>There isn’t a rule of economics that says better technology makes more, better jobs for horses. It sounds shockingly dumb to even say that out loud, but swap horses for humans and suddenly people think it sounds about right.

--CGPGrey, Humans need not apply.

You need a paradigm shift in your mind on why the modern world looks like it does. You don't need human consumers, you just need a consumer. Any system that allows you to get the hard resources you need to produce the hard/soft resources required is simply enough. Humans are fungible for anything else that can provide manual or intellectual labor.

As a thought excercize, just imagine a bunch of AIs/robots buying/selling/trading resources between each other. Where are humans required in this?

In this scenario, where the AI and robots no longer rely on human labor for maintenance and growth, their productive capacity exclusively serves the owning elite (including defending them with violence if necessary) and the rest of us are an inconvenient growth occupying land and consuming resources.

This is a scenario where the AI/capital owners complex has already survived the collapse of the consumer economy.

What is the point of producing all this stuff now?

Historically production was transactional. You give me something, I give you something. But along the way the average Joe ran out of things to offer in return. Businesses give, but increasingly fail to receive in kind. Apple, for example, produced in excess of $50 billion dollars worth of value that they've never been able to get anything in return for. In other words, they have effectively given away $50 billion dollars worth of stuff away for free and have shown no signs of wanting to stop.

At least there is no direct transactional value. There is social value. When you've given away $50 billion dollars worth of things, the masses start to idol you. That is why people, like those who oversee Apple, are willing to produce all that stuff. You get social access not afforded to the average Joe. You can do stupid Epstien-style crap without repercussions. You get to live a different life even when you aren't directly getting anything in return. That, no doubt, will remain the point of producing stuff in the future.

There will be no point, and the stuff that normal people use will become more expensive as resources are more and more directed to megaprojects that the capital class is interested in. More modern equivalents of pyramids and extravagant castles and less consumers goods.

Presumably the captains of the universe will want ever increasing luxuries and expressions of their power.

Why wouldn’t an advanced AGI robot, trained on human behavior, not want their own house and mode of transportation? Sure it’s basically kayfabe for them to ‘want’ the stuff we do but if we’re following the script of who will buy all the stuff, then the answer will be the robots I guess.

You think housing market are tough now, wait until you’re competing with 5 robot families who all have jobs you used to do.

You act like I produce the dollars I spend. No, the government does, who debases our currency to print more.

You are right.

The novel is a bit of a dark comedy sci-fi. And even though many details are surprisingly accurate (for a story written a decade ago) robots buying crap produced by robots is IMO meant to illustrate an absurd racket to inflate demand.

This is true about LLMs themselves but the developments behind them have been a boon for robotics. I’m mostly familiar with computer vision so I can’t speak to everything, but vision transformers (ViTs is the term to search for) have helped a ton with persistence of object detection/tracking. And depth estimation techniques for monocular cameras have accelerated from the top of the line raw cnn based models from just a few years ago; largely by adding attention layers to their model.

I agree that they’re not there yet but I don’t want to discredit the benefits of these recent advancements

While you're correct we still need a lot more, the advances in the past 5 years represent more than I've seen in most of my life.

Just look at the speed in which we can train a humanoid robot things now. We can send out a mo-cap human, get some data, and in few hours run a few hundred trillion simulations, and publish a kernel that can do that task relatively well.

LLMs allow us any perception at all. They feed vision to scene comprehension an then let the robot control part calculate a plan to achieve a goal. It's not very fast, and fine motor controls have a long way to go, but it is possible.