plissken

This! Yes. 100% agreed. Of maybe 98. I don’t know if the large data center model ends, nor if he knows that.

I do know that Nvidia topline growth story cannot grow more than it’s done with frontier models. It can though keep growing if local inference or anyway distributed inference becomes a thing. But if the market doesn’t believe that, as frontier labs IPO it’s Nvidia stocks that get sold otherwise.

Finally coming out of the bloody closet. Frontier labs and Nvidia are in death dance.

The latter sells mostly to the former, but their growth story needs local inference to keep up with valuation. Nvidia could reduce prices on a whim and boost local inference, but that would make its investment in frontier labs lose too much.

Frontier labs meanwhile need public markets money, and their story really needs local inference to be more expensive/extremely worst than what they offer.

One narrative takes money away from.one side and puts it in the pockets of the other.

I assumed cc details are validated. If all one need to do is to provide imaginative cc details then I need no further explanation

Now that’s something I would like to learn, how?