Peter Zeihan

The Chinese problem is that they've run out of people under age 50 and it's people under age roughly 45 that do the consuming and have the kits.

And there is no way that AI can help with consumption or child rearing.

So the robotic systems that the Chinese are working on, not that they're not important,

But AI only helps to a degree there.

AI cannot physically move things.

It can learn from systems.

It can design systems even to a degree, again, 80-20.

But it can't actually produce.

Artificial intelligence is a completely different technological suite from automation.

And even if automation could solve the production side of the equation as the Chinese run out of workers...

Robots don't pay taxes and they can't raise kids and they can't consume product.

And so even if the Chinese could maintain their production levels without people

they'd still be dependent on international trade that they can't guarantee and on the largest of the United States in the long term.

It doesn't change their core problem.

50 years, that is the right timeframe.

I haven't talked to anyone...

who's involved in Silicon Valley at all, who expects us to get general thinking AI before the 2040s.

And based on what's going on with the large language models, that keeps getting moved back.

This is not a technology that's leading us in that direction.

So, yeah, there are some... Considering that you can fit the entire algorithm set and training data for ChatGPT on a thumb drive that's about half the size of this...