that to 2000 when at least there was Amazon, eBay, Google Search, email. Web3 has way fewer functional applications, yet the supply side is completely self-congratulatory about the technology.

So here's my take: we're not repeating the dot com crash with AI. We're repeating it with Web3. AI has real demand pulling the supply side forward. Web3 is pure speculation on technology with almost no actual user adoption to justify the valuations.

If I'm right about this, then the US stock market probably corrects but doesn't crash like 2000. Bitcoin takes a moderate hit since it's correlated with equities. But altcoins? They're going through a long, painful shakeout—separating the projects with real utility from the garbage. This process might take years, not months. Don't believe anyone telling you the alt season is over or that we've hit bottom. Some projects might be finished, but plenty of others have way further to fall.

The real question isn't whether AI is a bubble. It's whether Web3 can actually build something people want to use. Until then, we're just watching a different version of the dot com crash play out in real time.