This is an I-Heart podcast.

Guaranteed human.

Quarzo Media.

Hello and welcome to Better Offline.

I'm your host, Ed Zittron.

Better Offline. Let's our slide. In the last episode, I walked you through how Generative AI has hit its ceiling,

and how big tech has burned billions of dollars chasing a technology

that really isn't going to give them the returns they got from the smartphone and cloud computing revolutions.

Not that that's going to stop them.

In this episode, I'm going to walk you through the potentially massive problems they have ahead,

the ones they've created by building these massive supercomputers,

the ones necessary to capture this imaginary demand of this kind of bullshit AI revolution,

and I'm going to try and guess what might happen next.

All right, so, the AI boom helped the S&P 500 hit record levels in 2024,

largely thanks to chip giant Nvidia, a company that makes both the GPUs necessary to train

and run generative AI and the software architecture behind them, as well as some of the server

parts too.

Part of Nvidia's remarkable growth has been its ability to capitalize on Kuda architecture,

the software layer that lets you do complex computing on GPUs, rather than simply using them to render video games or complex 3D images.

Someone's going to email and say that that's partially wrong, just bear with me.

And of course, one of Nvidia's other things they do is continually create new GPC going to email and say that that's partially wrong, just bear with me.

And of course, one of Nvidia's other things they do is continually create new GPUs to sell for tens of thousands of dollars to tech companies that want to burn billions of dollars on generative

...