This is an IHeart podcast.

Guaranteed Human.

This new year swaps someday with today.

You've already got the idea, so make it real in 26 with a pro-level website.

GoDaddy makes it easy with the lowest-priced website builder, guaranteed.

Ready to build a confetti-ready site in minutes?

New customers start for free at go-dadi.com.

That's go-dadi.com to get the lowest-priced website builder, guaranteed.

CawoMedia. Chosen by God perfected by science.

I'm Ed Zittron.

This is Better Offline.

And as I've written about many, many, many, many times and argued on this very podcast just as often, the large language models run by companies like OpenAI, Anthropic, Google and meta, are unprofitable and unsustainable, and the transformer-based architecture, they run on as peaked. They're running out of training data, and the actual capabilities of these models were peaking as far back as March 2024.

Nevertheless, I'd assumed, incorrectly, by the way, that there would be no way to make

them more efficient, because I had assumed, also incorrectly, that the hyperscalers, along

with Open AI and Anthropic, would be constantly looking

for ways to bring down the ruinous cost of their services. After all, Open AI lost $5 billion last

year, and that's after $3.7 billion in revenue, too, and Anthropic lost just under $3 billion in

2024. And in the last episode, I told you a little bit about Deepseek, by the way,

in this one we're going to get into, well, how fucked things might actually be. But what I didn't

wager was that potentially nobody was actually trying to make these models more efficient.

My mistake was, if you can believe this, being too generous to the AI companies, assuming

that they didn't pursue efficiency because they couldn't, and not because they couldn't be bothered.

...