AbstractH24

Feels to me like local models are an under-covered aspect of this whole AI boom.

If everything improves over time, at some point a good chunk of tasks won’t need to be done in data centers or be subject to the whims of a few frontier AI labs.

How close are we to that? Or is my thinking flawed?

Clearly AI companies are gradually shifting to focus on that.

In the 90s was there any effort to do that BEFORE the bubble burst?

I was a kid and remember outlines like the Priceline for groceries and insane IPOs, but not much else.

Always interested in parallels and what can be learned.

I understand that breakthroughs such as multiplexing increasing efficiency caused a huge glut in the amount of fiber laid. But what prevented demand from increasing to use it?

I ask, wondering what will be the thing that causes the demand for data centers and processing power to outpace supply as GenAI gets more efficient. Rather than people just using more GenAI.

It's reminiscent of many other things where people tell you to believe them, but can't let you see the evidence.

And so happens they are starting the process of going public at a time when the burn rate of these companies is nuts.

Don't get me wrong, I use LLMs constantly in my work every day now. They've changed so much of what I do. But also, we all know both the AI bubble and the economy at large are at the brink of a correction, and preventing that is requiring extreme things.