Tulips and LLMs

Are we in an AI bubble? It depends on how you define it.

Tulips and LLMs

Lately, a lot of us have been reminded of “the time of great happiness” (better known as the dot-com bubble).

I had a front-row seat. From 1997 to 1999 it was a joyride; in 2000 the wheels came off. The parallels to today are hard to miss. And yet we shouldn’t be afraid. The dot-com bust, in hindsight, was less a graveyard than a nursery. Yes, Webvan, Pets.com, and Toys.com all ended in tears. But the internet matured and birthed trillion-dollar enterprises (e.g., Amazon, Google) and created entire new sectors (e.g., Netflix, Uber). And it rewired global services.

AI is bigger than the Internet, and far more important. But, like 1999, no one knows how. The bets today are outsized, concentrated, and mostly wrong. They’ll burst, just as Pets.com did, because people are grafting old models onto new technology instead of asking the only real question: what does the new technology actually want?

Five years from now, the answers will look obvious. Today, they’re anything but. So yes, I’m bullish on the AI trade. But in the short run? There will be blood.

Agent of the week: Xbow

Automate vulnerability discovery and enhance security testing.

And one more thing…

Reply

or to participate.