After his co-host argued cheap open models could crash AI prices, Campbell compares today's AI labs to the dot-com era. His revenue and spending figures are rough, unsourced estimates offered as such.
And they're good enough. Quinn
…a race to the bottom, all these big companies that are currently propping up the world economy are going to tank. Right. And so they're in that, that's a big problem for the economy. Now, for people who want to take advantage of AI at a lower price, it's good because the prices are going to fall.
The bigger challenge you've got are these open AIs that are, you know, maybe bringing in 50 billion a quarter, but they're spending 100. Yeah. And so, and they're raising their rates to try and get those numbers to make sense. And it's driving customers over to the free models. And so you get into a situation where, you know, the debate we've always had is: are they Netscape or are they Google? Both existed during the dot-com boom. Netscape have vanished, having created the dot-com boom. And Google went on to be relevant. And at this point, with these numbers, they are all Netscape. The companies that are going to survive are the hyperscalers, the ones building out the data centers, Microsoft, Google, Amazon. They appear to be fine.
But the cost of new data centers, or the building of new data centers, is practically going to be non-existent. It'll probably stall. It's the ones that are, yeah, the ones that are existing and owned.
Well, and I've done this talk where I said what we're really seeing by those big three is the same thing that the telcos did in 2000. Buy up all of the locations where you can put cell towers so there will never be a cell tower competitor again.
Yeah.