Webvan did not make any money.
Kozmo.com cratered.

This did not mean that universal networking of computers was a bad idea, nor did it mean that grocery delivery was a bad idea. Both concepts survived the initial stupidities and implosion.

The wastes of money that were the first versions of these companies were irrelevant to the eventual successful variations, and the successful *important concept* that could not be controlled.

The *universally networked computers* concept was the big one. The yeah woo we can deliver groceries concept was a nice little capitalistic by-product.

I am going to predict right now, in a way that will make people hate me, that the neural-net-like-analogue of brain behavior that is LLMs is a concept that is worthwhile, important, revolution-creating, and going to survive the popping of the economic bubble created by VC stupidity and the destruction of whatever stupid companies get created by the likes of stupid people like saltman.

This is what will last; this is what will be worth while. Hold onto this concept, because it's going to be what is still here in five years, even if you hate it today.

IMO Anthropic is running away with this race right now; Google might pull out a mediocre second place. I’m not sure it's worth discussing third place.

I also don't have any idea what's going to happen in two months, so I am at peace with being wrong. Anyway…